The synoptic problem is the chief detective story within New Testament scholarship. Are the first three Gospels–Matthew, Mark, and Luke–literarily dependent upon one another?
They seem to be, since, when their text is arranged “synoptically” (with like verses next to one another in parallel columns), they seem to be saying the same things in more or less the same words and in more or less the same order. Many solutions have been proposed to account for the similarity–some solutions, such as Augustine’s, being proposed before anyone realized the synoptic problem was a problem.
Frans Neirynck is one of the leading “establishment” scholars in biblical scholarhip and is a professor at the Catholic University of Leuven in Belgium. Since he is commonly cited as an authority on the question of the synoptics, we can look at his summary of the history of the solutions:
“The `Augustinian’ hypothesis assumed the order of composition to be Matthew, Mark, Luke. For a period this was replaced as the leading theory by the Griesbach hypothesis (Matthew, Luke, Mark). The priority of Mark was first suggested at the end of the eighteenth century as an alternative to the traditional view of Matthean priority, leading to decisive debate in the 1830s to 1860s. As a result, the Marcan hypothesis became the predominant scholarly opinion.” [Frans Neirynck, “Synoptic Problem” in The New Jerome Biblical Commentary (Englewood Cliffs: Prentice Hall, 1990), 587. In quotations abbreviations of the titles of biblical books have been spelled out for ease of reading.]
Quite true, but let’s not draw the wrong conclusion. Marcan priority remains the predominant opinion among professionals and, therefore, among amateurs who read and accept, often uncritically, what professionals say, but we shouldn’t conclude willy-nilly that the hypothesis is correct. (We must keep in mind that truth is not determined by a show of hands.) More and more scholars are objecting to what, until recently, has been the scholarly consensus. Today we can see a move away from Marcan priority, coupled with a move away from late datings for the Gospels.
In this essay we will examine the rise of the Marcan priority hypothesis, the revival of the opposing Griesbach hypothesis, and the emerging influence of what might be termed the “Hebrew Gospels” hypothesis. The reader will come to see that Marcan priority, which is commonly termed an “assured result” of modern biblical scholarship, is no longer a sure thing at all.
Neirynck says that “the absence of Matthew-Luke agreement against Mark in terms of order can be interpreted in more than one way. It can be explained by Marcan priority, but also by any hypothesis that proposes Mark as a middle term. . . . But the real argument from order for Marcan priority is that the differences in Matthew and Luke can be plausibly explained as changes of Mark made according to the general redactional [editorial] tendencies and the compositional purposes of each Gospel.” [Ibid., 588.]
After charting the “Marcan order in Matthew” and the “Marcan order in Luke,” Neirynck presents the chief reason for holding to Marcan priority: “We have discussed the common order of the Triple Tradition pericopes [short extracts] and have explained differences from the Marcan order as editorial divergences by Matthew and Luke. The argument from order, as understood since Karl Lachmann (1835), constitutes the main reason for positing Marcan priority.” [Ibid., 589.]
We will look at Hans-Herbert Stoldt’s persistent critique of this argument later. But first let’s continue with Neirynck’s overview: “Our treatment of order suggesting the priority of Mark as a solution to the synoptic problem is incomplete, for the non-Marcan material shared by Matthew and Luke is still to be considered.” [Ibid., 590.] This material is thought to come from a source called “Q” (from the German Quelle, “source”). “Recent studies tend to confine Q material to (all the) passages attested in both Matthew and Luke. Too uncertain for consideration is the possibility that only Matthew or Luke preserved a passage from Q.” [Ibid.]
Neirynck also looks briefly at the arguments against the originality of Mark. First is source-critical methodology: “Scholars have offered general criteria for deciding which is the more ancient among parallel traditions. Sanders (Tendencies [Edward Sanders, The Tendencies of the Synoptic Tradition (Cambridge: Cambridge University Press, 1969).] has examined these criteria: increasing length and detail, diminishing semitism, and the use of direct discourse and conflation, as they occur both in the synoptic Gospels and in postcanonical material. His conclusion is that `the tradition developed in opposite directions’ and therefore `dogmatic statements’ on the basis of these criteria are never justified.” [Ibid., 272.].
William R. Farmer has argued, notes Neirynck, against the criterion of increasing specificity; he proposed a new criterion, Palestinian or Jewish provenance, but this could be nothing more than re-Judaization. “The message of Sanders’s Tendencies can be understood as a warning against generalization rather than an invitation to synoptic skepticism. More important than his negative conclusion is the recommendation to be alert to `the editorial tendencies of each particular writer.'” [Neirynck, 592.]
Neirynck briefly mentions B. H. Streeter’s five arguments for the priority of Mark (these too will be discussed later). The fourth of these is the primitive character of Mark as shown by that Gospel’s use of phrases likely to cause offense–phrases which are omitted or toned down in the other Gospels–and a roughness of style and grammar, including the preservation of Aramaic words. “More simply we can distinguish two dimensions in a single argument for Marcan priority: taxis or order . . . and lexis or style.” [Ibid., 593.]
The Minor Agreements
Then comes the awkward part for the Marcan hypothesis, the Minor Agreements. These are passages, all of them short (some only a word or two), in which Matthew and Luke agree as against Mark. Often the passages are renderings peculiar enough that one would think two writers would not both stumble on them. If Matthew and Luke both depend on Mark in the Triple Tradition, then how can they agree with each other here and differ from Mark? Neirynck lists four possible answers:
“1. Proto-Mark (or Urmarkus). Matthew and Luke used an earlier version of Mark, shorter than the canonical Mark (which accounts for negative agreements or common omissions) and different in wording (which accounts for coincidences in content, vocabulary, style, grammar).
“2. Deutero-Mark. The Marcan text which was used by Matthew and Luke differed slightly from our Mark because of textual corruption, revision, or edition.
“3. Common source. Matthew and Luke depend on a source other than Mark, perhaps a primitive Gospel or oral tradition.
“4. Luke’s dependence on Matthew. Luke, who follows Mark as his basic source in the Triple Tradition, was acquainted with and influenced by Matthew.” [Ibid.]
Neirynck weighs the Minor Agreements differently than do other writers: “Although they are cited as objection number one against the priority of Mark, it can be argued that often these agreements are in fact not so striking and that for most of the `significant’ agreements a satisfactory redactional explanation can be given.” [Ibid.]
The concluding portion of his study considers alternative solutions to the synoptic problem, including a modified two-source theory (which balloons, in the case of M. E. Boismard, into four original sources, three intermediate Gospels, and the final canonical Gospels, with lines of connection between most of these) and the theory of Luke’s dependence on Matthew, which produces three quite distinct sequences from B. C. Butler’s (Matthew-Mark-Luke), to Austin Farrer’s (Mark-Matthew-Luke), and to William R. Farmer and Bernard Orchard’s (Matthew-Luke-Mark). Neirynck says, ” In all three hypotheses, Luke borrowed the Double-Tradition material [that is, what Matthew and Luke have in common but Mark omits] from Matthew, and there is no need for a hypothetical sayings source. With regard to Mark, conflicting views are defended: absolute priority of Mark (Farrer), Mark’s dependence on Matthew (Butler), Mark as a secondary combination of Matthew and Luke (Farmer). [Ibid., 595.]
Neirynck ends his discussion of the synoptic problem with a short consideration of the renewed Griesbach hypothesis. His doing so is an acknowledgement that this hypothesis, although unable to claim a majority of scholars, seems to be the locus of recent critical activity: “The essential thesis of J. J. Griesbach (1789) that Mark conflated and alternatively depended on Matthew and Luke remains the same in the neo-Greisbachian Two-Gospel hypothesis” [Ibid.] that has been supported by Farmer and Orchard. “A Griesbach-approach commentary has been written by C. S. Mann [C. S. Mann, Mark, vol. 27 Anchor Bible (Garden City: Doubleday, 1986).] . . . The neo-Griesbachians differ from Griesbach in two ways. (1) Luke’s dependence on Matthew is explicitly affirmed and studied. (2) The relevance of patristic evidence is emphasized, particularly Clement of Alexandria’s reference to `the gospels with genealogies which are written first.’ [Preserved in Eusebius, Ecclesiastical History 6:14:5.] . . . Since Farmer’s book (1964), the Griesbach hypothesis has been discussed in numerous Gospel conferences, from Pittsburgh (1970) to Jerusalem (1984), with a great variety of published papers.” [Neirynck, 595.] But Neirynck does not let pass a chance to toss in a barb: “New developments in the theory can scarcely be noted.” [Ibid. Of course, the neo-Griesbachians probably say the same thing about the theory of Marcan priority.]
The German Development
Until the eighteenth century, nearly everyone, whether amateur or scholar, presumed the Gospels were written in the order found in the New Testament. Then, two centuries ago, the synoptic fact became the synoptic problem. From the time of the Fathers it had been clear that the first three Evangelists used many of the same words and composed their Gospels in much the same sequence–but not entirely the same words and not entirely the same sequence. This was not seen as warranting much investigation until the eighteenth century, when suddenly it became a concern, and the nineteenth century, when the concern turned into a problem.
By 1794 Johann Gottfried Eichhorn already had posed his famous alternative: “Either the three [synoptic] Gospel writers made use of one another, or they depended upon a common source.” [J. G. Eichhorn, Allg. Bibliothek der biblischen Literature, first edition, 766, cited in Stoldt, 3.] The race was on, and so far the clear winners have been exponents of the theory of Marcan priority, which holds not only that the Gospel of Mark was the first to be written, but that it was to a large extent the origin of Matthew and Luke.
Supplanted was the traditional theory, known as the Augustinian, which held that Matthew wrote first, followed by Mark as a conflator of Matthew, and that Luke wrote last. The rise of critical biblical scholarship, which may be said to have begun in Eichhorn’s era, coincided with the demise of what had been until then the almost-unanimous opinion of the few people who bothered to write about what later would be called the synoptic problem.
Here should be mentioned a position which has been revived in recent years and which seems to be gaining adherents: the Griesbach hypothesis. Johann Jacob Griesbach (1745-1812) summarized his position this way: “When Mark wrote his book he had in front of his eyes not only Matthew, but Luke as well, and he excerpted from them whatever he intended to preserve of the deeds, words, and destiny of the Savior.” [Ibid., 5-6.]
The Griesbach hypothesis achieved a certain celebrity in the early nineteenth century, and then it fell out of style as the Marcan hypothesis took hold. In recent years– say, from the publication of William R. Farmer’s The Synoptic Problem in 1964–it has made something of a comeback, but the main hypothesis, still, is the Marcan.
The history of the Marcan hypothesis consists of a series of forward steps combined with some backtracking. The way was led by Christian Gottlob Wilke, who published in 1838 a work titled The Ur-Evangelist. Wilke criticized earlier writers, including Eichhorn, who had posited vaguely some sort of ur-Gospel from which the canonical Gospels could have been drawn. The entire absence from history of any mention of such a work was, naturally enough, a substantial drawback to such a theory. Wilke found a way around it.
The ur-Gospel, he said, has been before our eyes all the time–it is Mark itself! “It would be hard to exaggerate the surprise of his contemporaries,” writes Stoldt. “This was a genuine academic sensation. No one before had given a thought to the possibility that the sought-after urGospel, the foundation of the canonical Gospels, was in fact one of the latter.” [Ibid., 30.]
Wilke’s thesis was not without problems. It was based on his notion that “Mark is always the one accompanied”–that is, a comparison of parallel passages demonstrates that Matthew or Luke copied Mark, as proved by the commonality of the language. Of course, this begged the question, since a critic could (and critics did) say that perhaps it was Mark who did the following, copying Matthew here and Luke there. Working with the three Gospels alone, Wilke was unable to sustain his position. To the rescue came Christian Hermann Weisse, who posited a sayings collection. He identified this collection as the logia mentioned by Papias. [Eusebius, Ecclesiastical History, 3:39: “So then Matthew composed ta logia in the Hebrew language.”].
This seemed to solve the difficulties apparent with using Mark as the sole source of Matthew and Luke. Weisse thought Papias must have been referring to a collection of Christ’s sayings and teachings, something along the line of an ancient Bartlett’s confined to quotations from one man.
But further reflection brought up a substantial difficulty: How to account for the common non-sayings material? If the sayings collection included all such matter, what one ended up with was, basically, Mark, which returned one to Wilke’s position. If the sayings collection did not include material other than dominical sayings, then one could not account for common narrative portions. [Stoldt, 47-58.]
Heinrich Julius Holtzmann attempted to resolve such concerns. It is he who is commonly thought to have brought the Marcan hypothesis to a state of completion in the form of the two-source theory. He did this in Die Synoptischen Evangelien, which appeared in 1863. Holtzmann postulated a historical source consisted of almost all of the canonical Mark, the short form of the Sermon on the Mount (Luke 6:20-49), and a few stray verses from John and Matthew, and he accepted a sayings collection or logia. According to Holtzmann, all three synoptics used these sources and therefore constitute variations of them, either by omission or addition. Mark, the shortest of the Gospels, contains little “historical” material. If some event did not appear in a particular Gospel, it is because it was omitted when a particular Evangelist copied from the historical source. Similarly for sayings: If a saying is missing from a Gospel, it is because the writer failed to copy it from the sayings source.[Ibid., 69-86.]
The weakness in Holtzmann’s theory was that it could not answer why a particular writer would leave out a particular passage. Holtzmann would say it was an “unimportant” passage, so Matthew, say, left it out. Then why did not Mark also leave it out? Mark “left out” the Sermon on the Mount, but he “included” a parable chapter (Mark 4:1-34) which is five verses longer than the Lucan Sermon on the Mount. He also included an apocalypse (Mark 13:5-37), which is three verses longer.
Why leave out one and keep the others? Are we to imagine Mark thought the Sermon on the Mount “unimportant”? Are the other passages “important” by comparison? Holtzmann’s alternative notion that the Sermon on the Mount was dropped to save time prompts the question, “Why not save more time and drop the other passages also?”
William R. Farmer has written that Holtzmann’s ” most weighty point . . . was the scientifically gratuitous but powerfully apologetic fact that it was conceded by all contemporary critics that Matthew was secondary to the eye-witness period. This meant that the Griesbach hypothesis, as well as the Augustinian and all others which made Matthew the earliest of the Gospels, simply could not provide a viable solution to the source problem. [William R. Farmer, The Synoptic Problem (Dillsboro: Western North Carolina Press,  1976), 37.].
Paul Wernle is credited with developing, in Die synoptische Frage (1899), an idea which since has been taken as self-evident by many scholars: The emergence of a longer work from a shorter is more easily explained than the emergence of a shorter work from a longer–enlargement is more likely to occur than abridgment. [Ibid., 105.] He continued this idea by saying a longer work may be not just an expansion of a shorter, but a combination of a shorter with outside sources. These sources themselves might form a series, one dependent on the other. Applied to the Gospels, Wernle called these multiple sources “Q” (he identified seven Q sources in all). Unlike the logia mentioned by Papias, Q, in its various forms, was not confined to sayings material, but included narrative material (thus logia would not be a proper term for it).
The perfection of the two-source theory came with Berhard Weiss (1827-1918), professor of New Testament at the University of Berlin. This “perfection” came not so much from a new insight–Weiss argued that Q could not be confined to sayings materials, but also must include narrative material, but he was not the first to suggest this–as from Weiss’s status as perhaps the top biblical scholar of his time. His approbation of the two-source theory gave the theory a momentum it might not otherwise have sustained.
The greatest drag on the theory all along had been the gratuitousness of Q: Aside from Papias’s line about the logia, there was nothing at all to suggest that a document (or collection of documents) such as Q existed. No one mentioned it, and its disappearance from history seems remarkable. If the synoptics were all dependent on it, how could it have been allowed to disappear? If an epistle as inconsequential as Philemon had been preserved carefully for eighteen centuries, why had not Q, an immeasurably more important document?
Streeter and Farmer
Although the argument for Marcan priority was developed first and almost exclusively by German scholars, in this century the argument has been carried in large part by English scholars, the chief among them being B. H. Streeter, who made five main points:
1. Matthew reproduces 90 percent of Mark, and Luke reproduces more than half, in language nearly identical with that of Mark. This is the argument from content.
2. In passages found in all three synoptics, Matthew or Luke or both are in verbal agreement with Mark, and they almost never agree against Mark. This is the argument from wording.
3. The ordering of incidents seems more original in Mark and is, on the whole, supported by Matthew and Luke. When one of the latter departs from the Marcan pattern, the other is found to support it. This is the argument from arrangement.
4. Mark’s language is rough. Matthew and Luke improve it. This is the argument from language.
5. The way in which material–both Marcan and non-Marcan–is distributed throughout the other synoptics suggests each was working independently with Mark plus some other sources.
Streeter outlined these considerations in The Four Gospels: A Study of Origins He rejected attempts to revive the Augustinian hypothesis: “The attempt has recently been made to revive the solution, first put forward by Augustine, who styles Mark a kind of abridger and lackey of Matthew, `Tanquam breviator et pedes-quus e jus.’ But Augustine did not possess a synopsis of the Greek text conveniently printed in parallel columns. Otherwise a person of his intelligence could not have failed to perceive that, where the two Gospels are parallel, it is usually Matthew, and not Mark, who does the abbreviation.
“For example, the number of words employed by Mark to tell the stories of the Gadarene demoniac, Jarius’s daughter, and the feeding of the five thousand are respectively 325, 374, and 235; Matthew contrives to tell them in 136, 135, and 157 words. Now there is nothing antecedently improbable in the idea that for certain purposes an abbreviated version of the Gospel might be desired, but only a lunatic would leave out Matthew’s account of the infancy, the Sermon on the Mount, and practically all the parables, in order to get room for purely verbal expansion of what was retained.” [B. H. Streeter, The Four Gospels: A Study of Origins (London: Macmillan 1924), 157.].
Streeter’s confident style naturally enough attracted followers and expanders, one of the most prominent of whom has been Joseph A. Fitzmyer, who terms Streeter’s fifth argument his weakest and who explains that the priority of Mark can be found in the more primitive character of the narrative, what might be termed its freshness and circumstantial character. Mark’s phrases are more likely to cause offense, his grammar and style are rough, and he preserves Aramaic words. [Joseph A. Fitzmyer, “The Priority of Mark and the `Q’ Source in Luke,” in Jesus and Man’s Hope (Pittsburgh: Pittsburgh Theological Seminary, 1970), 1:134-147.].
Proofs for Marcan Priority
The hypothesis of Marcan priority relies on half a dozen main proofs; they are basically duplicative of the proofs given by Streeter. Each has undergone development, each has its partisans and critics. The basic proof is what Karl Lachmann, in 1835, called the “proof from order.” [Karl Lachmann, De Ordine Narrationum In Evangelis Synopticis, cited in Stoldt, 135.]. Weisse had put it this way: “It is precisely this consideration which carries the ultimate, decisive weight in favor of our view on the mutual relationship of the synoptic Gospels.” [Berhard Weisse, Evangelische Geschichte, 1:68, cited in Stoldt, 135.]. William Wrede, writing in 1901, said that “the strength of the Marcan hypothesis lies specifically in the fact that the sequence of the narratives in Mark underlies the sequence in Matthew and Luke.” [Ibid., 136.]. This is a key point, but not one accepted by all.
Some ask whether there is really a common sequence, a common narrative thread that can be isolated and agreed upon by scholars, or is the truth what Stoldt, the chief present-day critic of the two-source hypothesis, claims, “that there is neither a continuous actual nor even a merely reconstructible common narrative line. . . . There are only shifting parallels between Mark on the one hand and Matthew and Luke on the other, sometimes with one, sometimes with the other”? [Ibid., 141]. The proof from order depends on, in Weisse’s words, the concurrence of the three synoptics being “always mediated by Mark.” [Ibid., 143.].
Stoldt argues that ” Weisse makes a serious error in logic. For if all three synoptic authors concur with one another, it does not necessarily follow that the concurrence of the other two is `always mediated by Mark.’ This is readily shown by the simple fact that one could just as well reverse the conclusion: Mark concurs with Matthew and Luke only as long as these two agree with each other, but, whenever they do not agree with each other, Mark concurs with neither. Consequently, the concurrence of Mark with these two is always mediated solely by the agreement of these two with each other.” [Ibid.].
This problem was recognized by B. C. Butler, who reproduces an example taken from E. A. Abbott’s The Fourfold Gospel: “Matthew and Luke are in the position of two schoolboys, Primus and Tertius, seated on the same form, between whom sits another, Secundus (Mark). All three are writing (we will suppose) a narrative of the same event. … . Primus and Tertius copy largely from Secundus. Occasionally the two copy the same words; then we have agreement of the three writers. At other times Primus (Matthew) copies what Tertius (Luke) does not.
“But Primus and Tertius cannot look over each other’s shoulders, and hence agreement of them `against’ Secundus is only an accident. As the same results (exactly) will follow, if Secundus copied from Primus (or Tertius) and was himself copied by Tertius (or Primus), we must hope that Abbott, who was headmaster of a famous school, was not illustrating from real life.” [B. C. Butler, The Originality of St. Matthew: A Critique of the Two-Document Hypothesis (Cambridge: Cambridge University Press, 1951), 66.].
The second main proof of Marcan priority is the proof from uniformity. It is based on the proof from order. Mark’s is the only Gospel which exhibits an unbroken narrative continuity. It forms a whole. By comparison, the other synoptics seem to lack continuity.Hans-Herbert Stoldt says, “When one juxtaposes the other two synoptic Gospels against Mark, they could scarcely be more different. Compared to Mark they appear to lack unity; one can even say that they are poorly composed. A major block of discourse is inserted into the middle of Matthew’s historical narrative. In Luke the genealogy is interpolated inorganically between the baptism and the temptation, interrupting the previous train of thought.” [Stoldt, 156-157.].
The proof from originality is repeated by nearly every advocate of Marcan priority. Mark gives the impression of having been composed on its own, at once, by one man; it does not seem to be a revision of another’s work, nor does it seem to be composed of sections so disparate in style that a reader would suspect them to have been lifted from earlier writings.
The proof from language is thought by some scholars to be the weakest of the several proofs since it is rather subjective. Mark’s writing is said to be the most immature, and this suggests it came first. Matthew and Luke are more polished, suggesting they came later, or at least suggesting that they may have borrowed from Mark but that Mark probably would not have borrowed from them (and then introduced rougher language).
The proof from doublets has been one of the main supports of the theory of Marcan priority, a doublet occurring when a variant reading of an incident is given. Matthew and Luke share five doublets; Matthew has six others proper to his Gospel, and Luke has three others proper to his. A doublet may arise when an author finds two sources describing what is apparently the same event in slightly different ways. He may incorporate both descriptions into his new work. The argument from doublets has been criticized on the grounds that Mark also contains doublets, implying the argument as a whole proves nothing either way. [Ibid., 173-184.].
The last of the major proofs is the one from the Petrine origin of Mark. If Mark had been a close follower of Peter’s, and if he wrote down Peter’s remarks about the Lord, [As reported by Papias, Fragment 39], then one could argue toward a relatively early date for Mark–at least relative to the other synoptics. Curiously, Stoldt takes considerable pains to counter this argument, even though it is not the key argument for Marcan priority, yet he does so unconvincingly. If Mark really were Peter’s follower, he says, then “one would have every right to expect that the figure of Peter would not only take a central position in the Gospel of Mark, but also, beyond that, would stand out considerably more than in the other Gospels. This, however, is not the case–on the contrary.” [Stoldt, 187.].
It is curious that Stoldt should say this since many writers have argued precisely the opposite from Peter’s lack of prominence in Mark. It has been said that if Mark had been Peter’s disciple, it is likely he knew a Peter who was self-effacing, quick to recount his own failings (such as the triple denial) and reluctant to recount anything which might be to his credit. A faithful disciple of such a teacher would tend to follow his teacher’s example (and wishes?) and minimize Peter’s role. In other words, Mark’s relative silence concerning Peter argues in favor of, not against, Mark’s having been in Peter’s company.
Role of William R. Farmer
The chief critic of the Marcan hypothesis in the English-speaking world has been William R. Farmer, who teaches at the Perkins School of Theology at Southern Methodist University. Farmer has been at the center of an increasingly influential, but still minority, school of thought. This school has been promoted through seminars, festivals, colloquia, and conferences, beginning around 1966, and in a series of works, many published by Mercer University Press. [For example, Edward Sanders, Jesus, the Gospels, and the Church: Essays in Honor of William R. Farmer (Macon: Mercer University Press, 1987); William R. Farmer, ed., New Synoptic Studies (Macon: Mercer University Press, 1983); Arthur J. Bellinzoni, Jr., ed., The Two-Source Hypothesis: A Critical Appraisal (Macon: Mercer University Press, 1985)].
In The Synoptic Problem Farmer gave new credence to the Griesbach hypothesis and criticized the received position of Marcan priority. This is not to say he was the first to find fault with the majority position. “In 1951, B. C. Butler called attention to the fallacy of the argument from order when taken on Streeter’s terms, and Butler’s analysis of the fallaciousness of Streeter’s reasoning at this point was cogently reiterated at Cambridge by G. M. Styler.” [William R. Farmer, The Synoptic Problem (Dillsboro: Western North Carolina Press,  1976), 50.].
Farmer summarized the development of the two-source hypothesis, looking at the period up to and culminating in Holtzmann, then focusing on the English endorsement of the German position. After an extended critique of Streeter’s arguments, Farmer looked at the intellectual climate which, he said, allowed the Marcan hypothesis to reign unchallenged for so many decades. He identified a kind of silent argument:
“Eventually a new argument was to develop which, though it never appeared in print, exercised a greater influence in sustaining belief in the priority of Mark than any that Streeter or anyone else has ever published. This argument may be formally stated as follows: `It is inconceivable that so many scholars could have been so wrong on such a fundamental point for such a long period of time.’ This is a powerful argument precisely because in practice it is irrefutable. Every scholar respects honest humility and sincere pietas in his colleague. The critic who sets his judgment against a consensus endorsed by the vast majority of experts during a long period of time does so at the risk of being guilty of academic arrogance and of losing the confidence of his colleagues.” [Ibid., 195.].
Bernard Orchard’s Spin
Farmer has not been the only well-known scholar writing against Marcan priority. On the Catholic side [William R. Farmer joined the Catholic Church a few years ago, but his writings cited in this article were written while he was a Methodist.] there has been Bernard Orchard, O.S.B., [See his article “The Evolution of the Gospels” in this issue of This Rock.]. a Benedictine monk of Ealing Abbey in London. In 1976 he published Matthew, Luke & Mark,[Bernard Orchard, Matthew, Luke & Mark (Manchester: Koinonia Press, 1976)] which was his first extended treatment of the problem. In that work he described himself as “follow[ing] up the work begun over forty years ago by Dom John Chapman, and continued by Bishop B. C. Butler and then by Professor William R. Farmer. . . . For them, and for me too, the two-document hypothesis and the priority of Mark are still only hypotheses, not infallible dogmas, and they have stood secure for so long chiefly because no one has been able to offer any satisfactory alternative.” [Ibid., vii.]. In Matthew, Luke & Mark Orchard tried to sketch such an alternative. He hoped to follow up quickly with two further volumes, but his plans changed, and eleven years later, with Harold Riley, an Anglican priest, he published The Order of the Synoptics: Why Three Synoptic Gospels?[Bernard Orchard and Harold Riley, The Order of the Synoptics: Why Three Synoptic Gospels? (Macon: Mercer University Press, 1987).].
Unlike most other commentators, Orchard and Riley did not feel compelled to confine themselves to literary or internal arguments. Such evidence is examined by Riley in the first section of the book. Then Orchard tackles the historical tradition, looking at the writings of early Christian writers and heretics, both before and after Papias.
Finally there comes a synthesis of the internal and external evidence, the authors concluding that Matthew was written between A.D. 30 and 44 and that if a date after 70 is posited, “then the problem of its relationships to the other Gospels becomes insoluble.” [Ibid., 275.]. They see the Gospels as having been “the reactions of the church’s leaders to the successive phases through which the pre-70 church is known to have passed–crises which can actually be pinpointed.” [Ibid.].
The “Hebrew Gospels”
John A. T. Robinson said he wanted to take a fresh look at the presuppositions used in dating the New Testament books, presuppositions which, he thought, had not been reexamined critically since the nineteenth century. This look convinced him the presuppositions were little more than prejudices.
He started from scratch and concluded that every book of the New Testament was written prior to the fall of Jerusalem in A.D. 70. The Gospels of Matthew, Luke, and even John he put as early as the forties. [John A. T. Robinson, Redating the New Testament (Philadelphia: Westminster, 1976)].
Robinson “worked from an exclusively historical methodology,” wrote Jean Carmignac in The Birth of the Synoptics ” I work with a methodology which is principally philological but historical on occasion.” [Jean Carmignac, The Birth of the Synoptics (Chicago: Franciscan Herald Press, 1987), 86.]. Carmignac, a Dead Sea Scrolls translator and an expert in the Hebrew in use at the time of Christ, reached conclusions similar to Robinson’s, but he came at the problem from a different angle. He translated the synoptic Gospels “backwards,” from Greek into Hebrew, and was astonished at what he found.
“I wanted to begin with the Gospel of Mark. In order to facilitate the comparison between our Greek Gospels and the Hebrew text of Qumran, I tried, for my own personal use, to see what Mark would yield when translated back into the Hebrew of Qumran. I had imagined that this translation would be difficult because of considerable differences between Semitic thought and Greek thought, but I was absolutely dumbfounded to discover that this translation was, on the contrary, extremely easy. Around the middle of April 1963, after only one day of work, I was convinced that the Greek text of Mark could not have been redacted directly in Greek and that it was in reality only the Greek translation of an original Hebrew.” [Ibid.,1.].
Carmignac, now deceased, had planned for enormous difficulties, but they did not arise. He discovered the Greek translator of Mark had slavishly kept to the Hebrew word order and grammar. Could this have been the result of a Semite writing in Greek, a language he didn’t know too well and on which he imposed Hebrew structures? Or could the awkward phrasings found in our Greek text have been nothing more than overly faithful translations (perhaps “transliterations” would be more accurate) of Semitic originals? If the second possibility were true, then we have synoptic Gospels written by eyewitnesses at a very early date.
Carmignac spent most of the next 25 years meticulously translating the Greek into Hebrew and making endless comparisons. The Birth of the Synoptics is a popular summary of what he hoped to publish in a massive multivolume set. Some consider it a disturbing book–which it is. Let’s consider just one example of what Carmignac discovered.
The Benedictus, the song of Zachary, is given in Luke 1:68-79. In Greek, as in English, the Benedictus seems unexceptional as poetry. There is no evidence of clever composition. But, when it is translated into Hebrew, a little marvel appears. In the phrase “to show mercy to our fathers,” the expression “to show mercy” is the Hebrew verb hanan, which is the root of the name Yohanan (John). In “he remembers his holy covenant,” “he remembers” is the verb zakar, which is the root of the name Zakaryah (Zachary). In “the oath which he swore to our father Abraham” is found, for “to take an oath,” the verb sh<>ba, which is the root of the name Elishaba (Elizabeth).
Carmignac asks, “Is it by chance that the second strophe of this poem begins by a triple allusion to the names of the three protagonists: John, Zachary, Elizabeth? But this allusion only exists in Hebrew; the Greek or English translation does not preserve it.” [Ibid., 28.]. He explains that “Hebrew has a great preference for plays on words, and it takes great pleasure in making reference to similar sounds, which facilitate the task of memorization. Another typical case is hidden in the Our Father (Matt. 6:12-13), in which the word `forgive’ corresponds to the root nasa, `debts and debtors’ to nashah, and `temptation’ to nasah. Is this yet another case of mere chance? Isn’t it reasonable to think that these words have been chosen by design in order to produce a sort of internal rhyme?” [Ibid.].
Carmignac gives many other examples, and he draws these conclusions about the dating of the synoptics: ” The latest dates that can be admitted are around 50 for Mark …. .. around 55 for Completed Mark, around 55-60 for Matthew, between 58 and 60 for Luke. But the earliest dates are clearly more probable: Mark around 42, Completed Mark around 45, (Hebrew) Matthew around 50, (Greek) Luke a little after 50.” [Ibid., 61.].
These dates are all approximate, of course, particularly those for Mark and Matthew, and they are the result of Carmignac’s mainly philological analysis. Carmignac draws a few other conclusions:
“1. It is certain that Mark, Matthew, and the documents used by Luke were redacted in a Semitic language.
“2. It is probable that this Semitic language is Hebrew rather than Aramaic.
“3. It is sufficiently probable that our second Gospel [that is, Mark] was composed in a Semitic language by St. Peter the Apostle” [with Mark being his secretary perhaps]. [Ibid., 87.].
Expanding on this last point, he says that ” it is probable that the Semitic Gospel of Peter was translated into Greek, perhaps with some adaptations by Mark, in Rome, at the latest around the year 63; it is our second Gospel which has preserved the name of the translator, instead of that of the author.” [Ibid.].
As he wrote The Birth of the Synoptics, Carmignac suspected his “scientific arguments [would] prove reassuring to Christians and [would] attract the attention and interest of non-believers. But they overturn theories presently in vogue and therefore they will be fiercely criticized.” [Ibid., back cover.]. They may also be, with Carmignac’s death, fiercely (and quite unjustly) ignored.
Claude Tresmontant is a member of the faculty at the Sorbonne. In The Hebrew Christ [Claude Tresmontant, The Hebrew Christ (Chicago: Franciscan Herald Press, 1989)]. he takes, like Robinson, something of a historical approach, but also, like Carmignac, something of a philological approach. He combines the two and comes up with a provocative thesis: All four Gospels were written in Hebrew first, then translated.
Hebrew Matthew was written shortly after the Resurrection, followed within a few years by a Greek translation. Hebrew John was written by A.D. 36, then also quickly translated. Luke came next, being written between 40 and 60, and Mark probably came last, being written between 50 and 60, but Mark may have squeaked in front of Luke.
Tresmontant sets aside what he regards as unwarranted historical presuppositions and tries to examine each Gospel cold, as it were. Throughout his book he notes the Evangelists’ references to persecution. Who were the persecutors? He suggests they probably were not Romans working under Nero’s orders after A.D. 64, since the New Testament almost never warns against Romans. The warnings are almost always couched in terms of the Jewish authorities, and anti-Christian sentiment was evident early on–witness the stoning of Stephen. It’s more likely, then, says Tresmontant, that the persecutions which form the backdrop to many passages in the New Testament occurred not late in the first century, but by the middle of it. Tresmontant, writing of Matthew, says:
” It does not date only from the end of the first century A.D. as the majority of exegetes hold today. All the indications, signs, and characteristics of the book we call the Gospel according to Matthew point to a very ancient period, a period only shortly after the momentous events of A.D. 30–certainly before the first joyous proclamation of the good news of Christ to the pagans and the uncircumcised which occurred around A.D. 36-40.
“There is absolutely nothing in this Gospel that would lead us to suppose that it was composed later; there is no text, nor any fragment of a text; there is not so much as a mark; there is nothing. The claim that the Gospel according to Matthew was only composed toward the end of the first century is a totally arbitrary claim. The only thing this claim has going for it is that fact that the majority opinion among exegetes today supports it. That is simply to say that the opinion rests upon nothing but itself.” [Ibid., 9.].
Making a pericope-by-pericope examination of much of Matthew, Tresmontant points to the exclusively Jewish provenance of the writing. This is not just a matter of style. It is also a matter of substance. Matthew writes for an audience composed of Jews, not Greeks. But the Church quickly ceased to be a solely Jewish affair, notes Tresmontant, which implies a very early dating for Matthew. Similarly the other Gospels are given early dates.
Along the way Tresmontant junks opinions which have been taken as a matter of course by many scholars. Luke’s Gospel, for instance, is often said to be the one with the best Greek. But Tresmontant says that, “contrary to the general belief that has served as the conventional wisdom for some time, the Gospel according to Luke is the least Greek of the four and the most Hebrew (except for its first phrase).” [Ibid., 24.].
He argues that ” it is probably because of these first four verses that scholars for a long time have praised the literary style of Luke and have considered Luke an especially well educated writer, a Hellenist, and the like. From verse 5 on, however–and practically to the end of this Gospel–the Greek style of Luke is no longer quite so natural. Rather, it is manifestly a Greek that has been translated, and translated from the Hebrew at that!” [Ibid., 105.].
As for Mark, Tresmontant says, “I do not hold with the majority opinion of scholars regarding the composition of Mark. I do not believe that Mark was the first Gospel written. I certainly do not think that it is older than Matthew. On the contrary, all the indications point to a very ancient date for the composition of Matthew. Not all of these same indications apply in the case of Mark.
“We need only think, for example, of the sayings of our Lord in which he indicated that he was sent to preach exclusively to the Hebrew people, to save only the lost sheep of Israel. The sayings concerning the sign of Jonah are similarly not to be found in Mark. Also, the text of Mark tends to eliminate some of the words or phrases which would have grated most harshly on the ear of the educated Greek.” [Ibid., 102.].
This is the way in which Tresmontant summarizes his position: “In asserting that documents written in Hebrew lie behind the present Greek text of our four Gospels, we are, of course, proceeding by way of a hypothesis. Here, however, we are dealing with a hypothesis that has to be true–indeed, has to be considered certain–because it is the only hypothesis capable of explaining all the features that we find in the Greek text of the Gospels.
“The contrary hypothesis is that a long tradition of oral preaching and transmission preceded the setting down in writing at a comparatively late date of the Greek text of the four Gospels as we presently possess it. The contrary position is the one preferred by the majority of biblical scholars today, as has been the case for more than a century.” [Ibid., 197.].
In the epilogue, Tresmontant says that, as he “became more familiar with the subject matter, I began to perceive difficulties. Then I encountered some simple impossibilities. Finally, the entire superstructure collapsed like the proverbial house of cards.
“The more I studied the Old Testament the more I began to recognize the Hebraic phraseology behind the Greek in each of the four Gospels. Eventually, I arrived at the conclusions that I have tried to set forth in this book: Matthew and John wrote the earliest Gospels; those of Luke and Mark were written later; all four of the Gospels, as well as some of the other New Testament books, were evidently translations into Greek from earlier texts originally composed in Hebrew.” [Ibid., 319.].
Tresmontant notes that “philosophical assumptions, like philosophical preferences and dislikes, have always played a considerable role in the great scientific debates and controversies of the past. In the history of biblical scholarship and exegesis, it is evident that such philosophical assumptions adopted prior to any objective exegesis have played an equally considerable role. Ernest Renan openly declared this to be the case in the famous preface to the thirteenth edition of his Life of Jesus.” [Ibid., 320.].
Tresmontant’s theory is too new to have been subjected yet to the critical examination required of any theory. In a way he uses John A. T. Robinson’s method of attack: Set aside presuppositions and examine the text raw–then explain what you see. His familiarity with the Hebrew of the Dead Sea Scrolls and his willingness to step on scholarly toes makes his a provocative thesis. If he is correct that the Gospels were Hebrew compositions first and that they therefore must have been written early, and if Matthew is more ancient in style than Mark, then not only has Marcan priority been disproved, but so has the existence of Q, which, generally, has been posited as necessary if the Gospels were written fairly late in the first century. If they were written within a decade or two of the events they report (or even within a handful of years, in the case of Matthew), Q seems superfluous.
Jean Carmignac and Claude Tresmontant worked parallel to one another, not in concert, not in tandem. Their conclusions are similar but distinct. Both hold to Hebrew originals, not just Hebrew scraps included among the (mainly Greek) documents used by the authors. Both hold to early datings. But Carmignac comes down in favor of Marcan priority (though not insisting on it), while Tresmontant insists the first Gospel to be written was Matthew’s.
No “Assured Results”
It is too early to say what response will be given to these French scholars. Carmignac is dead, and the dead can be ignored easily. Tresmontant is alive, but he writes with only a minimum of diplomacy and nuance; he may not be taken seriously, even if his ideas should be. However that may be, their books are signs the discussion remains open and lively. Indeed, “in recent years the discussion of the synoptic problem has been much more lively than it has been for a generation,” writes Lamar Cope. [Lamar Cope, “The Argument Revolves: The Pivotal Evidence for Markan Priority is Reversing Itself,” in William R. Farmer, ed., New Synoptic Studies (Macon: Mercer University Press, 1983), 143.].
“In the decade of the 1980s, most sensitive critics are aware that many of the arguments which led to the consensus that Mark is the earliest Gospel and that there was a source Q are, in themselves, circular or fallacious. . . . [Nevertheless], the great majority of New Testament critics trained in the milieu of the scholarship of the last half century will still retain the workability of the Mark-Q hypothesis in some general form as the best available solution to the problem of the relationships among the Gospels. . . . But whatever the critical and historical consequences may be, new evidence about the structure of the key passages points unmistakably toward the literary priority of the Gospel of Matthew over Mark and Luke. Unless this evidence is countered successfully, nothing–scholarly consensus, scholarly reputation, or vehement denial–can rescue the priority of Mark.” [Ibid., 143, 159.].
If the 1980s saw a discussion “more lively than it has been for a generation,” the 1990s may see one livelier still. Cope and the people clustering around William R. Farmer will continue to undermine Marcan priority. Scholars following the lead of Carmignac and Tresmontant will undermine late dating.
Scholars adhering to what is now the majority opinion will attempt to shore up their positions. It may be that what has been regarded as an “assured result of modern biblical scholarship” will be regarded as anything but assured at the turn of the millennium, or it may be that the upstart theories, by calling attention to themselves, will have invited precisely the scrutiny that will insure their downfall.