Harnack’s book reconstructed a text and taught later writers how to narrate the previous century. For a time the narration was simple. The fathers had been right. The Tübingen episode had been a fever. Zahn and Harnack had settled the wording. New work could be done on Marcion’s theology, on his role in the formation of the Catholic canon, or on the Western text, without reopening the edition. The simplicity was a historical claim, and the claim was wrong on two points: how the 1850s had ended, and how far the wording could still move.
This chapter traces that narration from the books that told it. Wilson and Blackman trusted Harnack. Couchoud, a French scholar, reopened Luke, and Loisy answered him. Knox wrote a canon book whose pages on the German debate became the English history of the 1840s and 1850s. Gregory and Tyson then used that history, and Klinghardt summarized the German books to similar effect. Chapter 3 has already set out the positions of Schwegler, Ritschl, Baur, Volckmar, and Hilgenfeld year by year. That sequence is the control. The English story will be checked against it.
After Harnack: Theology, Text, and a Few Dissenters
Robert Smith Wilson’s Marcion: A Study of a Second-Century Heretic (1933) was the first full-length treatment of Marcion in English. Wilson described a gospel whose tendency was to omit, more often in the later chapters than in the early ones, and he insisted that even when every difference ascribable to theology had been so ascribed, a residuum of variant readings remained. He repeated Harnack’s Western diagnosis and was more explicit than Harnack about what the Matthean and Markan assimilations implied: by Marcion’s time the three gospels had circulated together long enough for scribes of one to be influenced by habitual use of the others. Wilson did not re-edit the text. He showed what an English historical book looked like when it trusted Harnack and still felt leftover wording that theology could not explain.
Raschke, in a 1923 article and a 1924 book, argued that Mark was Marcion’s gospel and that Marcion’s party had authored Mark. The thesis earned almost no following. Raschke’s criticism of Harnack’s stance toward the sources was more useful than his conclusion. Auguste Hollard’s 1935 Deux hérétiques: Marcion et Montan simply said that Marcion’s gospel was Luke, excised, edited, and interpolated. C. S. C. Williams, in 1951, could still write that Harnack’s reconstructed text was probably as accurate as modern scholarship could provide, while admitting the difficulty of knowing how far Tertullian’s and Epiphanius’s quotations were exact. E. C. Blackman’s Marcion and His Influence (1948) was, in Knox’s judgment, the most valuable book on Marcion since Harnack. Blackman protested Harnack’s estimate of Marcion’s religious significance, agreed that Marcion had altered a Western Luke, and then said that the gospel and apostle had been carefully restored by Harnack. “Nothing more remains to be done except a revision of the critical apparatus.” He used that apparatus to argue that Marcion’s influence on the Old Latin was greater than Harnack had allowed and still very limited. He also accepted readings, in both gospel and apostle, whose best ground was the theology that would have required them. Schmid later struck several of the Pauline examples from the usable list. The most valuable post-Harnack English book still treated the reconstruction as finished.
The mid-century field ran in several languages and produced few new running texts. Leon E. Wright warned that the argument from silence is precarious for a text that survives only in its opponents’ citations. Barbara Aland, Gerhard May, and Karlmann Beyschlag, writing in the 1980s and early 1990s, named a new complete reconstruction as a desideratum. Two things pressed on the next editions: silence was unsafe, and Harnack’s page was no longer enough.
One French intervention did try to reopen the literary question. Paul-Louis Couchoud’s 1936 article “Is Marcion’s Gospel One of the Synoptics?” reasserted Baur’s 1847 position, praised Harnack’s collection, and said that Harnack’s convictions had shaped the edition. Couchoud identified Marcion’s gospel with something like the Proto-Luke of B. H. Streeter and Vincent Taylor, dated the Synoptics to about 135–145, and in 1939 printed an English translation of Marcion’s gospel as a section of The Creation of Christ. The translation rested on Harnack’s reconstruction and did not advance it. Alfred Loisy answered immediately and reasserted Marcion’s dependence on Luke. Georges Ory later continued Couchoud’s line. Knox, who did not hold the traditional view, still rejected the identification with Proto-Luke because Marcion’s gospel so evidently contained a large Markan element. The French revival belongs in the history as a reopening of the literary question rather than as a new critical text.
Knox’s 1942 Settlement That Was Not a Settlement
John Knox had already written about Marcion’s vocabulary in 1939. The book that formed later narration was Marcion and the New Testament (1942). Knox’s subject was the canon. The chapters on the gospel were, he said, particularly tentative and intended only to reopen a question that had been prematurely closed. The position he advocated was nearly Baur’s later one: a separate gospel, the basis of both Marcion’s book and canonical Luke, preceded the making of Luke–Acts. That gospel, he held, almost certainly lacked Luke’s first two chapters. Knox printed a table of Marcionite, non-Marcionite, and uncertain passages, warned that any such list could be only approximately correct, and then, by his own admission, took all specific wording from Harnack. He classified a whole pericope as present if an ancient witness quoted a few words of it. He placed a passage among the absences only when a writer explicitly said it was missing. Silence went into the uncertain column. On that last point he was right, and he was more careful than Harnack. A reader who opens Knox’s table beside Harnack’s Beilage IV is not comparing two collations. The English columns regroup Harnack’s verses. The new work is the grouping, not the Greek.
The damage lay in the history of research. Knox told the German decade as a story that ended in a new view. Schwegler and Ritschl had reaffirmed the contrary of the traditional position. Baur had agreed in the main. Volckmar and Hilgenfeld had held that Marcion’s gospel rested on a document very like our Luke, while acknowledging primitive readings. Ritschl and Baur had then modified their positions, Ritschl no longer deriving Luke from Marcion but not returning to the fathers, Baur agreeing that Marcion had emended his original but insisting that the original was not our Luke. “It is important to note,” Knox wrote, “that, contrary to many assertions, this controversy in Germany over the relation of Marcion to Luke ended not in the vindication of the traditional position which Hahn had defended but in the establishment of a new view which denied both that Luke was derived from Marcion and that Marcion was derived from the canonical Luke.” That view, he said, was held “not only by Ritschl and Baur but also, more conservatively, by Volkmar and Hilgenfeld.”
In 2008 Dieter T. Roth set those pages of Knox beside the German publications, and the argument of this section and the next, through the paragraph on Klinghardt, is his. He started from an observation of Ulrich Schmid’s: Knox’s summary nowhere mentions Volckmar’s 1852 book, in which Volckmar gave up the middle position and declared the fathers right. Roth added that Knox does cite Hilgenfeld’s 1853 article, which was written in answer to that book. Nor does the rest of Knox’s roster hold. Ritschl in 1851 left Marcion one certain original reading and otherwise a Luke like ours. The later redaction of Luke that Hilgenfeld posited was slight. Of the four names, only the Baur of 1851 held anything like the view Knox described. E. Earle Ellis and Martin Hengel, as Roth notes, had each tied Knox’s position to Baur’s. Knox revived Baur’s position, and his summary became the English account of the decade.
Knox was not the first to miss Volckmar’s retraction, as Roth shows. Franck in 1855 still called Volckmar’s 1852 position “mediating,” though Volckmar had by then given the mediating position up. Cassels, reediting Supernatural Religion, wrote as if the 1852 book only confirmed the article of 1850. Meyboom missed the difference too. Hilgenfeld’s own 1853 opening, with its wish to see substantial agreement, may have made the mistake easy. Knox’s version became the influential one because later English books cited it as if it were a report.
The Knox Narrative After Knox
H. Philip West Jr. cited Knox’s summary in 1967. R. Joseph Hoffmann’s 1984 Marcion: On the Restitution of Christianity tried to avoid approaching Marcion on the basis of Harnack’s conclusions, credited far less patristic evidence, and therefore did not reproduce the text of the gospel. He still argued, from an earlier date for Marcion and a late date for Luke, that Marcion’s evangelion was an Urlukas. Reviews by C. P. Bammel and Gerhard May rejected the revised biography. The literary conclusion had again been reached from dates outside the wording.
Andrew Gregory’s The Reception of Luke and Acts in the Period before Irenaeus (2003) builds its chapter on Marcion on Knox and says so. Gregory quotes Knox’s statement that the relation is not to be described “by the simple statement that Marcion abridged Luke or by the assertion that Luke enlarged Marcion” (Knox, page 110). He calls this a middle view and adopts it: Marcion, he concludes, did not draw on canonical Luke as a modern eclectic text presents it, and Luke was not derived from Marcion’s gospel (page 193). Roth’s reply is that the German decade offers that middle view one supporter, the Baur of 1851. Volckmar and Ritschl had ruled it out, and the few readings Hilgenfeld kept do not turn Marcion’s source into a different book.
Joseph B. Tyson, in Marcion and Luke-Acts (2006), dedicated to Knox, closes his own brief survey with Knox’s concluding paragraph on the German debate. Volckmar’s 1852 book is absent again, so Tyson can say that Volckmar maintained that both books were derivative. He has Ritschl end near Baur, though Ritschl had refused Baur’s 1851 position almost point by point. Tyson then invokes a “formidable scholarly tradition, which includes Baur, Ritschl, and Knox,” as ground for serious doubt about the fathers and for a theory that canonical Luke did not rest directly on Marcion’s gospel and was composed, “among other factors, in reaction to the preaching of Marcion” (page 85). Roth’s judgment is that the tradition is wrongly invoked, since Ritschl had left it in 1851.
Matthias Klinghardt’s essays of 2006 and 2008 reopened the case in German and English. Klinghardt, unlike Gregory and Tyson, cites all the major publications of the 1840s and 1850s, Volckmar’s 1852 book included. Roth’s objection is to how the New Testament Studies article of 2006 summarized them: as scholars who had agreed that both texts reworked a common original, and who had lost sight of the plain alternative between Marcion’s priority and Luke’s. The German books, as Roth reads them, do not bear that out. Ritschl’s 1851 article withdrew the 1846 thesis and would not follow Baur’s new Ur-Lukas. Volckmar’s 1852 monograph rejected his own 1850 mediation and declared the fathers confirmed. The priority question remained alive in their conclusions. Klinghardt’s other essay of 2006, on the law in Marcion and Luke, listed Ritschl 1846, Baur 1847, and Schwegler 1846 as mid-nineteenth-century support for reading Luke as a redactional expansion of the gospel Marcion used, without mentioning the revision and the retraction. His Novum Testamentum essay could say that European scholarship agreed on Marcion’s posteriority after Volckmar’s book and also list Ritschl among the “exegetical heavyweights” who had proposed the opposite, adding that “their critique of the traditional view has never really been disproved.” Ritschl himself thought it had been disproved, and gave up the thesis for that reason. A footnote in the essay does grant that the participants were “forced to correct their older views.”
Robert M. Grant’s path through this literature is a smaller version of the whole story. He first argued, against Knox, that Marcion had corrected Luke in the light of his doctrines. He later conceded that the correspondence of “changes” with Marcion’s theology does not prove that Marcion created them, because Marcion might have relied on a proto-Luke. He still found it unlikely that Marcion had merely discovered an older book, and he could not explain the modern certainty that the unanimous patristic claim was wrong. Leland Edward Wilshire’s 1974 discussion of a second-century Luke, and Knox’s own 1987 retraction of the stronger form of his vocabulary argument, belong on the same shelf. Knox kept the larger thesis and admitted that the recoverable text, in its detail, was too meager and uncertain to bear a positive linguistic proof. The story had begun to criticize its own linguistic proof. It had not yet corrected its German summary.
David S. Williams’s 1982 thesis and 1989 Journal of Biblical Literature article belong here rather than among the editions, because Williams did not intend to reconstruct the gospel. He intended to challenge the strict identification of Marcion’s text base with canonical Luke, and to do so by textual rather than theological means. His method, restricting the inquiry to “explicit correlated readings,” direct quotations attested by both Tertullian and Epiphanius, and excluding the Adamantius dialogue, was, as Schmid said, a controllable starting point that permitted almost only negative conclusions. Tyson, who liked the conclusion, said the appendix in which Williams set out his readings could not be regarded as an adequate reconstruction. Williams still changed the field. He forced later editors to treat Harnack’s wording as a set of claims, each of which had to be reargued from the sources.
Roth’s check can be repeated by anyone with the pages in hand. Knox’s own sentence is in Marcion and the New Testament. Volckmar’s retraction is in the 1852 preface and close. Ritschl’s withdrawal is in the 1851 article, pages 528–533, and Baur’s letter of 1 February 1851 confirms the split. Hilgenfeld’s short list of original readings is in the 1853 article. Gregory quotes Knox. Tyson quotes Knox and dedicates the book to him. Klinghardt’s 2006 and 2008 essays can be laid beside Ritschl 1851 and Volckmar 1852. Those pages already show that Knox did not restore a German settlement. He wrote an English one.
