General
1. Method. Every letter was translated directly from 300 dpi page images of the 1921 printing, not from OCR alone, following the same method used for New Studies on Marcion, and each was then verified line by line against those images before being set.
2. Density of the apparatus. Harnack’s notes throughout Beilage III are far more compressed and citation-dense than in New Studies, closer to a critical edition’s apparatus criticus than to discursive scholarly prose. This is why each letter was verified separately against the page images rather than in one pass at the end: at this density, negations and clause boundaries are the first things to be lost.
3. The reconstructed text itself. The verse-by-verse Marcionite text (set in bold verse numbers) is the primary payload of every letter and is translated with the most care. Where Harnack marks a reading as uncertain (a following “?”), that uncertainty is carried into the English. Where he states outright that a passage is unbezeugt (“unattested”) and gives his own judgment on whether it was probably present or absent, both the fact of non-attestation and Harnack’s judgment are preserved distinctly, since the two are not the same thing.
4. Greek verse and chapter numbering. Harnack cites by chapter and verse in the modern (canonical) numbering throughout, even where reconstructing a text that omits the verse in question; this convention is preserved.
5. Manuscript sigla (D, G, g, d, N*, Fgr*, etc.) are reproduced as printed throughout, without expansion, matching the convention used for similar sigla in New Studies. A reader cannot identify the specific manuscripts without Harnack’s separate Prolegomena, which fall outside Beilage III and IV and are not included in this volume.
Notes on Galatians
6. What this letter shows, in brief. Marcion’s Apostolikon retained the great majority of Galatians essentially as it stands in the canonical text. The passages Harnack judges certainly or probably absent are concentrated in three places: (a) Paul’s account of his early dealings with the Jerusalem apostles and the churches of Judaea (1:18–24, part of 2:6–9a); (b) the extended argument from Abraham’s faith and the law in chapter 3 (3:6–9, most of 3:10–12, all of 3:15–25); and (c) the Isaiah citation about the barren woman at 4:27–30. The single largest editorial intervention Harnack identifies is not a deletion but a rewriting: the allegory of Sarah and Hagar at 4:21–26, which in Marcion’s text culminated in a description of “the holy church” as “our mother,” built in part from language imported from Ephesians. Harnack himself registers real uncertainty about how much of the transmitted wording of that passage belongs to Marcion and how much to later corruption in Tertullian’s own text; this is flagged, not resolved, in the translation above.
7. Questions left open. Four items remain flagged as uncertain rather than corrected, because the source itself is ambiguous rather than the translation being clearly wrong: (a) Tertullian’s compressed Latin on 1:7 (the “Est autem etiam dei novi…” passage), where the compressed Latin admits more than one defensible construal: the present rendering follows Harnack’s syntax but should not be treated as eliminating the underlying ambiguity; (b) the 4:21–26 Sarah/Hagar allegory and its “ostensio”/διαθήκη problem, which is the single passage most likely to be quoted for its “holy church… our mother” reading and where the ambiguity in the transmitted wording remains material to any use made of it; (c) the exact accent on ἡ/ᾗ at 5:1a; (d) the unexpanded manuscript sigla, which require Harnack’s untranslated Prolegomena to identify fully. None of these affect the letter’s overall shape as summarized in note 6.
Notes on 1 Corinthians
8. A far denser apparatus than Galatians, but a similarly conservative reconstruction. 1 Corinthians runs to 18 printed pages of source (pp. 77*–94*) against Galatians’ 13, but the greater length is mostly apparatus, not additional excision: across chapters 1–3 and 15 the reconstructed text tracks the canonical wording closely, verse by verse, with only scattered, mostly minor word-order and small-word variants (many already independently attested for other Western witnesses, not distinctively Marcionite). The clearest exceptions are chapters 7–9 and parts of 10–14, which Harnack marks as attested only by “brief allusions” (kurze Anspielungen) rather than continuous quotation, meaning the surviving evidence tells us these chapters were present in some form, but not their exact wording, and Harnack explicitly flags ch. 7 as “perhaps abridged” (see the letter’s closing remark on p. 94*). Chapter 16 has no surviving attestation at all, positive or negative.
9. Distinctive readings worth flagging for a reader. Three points recur across the notes and are worth surfacing rather than leaving buried in the apparatus: (a) at 10:9, Marcion’s Apostolikon read “tempt Christ” where the canonical text has “tempt the Lord.” Epiphanius states this explicitly as a deliberate Marcionite change; (b) at 15:38 a later Marcionite correction (attested only as a marginal/interpolated note, not in the main line of transmission Harnack reconstructs) substituted “God gives it a spirit” for “God gives it a body,” evidently to avoid conceding a bodily resurrection. Harnack treats this as secondary, after Marcion’s own text, and the translation preserves that distinction; (c) at 15:49, the reading “let us bear [the image]” (hortatory) against “we shall bear it” (a future-tense promise) is explicitly called out by Tertullian himself as “praeceptive, non promissive,” a distinction the Latin marks grammatically that the notes above try to preserve with “let us” versus “we shall.”
10. Questions left open. (a) Tertullian’s compressed Latin at V, 10 on 15:45–47 (the “dominum enim posuit novissimum pro novissimo Adam” clause) is elliptical even by his standards and the rendering above follows Harnack’s syntax without claiming to eliminate the alternative construals the clause admits, the same caveat as for Galatians 1:7; (b) the precise sense of “ensouled” versus “soul” as the reading behind 15:44 (Harnack’s own uncertain emendation of Tertullian’s “anima” to “anima‹le›”) is preserved as Harnack states it, not resolved; (c) manuscript sigla are again left unexpanded, per the general convention in note 5.
Notes on 2 Corinthians
11. The shortest letter so far, and the least altered. 2 Corinthians runs only 7 printed pages of source (pp. 94*–100*, the shortest letter treated to date) and Harnack states outright, in his own closing remark, that “no larger excision is attested” in this letter, the strongest statement of textual conservatism given to any letter so far. Much of the letter (all of chs. 9 and most of 6, 12, 13) is attested only by isolated verses or brief allusions rather than continuous quotation, which is a property of what survives in Tertullian’s compressed second-half treatment of the letter (he says so himself, and Harnack notes it; see Translator’s Note 12), not evidence that Marcion himself cut those chapters down.
12. The chapters 7–11 gap, addressed directly by Harnack. Tertullian’s citations jump from 7:1 to 11:2 with no material from chs. 8–10 except an isolated 10:18. Zahn had argued Marcion deleted the intervening four chapters outright; Harnack’s note on p. 99* gives five separate reasons against this (Adamantius attests material from ch. 10 independently; there is a demonstrable physical gap in Tertullian’s own text at exactly this point, of unknown size; deleting four whole chapters is not credible against the pattern of what Marcion kept elsewhere in Romans; Tertullian treats the whole second half of 2 Corinthians unusually briefly; and a 73-verse gap of this kind would be the sort of thing Tertullian would ordinarily flag as such, as he does elsewhere). This is Harnack arguing against a specific prior scholarly claim (Zahn’s), not simply reporting an absence, and the translation preserves the argument in full rather than compressing it to a summary judgment.
13. Questions left open. (a) The dense note on 3:14–15 (p. 95*), which reconstructs a homoeoteleuton in Tertullian’s manuscript tradition and proposes wording for the lost material, is Harnack at his most speculative and philologically technical in this letter; the translation preserves his hedged, provisional phrasing (“one could imagine…”) rather than presenting the reconstruction as settled. (b) The extended five-point argument on p. 99* against Zahn’s proposed deletion of chs. 8–10 is translated in full rather than summarized, since it is Harnack arguing a specific scholarly point rather than simply reporting the apparatus; a reader working through Marcionite scholarship on this letter should read it as an argument, not a list. (c) Tertullian’s compressed Latin at 5:1–4 (p. 97*, “‹Loco› terreni domicilii nostri…”) is a polemically ordered argument, not a plain citation, and is flagged inline as provisional, since the clause admits more than one defensible construal. (d) Manuscript sigla remain unexpanded, per the general convention in note 5.
14. Where this letter’s apparatus is most treacherous. Two places in 2 Corinthians turn on the direction of a variant rather than its content, and both are easy to reverse in translation: the ἐκδυσάμενοι/ἐνδυσάμενοι crux at 5:3, which appears in both the reconstructed text and Harnack’s own note, and the “this is missing… against” formulation at 5:4. Both were checked against the page images in each place they occur, and the Latin citations surrounding them were compared clause by clause, since Harnack’s compression here makes it easy to drop a subordinate clause without the English reading as though anything were missing.
Notes on Romans
15. By far the most heavily excised letter so far. Romans runs 10 printed pages (pp. 100*–109*) and Harnack’s own closing tally states plainly that Marcion struck out or did not offer roughly six chapters’ worth of material: 1:17b; 1:19–2:1; 3:31–4:25; 9:1–33 (very probably, given the size of the gap in Tertullian rather than direct proof); 10:5–11:32; and all of 15 and 16. This is a different order of editorial intervention from anything seen in Galatians, 1 Corinthians, or 2 Corinthians, where the letters survive close to entire. A reader should not expect Romans to read as a near-complete letter with a few variants, the way the earlier three did; large stretches of the canonical text (above all Paul’s argument about Israel, election, and the law in chs. 9–11, and the Abraham/law argument at 3:31–4:25, paralleling the similar cut in Galatians 3) are simply not in Marcion’s Bible as Harnack reconstructs it.
16. The famous missing half-verse at 1:17. Harnack’s note states that Tertullian’s citation runs 17a directly into 18a with no gap, which as a matter of citation logic means the clause “as it is written, the righteous shall live by faith” (Hab. 2:4, quoted at Rom. 1:17b) must have been absent from Marcion’s text: not merely unquoted by Tertullian, but structurally missing between two verses Tertullian otherwise cites back-to-back. This is a striking absence given the verse’s later importance in Reformation-era readings of Paul, and is flagged here for a reader’s attention rather than left to be noticed only in the apparatus.
17. 3:21–5:1 as an active rewrite, not just a cut. As with the Sarah/Hagar allegory in Galatians 4, Harnack does not treat 3:31–4:25’s absence as a simple, silent excision: he argues (from Tertullian’s own follow-on commentary, which presupposes the wording) that Marcion actively composed a bridging clause, “then, law; but now, the righteousness of God through the faith of Christ,” fusing the end of 3:21 directly to 5:1 (“being justified, then, by faith in Christ, not by the law, let us have peace toward God”) to paper over the four-chapter gap. Harnack’s note calls “then, law” (“τότε νόμος”) Marcion’s own insertion. This is presented in the reconstructed text and notes as Harnack argues it, an editorial rewrite rather than a passive omission, matching how the Sarah/Hagar rewrite was flagged for Galatians.
18. Chapters 9–11: the gap, the exceptions, and Tertullian’s own argument against Marcion. Harnack finds no citation from ch. 9 at all in Tertullian, Epiphanius, or Adamantius, and calls the gap a “large gap, very probably encompassing the whole of chapter 9.” Chapter 10 surfaces briefly (vv. 2–4, on Israel’s zeal without knowledge and Christ as the end of the law), and then Marcion’s text jumps directly to 11:33–35 (the “O the depth of the riches” doxology) with everything in between, including the entire argument about Israel’s election and the remnant, missing. Harnack preserves and translates in full Tertullian’s own rhetorical argument that this juxtaposition makes no sense on Marcion’s own terms: 11:33’s exclamation about the riches of God’s wisdom reads, in context, as the climax of an argument about Israel and the law that Marcion had just finished deleting, and Tertullian mocks Marcion’s apostle for exclaiming about riches he has just denied the reader any means of seeing. This is Harnack’s own argument for why the material was cut for dogmatic reasons despite reading awkwardly once cut, not merely a report of absence.
19. A genuine ending at 14:23, and 16:25–27 as probably not Marcion’s, or anyone’s original, text. Harnack marshals Origen directly: Origen states that Marcion cut everything after 14:23 (with the doxology at 16:25–27 relocated or absent in “uncorrupted” copies), meaning Marcion’s letter simply ended at 14:23, though Harnack argues, against Zahn, that a short closing greeting (16:24, “the grace of the Lord be with you all”) plausibly followed even in Marcion’s shorter text, citing manuscript evidence (a Monza fragment) that 16:24 could stand immediately after 14:23. More striking still: Harnack does not treat 16:25–27 (the “now to him who is able to establish you” doxology) as authentic Pauline text that Marcion cut. He argues it is not attested by Irenaeus, Tertullian, or Cyprian either, calls the verses “spurious” in general, and suggests they were composed by later Marcionites (not Marcion himself) to give the letter a sense of closure, later migrating with corrections into the catholic textual tradition. This is a genuinely striking claim about the transmission history of Romans’ ending that goes well beyond “what did Marcion cut” into “what these verses actually are,” and is preserved here as Harnack argues it, not softened into a simpler statement about excision.
20. What remains open. (a) The 3:21–5:1 Tertullian citation (V, 13, “tunc lex, nunc iustitia dei…”) is, per Harnack’s own admission, a passage whose “true state of affairs… has not until now been properly seen through” even in the scholarship. The translation follows Harnack’s own proposed reading, which does not resolve the underlying Latin syllogism so much as choose among the constructions it will bear. (b) Several short apparatus notes in this letter (e.g., on 6:19–20, 8:6) involve reversed-pair variants (which noun goes with which verb across two half-verses) that are easy to mis-transcribe in translation; these have been checked against the images, but given this letter’s length and density they warrant an independent specialist re-reading. (c) Manuscript sigla remain unexpanded, per the general convention in note 5.
Notes on 1–2 Thessalonians
21. Short, lightly attested, and combined for a reason. Both letters together run only 4 printed pages (109*–112*): 1 Thessalonians gets 2 pages, 2 Thessalonians gets 2 pages, and Harnack states for both, separately, that “no larger excisions can be demonstrated.” They are treated together here because the source itself treats them as a matched pair. Epiphanius, notably, admits he took no excerpts from either letter when he made his original notes on Marcion’s Bible, and Harnack argues Epiphanius later mistook his own gap in note-taking for evidence that Marcion had “hopelessly falsified and perverted” these two letters, rather than for what it actually was: an oversight on Epiphanius’s own part decades earlier. This is a case where Harnack is defending Marcion’s text against a later, secondhand accusation of corruption that Harnack traces to a plausible error in a hostile source, not evidence in itself.
22. Marcion’s letter order, in passing. Epiphanius’s own words, quoted in Harnack’s note, place 1 Thessalonians fifth and 2 Thessalonians ninth in Marcion’s arrangement of the Apostolikon (matching their fifth and sixth positions, respectively, in the standard canonical order used in the apostle’s collection Epiphanius is comparing against), a detail about Marcion’s own sequencing of Paul’s letters worth surfacing for a reader interested in the Apostolikon’s structure, not just its wording.
23. The Thessalonian eschatology material (1 Thess. 4:15–17) survives with unusual fullness, including through a full parallel citation in the Dialogue (Dial. I, 25) that Harnack judges “very probable,” though not certain, to be Marcionite in origin. He flags this uncertainty explicitly rather than silently upgrading a probable attestation to a certain one, consistent with the convention followed throughout this reconstruction.
24. What remains open. (a) The Dial. I, 25 citation on 1 Thess. 4:16–17, while probably Marcionite, is explicitly marked by Harnack as not absolutely certain; this uncertainty is preserved rather than resolved. (b) The two Kroymann-attributed textual judgments on 2 Thess. 2:11 (the “instinctum”/“instinctus” question) are technical points of Latin textual criticism carried over as Harnack states them. (c) Manuscript sigla remain unexpanded, per the general convention in note 5.
Notes on Laodiceans [= Ephesians]
25. Marcion’s own title for this letter, and why it matters. Marcion’s copy of this letter was headed “To the Laodiceans,” not “To the Ephesians.” This is not a modern scholarly hypothesis but is attested directly by Tertullian, who states that “the heretics, and especially the Marcionites,” entitle it so, while the church’s own copies read “to the Ephesians.” Tertullian treats the difference as inconsequential (“it makes no difference about titles, since the apostle wrote to all”), but Harnack preserves the full run of ancient testimony on the question: Tertullian in two places, Epiphanius in three, and the Dialogue, which cites the letter under the name “to the Ephesians,” because it bears on a genuinely separate question in New Testament scholarship (the relationship of canonical Ephesians to a lost “letter to the Laodiceans” mentioned at Col. 4:16) that predates and is independent of Marcion. This reconstruction uses “Laodiceans” as the section heading because that is what stood at the head of Marcion’s own copy, per Harnack’s own choice of heading in the German original.
26. The longest of the shorter letters, and lightly excised. At 8 printed pages (112*–119*), Laodiceans runs longer than 2 Corinthians or 1–2 Thessalonians combined, but Harnack’s own closing tally states that no larger excisions can be demonstrated apart from (very probably) 1:21 and 6:2b–3, plus one substantial recasting rather than excision, at 5:28–32.
27. 5:28–32 as a reshaped passage, not a silent cut, paralleling the method already seen in Galatians and Romans. As with the Sarah/Hagar allegory in Galatians and the 3:21–5:1 splice in Romans, Harnack does not treat this passage as simply missing material: he argues Marcion’s text kept vv. 28–29 and 31–32 essentially intact but dropped v. 30 (“for we are members of his body”) entirely, and altered the wording connecting v. 31 to what precedes so that “instead of this” now refers forward to “the church” rather than back to Adam and Eve, turning a passage about marriage as an image of Christ and the church into one where the ecclesial reading is made explicit and immediate, rather than an analogy drawn out at the end. Harnack notes that this suppression of v. 30 is also independently attested as absent from some catholic manuscripts on Origen’s own testimony, complicating any simple claim that its absence is purely Marcionite, a genuine textual-critical uncertainty that is preserved here rather than resolved in Marcion’s favor or against him.
28. What remains open. (a) The extended note on the letter’s title (Translator’s Note 25) draws on Tertullian, Epiphanius, and the Dialogue across several separate citations that Harnack does not fully reconcile with each other (e.g., Epiphanius’s own list of Marcion’s letters versus his running commentary do not perfectly agree on how the Laodicean material was presented); this is preserved as an open source-critical question, not resolved into a single clean narrative. (b) The III, 8–9 note on the “hidden mystery” passage involves an unusually intricate argument about a single preposition (ἐν, “in”) that Tertullian says the heretic exploited grammatically; the translation follows Harnack’s own reconstruction of the maneuver, though the clause admits more than one defensible construal and the rendering should not be treated as eliminating that ambiguity, the same standing caveat given to the most technical notes in earlier letters. (c) Manuscript sigla remain unexpanded, per the general convention in note 5.
Notes on Colossians
29. The shortest letter in the reconstruction so far, and far shorter than earlier estimated. Colossians runs only 4 printed pages (119*–122*), considerably less than the running headers of the 1921 printing would suggest, and the boundary was confirmed directly against the page images rather than estimated. Harnack’s own closing tally states that only 1:15b, 16, and 17b amount to a “larger excision,” meaning this letter, like 2 Corinthians and 1–2 Thessalonians, survives close to intact.
30. 1:15–17, the “fullness” passage, cut for the same reason as similar material elsewhere. Tertullian’s argument turns on a single rhetorical question: if “all fullness” dwelling in Christ refers to anything, what could it be except “those things which Marcion took away”? That question is Harnack’s basis for concluding that the highest-Christology clauses of vv. 15–17 (the “firstborn of all creation,” the catalogue of thrones, dominions, principalities, and powers “created in him,” and the affirmation that “he himself is before all things”) were absent from Marcion’s text. This is the same category of excision seen at Ephesians 1:20–21 and elsewhere: material that asserts Christ’s role in the original, structured creation of the visible and invisible world was reliably in tension with Marcion’s separation of the creator god from the god revealed in Christ.
31. A subtle but real uncertainty at 1:22. Harnack argues, from Tertullian’s specific wording (“he fills up ... on behalf of his body ... through the flesh”), that Marcion’s text of 1:22 did not include the phrase “of the flesh,” meaning the reconciliation described there was “in his body” rather than “in the body of his flesh.” This textual point is also why Harnack judges that a separate citation in the Dialogue (which does read “in the body of his flesh”) is unlikely to derive from Marcion’s Bible despite an otherwise-unusual shared reading with it. The two witnesses point in different directions here, and Harnack’s own resolution, trusting Tertullian’s more precise citation practice over the Dialogue’s looser, conflated one, is presented as his judgment call, not as settled fact.
32. What remains open. (a) The III, 3–4 note on the Dialogue’s citation of “your life is hidden with Christ in God” flags multiple divergences from the standard text (word order, grammatical person, a unique ending) and Harnack explicitly states it “remains uncertain” whether the citation is Marcionite at all. This uncertainty is preserved rather than resolved either way. (b) The technical argument distinguishing Tertullian’s two prepositional readings at 1:19–22 (summarized in note 31) is dense even by this reconstruction’s standards and would benefit from independent review by a reader with strong Tertullian training. (c) Manuscript sigla remain unexpanded, per the general convention in note 5.
Notes on Philippians and Philemon
33. Philippians: another very short letter, and the last of the letters proper to receive verse-by-verse reconstruction. At only 3 printed pages (122*–124*), Philippians is attested even more sparsely than Colossians, mostly through Tertullian’s own summary paraphrase of the letter’s argument (explicitly labeled “in Referat,” i.e., “in summary/report,” rather than direct quotation, for 1:14–18) rather than continuous citation of wording. Harnack’s own closing judgment is unambiguous: “of ch. 4 nothing is preserved” and “no larger excision is demonstrable,” meaning the letter is thin on attested wording precisely because so little of it was ever quoted by the hostile sources this reconstruction depends on, not because Marcion is known to have cut it down.
34. The kenosis hymn (2:6–7) as a flashpoint, exactly as with Colossians 1:15–17. Tertullian states plainly that the Marcionites treated this passage as their strongest textual support for a docetic Christology, and that Marcion’s distinctive contribution was not cutting words but omitting two grammatical elements (the participle “having become” and the comparative “as”) from “being found in fashion as a man,” so that the verse could be read to mean Christ merely appeared human in form and bearing, without becoming human in substance. Harnack preserves Tertullian’s own gloss on the intended sense (“he was found (only) as a man, because his form and bearing were those of a man”) rather than paraphrasing it away, since this is presented as the passage Marcion’s own later followers (per Chrysostom and the Armenian writer Esnik, both cited independently by Harnack) treated as foundational to their Christology.
35. Philemon: the one letter Marcion is reported to have left entirely alone, but for a telling reason. Harnack devotes only a single paragraph of prose to this letter, because Tertullian, Jerome, and (on Harnack’s reading) Epiphanius essentially agree that Marcion did not alter it. The interesting textual-critical point Harnack draws out is why, on Tertullian’s telling: not that Marcion had no motive to tamper with it, but that its brevity as a private letter made tampering pointless or too conspicuous. Harnack goes further and explicitly rejects Epiphanius’s claim that nothing was excerpted from Philemon “because it lies entirely distorted,” the same kind of claim already shown to be unreliable for 1–2 Thessalonians and Philippians, calling it “groundless” and attributing it to the same underlying error (Epiphanius mistaking his own decades-old gap in note-taking for evidence of corruption). This is presented here as Harnack’s own argument, correcting a patristic source rather than simply reporting it.
36. The end of the letter-by-letter reconstruction, and what follows it. Philemon closes Harnack’s verse-by-verse treatment of the ten letters of Marcion’s Apostolikon (Galatians through Philemon, pp. 65*–124*, all now translated above). Harnack’s text continues immediately, on the same printed page, into a differently-structured section headed “C. Untersuchungen” (“C. Investigations”), a general excursus surveying Marcion’s text as a whole rather than letter by letter, running pp. 124*–156* and now translated in full below. This is a distinct genre of material from the reconstruction proper: continuous critical prose rather than a verse-by-verse reconstruction, with Harnack’s footnotes collected at the end of each chapter.
37. What remains open. (a) The extent to which Tertullian’s “in Referat” (summary) treatment of Phil. 1:14–18 should be read as closely reflecting Marcion’s actual wording, versus Tertullian’s own compressed paraphrase, is a genuine interpretive question that Harnack does not fully resolve; the translation preserves the distinction Harnack draws (vv. 15 and 17 as “altered,” the rest as summarized) without overstating its certainty. (b) Manuscript sigla remain unexpanded, per the general convention in note 5.
Notes on “C. Investigations” (the excursus following the letter-by-letter reconstruction)
38. What kind of text this is, and how the notes below are organized. Unlike Galatians–Philemon, “C. Untersuchungen” is Harnack’s own continuous critical prose surveying Marcion’s Apostolikon as a whole: an excursus, not a verse-by-verse reconstruction. It is translated page by page like the letters, with Harnack’s footnotes collected at the end of each chapter under Notes and Apparatus. Harnack restarts his footnote numbers on most pages; in this edition they are numbered straight through each chapter, and each note keeps the starred page on which it stands. Where a footnote’s text runs across a page break, the continuation is marked with an italic editorial note identifying which footnote it continues, following the same convention already used once in the unheaded opening (p. 126*–127*).
39. The unheaded opening and §§1–2 (pp. 124*–134*): the shape of Harnack’s overall argument. The opening survey (124*–127*) establishes six general findings, most importantly that the Greek and Latin Marcion-text agree closely with each other and both belong to the same family as the “Western” text (D, G, and the Old Latin/Vulgate tradition), and that only five places can be identified with real confidence as Marcion’s own dogmatic corrections showing up later in mainstream Western manuscripts. §1 (128*–129*) examines roughly 100 places where Marcion’s text agrees with the Western text against other recensions, and concludes that only five of these are plausibly Marcion’s own doing rather than inherited Western readings he happened to share. §2 (130*–134*) surveys roughly 100 further readings peculiar to Marcion alone that are not dogmatically motivated (word-order shifts, tense changes, singular/plural swaps, particle insertions) and concludes, deliberately negatively, that no consistent non-dogmatic editorial program can be detected in them at all: Marcion corrected the text only when his theology required it, not as a general practice of textual improvement.
40. §3 (pp. 135*–147*) and the Marcionite Prologues. This is the longest and most consequential subsection translated so far: it argues that the Latin Marcion-text and the Latin Western text are close kin (though probably not parent and child), and it presents, as its centerpiece, the full Latin text of the ten Marcionite Prologues to the Pauline epistles: short introductory notes, one per letter, preserved by accident of transmission in a majority of Vulgate manuscripts, which Harnack (following de Bruyne and Corssen) argues are demonstrably Marcionite in origin despite having circulated for over a millennium in the very “great Church” Bibles whose canon Marcion rejected. Each prologue is given here in Latin with an English translation immediately following it, so a reader can check the translation against the original directly. Harnack’s argument for their Marcionite origin rests on their shared vocabulary (especially “veritas”/“verus,” which he traces to Galatians, Marcion’s most important letter), their uniform hostility to “false apostles” identified with defenders of the Old Testament, and, most strikingly, a doctored quotation of Titus 1:14 in the Titus prologue that silently substitutes “the Jewish scriptures” for “Jewish myths,” a substitution only a Marcionite would have a motive to make. Harnack also addresses, and resists, the tempting hypothesis that Marcion himself wrote the prologues, concluding instead that they are the work of a somewhat later Marcionite, probably composed as connected running commentary (perhaps as part of the lost Antitheses) before being split apart into per-letter introductions.
41. The most technical stretch of this section: pp. 142*–147*. The second half of §3 works through several specific text-critical puzzles: the “in Rome” address of Romans, the Ephesians/Laodiceans title question, and above all the puzzle of why Romans 15–16 is absent from Marcion’s text at all, given that Harnack has just argued Marcion had no dogmatic motive to remove them. Harnack’s resolution (following Corssen) is that the chapters were already missing from the “DG” branch of the Western text Marcion used as his base, meaning their absence is not Marcion’s doing at all, but a feature of the source text he started from. This section closes with a full translation and analysis of the disputed doxology at Rom. 16:25–27, which Harnack argues was composed by later Marcionites (not Marcion himself) and only partially “catholicized” before entering the mainstream textual tradition, a claim about the transmission history of Romans’ ending that goes beyond anything argued in the Romans letter-section itself.
42. §4 (p. 148*–149*): why Galatians comes first. Against the Tübingen school’s old claim that Marcion’s letter order proved four letters “genuine” and the rest spurious, and against Zahn’s alternative claim that the order is broadly chronological, Harnack argues the order is really just two principles stacked: Galatians first, for substantive/theological reasons (it was, for Marcion, “the Magna Carta of his Christianity”), and then the remaining letters purely by length (Euthalian stichometry), not chronology or theological weight. Harnack gives the actual stichos-counts and letter-counts he is working from, which are preserved in the translation as a check on the argument.
43. §§5–6 (pp. 149*–152*): a chapter-division system, and why the Pastorals were excluded. §5 is a brief, genuinely inconclusive discussion (Harnack calls the traces “insufficient” to attribute a formal chapter-division to Marcion himself, though perhaps to Tertullian’s source text). This negative/uncertain conclusion is preserved as stated, not upgraded. §6 is a sustained argument against Zahn’s theory that Marcion excluded the Pastoral Epistles because he held a general principle that only letters addressed to congregations (not individuals) belonged in his Bible; Harnack instead argues, primarily from a single close reading of Tertullian’s own wording, that Marcion knew the Pastorals, accepted them as genuinely Pauline, but excluded them for a narrower, more mundane reason bound up with their “private,” administrative character, and that this is simply the most obvious inference from Tertullian’s text, not a principle requiring Marcion’s own explicit statement of it.
44. §7 (pp. 152*–154*): the argument that Marcion’s Apostolikon was the first bounded Christian canon. This is the most consequential single claim in the whole excursus. Harnack marshals testimony (chiefly Hippolytus, via Pseudo-Tertullian and Tertullian’s own scattered remarks) that Marcion explicitly rejected Acts and Revelation as products of the Old Testament God’s apostles, then argues at length, against the possibility that Marcion was simply cutting books from an already-existing catholic New Testament of 18–19 writings, that no such fixed catholic collection yet existed for Marcion to react against. Harnack’s key piece of evidence is negative: Marcion’s polemic (the Antitheses) argues against the Old Testament and against the original apostles, but never argues against a Christian “New Testament” as such, which Harnack says would have been unavoidable had one already existed with authoritative standing. The conclusion, stated by Harnack in the closing sentence of this subsection, is that Marcion, not the later catholic Church, was the first person in Christian history to realize the concept of a firmly bounded apostolic canon. This claim is translated in full and without softening, since it is Harnack’s own considered thesis and one of the most cited claims in modern Marcion scholarship.
45. §8 (pp. 155*–156*): loose ends: forged letters, a substitute for Acts, and non-Davidic psalms. This closing subsection is more miscellaneous: it addresses the Muratorian Fragment’s claim of forged Pauline letters “to the Laodiceans” and “to the Alexandrians” (Harnack argues the “Laodicean” charge rests on the Fragment’s author’s own error, confusing Marcion’s canonical Ephesians/Laodiceans with an unrelated later forgery; the “Alexandrians” reference remains unexplained, and Harnack says so plainly rather than speculating further); it identifies Marcion’s Antitheses, on the testimony of the late-4th-century writer Maruta of Maipherkat, as functioning as a deliberate substitute for the Book of Acts within Marcionite communities; and it closes by weighing two independent, partially conflicting testimonies (Maruta and the Muratorian Fragment) that Marcionite communities used non-Davidic psalms in worship, concluding cautiously that only the bare fact of non-Davidic psalms in Marcionite worship (not their authorship or the details of either report) can be treated as established.
46. What remains open. (a) One citation-list entry (footnote 1 to p. 130*) contains what is judged, after close examination of the page image, to be a lowercase roman numeral “iii” rendered in a way that closely resembles the word “in” at this resolution; it is translated as the plain chapter number “3” rather than left standing as an apparent stray word. (b) A citation on p. 142* reads “Rom. 1:20” where the argument concerns a variant that, in the standard verse numbering, occurs at Rom. 1:29; this is flagged inline as printed in the source (“sic”) rather than silently corrected. (c) The ten Marcionite Prologues are printed with the Latin alongside the English precisely because their brevity and importance invite readers to weigh the rendering against the source for themselves. (d) The reading “Mitiadis”/“Tatiani” in the Muratorian Fragment quotation on p. 156* is a famous textual crux in its own right (usually resolved as “Miltiadis,” referring to Miltiades, though Harnack himself here prefers “Tatiani,” referring to Tatian); Harnack’s own hedged uncertainty (“? ich halte noch immer …”) is preserved as stated rather than resolved in translation. (e) Manuscript sigla remain unexpanded, per the general convention in note 5.