Could the Gospel of Matthew be dependent on Luke's Gospel?

Question

Why is it that a possible Gospel dependency order of Mark-Luke-Matthew without requiring 'Q' is considered particularly unlikely, when compared to other possible solutions to the synoptic problem? Do any respected NT scholars favor (or give credence) to this solution?

Answer 1

You have asked about the order being Mark > Luke > Matthew. My answer below addresses the order of Luke being written before even Mark. This arrangement is called Lukan Priority. It is very much a minority opinion. As you know, the prevailing theory in New Testament studies is Markan Priority. It has the most support among scholars. A minority position (which I hold) is Matthean Priority, that Matthew wrote first then Mark then Luke. I am unaware of any scholar supporting John's Gospel being written first (though some might). A quick search did not turn up anything supporting that stance either.

Brad H. Young defends Lukan Priority, that Luke's Gospel is before even Mark. Young is a member of the Jerusalm School. Lukan Priority is not original with them, a scholar named William Lockton first proposed it. Then Robert Lisle Lindsey began writing on it independently of Lockton (that is, Lindsey had not read Lockton's proposal).

Lindsey also disagreed with the priority of Q. Instead, he proposed a source called A for "Anthology". A differs from Q because A is a full life of Jesus's life, teachings, and actions while Q is a sayings document.

Lindsey hypothesized this arrangement while attempting to translate the Gospel of Mark into Hebrew. He noticed there were many hundreds of times where Luke used a semitic phrase (a phrase that comes from Hebrew or other semitic language) but Mark did not. As it makes more sense for the semitisms to be present in the older documents than the younger, Lindsey formed the following arrangement:

1. There was an original Hebrew Gospel.
2. There was a collection of poems, stories, and accounts of Jesus' life. This is called A and was probably in Greek.
3. There was also a lost Gospel, R, in Greek.
4. Luke used A and R in compiling his Gospel.
5. Mark used Luke with little regard for A and no reliance on R.
6. Matthew used Mark and A but not Luke's Gospel.
7. Luke and Matthew were unaware of the other's gospels.

David Flusser (1917-2000), the other founding member of the Jerusalem School of Synoptic Research, also agrees at least in part with this source arrangement. Flusser believed that Hebrew sources predate all three synoptic gospels. Brad Young was his student so it comes as no surprise that Young also supports Lukan Priority. David Bivin and Roy B. Blizzard Jr might also support it.

A very detailed blog series on Lukan Priority (at the very least it raises good questions about Markan Priority) is available here.

Answer 2

In my understanding, the key arguments put forward for the order Mark > Luke > Matthew (i.e., for "Matthean posteriority") are:

• the literary observation that Matthew appears to collect, collate, and develop traditions found in Luke (e.g., what appears in Matt 5-7 in the "Sermon on the Mount" is found at various points, and in a more "primitive" form at various places in Luke);
• at points in the "double tradition" (the common "Matthew-Luke" material) Luke's version appears to be older than Matthew's; and
• Matthew's "theology" appears to resonate with features of later documents (John's gospel, the Didache), making it post-date Luke.

These broad categories have finer subdivisions in the specialist literature (see below). On the other hand, arguments against this ordering include:

• Matthean elements in the double tradition point towards his priority over Luke (and this order has had a recent advocate in Mark Goodacre, and see the critique of Goodacre's proposal by John Kloppenborg);
• it negates any kind of influence of "oral" traditions on the synoptics, and requires our texts to be explained solely in terms of literary dependence (which is thought to be unlikely by many); and
• Matthew's schematic organization is a general tendency, and does not necessarily display a selective systematization of Lukan material.

Further Reading

The two main scholars who espouse this view are:

• Martin Hengel (giant of 20th C. NT scholars) in an appendix to his The Four Gospels and the One Gospel of Jesus Christ (T & T Clark, 2000) (n.b. this is not a "full blown" version of the hypothesis, as he still allows for other sources); and
• Ronald V. Huggins, "Matthean Posteriority: A Preliminary Proposal", Novum Testamentum 34/1 (1992): 1-22 (not so well-known, but published in one of the premier NT academic journals).

There is also an helpful write-up found in a substantial unpublished (student?) essay by Tim Lewis, "Solving the Synoptic Problem For Students?". Mention must also be made of Bartosz Adamczewski, Q Or Not Q?: The So-called Triple, Double, and Single Traditions in the Synoptic Gospels (Peter Lang, 2010) - so far as I'm aware the only recent monograph-length attempt to argue for "Mk > Lk > Mt". Unfortunately, it spawned what has been called "the most negative book review known" in New Testament studies. Caveat lector!

Answer 3

Let's consider for a moment what the Farrer (Mt used Mk, Lk used Mk and Mt) and Wilke (Lk used Mk, Mt used Mk and Lk) theories suggest that the third evangelist in each case did. (For what it's worth, I would regard Kloppenborg's layered Q as a nuanced form of Wilke: he puts the sayings material in the Lucan order, then adds in some para-Marcan material.)

Under the Farrer hypothesis, what Luke does is relatively simple:

0) Let's ignore the infancy narrative for now – it differs in style and substance from the rest of the Gospel and could well be secondary under any hypothesis.

1) In 3.1-4.15, Luke tries to reconcile Mk and Mt – two baptisms would not work.

2) The Rejection at Nazareth is intentionally displaced.

3) From 4.31-6.16, Luke basically follows Mk 1.21-3.19. This is a reasonable choice – the Matthaean parallels differ substantially in order in this part of the gospel.

4) From 6.17-7.10, Luke follows Mt 5.1-8.13, ignoring the Healing of the Leper (as it has already been used in its Marcan context) and setting to one side a lot of the sayings material. It is dangerous to speculate as to the motives of 1st Century evangelists, but despite the position of the Sermon on the Mount in cultural kudos, I feel a good deal of sympathy for Luke's abbreviation: one would never want to read something that long in church, let alone preach on such a text.

5) In 7.11-35, Luke is showing his intentions again. He brings forward the Inquiry of John the Baptist, as using it in his next Matthaean block would put it after John's death. Then there's some special material (the Widow of Nain's Son) and the deliberately relocated Anointing.

6) In 8.1-9.50, Luke returns to following Mk 4-9. The Great Omission of course occurs in this block, eliminating the duplicate Feeding miracle, Jesus calling a Gentile woman a dog, and the strange observation on sewage.

7) The special incident in the Samaritan village (9.51-56) sits on the most notorious seam in Luke.

8) Immediately, Luke returns to following Matthew. The sayings on following Jesus (Mt 8.18-22/Lk 9.57-62), a blatant skip over material used in its Marcan context, the harvest labourers sayings (Mt 9.37-38/Lk 10.2), a reworked Mission discourse with almost all the Marcan material removed and some sayings set aside, a skip over the relocated Inquiry of John, then those Woes against the Cities (Mt 11.20-24/Lk 10.13-15), a special piece on the Fall of Satan (which is clearly the key to understanding Luke's discourse), Jesus' Thanksgiving to the Father (Mt 11.25-27/Lk 10.21-22), and then the most wonderful piece of Lucan craftsmanship: relocating the Blessed are your Eyes saying from a dangling location in the Parables Discourse once the Marcan material is removed to be a fitting finale to his Mission Discourse.

9) At this point Luke makes another deliberate relocation: the Greatest Commandment, expanded upon with that famous parable and that strange story about Martha (I'm with Caird on that one: it's a warning against self-concern).

10T) First terminal block. Luke begins to re-use the sayings material from other sections. In the terminal blocks before returning to Matthew's order, it is always material transferred from sections 4-5 above (the Sermon and the material following it). In this instance it's the Lord's Prayer and the Ask, Seek, Knock bit.

11) Luke then returns to Matthew's order with the Beelzebul Controversy (Mt 12.22-45/Lk 11.14-26). Luke once more shortens the long discourse, pulling out the Sign of Jonah and Sin against the Spirit material to bracket the next section (12A). As Matthew's conclusion (12.46-50) is Marcan material, Luke creates his own (11.27-28) on a similar theme, perhaps influenced by his clever conclusion to his Mission Discourse.

12A) Luke's non-terminal sections use Matthaean material in a specific order: they draw first from sections 4-5 (Sermon and Sequel), then from sections 15-16 (stretching back from the Eschatological Discourse to the Lost Sheep), then concluding with material from the Mission Discourse. The effect is broadly to refocus didactic and eschatological material on Mission. Lk 11.29-12.12 is the first such section.

12B) This is followed, with a short introduction (Who made me judge?), by another section drawing from the three Matthaean sections in turn (Lk 12.13-53). In this section, Luke betrays his modus operandi: Lk 12.35-38 is paralleled in Mark's Eschatological Discourse, not Matthew's.

12T) The second terminal section (12.54-13.9). Once more, the Matthaean material comes from sections 4-5.

13) The special healing of the woman bent double provides a new scene for rejoining Matthew's order with the non-Marcan part of chapter 13 (the Parables Discourse). After two parables, Luke decides he has had enough. Maybe those tares were too eschatological for his de-eschatologizing programme. I feel somewhat sadder that Luke omitted the ending about things old and new, but I am not a first-century evangelist.

14A) In 13.22-14.33, Luke resumes his pattern of drawing from each of the Matthaean blocks in turn, with short bursts of special material to make it fit together (13.31-33; 14.1-14).

14B) In 14.34-15.32, Luke has now exhausted his stock of sayings from the Mission Discourse. So the saying on salt from the Sermon (complete with ironic "he who has ears to hear, let him hear!"), followed by the Lost Sheep from later in Matthew, is followed by a sort of narrative anaphora: next it's a Lost Coin; then a Lost Son!

14T) If the previous section is Luke at his best, this one is Luke at his worst. It's a terminal block, so the Matthaean material comes from sections 4-5. What it actually consists of is the ethically dubious parable of the shrewd steward by way of an over-long introduction, a morass of left-over sayings that don't really fit together in 16.14-18, then the memorable Rich Man and Lazarus Parable, which would fit much better without the sayings blob. The crassness of the divorce saying's situation is particularly marked; seeing as Luke completely omits Mk 10.2-12, it is tempting to speculate...

15) Luke 17.1-4 is where his two sources finally are at the same point synoptically (Mk 9.42-48/Mt 18.6-22). It's a definite mix of the two, but Luke has been utterly ruthless in his selection of materials. To finish it off, Luke sees Matthew's parable of the Unforgiving Servant (Mt 18.23-35) and instead provides on of Unprofitable Servants (Lk 17.5-10).

16) Luke then takes Matthew's order, but continues to omit material that will occur in the Marcan context. So once Jesus starts journeying towards Jerusalem (Mt 19.1-2/Lk 17.11), everything to midway through Mt 24 has to be skipped. The Healing of the Ten Lepers fills the vacuum. In 17.20-37, Luke then uses the non-Marcan pieces of Matthew's Eschatological Discourse. Whereas Matthew's Discourse then ended with thoroughly eschatological parables of the Virgins, the Talents, and the Sheep and the Goats, Luke replaces these with de-eschatologized parables of the Widow and the Judge and the Tax Collector and the Pharisee.

17) For the remainder of chapter 18, Luke rejoins Mark.

18) In the first half of chapter 19, the seam before the triumphal entry, Luke once more has special material (Zacchaeus). This provides a suitable setting for those Talents/Pounds that he had replaced at the end of the First Eschatological Discourse.

19) Luke once more follows Mark's narrative, with some omissions, from the Triumphal Entry to the end of the (Second) Eschatological Discourse.

20) The Passion Narrative depends on both Mark and Matthew, but most of all exhibits an independent streak, as Luke's theology of the death of Jesus is so different from the other two Synoptics.

The Wilke theory (and its modified "Q" form with a proto-Luke with all the Marcan blocks and the Passion Narrative removed) does not allow such a neat overview of how Matthew wrote his gospel with Mark and Luke/proto-Luke in front of him. The unscrambling of the blocks of Luke's Central Section that do not match Matthew in order is much messier than assembling them from three groups of sayings. Folding the doubletted Mission, Beelzebul, Parables, and Eschatological Discourses in between into single discourses is harder work than Luke using Mark and Matthew in turn. And it beggars belief that with his two main sources agreed in order from the Call of the Disciples to the Feeding of the 5000, he would still have gone his own way and created a curious order with a fronted Sermon, two Marcan strands, and a non-Marcan strand as he did.

So in terms of thinking about how the last evangelist would have written their gospel from two sources, the Farrer order is the easier hypothesis.

Answer 4

Lots of scholars look for alternatives to the traditional Mark-Q priority, but in my view without success. Dennis R. MacDonald wrote a well-researched thesis in Two Shipwrecked Gospels that Luke knew not only Mark, but also Matthew. That would have caused an even bigger stir among critical scholars than his book, The Homeric Epics and the Gospel of Mark, but after carefully studying his arguments I came away unconvinced. There seems to be no direct link from Matthew to Luke or from Luke to Matthew.

There should be no doubt that Luke is, in part, dependent on Mark as a direct source. The 'Missing Block' alone proves that. Another of the many proofs is put forward by John Dominic Crossan, who demonstrates movement of Markan intercalations from Mark to Luke. A different set of Markan intercalations flows to Matthew, and this alone makes it highly improbable that the author of Matthew did not rely directly on Mark for much of his narrative. The best explanation for non-Markan agreements between Matthew and Luke remains the hypothetical 'Q' document, although other alternatives as a secondary source remain worth considering.

Not only is Matthew directly dependent on Mark for much of its narrative content, but when Matthew and Luke are in agreement on sayings, they differ with context. For example, Matthew has the Sermon on the Mount, while Luke has the Sermon on the Plain. This is what you would expect if both authors were relying on a common Sayings Gospel (Q) and each had to create his own context for what Jesus said, but not if Matthew was developed from Luke because in this case Matthew could remain true to Luke as he does with Mark. There are many instances where Matthew and Luke have the same sayings and they are in the same sequence, but placed in different contexts. The agreement on sequence confirms that the two authors are working from a written, not oral source text, and this is consistent with Q, and the disagreements on context confirm that the source text is not a narrative. All this leads us away from Mark>Luke>Matthew.

Matthew's Gospel could not be more different than Luke's in respect to the resurrection appearances. It is also very different in respect to the empty tomb, but I accept these differences as theological. Getting back to the resurrection appearances: We know that Mark originally ended at 16:8 with the young man telling the women that Jesus was risen and they fled in terror, telling no one. The authors of Matthew and Luke both wanted to prove that Jesus had risen, which meant they needed resurrection appearances, but of course Mark was not giving them a lead. Matthew has Jesus meet the women and give them a message for the disciples to meet him in Galilee, which the disciples did do. Luke has Jesus meet the two on the road to Emmaus, an implied appearance to Peter and then to the eleven at a meal in the upper room. He then went with them out to Bethany where he rose up to heaven. I am sure that Luke's resurrection ending is original to the Gospel, so why would the author of Matthew write a contradictory ending?

There are several clues that Matthew is slightly older than Luke. One of them, noticed by critical scholars, is its place on the continuum of progressive antisemitism. It is more antisemitic than Mark (which is more so than Paul) but less antisemitic than Luke, with John quite scathing of Jews in general, not just of the Pharisees etc as in the earlier gospels. because of this and other clues, Matthew is generally dated at least ten years before Luke,making dependence on Luke impossible.