Double Tradition Doublets in the ‘Main Sequence’

At the beginning of Double Tradition Doublets in Mt 10: The Mission of the Twelve it is noted that Kloppenborg comments (from the point of view of the Mark-Q hypothesis) “That Q was written first is also suggested by the fact that Matthew and Luke concur in the relative sequence of more than 30 percent of the double tradition pericopae,” even though (as he also notes in Excavating Q) Matthew and Luke very rarely agree on the placement of the double tradition text relative to Mark. He then continues: 

Second, if one does not measure sequential agreement of these Matthew-Luke materials relative to Mark, but relative to each other, approximately one-third of the pericopae, accounting for almost one-half of the word count, are in the same relative order. That is, in spite of the fact that Matthew and Luke place double tradition materials differently relative to Mark, they nonetheless agree in using many of the sayings and stories in the same order relative to each other.

The double tradition consists of approximately 4,500 words, comprising 24.5% of Matthew and 23% of Luke, so Kloppenborg is above referring to sayings and stories in the half of this double tradition text that are placed in the same order (or sequence) in Matthew and Luke, although not so relative to Mark, that will here be referred to as the ‘Main Sequence’ of double tradition text. 

The ten double tradition doublets/formulas discussed below (none with a half in Matthew 10) all have one half in this main sequence of double tradition text, and include doublets identified by Hawkins as Nos. 1, 9, 10, 20 in Matthew and No. 3 in Luke, together with several formulas and similarities in Mathew and Luke almost all of which are also noted by Hawkins.

Mk -, Mt 3:7 // 12:34 // 23:33, Lk 3:7 – Generation of vipers (Hawkins: Formula 1 in Matthew) 

There is no mention of a ‘generation of vipers’ in the Old Testament (OT), and there is no evidence to indicate that aMark knew this phrase, so depending on the assumed synoptic hypothesis the origin of the double tradition verses Mt 3:7 / Lk 3:7 is either aMatthew or aLuke, or another (unknown) source. Mt 12:34a and 23:33 are each one of several Matthean additions to otherwise double tradition passages at Mt 7:15-20,12:33-37 / Lk 6:43-45 and Mt 23:4,29-36,13 / Lk 11:46-52 respectively. As in Mt 23:29b Jesus is speaking to ‘scribes and pharisees’ it is almost certain that Mt 12:34a and 23:33 are Matthean and are intended to show that Jesus is here speaking to people of similar character to those in Mt 3:7.

Mt 3:7 / Lk 3:7 are part of longer double tradition parallels at Mt 3:7-12 / Lk 3:7-9, 16-17, although with Mt 3:11 / Lk 3:16 also having a parallel at Mk 1:7-8. This appears to make Mk 1:7-8 / Mt 3:11 / Lk 3:16 triple tradition text, but the differences in order between Mk 1:7-8 and Mt 3:11 / Lk 3:16 suggest that one is not the source of the other, and that instead this is more likely to be a Mark-source overlap.

If aMatthew created the phrase in Mt 3:7 then aLuke chose to only use one of the three uses of the phrase that he saw in Matthew, but if instead the phrase came from a second source (i.e. other than Mark) then aLuke included a parallel in his gospel at Lk 3:7. However, he chose not to do so in the other places in which aMatthew used the phrase in what are extensions to otherwise double tradition passages that, presumably, were also in the same second source, whether Q, Early Luke, or any other. It is possibly that the ‘dislocation’ of verses in Matthew when compared with Luke (or vice versa) in the latter two passages may be associated with the lack of parallels in Luke to several verses in Matthew.

Mk -, Mt 3:10 // 7:19, Lk 3:9 - Trees bringing not forth good fruit (Hawkins: Doublet in Matthew No. 20)

Here there are two different versions of what appears to have originally been the same saying: One is at Mk 8:33 / Mt 16:23, and the other is the double tradition passage at Mt 4:10 / Lk 4:8, with Mt 16:23 // 4:10 being the doublet. On the assumption of any synoptic hypothesis in which aMatthew knew Mark then Mk 16:23 has Mk 8:33 as its source, while the source of Mt 4:10 / Lk 4:8 depends on whether either Matthew or Luke depend on the other, or not. If neither aMatthew nor aLuke knew the other’s gospel then there must have been a non-Markan source for these verses. 

If the source of Lk 4:8 is Mt 4:10, as on the MwQH, then there appears to be no need to propose that Mt 4:10 and 16:23 have different sources. Instead, it would appear perfectly reasonable to suggest that Mt 4:10 is just an edited version of Mk 8:33 / Mt 16:23, if it were not for the fact that this would mean that aMatthew wrote his edited version many chapters before adding a much closer copy of Mk 8:33 at Mt 16:23, or alternatively that he backtracked through his gospel in order to add Mt 4:10 later. Following his list of formulas in Matthew, on pages 137-138 Hawkins suggests a possible solution: 

A careful examination of such cases certainly leaves the impression that the mind of Matthew was so familiar with these collocations of words that he naturally reproduced them in other parts of his narrative, besides the places in which they occurred in his sources. It is to be observed that these apparent reproductions often occur earlier in the Gospel than do the apparently original occurrences of the formulas, which seems to indicate that Matthew drew them from his memory of the sources and not from documents before him.

Hawkins appears to be suggesting that here Mt 4:10 is based on aMatthew’s memory of Mk 8:33, while he later referred directly to the text of Mark when writing Mt 16:23, which would seem to be a very strange procedure to follow, especially as this phenomenon occurs in several places in Matthew. A particular issue here is the variants in Mt 4:10b and Lk 4:8b, both of which suggest that the inclusion of the phrase “Get thee behind me, Satan” is not secure. The NET Bible has the following comments: 

Re. Mt 4:10: The majority of later witnesses (C2 D L Z 33 œ) have “behind me” (ὀπίσω μου; opisw mou) after “Go away.” But since this is the wording in Matt 16:23, where the text is certain, scribes most likely added the words here to conform to the later passage. Further, the shorter reading has superior support (א B C*vid K P W Δ 0233 À1,13 565 579* 700 al). Thus, both externally and internally, the shorter reading is strongly preferred.

Re. Lk 4:8: Most mss, especially the later ones (A Θ Ψ 0102 Ë13 Ï it), have “Get behind me, Satan!” at the beginning of the quotation. This roughly parallels Matt 4:10 (though the Lukan mss add ὀπίσω μου to read ὕπαγε ὀπίσω μου, σατανᾶ [{upage opisw mou, satana]); for this reason the words are suspect as a later addition to make the two accounts agree more precisely. A similar situation occurred in v. 5.

Because the words in Lk 4:8b appear to be a later addition these and the corresponding words in Mt 4:10b may not have originally been in their current form. 

On the MwQH aMatthew either included an edited version of Mk 8:33 at Mt 4:10 and later added a more faithful version at Mt 16:23, or backtracked in order to add Mt 4:10 after having already written Mt 16:23. In either case aLuke ignored Mt 16:23 and may not have seen any reference to Satan in Mt 4:10. 

On the Mark-Q hypothesis both aMatthew and aLuke included a version of Q 4:8, and aMatthew then later included a version of Mk 8:33 in his first passion prediction, while aLuke did not. On this hypothesis it is likely that Q 4:8 (like Lk 4:8) did not include any mention of Satan, and that both Mt 4:10b and Lk 4:8b were later added to match Mt 16:23. 

On the MwEL hypothesis there are two possibilities:

• • Early Luke included a parallel to Mk 8:33 at ELk 4:8, and aLuke included that as Lk 4:8;

  • There was no parallel in Early Luke, and instead aLuke included a parallel to Mt 4:10 in a similar way to on the MwQH hypothesis.

In either case aMatthew placed his version of Mk 8:33 at Mt 16:23, which aLuke appears to have ignored.

Mk 5:35, Mt 8:7-8, Lk 8:49 // Lk 7:6 Troubling the master (Hawkins: Formula 4 in Luke) 

These verses connect two different stories about people who are close to death, the first being the daughter of Jairus (Mk 5:22-24, 35-43 / Lk 8:41-42,49-56), and the second the centurions servant (Mt 8:5b-13, Lk 7:2-10). In both cases Jesus was called for and someone then remarked that he should not be troubled, with the similarities suggesting that both passages had their origins in one original event. Assuming Markan priority Mark is the source of the story of the daughter of Jairus. Then:

• • On the MwQH aMatthew shortened and completely re-cast what he saw in Mk 5:22-24, 35-43. aLuke then chose to use both passages in his own gospel, also shortening (but not re-casting) what he saw in Mark, but in contrast considerably expanding on Mt 8:5b-13.

  • On the Mark-Q hypothesis Q is the source of the story of the centurion’s servant. aMatthew chose not to use large portions of what he saw in Mark (Mk 5:23a, 24b,35-37, 40b, 41b, 42b-43), and also shortened what he saw in Q 7:2-10.

  • On the MwEL hypothesis Early Luke contained a story of the centurion’s servant on which the versions in Matthew and Luke are both based.

Mk 1:2-3, Mt 3:3 // 11:10a, Lk 3:4 // 7:27a – This is he … Preparing the way (Hawkins: Quotations in Matthew, Luke)

Hawkins does not note these verses in his lists of either doublets or formulas, but instead notes Mt 3:3a in his section on quotations from the OT in Matthew under the heading “Quotations occurring in the course of the double or triple narrative, and found also in Mark or Luke or both of them.’ In general quotations from the OT do not impact the synoptic problem, and this would appear to be the case in this example as in all three synoptic gospels John the Baptist quotes from both Malachi and Isaiah with regard to Jesus. Patton raises the same point, commenting that: 

Streeter’s suggestion that Mk i, 2-3, is from Q seems unjustifiable. Vs. 3 is an Old Testament quotation which Matthew, Mark, and Luke all have in common. If it stood originally in Mark and is not to be regarded as a later addition, there is no occasion for the assumption of Q. Vs. 2 could hardly have stood in its present place when Matthew and Luke used Mark. It occurs in another connection in Matthew and Luke (Mt xi, 10; Lk vii, 27), and was probably copied from there into its present place by a later hand.

However, while in Mark John utters both quotes just before the baptism, in both Matthew and Luke he only quotes Malachi later, when in prison (Mt 11:10b / Lk 7:27c). It appears that either Mk 1:2-3 has been split into Mt 3:3 / Lk 3:4 and Mt 11:10 / Lk 7:27 respectively, or that Mk 1:2-3 is a later combination of the two different quotes. If Mk 1:2a originally contained “in Esaias the prophet” this would suggest that Mk 1:2b, containing a quote from Mal 3:1, did not originally exist, while if “in the prophets” was original it would suggest that it did. Wieland Willker discusses the issues in Mark (TVU 2) (re-formatted for readability, and for more see Mark 1:1-3 - The Short Beginning): 

If Isaiah was originally in Mk, it might have been changed to "in the prophets" because the first part is from Malachi and only the second part is from Isaiah. This is the traditional explanation of the NA reading. Perhaps it was a reaction to Porphyry's assault? Note that Malachi is never mentioned in the NT by name.

On the other hand if "in the prophets" was originally in Mk, it might have been changed to Isaiah to be more specific. This happened also at other places e.g. at Mt 13:35 where "through the prophet" has been changed to "through the prophet Isaiah" even though the word is NOT from Isaiah… Isaiah was certainly the best known and most important prophet for the NT writers. The change to "Isaiah" could also simply be a harmonization to Mt/Lk.

A third alternative, not backed up by manuscript evidence though, is that the beginning of Mk is (as the ending) somewhat corrupt. This has been argued by several authors … Some argue that verses 2-3 are an early gloss. Compare E. Güting (TC Mark, 2005, p. 53-55): "The beginning of the Gospel of Jesus Christ was John the Baptizer, who, in the desert, was proclaiming a baptism of repentance ...".

If verses 2-3 are a harmonization to Mt/Lk, it is a rather sophisticated one, because verses 2 and 3 come from different places.

Lachmann suggested that the two verses are the result of a conflation… Güting suggests that the verses have been added from Q. At least verse 3 is in Q... That both Mt and Mk don't have the Malachi part of the LXX quotation at this place is a Minor Agreement, too. Both cite it later at Mt 11:10/Lk 7:27, but not here. Did they both omit it, because it is not from Isaiah? Did they read it at all in their copy of Mark? Did they have a different source? Q?

From the above it should be clear that it is not in any way obvious how what is generally believed to have originally been in Mark (Mk 1:2-3) should have become split across two widely separated events in both Matthew and Luke. Assuming that what we see as Mk 1:2-3 preceded the parallels in Matthew and Luke, the different versions of these quotes were created in different possible ways depending on the synoptic hypothesis: 

• • Assuming no knowledge of each other, aMatthew and aLuke independently split what they saw in Mark in the same manner, with both including the Isa 40:3 quote in their respective baptism narratives and Mal 3:1 in their tributes to John the Baptist, with the tributes themselves being double tradition material as they have no parallel in Mark. For this to be in any way feasible there must have been another source for the common non-Markan material on John the Baptist, and on the Mark-Q hypothesis Q is that source, while on the MwEL hypothesis it is Early Luke;

  • Alternatively, aMatthew split what he saw in Mark, placing the Mal 3:1 quote in his tribute to John the Baptist, and aLuke followed Matthew (as on the MwQH), leaving unanswered the question of where aMatthew obtained the rest of his material on John the Baptist in Mt 11:2-9, 11.

As mentioned by Willker there is also the possibility of a harmonization, and the suggestion of Q (or indeed a different second source) is also raised above. However, it is known that Marcion’s gospel did not include the baptism narrative, but Tertullian refers to most of what we know as Lk 7:18-28, and both Tertullian and Epiphanius do quote a version of Mal 3:1 at the point in their respective discussions of Marcion’s gospel at which John the Baptist is being discussed, so indicating that Marcion’s gospel did contain at least some of the tribute to John the Baptist. Nevertheless Tertullian does not include “before me” in his quote from Mal 3:1 / Mk 1:2b / Lk 7:27b (while Epiphanius actually quotes even less), as is also not present in Mk 1:2 and Lk 7:27 in some mss. If Marcion’s gospel was an early version of Luke then this would suggest that it could be the source of Mt 11:10 / Lk 7:27.

Mk 2:15b-16, Mt 9:10b-11 // 11:19a, Lk 5:29b-30 // 7:34 // 15:1-2 - Eating and drinking with publicans and sinners (Hawkins: Similarity in Luke)

After having identified and commented on the doublets he sees in Luke, Hawkins suggests two more similarities in Lukan verses, neither of which he thinks constitute doublets, but in regard to the second one he writes: “Also the narratives in Luke v. 29, 30 and xv. 1,2 are remarkably similar,” with both passages referring to Pharisees and scribes and containing the phrase “publicans and sinners." In the Bible this phrase exists in nine places, all of which are in the synoptic gospels: Six are in the triple tradition passage at Mk 2:15-17 / Mt 9:10-13 / Lk 5:29-32, two are in the double tradition passage at Mt 11:16-19 / Lk 7:31b-35, and one is in Lk 15:1-3, which provides a unique beginning to the otherwise double tradition parable of the hundred sheep at Mt 18:12-13 / Lk 15:4-7. 

The use of this phrase nine times in the synoptic gospels suggests that it is a common formula or saying, especially as it is used three times in just two verses in Mark. Willker (Mark: TVU 33, 34, 35, 36) records multiple different variant readings in four places in just Mk 2:15-16, ending by noting that for Mk 2:16 “External analysis is hampered by the fact that heavy harmonizations took place in this verse,” which make it hard to determine the original reading, and in particular whether Mk 2:16 originally mentioned drinking. As Mark mentions both eating and drinking only in this one verse, while Matthew has both in five places and Luke in ten, it seems likely that Mk 2:16 originally did not mention drinking. 

References to eating (with or without drinking) combined with publicans and sinners are found in six verses, as shown above. As nowhere does the combination appear more than three times in any gospel it appears reasonable to consider this be a doublet in Matthew and a triplet in Luke. However, Lk 15:1-2 differs in format, does not referring to drinking, and is the only reference that has no parallel in either Mark or Matthew (as is also the same for most of Lk 15). On this basis it is possible to consider Lk 15:1-2 to be a parallel, but perhaps not part of a triplet in Luke. 

• • On the MwQH these verses can be explained as aMatthew creating a parallel of Mk 2:15-16 at Mt 9:10-11, and aLuke also doing so at Lk 5:29-30. However, what is then harder to explain is why both aMatthew and aLuke chose not to include parallels of both Mk 2:15c and 16b, or alternatively why they may not have seen these pieces of text. The opposite issue applies to Mt 11:19a and Lk 7:34: What was the source of Mt 11:16-19, and why did aLuke then write a very close parallel to these non-Markan verses at Lk 7:31b-35?

• On the Mark-Q hypothesis there is no explanation for how or why aMatthew and aLuke independently excluded the same portions of Mk 2:15-16 from their parallels, while here Mt 11:19a / Lk 7:34 come from Q. On this hypothesis Lk 15:1-2 most likely has Mk 2:15-16 is its source, and is used to provide a preface to the Lukan version of Q 15:4-7.

• On the MwEL hypothesis the source of Mt 11:19a / Lk 7:34 is Early Luke, which also contained a parallel to Mk 2:15-16 that was then ignored by aMatthew but used by aLuke as Lk 15:1-1.

Mk 2:8, Mt 9:4 // 12:25a, Lk 5:22 //11:17a – Knowing Their Thoughts (Hawkins: Formula Peculiar to Matthew, 4)

Hawkins does not mention any of the four Lukan parallels all of which also refer to Jesus knowing or perceiving people’s thoughts, shown ordered here on the basis of their parallels in Mark and/or Matthew: Hawkins records this as a formula “Peculiar to Matthew,’ at Mt 9:4 // 12:25, and he identifies the parallels in these verses as: 

Mt 9:4a knowing their thoughts εἰδὼς … τὰς ἐνθυμήσεις αὐτῶν

Mt 12:25a But knowing their thoughts εἰδὼς δὲ τὰς ἐνθυμήσεις αὐτῶν

These are almost identical, but Hawkins does not mention the variant in Mt 9:4a in which the majority of mss read ἰδὼν (seeing / perceiving) instead of εἰδὼς (knowing). In his discussion of Mt 9:4 Willker (Matthew, TVU 98) notes that ἰδὼν is used 12 times in Matthew but that εἰδὼς is used only once, in Mt 12:25, and then writes: 

Interestingly the same variation occurs in Mt 12:25 with similar support. It is possible that the 12:25 variation is a conformation to 9:4, but it is not clear which reading is original.

It could be said that thoughts cannot be seen, only known, except in a figurative sense.

εἰδὼς could be a (partial) harmonization to ἐπιγνοὺς in the parallels [i.e. in Mk 2:8a / Lk 5:22a], but that's rather improbable.

It is possible that the variation is at least in part accidental, because εἰ and ἰ are pronounced alike (compare the ἰδὼς variant).

Hawkins does not mention any of the four Lukan parallels all of which also refer to Jesus knowing or perceiving people’s thoughts, shown ordered here on the basis of their parallels in Mark and/or Matthew:

Lk 5:22a    But Jesus, perceiving their thoughts ἐπιγνοὺς δὲ ὁ Ἰησοῦς τοὺς    διαλογισμοὺς αὐτῶν

Lk 11:17a    But he, knowing their thoughts              αὐτὸς      δὲ εἰδὼς αὐτῶν τὰ διανοήματα

Lk 9:47b            he knew the thoughts                                                 εἰδὼς      τὸν      διαλογισμὸν

Lk 6:8a      But he knew their thoughts                     αὐτὸς       δὲ ᾔδει        τοὺς    διαλογισμοὺς αὐτῶν

If these parallels all contained the same word for knowing/perceiving thoughts they could be considered to be a quadruplet, but instead they contain three different words for perceiving: Lk 5:22a has ἐπιγνοὺς (c.f. Mk 2:8a); Lk 11:17a and 9:47b have εἰδὼς (c.f. Mt 12:25a); and Lk 6:8a has ᾔδει. On this basis Hawkins is correct in not identifying any of these four verses as containing a doublet or formula, as the closest to a formula is in Lk 11:17a / 9:47b: εἰδὼς … διανοήματα / εἰδὼς … διαλογισμὸν. 

Lk 9:47b and Lk 6:8a have no parallels in Mark and Matthew, and none of the Lukan verses have ἰδὼν (seeing) as in Mt 9:4a. In addition, all four verses in Luke use a different word for ‘thoughts.’ Although the sense may be the same as in Matthew, the words used in the parallels in Luke are different (Mt 9:4a / Lk 5:22a, Mt 12:25a / Lk 11:17a). 

As both Mk 2:8a and Lk 5:22a have ἐπιγνοὺς it would be natural to see this also in Mt 9:4a. Consequently, the use of either ἰδὼν or εἰδὼς instead in Mt 9:4a would have been a deliberate choice by aMatthew, unless perhaps he saw either word in another source. As εἰδὼς is safe in Mt 12:25a (its only secure use in Matthew) while ἰδὼν occurs in 11 other places in Matthew, the use of εἰδὼς in Mt 9:4a would be more likely to be a conformation to Mt 12:25a. This then leaves the question of why Mt 12:25a, uniquely, has εἰδὼς instead of ἰδὼν. 

• On the MwQH aMatthew saw ἐπιγνοὺς in Mk 2:8a but decided to use ἰδὼν in his parallel at Mt 9:4a instead, and then used εἰδὼς in the other half of his doublet at Mt 12:25a. aLuke then followed Mk 2:8a by using ἐπιγνοὺς in Lk 5:22a, used εἰδὼς from Mt 12:25a in Lk 9:47b and 11:17a, but in Lk 6:8a he used ᾔδει. As εἰδὼς appears only two times in Mark and once in Matthew there is no obvious reason why aLuke would use it in Lk 9:47b and 11:17a in preference to what he had already used in Lk 5:22a.

• On the Mark-Q hypothesis aMatthew also saw ἐπιγνοὺς in Mk 2:8a but used ἰδὼν from Mt 9:4a instead, while aLuke did not know Mt 9:4a and so simply followed Mk 2:8a. Mt 12:25a and Lk 11:17a are part of the double tradition and so both have εἰδὼς from Q 11:17a. aLuke also used εἰδὼς in Lk 9:47b, but in Lk 6:8a used ᾔδει for an unknown reason.

• Because on both the MwQH and the Mark-Q hypothesis there is only one formative stage for Luke it is hard to explain why Mt 12:25a, Lk 9:47b and 11:17a all use the same word for Jesus knowing/perceiving thoughts, while Lk 6:8a uses a word that is not used anywhere else in the synoptics to mean someone knowing the thoughts of others. However, on the MwEL hypothesis not only is there an additional source in Early Luke, it is possible that Early Luke is the source of readings that we see as Sondergut Lk because aMatthew chose not to use them, for example ᾔδει in Lk 6:8a, and others as double tradition text because aMatthew did choose to use them, e.g. εἰδὼς in 12:25a.

Mk 9:39-40, Mt 12:30, Lk 9:50 // 11:23 – He that is not against me (Hawkins: Similarities in Luke) 

In a similar fashion to what we see in Mk 8:33, Mt 4:10 // 16:22-23, Lk 4:8 (Hawkins’ Formula 9 in Matthew), here there appear to be two different versions of the same saying: One is at Mk 9:40 / Lk 9:50c, and the other is an expanded version in double tradition text at Mt 12:30 / Lk 11:23, but in this case the doublet is in Luke, at Lk 9:50bc // 11:23. 

• • On the MwQH aLuke includes the parallel to Mk 9:40 even though he (presumably) saw that aMatthew did not include it, but then later adds a parallel to Mt 12:30 even though it has no parallel in Mark

  • On the Mark-Q hypothesis this is a Mark-Q overlap, but here with aLuke rather than aMatthew choosing to include a parallel to Mk 9:40 in addition to the parallel to Q 11:23.

  • On the MwEL hypothesis the flow of the text depends on whether there was a parallel to Mk 9:40 in Early Luke:

    • If there was a parallel at ELk 9:50c then aLuke used it at Lk 9:50c, otherwise he created a parallel to Mk 9:40;

    • If there was a parallel at ELk 11:23 then this was the source of Mt 12:30 / Lk 11:23.

Mk 8:12-13, Mt 16:4 // 12:39, Lk 11:29 - The sign (of Jonah) (Hawkins: Doublet in Matthew No. 9) 

See also Mk 8:11-12, Mt 12:38-39 // 16:1-2, Lk 11:16 - Seeking a sign from heaven // Jonah (Hawkins: Doublet in Matthew No. 18). Hawkins notes: 

For the introductory narratives, see No. 18 ... below. And observe that Mt A [12:39] and Lk, Mt B [16:4] and Mk, are respectively followed by similar contexts; for the mention of Jonah is not enlarged upon after Mt B as it is after Mt A and Lk.

The parallels in Mk 8:11-13 / Mt 16:1-4 both directly follow the feeding of the four thousand in Mk 8:1:1-10 and Mt 15:32-39 respectively (a passage that is part of the Great Omission in Luke), and so are clearly in the same context, making it all but certain that, assuming Markan priority, Mt 16:4 depends on Mk 8:12-13. Then, as Mt 12:40-42 and Lk 11:30-32 (the explanation of the sign of Jonas) are obvious parallel passages, it is clear that the preceding verses (Mt 12:39 and Lk 11:29) are also part of these parallels. In addition, as Mt 12:38, 40-42 and Lk 11:29a, 30-32 have no parallel in Mark it is most likely that Mt 12:39 and Lk 11:29b also have no parallel in Mark, i.e. that Mk 8:12b is not part of a Markan version of this passage, and is therefore not in the same context as Mt 12:39b / Lk 11:29c. 

Under the heading ‘The Synoptic Tradition on the Sign of Jonah,’ and after stating that “Mark 8:11-13 is a conflation of Matthew 16:1-5 with Luke 11:16 and Luke 11:29,” [i.e. with Mark knowing both Matthew and Luke] William Reuben Farmer writes: 

Only the Griesbach hypothesis enables one to unravel satisfactorily the redactional history of Mark 8:11-13. Luke tended to avoid doublets (compare his avoidance of a second story of a miraculous feeding of the multitudes); therefore, while he has a parallel account of Matthew 12:38-42, there is only a trace of the influence of the doublet in Matthew 16:1-4 upon his text (Luke 11:16). Mark included Matthew 16:1-4 in a section of his Gospel where he was incorporating something of all the literary units in the parallel section of Matthew. But in so doing he compared the Lucan parallels in Luke 11:16 and 29, and conflated the first part of Matthew’s text with Luke 11:16 and the second part with Luke 11:29, ending with Matthew’s transitional sentence leading to the saying about the leaven of the Pharisees (Mk. 8:11 // Mt. 16:1 // Lk. 11:16; Mk. 8:12 // Mt. 16:1b // Lk. 11:29; Mk. 8:13 // Mt. 16:4a, 5a).

It is very difficult to place Mark at the beginning of this redactional process, making his account the source of all the Synoptic parallels. Streeter (Oxford Studies, p. 177) appealed to the notion of Mark’s knowledge of Q to explain Mark 8:12. He rightly regarded Luke 11:29 as the proper parallel to Matthew 12:39, but he was mistaken in making Mark 8:12 parallel to Matthew 12:39. In that particular verse the agreements are deceptively equal between Mark and Matthew 12:39 on the one hand and Mark and Matthew 16:4 on the other. But outside that verse, it is quite clear that the literary kinship of the Marcan passage is closer to Matthew 16:1-5.

There are a number of problems here. The first is Farmer’s belief that aLuke avoided the feeding of the four thousand, because avoidance assumes that aLuke actually saw it, while even on the Griesbach hypothesis aLuke could have seen a damaged copy of Matthew, and of course he did not know Mark. Then, even if aLuke did see and “avoid” a second feeding story, the suggestion that aLuke avoided doublets in general implies that he saw both halves (in Matthew, not Mark) and chose to use just one half in several places where Matthew has a doublet. Farmer uses the feeding of the four thousand to support Griesbach, but ignores the fact that if aLuke saw Mark as well as Matthew then his ‘avoidance’ of doublets could be explained as simply following Mark instead of Matthew in those places.

Farmer then provides his reasoning as to how (on the assumption of the Griesbach hypothesis), Mark used Matthew and Luke 11:16, 29, followed by stating that “it is very difficult to place Mark” first here, without suggesting why. He notes that Streeter suggests that Mk 8:12 shows Mark had knowledge of Q, but in a review of Mark and Q. A study of the Overlap Texts by Harry T. Fleddermann, Christopher Tuckett notes: 

Nor does Fleddermann note the widely accepted seam in terms of tradition history between Q 11,29-30 and Q 11, 31-32. (A reference to “this generation” is almost certainly present already in Q 11,30: one does not need Q 11,31-32 to explain Mark’s wording…)

Farmer then agrees with Streeter that Mt 12:39 and Lk 11:29 are parallels, but not that Mk 8:12 is parallel to Mt 12:39. This writer agrees with Farmer in both instances, but sees no reason why this might support the Griesbach hypothesis but not others. 

On the MwQH it is easy to see how Lk 11:29 could contain a parallel to Mt 12:39, and how aMatthew could have created different versions, with the later Mt 16:4 not followed by a version of the explanation from Mt 12:40-42. However it is not so easy to explain why in Mt 12:39 // 16:4 the question in Mk 8:12b was changed into a statement. It should be noted that because Mk 8:12 is parallel to the Great Omission in Luke this could indicate that aLuke did not see this verse, and if so that would explain why Lk 11:29 follows Mt 12:39 rather than Mk 8:12. It is odd that in Mt 16:4b Jesus names the sign for a second time (an addition from Mt 12:39b) rather than simply reminding the Pharisees that he had given them the sign of Jonas earlier, while if this was instead a different group of people why would he not explain the sign to them in Mt 16:4 as he had previously done in Mt 12:39? 

Patton has two sets of comments on these verses and clearly sees a form of Q here. With regard to Mk 8:11-13 he states: “Doublets in both Matthew and Luke indicate the presence of this section in both Mark and Q,” and then when discussing Mk 8:12 specifically he writes: 

On the ground of Matthew’s having doublets for this saying (Mt xii, 39; xvi, 4) and Luke a parallel to it (Lk xi, 29), it may without further consideration be assigned to Q. The agreement of Matthew and Luke, and the agreement of Matthew’s doublets, in adding “Except the sign of Jonah,” may be taken to indicate the difference here between Mark’s Q and the later recensions.

On the Mark-Q and MwEL hypotheses this situation simply requires a parallel in Q or Early Luke (respectively) similar to what we see in Mt 16:4 // 12:39, but re-located to a different point in the narrative in Q or Early Luke respectively. On the MwEL hypothesis aMatthew would then have seen both Mk 8:12-13 and a parallel in Early Luke (neither containing the reference to Jonas), and chosen to include both in Matthew. In addition he chose to add the OT reference to Jonah in both places, to the first of which he also added an explanation. On the MwEL hypothesis aLuke would have then chosen the longer version he saw in Mt 12:39-42, but some scribes/copyists of Luke also knew the short version from Early Luke (based on Mk 8:12), so accounting for the variants in Luke 11:31-32 found in some mss. 

Given the above it is highly likely that Mt 16:1-4 depends on, and is an expansion of, Mk 8:11-12. However, it is hard to see why aMatthew would refuse to give this group of people the explanation given earlier, in Mt 12:40-42. Even if Jesus is addressing the same group of Pharisees, Sadducees and scribes why would he not just say “I’ve already given you the explanation’? Instead, it seems likely that there were two different records of the request for a sign, a shorter one known to aMark, and a longer one in another source. 

Because Lk 11:31-32 are swapped when compared with Mt 12:41-42 the original order of these verses is not clear, and it should be noted that Lk 11:32 is not present in either D or d (respectively the Greek and Latin of Codex Bezae). Regarding Lk 11:32 Willker (Luke, TVU 204) writes: 

Mt has the same verse in identical wording. But the interesting fact is that Lk has the two verses reversed. Mt has the more logical order because in the preceding verses Jonah is the topic. It would be natural to end with "something greater than Jonah is here!" and then go on with the queen of the south. It is possible that the omission by D is original and that some early scribe added the verse as a harmonization to Mt, but added it at the wrong place. But this is rather improbable.

On the other hand it is possible that in an ancestor of D the verse has been labeled for omission and transfer before verse 31. This lead accidentally to complete omission.

On the MwQH (with Mt 12:41-42 being the source of Lk 11:31-32) there is no obvious reason either to swap these two verses or to omit Lk 11:32. However, if these two verses were originally two separate sayings in another source (such as Q) then it is possible that aMatthew and aLuke would make different choices regarding their addition and their placement. Both Tertullian and Epiphanius record that in Marcion’s gospel Jesus did not give a sign, i.e. that it did not contain any parallel to Lk 11:29d-32, as is also the case in Mark, and Epiphanius states that Marcion’s gospel also did not contain any parallel to Lk 11:31-32. While Marcion could have cut Lk 11:29d-32 out of his gospel there is no obvious reason why he would do so, and instead as neither Tertullian nor Epiphanius indicate that the generation was ‘evil,’ it is likely that here Marcion’s gospel followed what we see in Mark. 

Note that here the text of Marcion’s gospel could have been a source for Mt 12:39 and Lk 11:29, and because both Tertullian and Epiphanius are clear that it did not include any reference to the sign of Jonah, it is very close to Mk 8:12, and so could have had that verse as a source.

Mk 13:21, Mt 24:23 // 24:26, Lk 17:23 – Rumors of the Coming of Christ/the Son of Man … (Doublet)

Hawkins does not identify Mt 24:23 / 24:26 as a doublet, possibly because although the subject matter and “believe it not” is common to both verses, Christ is not directly referred to in Mt 24:26. 

• • On the MwQH aMatthew wrote two different forms of the same saying at Mt 24:23, 26, and aLuke used a version of the latter at Lk 17:23;

  • On the Mark-Q hypothesis Mk 13:21 is the source of Mt 24:23, and Q is the source of Mt 24:26 / Lk 17:23;

  • On the MwEL hypothesis Mk 13:21 is the source of Mt 24:23, and Early Luke is the source of Mt 24:26 / Lk 17:23.

Mk 4:25, Mt 13:12 // 25:29, Lk 8:18 // 19:26 - Whosover hath to him shall be given … (Hawkins: Doublet in Matthew No. 10, Luke No. 3) 

Here there are two doublets: Mt 13:12 // 25:29 and Lk 8:18 // 19:26, regarding which Hawkins writes: 

The three passages have been printed as parallel, though Matthew A [13:12] occurs before, and Mk and Lk A [8:18] after, the explanation of the parable of the Sower.

Hawkins notes that the Markan contexts for Mt 13:12 and Lk 8:18 are slightly different, but he does not note that Mt 25:29 and Lk 19:26 have no corresponding Markan context, i.e. neither derive from a portion of Mark that contains a parallel of these verses. Instead, both are part of the double tradition and so both are edited versions of the earlier halves of the doublets, one is an edited version of the other, or they both have an unknown source. 

• • On the MwQH aMatthew creates a version of Mk 4:25 into which he inserts a small portion of Mk 4:24, and then adds a close copy of the same text at Mt 25:29. aLuke then creates copies of both verses, but even though he saw Mt 13:12 and 25:29 he for some reason removes both references to abundance, and instead adds a different portion of Mk 4:24 to Lk 8:18.

  • On the Mark-Q hypothesis Mt 13:12 and Lk 8:18 are created from Mk 4:24-25 as on the MwQH (except that aLuke did not see Matthew), while Q is the source of both Mt 25:29 and Lk 19:26, with an addition from Mk 4:24 / Mt 13:12b in the case of Mt 25:29.

  • On the MwEL hypothesis ELk 8:18 is a close copy of Mk 4:24-45, excluding those portions of Mk 4:24, which, on the basis of the variants indicated above, may not have originally been in Mark, or perhaps in aELk’s copy of Mark. Lk 8:18 then follows ELk 8:18, while aMatthew seems perhaps to have seen a more complete version of Mk 4:24, but ignored Mk 4:24ab because he is continuing an existing saying. Because aMatthew sees both Mk 4:24-25 and ELk 8:18 he creates a second version of this verse at Mt 25:29, which aLuke then also uses although he excludes the reference to ‘abundance’ because it is not in either Mark or Early Luke.

Summary – Double Tradition Doublets in the Main Sequence 

As previously noted Kloppenborg records that approximately half of the double tradition text is the same order (or sequence) in Matthew and Luke, referred to here as the ‘Main Sequence’ of double tradition text.

The table below records all the double tradition doublets (including formulas) just discussed in detail above. In each case in the ‘Matthew’ and 'Luke' columns any doublet half with a parallel in Mark in the same context is listed first, with the half in double tradition text second. For identification purposes D denotes a doublet, F a formula, S a ‘similarity,’ Q a quotation, and P a formula ‘Peculiar to Matthew’ according to Hawkins. If a doublet is given a number by Hawkins then it is identified as such. If a formula has a number then that is its position in Hawkins’ corresponding list. There are no doublets/formulas here in Mark, but with two exceptions the halves of the doublets/formulas that are not in double tradition text have a parallel in Mark in the same context. This means that on hypotheses in which Matthew and Luke had a non-Markan source there is a Mark-source overlap (a Mark-Q overlap on the Mark-Q hypothesis) in all but two of these doublets/formulas. 

As in Summary - Double Tradition Doublets in the Mission or the Twelve verses in the double tradition are shown with a colored background. Where they have a parallel in the same gospel (i.e. there is a doublet) and that parallel has a direct parallel in Mark then this is a strong indication that these verses do not have a direct parallel in Mark, e.g. on the Mark-Q hypothesis the source of these verses is Q. However, other sources are possible, for example on the MwQH any double tradition verse in Luke can have Matthew as a source. Verse numbers in red mark doublets in which the doublet half with a direct parallel in Mark is located after the half in the double tradition.

The first two doublets/formulas shown above (the parallels at Mt 3:7-10 / Lk 3:7-9) are unusual. These verses have no parallels in Mark, but nevertheless both passages are inserted into the text of their respective gospels at exactly the same point in the Markan narrative. This indicates either that aLuke or aMatthew knew where the other had inserted the parallel text into his gospel relative to the Markan text, or that if Matthew and Luke were independent both had a ‘locator key’ that allowed them to insert their parallels into the Markan text at exactly the same place. This ‘key’ is Mk 1:7-8, which is parallel to Mt 3:11 / Lk 3:16 (and so on the Mark-Q hypothesis is a Mark-Q overlap) and so allowed aMatthew and aLuke to insert Mt 3:7-12 and Lk 3:7-9, 16-17 respectively into their narratives aligned with the same verses in Mark.

There are no doublets/formulas here in Mark, but with the above two exceptions the halves of the doublets/formulas that are not in double tradition text have a parallel in Mark in the same context. This means that on hypotheses in which Matthew and Luke had a non-Markan source there is a Mark-source overlap (a Mark-Q overlap on the Mark-Q hypothesis) in these doublets/formulas, five of which H.T. Fleddermann records as ‘Overlap Texts’ in ‘Mark and Q:’ 

Mark 1,2 (Matt 11,10 par. Luke 7,27)                The Messenger

Mark 4,25 (Matt 25,29 par. Luke 19,26)             To One Who Has Will be Given

Mark 8,11-13 (Matt 12,38-42 par. Luke 11,16.29-32) Demand for a Sign

Mark 9,40 (Matt 12,30 par. Luke 11,23)              On Tolerance

Mark 13,21 (Matt 24,26 par. Luke 17,23)          Rumors of the Coming

In ‘Demand for a sign’ Fleddermann references both Luke 11,16 and 11,29-32. Because Lk 11:16 is only a parallel to some of these verses this author notes this as two sets of doublets, recorded as: 

As indicated in Summary – Double Tradition Doublets in the Mission of the Twelve the recognition that these four sets of double tradition parallels are recorded elsewhere as Mark-Q overlaps means that the remaining sets of doublets shown above could also be viewed as Triple Tradition / Double Tradition overlaps, and so possibly also (on the Mark-Q hypothesis) as Mark-Q overlaps. 

Again, these sets of doublets appear to be structurally no different to those noted by Fleddermann as Mark-Q overlaps, and so on that basis it appears that Fleddermann should have also included the following overlaps in his analysis: 

Mark 2,8 (Matt 12,25 par. Luke 11,17) Knowing their Thoughts

Mark 2,15b-16  (Matt 11,19a par. Luke 7,34) Eating/Drinking – Publicans and Sinners

Mark 5,35           (Matt 8,7 par. Luke 7,6) Troubling the Master

Mark 8,33           (Matt 4,10 par. Luke 4,8) Get Thee Behind Me

As noted in the previous section the fact that these sets of doublets fit the same ‘pattern’ as those already identified as Mark-Q overlaps does not mean that any or all of them must be Mark-Q (or other source) Overlaps. However, again it provides support for the view that Matthean and Lukan parallels to Mark in ‘Markan blocks’ or a ‘Markan context’ have Mark as their source, while those in the Double Tradition have a different source. Whether that source was Q, an early version of Luke, something derived from an Aramaic source, or something else, is a different issue.

NEXT: Other Double Tradition Doublets