In order to assess the importance of the Marcionite Gospel for the New Testament we must determine the editorial relation between this gospel and Luke: this is the basic problem for everybody dealing with the Marcionite Gospel, no matter whether for literary, historical or theological reasons. As I have argued in some detail elsewhere, I strongly believe that the direction of the editorial process linking the two texts runs from the Marcionite Gospel to Luke.Footnote ¹ Only under this basic assumption does the full impact of the Marcionite Gospel become visible: with regard to the emergence of the gospel tradition, the understanding of the New Testament and its textual history, and many other – hitherto unanswered – questions.Footnote ²

Determining the direction of the editorial process is a matter of source criticism, which used to be one of the basic exegetical tools. Although it has been a while since gospel studies were seriously engaged in source criticism, the ground rules of source criticism are straightforward and well known. The basic criterion for determining the direction of the editorial process is editorial plausibility. This criterion can be applied to every single editorial alteration: for which direction and for which text is an alteration more plausible? However, editorial plausibility must also be applied to the text as a whole. This is as simple as it is fundamental: when an editor makes changes to an older text, these changes should be governed by a coherent editorial concept; otherwise, such an editor would appear to be a literary ‘crank’, to borrow the famous idiom by B. S. Streeter.Footnote ³ While such ‘cranks’ may have existed, it is impossible to reconstruct their editorial decisions and concepts as they are completely random.

Determining the greater editorial plausibility with respect to the relation between the Marcionite Gospel and Luke is an undertaking that renders an unambiguous result: in almost every single instance the direction of the editorial process runs from the Marcionite Gospel to Luke. Some passages – such as the beginning of the gospelFootnote ⁴ or the account of the Last Supper – confirm this editorial direction beyond any doubt. True, there are indeed a few examples where the editorial process could run in either direction, but none of these examples requires, or even suggests, a reversal of the Marcionite priority.

The result of the search for a coherent editorial concept is even more compelling: the Marcionite Gospel reveals no such concept. Critics from TertullianFootnote ⁵ to HarnackFootnote ⁶ and beyondFootnote ⁷ have in fact noted (with a greater or lesser degree of astonishment) that Marcion's alleged changes to Luke do not exhibit the theological intentions attributed to Marcion. Even J. E. Chr. Schmidt, one of the early critics in the wake of Semler, exclaimed most indignantly that had Marcion really altered the Gospel of Luke, then ‘he would have altered against his own intentions’.Footnote ⁸ While Harnack and other supporters of the traditional view acknowledged this contradiction, they only conceded that Marcion's alleged revision might have been inconsistent here and there, and they still claimed that Marcion was an unaccountable ‘crank’ who altered randomly without following an editorial concept. That argument is methodologically inadmissible: it bases a theory (‘Marcion altered Luke’) on a particular premise (‘He did so for theological reasons’) but then dispenses with this very premise because it does not fit the evidence. Methodologically, this is an absolute no-go. If we look for a coherent editorial concept explaining the differences between the two texts, we find the natural explanation in Luke: most of the changes become editorially persuasive once they are seen as alterations and additions that were made by Luke to the Marcionite Gospel.

This means: Luke edited the Marcionite Gospel. Since there is no need to postulate any intermediary stage, this relation must be seen as Luke's direct literary dependence on the Marcionite Gospel. We should not be surprised that this solution contradicts the patristic writers, who unanimously claim the priority of Luke. This simply confirms one of the basic rules of historical criticism: do not believe your sources’ claims only because they tell you so!Footnote ⁹

The most obvious consequence of the priority of the Marcionite Gospel over Luke relates to the Synoptic Problem: when taking this ‘pre-Lukan’ gospel into account, the model of the inter-gospel relations changes profoundly. Most remarkably, this model disposes of the need for ‘Q’: the Two-Source Theory becomes entirely redundant, and the other models in discussion – such as the Farrer–Goulder–Goodacre hypothesis or the Neo-Griesbach Theory – are irrelevant.

The outcome of my own source-critical analysisFootnote ¹⁰ is presented in Fig. 1.

Figure 1. The development of the gospel tradition

The model is one of editorial direction and expansion: a new text is added during each single stage of the editorial expansion; each individual text is utilising all existing pre-texts as sources. The tradition begins with the Marcionite Gospel and ends with Luke. These literary dependencies among the gospels change many issues relevant to their interpretation. Since I cannot explain any of these issues here, I shall simply point out a few wide-ranging consequences.

The first important aspect is the remarkable uniformity of the tradition: the Marcionite Gospel appears to be the root from which the whole gospel tradition emerges and with which all later stages remain closely connected. Obviously, every subsequent stage had knowledge, and made use of, all available previous stages of this development. This uniformity leads to numerous consequences, including the inquiry about the historical Jesus. Whereas the Two-Source Theory assumes two independent origins, namely Mark and ‘Q’, which allegedly validate each other and thereby claim a certain reliability, this model involves no such thing as an independent source. The search for the ‘historical Jesus’, therefore, becomes a completely different, if not an impossible, task.

Secondly, John is included in this model. From the beginning, the source-critical separation of John from the Synoptics was artificial and arbitrary, because it was based on aspects of style and content rather than on keen literary observations. If we look at the literary evidence, there is little doubt that John is central to this integral network of gospels. The most obvious example is the passion narrative where Luke and John, time and again, agree with each other in opposition to Mark and Matthew.Footnote ¹¹ This is one of the features that never received adequate attention because of the dominance of the Two-Source Theory with which it is completely incompatible.Footnote ¹²

Finally, this model is based exclusively on literary observations: the literary dependency is fully sufficient for explaining all aspects of the evidence. This has a number of implications. On the one hand, there is no need for employing oral traditions in order to explain the literary evidence. According to the principle of Ockham's razor, the oral tradition is eliminated as a significant formative factor of the gospel tradition. On the other hand, the closely meshed and highly literary relations between the single stages within this model prove that gospel writing was a sophisticated task that required concentrated desk-work of textual redaction: our gospel writers were highly skilled editors.

With regard to this closely knit literary tradition, we also have to acknowledge that our customary time frame of the gospels’ emergence is outdated.Footnote ¹³ The model presented here does not necessitate, or even indicate, longer periods between the single stages of the gospel tradition. I believe it entirely plausible that the overall emergence of the gospels could have been completed within a very short period of time. In view of all this, I would like to point out that, although it is not impossible that a gospel existed before the middle of the second century, there is simply not even the slightest shred of evidence for any written gospel prior to that time.

The priority of the Marcionite Gospel also modifies our understanding of the early transmission of the New Testament text. One of the riddles posed by the Marcionite Gospel is the fact that many of the differences vis-à-vis Luke have analogies in the variants of canonical Luke. To be precise: of the well over 500 differences noted for the Marcionite Gospel, no less than three quarters show up as variants within the manuscript tradition of Luke.Footnote ¹⁴

How is that possible? There is no way that, of all things, the gospel of the much-despised arch-heretic Marcion could have had such a broad and sweeping impact on the canonical manuscript tradition. The sheer number of these variants, therefore, confirms the priority of the Marcionite Gospel: this gospel was not the arbitrary product of a mean-spirited heretic but, quite simply and obviously, an older text utilised by many, including Marcion himself. And that text was, quite simply and obviously, edited by Luke. Therein lies the striking explanation for how the manuscripts of the older gospel influenced the text of the revised gospel.

A striking illustration of this phenomenon is the account of the Last Supper: according to Tertullian's testimony, the Marcionite Gospel had the account's ‘short text’.Footnote ¹⁵ This so-called short text, however, also turns up in the so-called ‘Western’ manuscripts, and it constitutes one of the ‘Western Non-interpolations’. While there is no doubt that the shorter version of the text is older than the longer version, the longer text actually represents that of canonical Luke. It is, therefore, necessary to distinguish between two editions of the same text: the earlier gospel, which was part of the collection of eleven writings known to be used by the Marcionites, and the gospel's edited version – namely, our canonical Luke, which is part of the New Testament.

This means, however, that in many cases the presumably older readings do not belong to Luke but to the older Marcionite Gospel: while they are not part of the New Testament, they are, nevertheless, part of one of its sources. All the sophistication employed by textual criticism for determining the oldest variants is of little use when the sought-after text is in fact a younger, secondary phenomenon. This insight applies to the other gospels as well: the evidence suggests that these gospels existed in older versions, and that they, too, were edited as they became part of the New Testament. Many of the older variants of these other gospels also belong to the pre-NT stage.Footnote ¹⁶

It is a completely new task to establish the text of the New Testament rather than a presumably oldest text which contains readings from the antecedents of the New Testament writings. This task is challenging and requires an entirely new methodology. Nevertheless, distinguishing between two editorial stages of the same text allows us to understand the early history of the textual transmission: taking the revised canonical editionFootnote ¹⁷ into account provides a systematic explanation for many variants. ‘Systematic’ means that we do not have to postulate a seemingly endless series of individual and uncertain alterations of the text. Instead, we can attribute the origin of many of the variants to this single important step within the emergence of the canonical edition.

If the Marcionite Gospel is seen as an alteration of Luke, it has almost no relevance for the New Testament; this was the view of Harnack, who never examined the fundamental question of the direction of the editorial process.Footnote ¹⁸ Assuming the priority of the Marcionite Gospel, however, reveals its innovative potential for understanding the New Testament: on the one hand, it provides answers to many old questions, some of them seemingly resolved a long time ago, others fiercely debated for many decades. On the other hand, the priority of the Marcionite Gospel promotes the most noble and important task of scholarship: it poses a whole new set of questions that did not exist before.