Investigations I: Categories of Textual Variants

C. Investigations [pp. 124*–156*]

[Continuous critical excursus following directly on the letter-by-letter reconstruction, on the same printed page as the close of Philemon. Unlike the letters above, this section is Harnack’s own connected scholarly prose, not a verse-by-verse reconstruction with separate notes; it is translated here as continuous text, page by page, with Harnack’s footnotes collected at the end of the chapter under Notes and Apparatus. It falls into an unheaded opening discussion (pp. 124*–127*) and eight numbered subsections (pp. 128*–156*).]

[p. 124*] C. Investigations.

The judgment of the Church Fathers has proven sound; Epiphanius is right when he says, Haer. 42, 9: “He has also letters from [p. 125*] him, ten of the holy apostle, which alone he makes use of, yet not of all the things written in them, but he has cut away some of them, and altered other sections.”¹

The critical apparatus yields the following:

(1) Marcion has intervened, at far more numerous places than Tertullian indicates,² in the transmitted text of the Pauline epistles.

(2) The interventions consist of large deletions and of small ones — but often very incisive ones, some even turning the sense of the text into its opposite — corrections and deletions demonstrable, for the most part, as serving his own peculiar doctrine. On the basis of the fragmentary material transmitted to us, large deletions cannot be demonstrated in 1 and 2 Corinthians, in the two Thessalonian letters, in Philippians and Philemon; in Ephesians only the absence of 1:21; 6:2b.3, and heavy abridgments in 5:22ff.; in Colossians only the absence of 1:15b.16.17b. In Galatians, by contrast, [material] was absent at 1:18–24 (probably); 2:6–9a (probably); 3:6–9.10b.11a.12a.14a.15–25; 4:27–30; in Romans at 1:17b; 1:19–2:1; 3:31–4:25; 8:19–22; 9:1–33 (probably entirely); 10:5–11:32; 15, 16. Since, however, Tertullian, in his discussions in the course of his polemical treatise (his polemic on Ephesians forms an exception), becomes ever more erratic (because he shuns repetitions), much more may have been absent in Marcion even in the letters treated last than we are able to demonstrate.

The smaller tendentious deletions, corrections, and reshapings are found at: Gal. 1:1.7; 2:(1).9.10; 4:4.8.21–26(!); 5:14; 6:17. 1 Cor. 1:18; 3:17; 10:11.19; (6:13); 15:3.20.38.45. 2 Cor. 1:3; 2:15; 3:14; 4:10.11.13; (5:10); 7:1. Rom. 1:16.18; 6:9.19; 7:5; 10:3; 11:33; 12:19. 1 Thess. [p. 126*] 2:15; 4:4.16; 5:23. 2 Thess. 1:8; 2:11. Ephes. 2:2.(11).14.15.20; 3:9; 5:22.28.31; 6:2.3.4. Col. 1:19.22; 2:8; 4:14; Phil. 1:15.17; 2:7; 3:9. All these tendentious omissions and corrections I leave aside here, since they have been treated in the presentation [above].

(3) The Greek Apostolikon of M. (Orig., Adamant., Epiph., etc.) agrees excellently, on the whole, with its Latin counterpart (Tert.); the not-very-numerous demonstrable exceptions are explained partly by the necessary differences between an original and its translation,³ partly by the slight alterations that M.’s text suffered within M.’s own church.

(4) Alongside the recognition of M.’s severe tendentious interventions in the text, the most important insight regarding the nature and character of the text is this: that the Greek and Latin text of M. is, in each case, a twin brother of the bilingual text DG, and therefore also a blood relative of the texts of the Itala and Vulgate (as against the African text), as well as of the texts as found in Tertullian (in his other writings),⁴ Irenaeus (Latin), Novatian, Ambrosiaster,⁵ Lucifer, Origen (Latin), Augustine, etc. The kinship with DG — and precisely at the points where these agree with each other⁶ — is in this respect striking to the highest degree; in all, it concerns some 98 places.⁷ Along [p. 127*] with these, the Greek and Latin M-text are the oldest witnesses we possess for the “Western” text of the Pauline epistles — hence the most venerable documents for it. That these texts also display a not-inconsiderable kinship with syr^cu on the one hand and with the oldest Egyptian text on the other,⁸ cannot surprise us, given what we otherwise know about the influence of the Western text on these in earlier times. The text has also exercised influence, where M. represents it, on the Antiochene text — though a lesser one.

(5) M. accordingly did not recense his Apostolikon in Pontus or Asia, but only in the West. The text that lay before him was thus the text of the Roman community shortly before the middle of the 2nd century.

(6) A very large series of peculiarities of the two M-texts, as against the modern text-recensions, is accordingly not Marcionite, as was formerly mistakenly assumed, but rather primitive-Western [ur-abendländisch].

On this newly won foundation, however, two questions now arise: (a) has M., beyond the tendentious reshaping of the Pauline letters, also otherwise corrected the text and permitted himself non-dogmatic interventions — and, if this question is to be answered affirmatively, from what motives? (b) What relationship exists between the Latin Marcion-text and the Latin Western text? Has the former influenced the latter, or is it even its foundation? But ahead of these two questions yet another must be placed first: namely, whether among the places at which M. is a witness to the Western text, there are already some that are specifically Marcionite, because they express his own peculiar heretical tendencies.

[p. 128*] 1. The character of the ca. 100 places at which the Marcion-text agrees with the Western text against the other principal recensions.

All these places, with very few exceptions, are entirely unimportant in content, represent an essentially indifferent stylistic uniform, and are neutral with respect to M.’s peculiar dogmatic tendencies — with the exception of 9 places; but even for half of these the [dogmatic] tendency is doubtful. (1) Gal. 1:6: M. and F^gr*Gg Tert. Cypr. Lucif. Victorin. lack “of Christ” after “grace,” and this could be a dogmatic correction of M., who, given his relative Modalism, would not have wanted it granted that God had called us “in the grace of Christ”; but elsewhere he has let similar things stand. One may therefore by no means assert with certainty that a specifically Marcionite reading is present here. (2) The famous “nor” [οὐδέ] in Gal. 2:5 could be a falsification by M.; only, as is well known, it is offered (against D*d Iren. Tert. Victorin. Ambrosiaster) by such an overwhelming majority of witnesses (Gg too) and is also, on the merits, so much more probable, that it must count as the original reading. (3) “Blessing” [εὐλογίαν] with D*Gdg Ambrosiaster, Vigilantius, in Gal. 3:14, in place of “promise” [ἐπαγγελίαν], could be claimed as a dogmatic correction; but it is nearer at hand to assume an error, since “blessing” [εὐλογία] immediately precedes. (4) The insertion of “in you” in Gal. 5:14, “the law is fulfilled ‹in you›” (so too D*Gdg go Ambrosiaster, etc.), is probably tendentious. (5) The excision of “first” beside “to the Jew” in Rom. 1:16, with BGg, is Marcionite. (6) That Rom. 15 and 16 were also absent in DG, as in M., cannot count as a tendentious intervention, since this motive is simply insufficient to explain the deletion of these two chapters. (7) The reading “who killed their own prophets,” in 1 Thess. 2:15, with D^bc syr go KL and the Fathers, in place of “the prophets,” is probably Marcionite, and (8) so too is the “to us” in Ephes. 4:6, in the sentence “Father of all, who is over all and through all and in all of us” (with DGKLdg vulg. Firmil. etc. etc.); for M. could not let it stand that the good God pervades the whole cosmos. (9) The absence of “and shall be joined to his wife” in Ephes. 5:31, in the [p. 129*] manuscripts of Origen and Jerome, and other alterations here, are the influence of M.⁹

There are thus only the 5 places at Gal. 5:14; Rom. 1:16; 1 Thess. 2:15; and Ephes. 4:6; 5:31 that may be claimed with high probability as Marcionite tendentious corrections.¹⁰ Hence: the agreements of the M-text with the Western text are, with few exceptions, to be regarded as stylistic (etc.) peculiarities of the Western text in which the M-text shares. Subsequent influence of this text upon that one, and upon the [wider textual] tradition generally, was not entirely absent, but was, so far as the material before us allows conclusions, very slight (see further on this below).¹¹

[p. 130*] 2. The character of the readings found in Marcion alone, apart from the dogmatic-tendentious interventions.

The great mass of readings peculiar to M. has been called forth by his dogmatic tendency. But apart from these there are found about 100 places,¹² at which M. offers singular readings that are dogmatically almost or wholly indifferent. In judging them one must bear in mind that Tertullian’s statements consist, alongside literally exact citations, of reports that, while good, are not always correct down to the last detail, and that Epiphanius, in reproducing Marcionite texts which he [p. 131*] adduces in order to prove from them that they attest catholic doctrine, has not always been careful,¹³ and that Adamantius has his citations at second hand. To this must be added, further, the possibility that M. may, here or there, alone have preserved the original text, which has been lost among all the other witnesses — he is, after all, our oldest witness for the text of the Pauline epistles. We shall therefore have to assume that, of the c. 100 singular and yet not dogmatically-tendentious variants of M., a considerable number would probably fall away if we possessed M.’s original text, and that, conversely, some Marcionite readings perhaps preserve the original text (against the other Western witnesses).

If one now tries to arrange these places by category, it emerges that by far the larger half of the variants are of a kind that, in their nature, cannot be distinguished from the variants that the manuscript tradition offers everywhere in general. They concern (1) word order within the sentence, (2) differences of tense, (3) interchange of singular and plural (οὐρανός > οὐρανοί [“heaven” > “heavens”], etc.), (4) interchange of synonyms, or the omission of one member in a series of synonyms (κύριος > θεός [“Lord” > “God”], etc.), (5) prepositions (πρός > εἰς, διά > ἐν, etc.), (6) particles (ἀλλά > καί, οὖν > δέ, οὔτε > οὐδέ), or the insertion or omission of such, (7) the insertion or omission of pronouns (αὐτοῦ [“his”], etc.), and similar matters. After setting aside these nearly indifferent cases, a brief survey shall show what still remains here:

In Galatians, 1:7a shows a strengthening of the thought through an inserted πάντως [“altogether”], and 1:8 through an inserted ἄλλως [“otherwise”]; probably also ὑμᾶς θέλοντες μεταστρέψαι εἰς ἕτερον εὐαγγέλιον τοῦ Χριστοῦ [“wishing to turn you to a different gospel of Christ”], for θέλοντες μεταστρέψαι τὸ εὐαγγέλιον τοῦ Χριστοῦ [“wishing to turn the gospel of Christ”], is to be judged a strengthening. 2:2 shows a logical improvement of the thought: ἔδραμον ἢ τρέχω [“I had run, or am running”] > τρέχω ἢ ἔδραμον [“I am running, or had run”]. 2:20 shows a strengthening of the thought through the statement “the Son of God, who bought me,” instead of “the Son of God, who [p. 132*] loved me.” 3:10 shows ὑπὸ νόμον [“under the law”] for ἐξ ἔργων νόμου [“from the works of the law”] — a clarification. 4:3 shows the opaque prefixing of the words drawn from 3:15, ‹ἔτι› κατὰ ἄνθρωπον λέγω [“‹still› I speak according to man”], before the words ὅτε ἦμεν νήπιοι, ὑπὸ τὰ στοιχεῖα τοῦ κόσμου ἦμεν δεδουλωμένοι [“when we were children, we were enslaved under the elements of the world”]; the ἔτι [“still”], otherwise unattested, is a stylistic addition. 4:24f. shows the opaque insertion of Ephes. 1:21 at this point. 6:17 shows a logical sharpening of the thought through the insertion of εἰκῇ [“for nothing/without cause”] into the sentence κόπους μοι μηδεὶς παρεχέτω [“let no one trouble me”]. One more word on 4:24f.: this passage is remarkable because, alongside an incisive tendentious intervention, two further measures are found here that are otherwise without parallel in M. — namely, besides the just-mentioned insertion from Ephesians, the other, freely composed insertion εἰς ἣν ἐπηγγειλάμεθα ἁγίαν ἐκκλησίαν [“to which we have pledged ourselves, (the) holy church”]. The attempt to relieve M. of these insertions seems hopeless, and so one must accept that M., as a textual critic, permitted himself at this one place something he otherwise never allowed himself. Only if he wanted this passage for a special purpose can this elaboration be understood. Was it — bringing out as sharply as it does the contrast of the two “demonstrations” [Ostensiones] (the dispensation of bondage, which led from Sinai to the synagogue, and the otherworldly dispensation, which led to the holy church, our mother) — perhaps intended to serve as a sacred formula at admission into the Marcionite church?

In 1 Corinthians, 1:11 shows ἤκουσται [“it has been heard”] > ἐδηλώθη [“it has been made known”] — opaque; though it is doubtful to me whether a variant should be assumed here at all. 1:18 shows a logical improvement of the antithesis μωρία > δύναμις [“folly (and) power”] through the formulation μωρία > δύναμις καὶ σοφία [“folly (and) power and wisdom”]. 1:28 shows a strengthening of the thought through the insertion of τὰ ἐλάχιστα [“the least things”] between τὰ ἀγενῆ [“the base things”] and τὰ ἐξουθενημένα [“the things that are despised”] (probably with an eye to 1 Cor. 15:9). 3:3 shows the deletion of the words καὶ κατὰ ἄνθρωπον περιπατεῖτε [“and walk according to man”] (opaque). At 3:20 (see also 2:16), M. appears to have inserted the correct Septuagint text — or rather, a peculiarity of the Western text is to be noted here (see Lietzmann, Römerbrief², p. 15). 6:13 shows the addition ὡς ὁ ναὸς τῷ θεῷ καὶ ὁ θεὸς τῷ ναῷ [“as the temple (belongs) to God, and God to the temple”], which is so improbable for M. that one must suspect he has here preserved an original sentence that was very early lost [p. 133*] . At 15:20, M. writes Χριστὸς κηρύσσεται ἐκ νεκρῶν ἀναστάναι [“Christ is proclaimed to have risen from the dead”] > Χρ. ἐγήγερται [“Christ has been raised”]. That he substitutes “has risen” for “has been raised,” with respect to Christ, can also be documented elsewhere and is tendentious (see below); but why he did not simply write Χρ. ἀνέστη [“Christ rose”] is obscure. 15:25 and 29 I leave aside, since I am not entirely certain that he really did omit πάντας [“all”] there and ὅλως [“at all”] here; if he did omit them — they are not necessary — then he judged the strengthenings of the thought found there to be superfluous, whereas elsewhere he underlines them (see above).

In 2 Corinthians, M. must have taken particular pains over 3:18. I set the two versions side by side:

Original text: “But we all, with unveiled face, beholding as in a mirror the glory of the Lord, are being transformed into the same image, from glory to glory, even as from the Lord, the Spirit.” [Ἡμεῖς δὲ πάντες ἀνακεκαλυμμένῳ προσώπῳ τὴν δόξαν κυρίου κατοπτριζόμενοι τὴν αὐτὴν εἰκόνα μεταμορφούμεθα ἀπὸ δόξης εἰς δόξαν, καθάπερ ἀπὸ κυρίου πνεύματος.]

Marcion: “We, now with unveiled face, beholding Christ as in a mirror, are being transformed into the same image, from glory — [that] of the Lord — to glory, even as from the Lord of spirits.” [Ἡμεῖς ἤδη ἀνακεκαλυμμένῳ προσώπῳ τὸν Χριστὸν κατοπτριζόμενοι τὴν αὐτὴν εἰκόνα μεταμορφούμεθα ἀπὸ δόξης τοῦ κυρίου εἰς δόξαν, καθάπερ ἀπὸ κυρίου πνευμάτων.]

The changes may be described as logical improvements: thus one can understand the omission of πάντες [“all”], the inserted ἤδη [“now/already”] (cf. on Rom. 3:21; 8:9), the interchange of “Christ” for “the glory of Christ,” the sharpening of δόξα [“glory”] by τοῦ κυρίου [“of the Lord”] in order to establish exact correspondence with the close, and the interchange of πνεύματος [“Spirit,” singular] with πνευμάτων [“spirits,” plural] — to be understood accordingly. — The pedantic addition of τοῦ σώματος [“of the body”] at 5:4, to ἐν τῷ σκήνει [“in the tent”], serves to clarify the sense.

In Romans, 1:18 again has πᾶσαν [“all”] omitted (cf. on 1 Cor. 15:25; 2 Cor. 3:18); it could fall away as pleonastic, as at 1 Cor. 15:25. In joining 3:21 to 5:1, M. sharply brought out the “then” and “now”: τότε νόμος, νυνὶ δικαιοσύνη θεοῦ [“then, law; now, the righteousness of God”], and, to secure the concept, added τοῦ Χριστοῦ [“of Christ”] to διὰ πίστεως [“through faith”]. At 5:21, αἰώνιον [“eternal”] is omitted; the contrast of death and life thereby becomes tauter. At 8:9, νῦν [“now”] is added in the interest of precision (see above).

In 1 Thessalonians, at 4:16, ἐσχάτῃ [“last”] is added to σάλπιγξ [“trumpet”] out of pedantry. At 5:23, the addition of καὶ σωτῆρος [“and Savior”] to τοῦ κυρίου [“of the Lord”] is noteworthy; M. thus laid, in a [p. 134*] solemn formula, weight on the fact that this concept was not missing.

In Ephesians, at 1:13: did τῆς σωτηρίας ὑμῶν [“of your salvation”] after εὐαγγέλιον [“gospel”] really go missing in M.? One does not see why he should have omitted it — did he perhaps want to abbreviate? At 2:14f., τὸ μεσότειχον τοῦ φραγμοῦ λύσας, τὴν ἔχθραν ἐν τῇ σαρκὶ αὐτοῦ [“having broken down the middle wall of partition, the enmity, in his flesh”], M. changed to τ. μεσ. τῆς ἔχθρας ἐν σαρκὶ λύσας [“having broken down the middle wall of the enmity, in the flesh”]. Here an abbreviation is certainly present. At 6:11, if M. really wrote ἐν ᾗ στῶμεν [“by which we may stand”] for πρὸς τὸ δύνασθαι ἡμᾶς στῆναι [“so that we may be able to stand”], the reason for the change is opaque.

In Philippians, at 1:18, οὐδέν μοι διαφέρει [“it makes no difference to me”], if M. wrote it so, is a mere stylistic correction. At 3:7, νῦν [“now”] is added; at 3:9, ἤδη [“already”] is added — see similar cases above.

One sees that generic motives can scarcely be derived from these not-very-numerous variants. Even if M. has, at a few places, underlined [the thought], at a few places made logical improvements, and elsewhere added time-particles that strengthen the thought, and so on, still these places, measured against the extent of the texts, are so few in number that one may not speak of definite intentions that determined him to any considerable degree. The investigation therefore ends here negatively — that is, beyond the facts that M.’s text, in both Greek and Latin, belongs to the Western tradition, and that he corrected the texts out of dogmatic tendencies, nothing definite can otherwise be established about his intentions. Probably he had no such intentions at all, even if here and there, since he was correcting at all, he may have made an occasional “improvement” — emendations [Diorthosen] of the kind textual critics of that time otherwise permitted themselves. Yet a large part of these may already have belonged to the Western text he took as his base, even if we are no longer able to demonstrate this.¹⁴ That he [p. 135*] has anywhere, at the places cited, preserved the original text is, so far as I can see, improbable; unless perhaps at 1 Cor. 6:13. — Foreign material he has nowhere mixed in.

3. What relationship exists between the Latin Marcion-text and the Latin Western text? Has the former influenced the latter? Is it even its foundation?

Above (p. 129) five places were demonstrated at which readings of M. — that they are such is shown by their tendency — are to be found in the tradition of the Western text.¹⁵ Since the Greek and Latin M-text itself belongs to the Western tradition, the question arises whether the Latin M-text is not the first translation of the Pauline epistles into Latin, on which the catholic translation (Western Latin) is based. A priori this is quite possible; for the need for a Latin translation of the Pauline epistles must have been still stronger among the Marcionites, who did not know Greek, than in the Latin-speaking great Church. Lietzmann¹⁶ therefore thought that the problem here at hand “perhaps finds its solution most simply through the assumption that, in the middle or second half of the 2nd century, when the official Roman church still spoke Greek, Marcionite preachers, seeking to win disciples for their teaching also among the Latin-speaking population of Rome, first translated into Latin, for this propagandistic purpose of theirs, the Pauline text especially dear to their hearts; the catholic church then took over this translation and assimilated it to its own text, yet was not able to erase the traces of its origin everywhere.”