The relationship is stronger than a resemblance of themes. Extended stretches of the two works reproduce the same Latin exposition, including jokes, explanatory glosses, rhetorical questions, and a shared mistake. Biblical quotations form only part of the agreement. The evidence points to a common Latin drafting or rewriting stage, whether one work borrowed from the other or both used a common draft.
That result must be distinguished from the earlier history of the argument. Several of the interpretations have close precedents in Justin’s Dialogue with Trypho. Justinian ancestry can explain the older exegetical material; it does not, by itself, explain the nearly identical Latin in the two Tertullianic works.
1. The extent of the exact agreement
I compared all fourteen chapters of Adversus Iudaeos with all twenty-four chapters of Adversus Marcionem III. A conservative search for at least twelve consecutive matching Latin words found 53 maximal matching occurrence-pairs. Together they cover 965 word positions in Marcionem III and 948 in Iudaeos, counting overlapping coverage only once in each work.
These exact runs occur in Iudaeos 9, 10, 12, 13, and 14, and in eleven chapters of Marcionem III: 7, 12–14, 16–21, and 23. These are descriptive counts for the printed texts. They include biblical quotations, and they do not measure the full extent of paraphrase, adaptation, or source dependence.
For this count, capitalization, punctuation, ligatures, i/j, and u/v were normalized. Marked editorial supplements and deletions interrupt a match. The apparatus and modern website material were excluded. “Maximal” means that an exact run cannot be extended further in both places; it does not mean that every run is independent evidence. The different word-coverage totals reflect repeated occurrences and overlapping matches.
The close readings below explain why the non-biblical agreements matter. Latin is retained in both columns, with English translations. Small differences remain visible.
| Case | Adversus Iudaeos | Adversus Marcionem III | What the comparison shows |
|---|---|---|---|
| 1. Two conditions of one Christ | Adversus Iudaeos 14.1 Duos dicimus Christi habitus a prophetis demonstratos English: We say that two conditions of Christ were shown by the prophets. | Adversus Marcionem III.7.1 Duos dicimus Christi habitus a prophetis demonstratos English: We say that two conditions of Christ were shown by the prophets. | The common sentence introduces humble and glorious appearances of the same Christ. Its placement changes: near the end of Against the Jews, early in Book III. The anti-Marcionite introduction adds the heretic alongside the Jew. |
| 2. The shared conversational reply | Adversus Iudaeos 9.21 Certe, inquis. Hanc prius dicimus figuram futuri fuisse. English: “Certainly,” you say. We say that this was beforehand a type of what was to come. | Adversus Marcionem III.16.4 Certe, inquis. Hanc prius dicimus figuram futuri fuisse. English: “Certainly,” you say. We say that this was beforehand a type of what was to come. | A reply and a first-person explanation travel with the argument. This is reuse of a staged discussion, not merely agreement in a biblical verse. The two texts cannot be treated as independent records of the same exchange. |
| 3. Humble first advent, exalted second advent | Adversus Iudaeos 14.3 Quae ignobilitatis argumenta primo adventui competunt sicut sublimitatis secundo English: These indications of lowly status belong to the first advent, as those of exaltation belong to the second. | Adversus Marcionem III.7.3 Quae ignobilitatis argumenta primo adventui competunt, sicut sublimitatis secundo English: These indications of lowly status belong to the first advent, as those of exaltation belong to the second. | An identical explanatory sentence, apart from punctuation. It reconciles the contrasting prophecies and supplies an answer to the distinction between two Christs. |
| 4. Meaning rather than the sound of Emmanuel | Adversus Iudaeos 9.2 uti non solum sonum nominis spectes sed et sensum English: so that you consider not only the sound of the name, but also its meaning | Adversus Marcionem III.12.2 uti non solum sonum nominis spectes, sed et sensum English: so that you consider not only the sound of the name, but also its meaning | This is shared commentary on the name, rather than the biblical name itself. The surrounding discussion distinguishes the Hebrew name from its interpretation, God with us. |
| 5. Disbelief becomes a charge of Gospel excision | Adversus Iudaeos 9.10 immo reddite veritati quae credere non vultis English: rather, give back to the truth what you are unwilling to believe | Adversus Marcionem III.13.6 immo redde evangelio veritatis quae posterior detraxisti English: rather, restore to the Gospel of truth what you, a latecomer, removed | The corresponding position in the argument changes from an appeal to belief, addressed in the plural, to a singular accusation of removing material from a Gospel. The inherited example remains the Magi and the infant Jesus. |
| 6. The Magi as spoils taken from idolatry | Adversus Iudaeos 9.12 spolia sunt facti Samariae id est idololatriae, credentes videlicet in Christum English: they became the spoils of Samaria—that is, of idolatry—by believing, of course, in Christ | Adversus Marcionem III.13.8 spolia sunt facti Samariae, id est idololatriae, credentes videlicet in Christum English: they became the spoils of Samaria—that is, of idolatry—by believing, of course, in Christ | The match includes a particular allegorical identification and an explanatory gloss. The longer surrounding passage also shares the gifts, the guiding star, and Damascus’s geographical classification. Justin, Dialogue 77–78, supplies a close earlier argumentative parallel. |
| 7. Christ becomes explicitly the Creator’s Christ | Adversus Iudaeos 9.20 Sic bellipotens et armiger Christus est English: Thus Christ is mighty in war and armed. | Adversus Marcionem III.14.7 Sic bellipotens et armiger Christus creatoris English: Thus the Creator’s Christ is mighty in war and armed. | The anti-Marcionite wording adds the decisive identification with the Creator. The surrounding allegory of weapons and spoils remains. A small substitution retargets a much larger inherited argument. |
| 8. A new law becomes the Gospel | Adversus Iudaeos 9.22 idque non per Moysen id est non per legis disciplinam, sed per Iesum id est per novae legis gratiam English: and this was to occur not through Moses—that is, not through the discipline of the Law—but through Jesus—that is, through the grace of the new law | Adversus Marcionem III.16.4 idque non per Moysen, id est non per legis disciplinam, sed per Iesum, per evangelii gratiam English: and this was to occur not through Moses—that is, not through the discipline of the Law—but through Jesus, through the grace of the Gospel | The stable structure carries the contrast between Moses and Joshua/Jesus. The explicit explanatory phrase new law changes to Gospel. This is adaptation of phrasing within a common typological exposition. |
| 9. The near-identical Isaac explanation | Adversus Iudaeos 10.6 Christi exitum iam tunc denotabat in victimam concessi a patre lignum passionis suae baiulantis English: he was already then foreshadowing the death of Christ, given by the Father as a victim and carrying the wood of his own passion | Adversus Marcionem III.18.2 Christi exitum iam tunc denotabat, in victimam concessi a patre et lignum passionis suae baiulantis English: he was already then foreshadowing the death of Christ, given by the Father as a victim and carrying the wood of his own passion | The addition of et is small; the distinctive explanatory construction is stable. The context is Isaac carrying the wood. This example is not classified as a completely identical Latin sentence. |
| 10. The rhinoceros, the minotaur, and the cross | Adversus Iudaeos 10.7 Non utique rhinoceros destinabatur unicornis nec minotaurus bicornis, sed Christus in illo significabatur, taurus ob utramque dispositionem, aliis ferus ut iudex aliis mansuetus ut salvator, cuius cornua essent crucis extima English: No one-horned rhinoceros or two-horned minotaur was being designated; Christ was signified in it, a bull in respect of both dispensations, fierce to some as judge, gentle to others as saviour, whose horns were the extremities of the cross. | Adversus Marcionem III.18.3 non utique rhinoceros destinabatur unicornis nec minotaurus bicornis, sed Christus in illo significabatur, taurus ob utramque dispositionem, aliis ferus ut iudex, aliis mansuetus ut salvator, cuius cornua essent crucis extima English: No one-horned rhinoceros or two-horned minotaur was being designated; Christ was signified in it, a bull in respect of both dispensations, fierce to some as judge, gentle to others as saviour, whose horns were the extremities of the cross. | A long shared explanation includes ridicule, the judge/saviour distinction, and an anatomical interpretation of the cross. The sequence continues across the modern subsection boundary in Marcionem III.18.4. This is among the clearest evidence of a common Latin composition. |
| 11. Moses’s posture: one added syllable | Adversus Iudaeos 10.10 genibus positis et manibus caedentibus pectus et facie humi volutante English: with knees bent, hands beating the breast, and face bowed down to the ground | Adversus Marcionem III.18.6 genibus depositis et manibus caedentibus pectus et facie humi volutante English: with knees lowered, hands beating the breast, and face bowed down to the ground | Positis becomes depositis. Both ask why Moses instead prayed with outstretched hands during Joshua’s battle, then explain his posture as a figure of the cross. Justin, Dialogue 90, already has the unusual objection about the proper posture of prayer. |
| 12. A royal badge carried on the shoulder | Adversus Iudaeos 10.12 Quis omnino regum insigne potestatis suae humero praefert et non aut capite diadema aut manu sceptrum English: What king at all carries the emblem of his power on his shoulder, rather than a diadem on his head or a sceptre in his hand? | Adversus Marcionem III.19.2 Quis omnino regum insigne potestatis suae humero praefert, et non aut capite diadema aut manu sceptrum English: What king at all carries the emblem of his power on his shoulder, rather than a diadem on his head or a sceptre in his hand? | The rhetorical question interprets Isaiah’s government upon the shoulder as the cross. It is a shared authorial inference, not a quotation from Isaiah. |
| 13. Christ becomes Lord in the vineyard accusation | Adversus Iudaeos 13.26 Fecerat enim spinas ex quibus Christum coronaverat et non iustitiam sed clamorem quo in crucem eum extorserat English: For it had produced thorns with which it crowned Christ, and not justice but the outcry with which it forced him onto the cross. | Adversus Marcionem III.23.3 Fecerat enim spinas, ex quibus dominum coronaverat, et non iustitiam, sed clamorem, quo in crucem eum extorserat English: For it had produced thorns with which it crowned the Lord, and not justice but the outcry with which it forced him onto the cross. | The polemic identifies Israel with the vineyard. Christum changes to dominum, while the sentence and the following sequence about withdrawn gifts and devastation remain closely parallel. A reference to the Bethesda pool in Iudaeos is absent from the corresponding Marcionem paragraph. |
| 14. An identical concluding challenge | Adversus Iudaeos 14.14 Haec aut prophetata nega, cum coram videntur, aut adimpleta, cum leguntur; aut si non negas utrumque, in eo erunt adimpleta in quem sunt prophetata English: Deny either that these things were foretold, when they are visible before us, or that they were fulfilled, when they are read; or, if you deny neither, they will have been fulfilled in the one of whom they were foretold. | Adversus Marcionem III.21.4 Haec aut prophetata nega, cum coram videntur, aut adimpleta, cum leguntur, aut si non negas utrumque, in eo erunt adimpleta in quem sunt prophetata English: Deny either that these things were foretold, when they are visible before us, or that they were fulfilled, when they are read; or, if you deny neither, they will have been fulfilled in the one of whom they were foretold. | Twenty-four consecutive words agree after the stated normalization. A sentence at the close of Iudaeos appears inside the continuing anti-Marcionite discussion. Again, the shared material includes its argumentative voice. |
2. A shared error ties the Latin treatments together
The Joseph passage supplies a particularly useful control. Both texts attach Deuteronomy 33:17 to Joseph’s blessing “by his father”:
| Iudaeos 10.7 | Marcionem III.18.3 |
|---|---|
Nam et benedicitur a patre in haec verba Ioseph English: For Joseph too is blessed by his father in these words. | cum benedicitur a patre etiam in haec verba English: when he is blessed by his father also in these words |
The scriptural blessing is Moses’s, not Jacob’s. Justin, Dialogue 91, correctly introduces it as God speaking through Moses. The two Latin works therefore share both the same long interpretation and an apparent conflation of Joseph’s paternal blessing with the Mosaic blessing.
This shared error is evidence for the relationship between the Latin texts. It does not establish that Justin made the mistake: his surviving treatment does not. Nor does it decide which Latin work came first. What it establishes is a closer textual relationship than the independent selection of the same Old Testament verse. Justin, Dialogue 91.
3. The clearest retargeting: unbelief becomes textual mutilation
The most revealing change is the Magi passage. In Iudaeos 9.10 the audience is told to give the truth what it refuses to believe. In Marcionem III.13.6 the corresponding demand becomes: restore what you removed from the Gospel of truth. A singular Marcion replaces the plural audience, and a charge of excision occupies the place of a charge of disbelief.
The inherited proof remains the same: the infant receives the strength of Damascus and the spoils of Samaria through the Magi’s gifts and conversion. These Magi belong to Matthew’s infancy narrative, not ordinary Luke. The comparison shows how an established Matthean proof could remain in place while the opponent and the accusation changed.
This is a concrete clue to the formation of an anti-Marcionite argument. A claim framed as Gospel restoration stands inside a passage with a very close anti-Jewish counterpart. Its presence alone cannot tell us that the adapter independently inspected a Marcionite Gospel containing, or previously containing, the Magi.
4. The same blocks are put in a different order
| Iudaeos block | Corresponding Marcionem III block | Editorial significance |
|---|---|---|
| 14.1–10: two conditions and two advents | 7.1–8 | A late explanation in one work becomes an early framework in the other. |
| 9.2–20: Emmanuel, infant, Magi, figurative weapons | 12–14 | Much of the local sequence remains, with anti-Marcionite changes to its address and conclusions. |
| 9.21–25: the name Jesus and Joshua | 16 | The typological exposition is retained while its terminology is adjusted. |
| 10.6–16: wood, horns, Moses, serpent, suffering | 18–19 | A substantial passion-and-cross dossier survives with both close verbal matches and expansions. |
| 12.2 and 14.14: Gentile fulfilment and a concluding challenge | 20 and 21.4 | Material dispersed in Iudaeos is incorporated within a continuing argument. |
| 13.25–27: withdrawal of gifts and devastation | 23.2–4 | An extended polemical sequence is shared, with some details omitted or rephrased. |
This arrangement helps explain why the relation can be missed if only entire chapters are compared. The borrowing or common drafting operates in blocks. A compiler can preserve sentences with considerable fidelity while changing the architecture of the work.
5. What can be traced back to Justin?
There are close earlier parallels for the Magi and Isaiah’s spoils (Dialogue 77–78), Joshua’s name as a type of Jesus (75, 113), Moses’s outstretched arms and the horns of the cross (90–91), and two advents signified by the goats (40, 111). These correspondences make a Justinian ancestry for important parts of the dossier plausible.
The Latin comparison establishes something more immediate: the two surviving treatments share a developed wording and presentation. Independent use of Justin alone does not adequately explain the long Latin agreements, the same conversational replies, and the shared error in the Joseph passage. A common Latin redactional stage is the better description of the observable evidence. Justin’s surviving Dialogue, his lost anti-Marcionite work, and a subsequent dossier remain different possible vehicles for the older material.
6. Which work came first?
The verbal overlap is secure; its direction is disputed. Evans discusses the proposal that the later part of Iudaeos was taken from Marcionem, and reports Quispel’s attempt to connect its maker with the unauthorized copyist described in the preface. Tränkle and, later, Geoffrey Dunn instead defend Tertullianic material in Iudaeos as earlier than its reworking in Marcionem. A common draft is another possible account of the relationship.
The substitutions examined here are compatible with adaptation toward an anti-Marcionite setting. They do not settle priority by themselves: an adapter moving the other way could remove the Marcionite specifications. Secure direction requires the whole structure, the manuscript evidence, and the passages peculiar to each work. Evans, introduction, sections iv–v; review of Dunn’s 2008 study.
For the larger three-recension hypothesis, the result is substantial. Book III contains identifiable reused Latin composition, and the final voice has fitted that material to a new argumentative setting. The further claims—that the adaptation first entered in the third recension, and that Irenaeus authored a preceding Luke-and-Paul version—require evidence beyond the verbal parallels.
Concordance of all 53 exact runs of twelve or more normalized words
Each row identifies one maximal occurrence-pair. Some share words or passages with other rows. The inventory includes biblical quotations; the interpretive case above rests especially on the non-biblical wording.
| Iudaeos locus | Marcionem locus | Consecutive matching words |
|---|---|---|
| 9.2 | III.12.2 | 17 |
| 9.3 | III.12.4 | 13 |
| 9.4–9.5 | III.13.1–III.13.2 | 17 |
| 9.6 | III.13.3 | 15 |
| 9.7 | III.13.4 | 17 |
| 9.8 | III.13.5 | 15 |
| 9.10–9.11 | III.13.6–III.13.7 | 17 |
| 9.11 | III.13.7 | 15 |
| 9.12 | III.13.8 | 12 |
| 9.12 | III.13.8 | 34 |
| 9.12–9.13 | III.13.8–III.13.9 | 28 |
| 9.16–9.17 | III.14.1 | 14 |
| 9.17 | III.14.2 | 14 |
| 9.20 | III.14.7 | 17 |
| 9.21–9.22 | III.16.3–III.16.4 | 30 |
| 9.22 | III.16.4 | 19 |
| 9.25 | III.16.6 | 17 |
| 9.28 | III.17.4 | 12 |
| 10.5 | III.18.2 | 12 |
| 10.7 | III.18.3–III.18.4 | 49 |
| 10.8 | III.18.4 | 15 |
| 10.8 | III.18.4–III.18.5 | 19 |
| 10.9 | III.18.5 | 12 |
| 10.10 | III.18.5–III.18.6 | 12 |
| 10.10 | III.18.6 | 13 |
| 10.10 | III.18.6 | 17 |
| 10.10 | III.18.6–III.18.7 | 24 |
| 10.11–10.12 | III.19.2 | 22 |
| 10.12 | III.19.3 | 13 |
| 10.13 | III.19.5 | 19 |
| 10.13–10.14 | III.19.6 | 16 |
| 10.14 | III.19.6 | 15 |
| 12.2 | III.20.4–III.20.5 | 22 |
| 13.25–13.26 | III.23.2–III.23.3 | 25 |
| 13.26 | III.23.3 | 14 |
| 13.26 | III.23.3 | 17 |
| 13.26–13.27 | III.23.3–III.23.4 | 12 |
| 13.27 | III.23.4 | 16 |
| 13.29 | III.23.6–III.23.7 | 14 |
| 14.2 | III.7.2 | 13 |
| 14.2 | III.7.2 | 22 |
| 14.2 | III.17.1 | 17 |
| 14.2 | III.7.2 | 15 |
| 14.2–14.3 | III.7.2–III.7.3 | 30 |
| 14.3–14.4 | III.7.3–III.7.4 | 18 |
| 14.4 | III.7.4 | 20 |
| 14.6 | III.7.6 | 13 |
| 14.7 | III.7.6 | 17 |
| 14.9–14.10 | III.7.7–III.7.8 | 24 |
| 14.10 | III.7.8 | 13 |
| 14.10 | III.7.8 | 14 |
| 14.10 | III.7.8 | 24 |
| 14.14 | III.21.4 | 24 |
Texts and references
- Adversus Iudaeos: Herman Tränkle’s Latin text (1964), with subsection references supplied from CCSL, transcribed by Roger Pearse. The site explains its editorial brackets. Thelwall’s older English translation supplies a further context check.
- Adversus Marcionem III: Ernest Evans’s Latin text (1972), Book III, compared with the sectioned Latin transcription. English renderings of the displayed excerpts are supplied here.
- Justin, Dialogue with Trypho 40, 75, 77–78, 90–91, 111, 113: complete English text. The particular claims concerning Justin are argumentative parallels, not a claim that the surviving Latin is Justin’s wording.
- Ernest Evans, Tertullian: Adversus Marcionem (1972), introduction, sections iv–v; G. Quispel, De bronnen van Tertullianus’ Adversus Marcionem (1943), as discussed there; Geoffrey D. Dunn, Tertullian’s Adversus Iudaeos: A Rhetorical Analysis (2008), particularly the priority question, discussed in the contemporary review.
Textual control: the numerical audit concerns printed-text agreement, not a new manuscript collation. Editorial supplements are not counted as independently transmitted matches. The 53-run inventory is one defined search at one threshold; the fourteen close comparisons show exact and near-exact reuse without treating every difference as a new source.
