The Synoptic Problem

Return to Index

This comprehensive study was completed for my students at UTC, Bangalore in August, 1976.  

TABLE OF CONTENTS:
          I.  The problem
                A.    Why is there a problem?
                B.    What is the problem?
        II.  The data and their significance
                A.  The individual data and the directions they point toward a 'solution'.
                B.  Summary (and evaluation) of the probable directions pointed by the various data.
      III.  The likely solution to the problem
                A.  Summary of the results  obtained in Section II.
                B.  A probable solution to the synoptic problem.
        Postscript:  Why bother to solve it?  What difference does it make?

Symbols used:
            Mk            The Gospel according to Mark (& all the material in it0
            Ur-Mk       'Original Mark', a (possible) earlier gospel, the basis of Mk
            Mk I         An edition of Mk without Mk 6.45-8.26
            Mk II        Our canonical Mk, Mk 1.1-16.8
            Mt            The Gospel according to Matthew
            Lk             The Gospel according to Luke
            Q               Basically, the material common to Mt & Lk, but not found in Mk
            M              The material peculiar to Matthew (over 1/5 of Mt)
            L                The material peculiar to Lk (over 1/3 of Lk)
            par.            Parallel sections in the other gospels (also indicated by //)

Evaluation of data:
            s.e.            Data constitute strong evidence, usually decisive.
            e.               Data constitute evidence, but not as decisive.
            p.               Data can be explained by this possibility, but the data are not evidence for it as such. 

I.    THE PROBLEM
        A.    Why is there a problem?
                1.  In all 3 gospels (Mt, Mk & Lk) the content and arrangement of the material they have in common (the 'triple-attested tradition') are closely related.
a)  Linked together are the Baptist's coming, Jesus' baptism & temptation, and Jesus' appearance in public.
b)  Jesus is active almost solely in Galilee until his passion, and his public activity is in the same sequence in all 3 gospels (unlike the 4th Gospel).
c)  His going to Jerusalem, his activity there, and his trial are similar in all three.
d)  All three close with his crucifixion and resurrection.
                2.  Details of style and language are the same in all three or in two of them against the third.
a)  Large sections are word-for-word the same.
    i.  In all 3: e.g. Mk 1.40-45 par. (healing of leper)
    ii.  In material found only in two gospels:
        e.g. Mt 4.18-22a // Mk 1.16 (call of disciples)
        e.g. Mk 1.21-25 // Lk 4.31-35 (man with unclean spirit in synagogue on Sabbath)
b)  In material common to all, two gospels are often in broad agreement against the third.
        e.g., Mt 20.24-28 & Mk 10.41-45 (about James & John; Son of Man giving life as a ransom) vs Lk 22,24-27.
                3.  Much material dealing with sayings of Jesus is common to Mt & Lk but not found in Mk (this is the so-called Q material - about 200+ verses0.  T. W. Manson, The Mission and Message of Jesus, grouped the material as concerning:
                            a) Jesus and John the Baptist,
                            b) Jesus and his disciples,
                            c) Jesus and his opponents,
                            d) Sayings about the last things.
                4.  Mt & Lk also have substantial materials peculiar to each of them.
M:  300 verses, including birth & infancy stories (Mt 1-2), narratives (e.g. Peter's walking on water, Judas' suicide, Pilate's hand-washing), and much teaching (e.g. large parts of Sermon on Mount).
L:  400+ verses, including birth & infancy stories (Lk 1-2), narratives (e.g. sermon at Nazareth, woman who was a sinner, walk to Emmaus), and a large part of Lk 9.51-18.14 (travel toward Jerusalem, including much teaching, e.g. parables of Good Samaritan, Prodigal Son, and the Pharisee and the Publican).
            B.   What is the problem?
From what we have seen, each of the three gospels, though written independently, much have drawn much of its materials from a source, or sources, available to one or both of the others.  To discover these sources is the task set by the synoptic problem, and it is a task set by the text of  the gospels.

II.     THE DATA AND THEIR SIGNIFICANCE
     A.    The individual data and the directions they point toward a 'solution'
        1. Of the 661 verses which belong to Mk 1.1-16.8, the material of all but about 30 verses is to be found in Mt &/or Lk (ca. 606 in Mt; 300+ in Lk).  (Of the material common to Mk and also MT &/or Lk, 8.189 of the 10,650 words of Mk are found in the other two: 7,768 in Mt, 7,040 in Lk).
                This shows Mt, Mk & Lk depend upon each other or upon a common source, now lost (s.e.)
                By itself it does not indicate that Mt & Lk copied from Mk, for it could be that Mk copied common material form Mt & Lk (as argued by W. Farmer, The Synoptic Problem [1964]). (p.)
        2.  Material common to Mt, Mk & Lk:  In the order of the pericopes (sections) Mt, Mk & Lk usually agree; where they do not, Mt & Mk agree against Lk, or Mk & Lk agree against Mt.  Mt & Lk never agree against Mk (there are 3 single isolated verses and one OT citation where Mt & Lk agree in their order against Mk, but this is all):

Order of Pericopes of Triple-attested Tradition
Lk, Mk & Mt agree in sequence the vast majority of the time

                This excludes as highly improbable that Mt used Lk or Lk used Mt (s.e.)
                 It is a strong argument that Mk is the middle term between Mt & Lk (s.e.) since the agreement in order is always with Mk.
        3.  There are many strict agreements in precise language between Mt, Mk & Lk (see I.A.2.a above).
                This excludes the possibility that they merely have oral tradition in common (s.e.).  (This argument is strengthened by the fact that the Jesus-tradition is cited quite fluidly until the end of the 2nd century, i.e. long after all four gospels written.)
                This means that there is a literary (i.e., written) base in common (s.e.)
        4.  Mk's Greek is colloquial or Semitic in style, whereas Mt & Lk present the same material in better Greek.
e.g., Mk 2.4: κράβαττος (camp bed, pallet); Mt uses κλίνη , Lk uses κλινίδιον (from these we get 'recline') - better words for 'bed'
                This points toward Mt & Lk being dependent upon Mk (or something like Mk) (p.)
                It points away from Mk being drawn out of Mt & Lk  (i.e., if Mk found good Greek, he would not have made it worese) (e.)
        5.  Stories which are coherent in Mk are not always as coherent in the shorter versions in MT & Lk.
e.g., in story of John's passion (Mk 6.17-29)
        6.19:  Herodias has a grudge against John
        6.20:  Herod fears John, keeps him safe, hears him gladly.
        6.26:  Herod is sorry he has to behead John.
in Mt:
        14.5:  Herod wants to put John to death
        14.9:  Herod is sorry he has to behead John (i.e. inconsistent)
e.e., in Jersus before Pilate (Mk 15.6 ff.)
        15.6:  Governor's custom top release one prisoner
        15.11:  Crowd wants him to release Barabbas, not Jesus
in Lk:  Omits the custom of 15.6
        23.18:  They cry, 'Away with this man [Jesus], release Barabbas (i.e., Lk's version is understandable only on basis of Mk's)
                This points to: Mt & Lk being based on Mk (or Ur-Mk) (e.)
        6.  Mk 1.1-6.44; 8.27-14.52; 16.1-8 is based in continuous sequence on Jewish lectionary (see Mark and the Triennial Lectionary);  Mk 6.45-8.26 is based on Joshua (= 'Jesus' in Greek)  (see Mark 6.45-8.26 in parallel with Joshua  &  Mark 6.45-8.26 and the Gentile Mission); Mk 15 is the heart of the Passion (fixed before Mk was written) to be read at Passover.
            Therefore, Mk 6.45-8.26 appears to be a later insertion
            Lk does not produce any of Mk 6.45-8.26.
            Mt reproduces almost all  of Mk 6.45-8.26 and in sequence with the Markan material that precedes and follows it, and Mt appears to adapt this (Joshua-based) material to fit Jewish lectionary sequence.
                This shows that:
(1)  There were two editions of Mk. the earlier one not having Mk 6.46-8.26 (s.e.) (i.e. Mk I & Mk II)
(2)  Lk used this earlier edition or an Ur-Mk that was like it (e.).
(3)  Mt used the later edition, Mk II (s.e.)
(4)  Mk did not use Mt, for it makes sense for Mt to adapt Mk 6.45-8.26 to the lectionary, but it makes no sense for Mk to dislocate this material from the lectionary base (s.e.)
(5)  Heart of Passion Narrative (Mk 15) fixed before Mk (e.)  (Note: Mt re-works it to fit lectionary)
(6)  Lk did not use Mt (e.) for same reason he did not use Mk II.
        7.  In some places Mt, Mk & Lk all seem to be dependent upon an earlier written source.  E.g.:

We note that Mk's 'baptize with Holy Spirit' is a change from the threshing (judgemental) imagery of Mt & Lk.
Note: all three agree in lectionary setting for Temptation,, namely Gen 33.18-35.8 (T1 for Mt & Mk, N1 for Lk).
Since (1) additions common to Mt & Lk come at same place, & (2) Mt & Lk agree (vs Mk) that John's 'Coming One' would be a judge, and (3) Mt, Mk & Lk all use same 'lctionary' location for temptation, but Mk's version is very short,
            Therefore, indicates: Mk, Mt & Lk all dependent (in part, at least) on a common source earlier than Mk (Ur-Mk), and Mk has shortened it (s.e.)
        8.  Materials common tp Mt, Mk & Lk:  There are a few (but not too few) passages in which Mt & Lk agree in wording or content against Mk.
(a)   Some can be explained as Mt & Lk choosing the same common words to replace Mk's words (i.e., merely coincidental agreement)), but others are more striking.   (J. Hawkins, Horae Synopticae, 2nd ed., 1909, pp. 210 f., gives 21 striking passages.)
(b)  Some may be explained as the influence opf one gospel on another in the process of trasmission (i.e., copyists making chnages).
(c)  But others still remain that are hard to account for to everyone's satisfaction on the basis of either (a) same choice or (b) copyist's changing one to the other.  (B. H. Streeter, The Four Gospels, 1924, pp. 295 ff., argued for arguments (a) & (b) as accounting for all the passages.  R. H. Fuller, A Critical Introduction to the New Testament, 1965, p. 71, agrees.  Others do not.)
            Therefore, this suggests that Mt, Mk & Lk used Ur-Mk (e.)
        9.  Materials common to Mt, Mk & Lk:
(a)  occur mostly in (five or so) large solidly Markan blocks in Lk;
(b)  occur intimately fused with other material in Mt.
            This argues for:
                        (1)  Mk (or Ur-Mk) as the middle term between Mt & Lk (s.e.)
                        (2)  Mt as more closely related to Mk (or Ur-Mk) than Lk is related to Mk (or Ur-Mk). (e.)
        10.  "The Gospel of Mark differs decidedly from its fellows [i.e, Mt & Lk] in that it seems to be an abridgement, a digest of material known to its author but utilized only in part."  (C. C. Torrey, Our Translated Gospels, 1936, p. 261.)
            This argues for extensively organized pre-Markan material, pehaps even an earlier gospel (our Ur-Mk) (e.)
            This might also be explained by Mk's being an 'agreed summary' of Mt & Lk (W. Farmer) (p.)
        11.  Mk's "picturesque details" are not found in Lk or Mt.
e.g., young man (νεανίσκος ) who runs away naked at Jesus' arrest.  (These are sometimes taken as showing that Mk is earlier, but they can often be explained, like the νεανίσκος, as being 'theologically loaded' and hence not necessarily early or primitive.  On νεανίσκος in Mk 14.51; 16.5, cf. R. Scroggs & K. I. Groff, 'Baptism in Mark: Dying and Rising with Christ', JBL 92 (1973), pp. 531-548, and Mk 16.1-8 in Mark and the Triennial Lectionary.
            Therefore, these prove nothing.
        12.  Mk gives peoples' names and Aramaisms not found in Mt or Lk.
        W. Farmer, The Synoptic Problem, argues these are traits found in 2nd century apocryphal gospels, therefore they tend to show that Mk is late.
        But many other scholars say that these are traits of primitiveness in Mk that would be removed in later (more-Greekish) writings.
        In view of this mixed assessment, it is wiser to take this material as not positive 'proof' in any direction.
        13.  Words in any given section of Mk usually occur in Mt & Lk or in Mt or in Lk.
            This argues for:
                   (1)  Direct relationship between Mk & Mt, and between Mk & Lk (s.e.)
                   (2)  By itself, it could be either Mt & Lk draw on Mk (or Ur-Mk), or Mk is an 'agreed  synopsis' of Mt & Lk (as per W. Farmer) (p.)
        14.  Pierson Parker, The Gospel Before Mark, 1953, pp. 52-60, finds 16 cases where Mk clearly appears to shorten or summarize material, and
                    in Mt all 16 parallel passages are extended, while
                    in Lk only 2 parallel passages are extended.
            This argues that:
                    (1)  Mk is either dependent upon Mt (p.), or
                    (2)  Mk is dependent upon a source common to Mt (Ur-Mk?) (p.)
                    (3)  But Mk is not dependent on Lk or Lukan source . (e.)
        15.  Style:  Cruder Aramaisms in Mk (in vocabulary & also details of style & construction) than in Mt & Lk.
Also, the 8 Aramaic words in Mk (e.g. 5.41: talitha cumi; 7.11: corban; 7.34: ephphatha) are all lacking in Lk & only 2 are in Mt (27.33, 40).        
            This argues for Mk not being dependent on Mt or Lk (e.)
        16. Style:  Of expressions which occur at least 3 times in Mk:
37 expressions are as common to Mt's parallels as they are to Lk, Jn & Acts combined.
5 expressions only are as common to Lk's parallels as they are to Mt, Jn & Acts combined.
            This argues:
                    (1)  Mt & Mk are more closely related than Lk & Mk. (s.e.)
                    (2)  Mk is not an 'agreed synopsis' of Mt & Lk (W. Farmer's position) or Mk would reflect Lk's usage more (e.)
        17.  Mk's survival:  The fact that Mk has survived at all in the face of the fullness of Mt & Lk and in the face of Mt's great popularity (in eucharistic lectionaries, etc.), means that Mk had an importance of its own that prevented its being displaced.
        If Mk was an 'agreed summary' of Mt & Lk, this would place Mk as late first century or early 2nd, and it would be harder to account for Mk's survival as a canonical gospel, since Tatian's Diatessaron (in Syriac, ca. 150 CE), a conflation of all 4 Gospels, was suppressed in the 5th century CE.
            This argues:
                    (1)  Most likely, Mk was written earlier than Lk or Mt, & its use became so widespread it could not be lost (e.).
                    (2)  Mk's survival as canonical gospel makes unlikely W. Farmer's position that Mk was 'agreed summary' of Mt & Lk. (e.)
        18.  The so-called Q material in Mt & Lk:  Much of it is verbally identical (see I.A.3 above for outline of contents).
a.  Word-for-word agreement (e.g., Mt 3.7-19; 7.7-11; 11.4-6; 12.43-45; 24.45-51 & Lukan parallels).
b.  Rather slight agreement in some parts (e.g., Mt 10.26-33; 25.14-30 & Lukan parallels).
c.  Common vocabulary in all such sections is over 50%  (de Solages, A Greek Synopsis of the Gospels, 1959, p. 1047, & R. Morgenthaler, Statistiche Synopse, 1971, p. 165; C. E. Carlston & D. Norlin, "Once More--Statistics and Q", Harvard Theological Review 64, 1971, pp. 71, 77, estimate the verbal agreement at 71% -- 27% higher than in the Mk material Mt & Lk have in common.)
            This argues for:
                    (1)  Mt & Lk using a common source or sources, at least part being in a written document now lost (s.e.).
                    (2)  The closer verbal agreement between Mt & Lk in Q material than in Mk material argues against Mk being an 'agreed summary' of Mt & Lk (e.).
                    (3)  Lss likely (because of the difference in verbal agreement in Mt & Lk regarding Q & Mk materials) is the possibility Mt copied Lk or Lk copied Mt (p.).
                    (4) Some Q material was oral &/or circulated in two versions, one reaching Mt & one reaching Lk (e.).
        19.  Q material:  at times Mt & Lk diverge in their wording in ways that are contrary to their own emphases, which strongly suggests that these wordings are the forms in which the tradition reached them.
e.g., Mt does not repeatedly emphasize 'Holy Spirit', but Mt 11.28: 'But if it is by the Spirit of God that I cast out demons, ... then the kingdom of God has come upon you.'
Lk emphasizes Jesus as constantly guided and empowered by the Spirit, but Lk 11.20: 'But if it is by the finger of God that I cast out demons, then the kingdom of God has come upon you.'
except for 'Spirit' and 'finger', these are identical, but 'Spirit' and 'finger' occur in the opposite gospels from those we would expect.
            Therefore, probably Mt found 'Spirit' in his source & Lk found  'finger' in his, so that:
                    (1)  It is quite possible, even probable, that the Q document circulated in at least two forms, one used by Mt, one used by Lk (p./e.).
                    (2)  The data could also be explained by Q being partially oral (rather than all written) (p.e.).
        20.  Q & Mk material:  How their use differs in Mt & Lk.
In Mt: Mk + Q + M intimately mixed together.
In Lk: Q + L mixed together, but Mk appears primarily in large blocks without any Q or L mixed in (most of the material not in Mk occurs in Lk 6.20-8.3 and 9.51-18.12).
            This argues: for direct relationship between Lk & Mk (or Ur-Mk) and against direct relationship between Mt & Lk (e.).
            It also argues for: Q + L materials combined before Lk took up the materials from Mk (i.e., it agrues for Proto-Luke) (e.).
        21.  Q: Despite the different distribution of Q material, the order of it is largely the same in Mt & Lk. 
(See W. G.Kümmel, Introduction to the New Testament (1966), 51-53, Revise edition (1975), 65-67, for table showing this.)
            This argues: (1) for separate existence of Q (e.), and (2) against Mt using Lkl or Lk using Mt, & not Q (e.).
        22.  Q & Mt:
(1)  When Q sections are removed from Mt, remainder often makes good sense & often is even more coherent than before.
(2)  The Q passages have a characteristic vocabulary, style, & subject matter as against the rest of Mt (and even a christology which Mt has to change to match his own -- cp. Mt 11.19 to more primitive Lk 7.35; Mt 23.34 to more primitive Lk 11.49, remembering that for Mt Jesus is Wisdom, while in Q he is only a child of Wisdom).
            This argues: for Mt drawing on a pre-Matthaean source, i.e. Q (e.), and against M. D. Goulder, Lection and Midrash in Matthew (1974), who sees Matthew working only from Mk and .making up' all the rest on basis of Mk, and then Lk using Mt (e.).
        23.  Q:  Material; has characteristic vocabulary, style & theology.
(see, e.g., H. E. Todt, The Son of Man in the Synoptic Tradition (1963) and M. Jack Suggs, Wisdom, Christology and Law in Matthew's Gospel (1970).)
            This argues for separate existence of Q as written document (s.e.).
        24.  Mt & Mk:  In Mt there is idiomatic Semitic parallelism in passage also found in Mk, where Mk has only summaries lacking the parallelism.
            Either:  (a) Mk has abbreviated Mt (p.) or (more likely) Mk has abbreviated pre-Matthaean materials (Ur-Mk) (p.), and (b) Mt also dependent upon Utr-Mk (p.).
            Or:  Mt has made a poetic expansion of Mk or pre-Markan materials (Ur-Mk) (p.)  (Goulder's position: see 22.)
        25.  Style:  Hebraizing features in Lk (with close similarities to LXX), but Aramaizing features in Mt & Mk:
'And it came to pass' (καὶ ἐγένετο or ἐγένετο δὲ ) -- this has a Hebrew equivalent, but not an Aramaic one.  (6x in Mt, 4x in Mk, 42x in Lk, 0x in Jn).
 ἐν τῷ + infinitive: seems Hebrew rather than Aramaic.  (1x in Mt, 1x in Mk, 25x in Lk, 0x in Jn).
ἀφεῖς or καταλιπών with terms signifying departure -- apparent Aramaism, occurs several times in Mt & Mk, not at all in Lk or Jn.
            Indicates: 
(1) Mt & Mk are more closely related than Mk & Lk (s.e.);
(2) Mk is not an 'agreed synopsis of Mt & Lk' (Farmer's position -- see  10.), or Mk would probably refelct Lks usage more (e.).
        26.  Pre-Matthaean catechizing unit in Matthaean community:
Mt 5.1-7.29 (Sermon on Mount) is chiastic against Mt 22.15-24.35 (which includes the Woes of Mt 23), see The Chiasm of Matt 5.1-7.29 against 22.15-24.35
Mt 5.1-7.29 and 22.15-24.35 placed together make a continuous lectionary sequence in Mt (e.)
Mt has placed 5.1-7.29 in his first year of lectionary sequence & 22.15-24.35 in his third year of lectionary sequence (e.).
        The core of this unit appears to be the (Q) beatitudes & woes as found in Lk 6.20-26 (4 beatitudes, 4 woes), built up with Mk, M & Q materials (Mt 5-7: M + Q; Mt 22.15-46: Mk; Mt 23: M + Q, with a bit of Mk; Mt 24.1-35: Mk, with abit of M & Q).
        This appears to be a Two-ways document, a Way of Life & Light (Mt 5-7) and a Way of Darkness &B Death (Mt 23 3especially), like the Two-Ways document found in the Didache and the Epistle of Barnabas (the 2-Ways section in Did. & Ep. Barn. probably dating from ca. 70 CE).
            This argues: Mt 5.1-7.29; 22.15-24.35 is a pre-Matthaean unit, probably used for catechizing (e.)
            This suggests:  (1)  Since second half is Mk in Mt 22 & 24, that Mt drew on Ur-Mk (p./e.), or (2)  catechism-maker could have drawn on Mk while while Mk was still used as the gospel for the Matthaean community (p./e.).
        27.  Relation of Mt & Lk regarding Q material:  Material common to Mt & Lk  (Q) never appears at the same point in relation to the franework they have in common with Mk, apart from the Baptist texts and temptation stories (see 7 above).
            This argues:  (1) against any direct relationship between Mt & Lk (s.e.), and (2) for independent existence of the Q material apart from Mk (s.e.) and also apart from Ur-Mk (e.). 
[Note: this is only e. -- not s.e. -- with regard to Ur-Mk, since to be s.e. one would have to show clearly (a) Ur-Mk existed, and (b) both Mt & Lk used it.]
        28.  Mt, Lk & Q:  Mt & Lk alternate in offering what scholars consider to be the original form of the material they have in common apart from Mk (i.e., Q).
(Shown by J. Schmid, "Mt und Lk, Eine Untersuchung des Verhältnisses ihrer Evv.", Biblische Studien 22, 2-4, 1930, pp. 183 ff., and L. Vaganay, Le problème synoptique, 1954, pp. 293 ff.)
            This argues: (10 against direct relationship between Mt & Lk (s.e.), and for independent existence of Q material, perhaps in 2 versions (one in Lk's church and one in Mt's church) (e.).
        29.  Mt, Mk, Lk & Q:  There are a string of sayings which occur twice in Mt & Lk, once in a setting which Mk has also, a second time in a setting found only in Mt & Lk.
E.g., "He who has, to him shall be given" (Mt 13.12; Mk 4.25; Lk 8.18; 2nd time: Mt 25.29; Lk 19.26)  (For list of passages see Kümmel, Intro. to NT, rev. ed., p. 67).
In the whole of Mk only one saying occurs twice (Mk 9.35; 10.43 f.).
            This argues Mt & Lk used in common another source other than Mk, namely Q (s.e.).
        30.  Passion narrative:  All four Gospels are basically closest together in the Passion narrative, but there is no clear evidence than Jn knew Mt, Mk or Lk (although in some things, e.g., Feeding of 5,000, Jn is closest to details of Mk's tradition).
            This argues for passion narrative as relatively 'fixed' earlier than any other part of the Jesus-tradition (s.e.).
        31.  Some suppose that there is 'higher Christology' in Mk or Mt (less often in Lk), and therefore they argue that this 'proves' one gospel is likely to be later than another (because of its 'higher Christology').
(1)  The christological and soteriological models are different, but it is not self-evident that any one of the gospels clearly represents a 'later' development than any other.
(2)  These differences may well be different models for different localities rather than for different times.
            Thus this proves nothing.
        32.  Mt's and Lk's supposed 'respect' for he disciples compared to Mk is often cited as an argument that Mk is earlier than Mt or Lk.
But this is basically a question of difference of theological models and the evangelists' purposes rather than 'respect'; for the disciples as such.
(E.g., Mk: 'Messianic Secret' = Jesus is God's Son, and 'faith' = understanding plus trusting obedience;  Mt: 'Messianic Secret' = Jesus is the Christ.  In Mt it is no secret that Jesus is God's Son, and 'understanding' is a presupposition of 'faith' rather than a part of it.  Such changes as these mean that Mt castigates the disciples on different grounds (namely,  lack of obedient trust) than Mk does.
            Thus this proves nothing.
        33.  Mt. Mk & Lk -- Lectionary differences.   [This is under continued investigation in 2005, hence omitted here.]
        34.  Material common to Mt & Lk (& Papias' statement):
(1)  This material is almost completely sayings of Jesus.
(2)  But it also includes the narrative framework of Mt 3.1-4.11 // Lk 3.1-4.13 // Mk 1.1-13, plus the story of the centurion, Mt 8.5-13 // Lk 7.1-10.
(3)  Papias' (ca. 60-130 CE) statement, recorded by Eusebius (H.E. III,xxxix.16) that "Matthew composed the oracles (τὰ λόγια ) in the Hebrew language [?=Aramaic], and each one interpreted them as he was able" may well refer to Q, since τὰ λόγια can mean 'oracular utterances'.  Since our Mt is essentially a Greek work based on Greek sources, τὰ λόγια is most unlikely to refer to our Mt.
(4)  T. W. Manson, The Teaching of Jesus, 2nd ed., 1935, pp. 27 ff., shows
        (a)  An Aramaic sayings-source, differently translated (or mis-translated) would explain many of the Q-material differences in Mt & Lk..
        (b)  The only part of the centurion's story that is the same in Mt & Lk is the sayings part, so that Q very likely had no narrative materials, & Q is likely the Logia compiled by Matthew (in Aramaic, not Hebrew).
            This suggests Ur-Mk as the source of the narratives common to Mt & Lk not found in Mk (p.).
            This suggests Q = The Logia compiled by Matthew (p.).
            This supports Q as having no narrative material (e.).
            This supports variations between Mt & Lk as being due  (at least partly) to differing translations of logia written in Aramaic (e.).

B.  Summary (and Evaluation) of the directions pointed to by the various data

        Passion Narrative
            Passion narrative as first part of Jesus-tradition to reach fixed form (which is agreed by all scholars).
                6 (e.)        Mk 15 (heart of Passion) is not in lectionary sequence.
                30 (s.e.)    All four gospels agree most in thye Passion narrative.

        Inter-relation of Mt, Mk & Lk
           Mt. Mk & Lk have written source(s) in common - proven
                1 (s.e.)    Content and arrangement of common materials closely similar
                3 (s.e.)   Much sayings material common to Mt & Lk but not Mk
                7 (s.e.)    Parallel narratives & additions (Baptist to Temptation)
                13 (s.e.)  Mk's words occur in Mt & Lk, or Mt, or Lk
            Mk (or Ur-Mk) as middle term between Mt & Lk - proven
                2 (s.e.)    Mt. Mk & Lk agree in order, or Mt & Mk, or Mk & Lk
                5 (e.)       Mk's coherent stories not always coherent in Mt or Lk
                6 (s.e.)    Mt related to Mk II; Lk related to Mk I
                9 (s.e.)    Mk material in blocks in Lk, fuswed with M & Q in Mt
                13 (s.e)   Words in Mk usually in Mt & Lk or Mt or Lk
                15 (p.)    Cruder Aramaisms in Mk than in Mt or Lk
                17 (e.)    Mk has survived as canonical gospel
                20 (e.)    Q & Mk material used differently by Mt & Lk
            Mk not an 'agreed synopsis' of Mt & Lk - almost certainly proven 
               Possibility Mk is an 'agreed synopsis' (W. R. farmer's hypothesis)
               |     2 (p.)     Mk's order agrees with Mt & Lk or Mt or Lk
               |     10 (p.)   Mk seems to be an abridgment of materials partly used
             or    13 (p.)   Words in Mk occur in Mt & Lk or Mt or Lk
               |     24 (p.)    Mt has idiomatic Semitic parallelism where Mk does not
               |    [33 (p.)    Mk as a 1-year lectionary to replace 3-year lectionary (Mt & Lk) - questionable]
               High probability Mk is not an 'agreed synopsis' of Mt & Lk
                    4 (e.)     Mk's Greek colloquial; Mt's & Lk's is not
                    15 (e.)    Cruder Aramaisms in Mk than in Mt & Lk
                    16 (e.)    Mk's common expressions often in Mt, seldom in Lk
                    17 (e.)    Mk has survived as canonical gospel
                    18 (e.)    Q material in Mt & Lk -- why not used if Mk is 'agreed synopsis'?
                    25 (e.)    Mt's & Mk's Aramaisms vs Lk's Hebraisms -- why not in Mk?

        Mark
            Mk existed first as Mk I (lacking 6.45-8.26) then Mk II (with 6.46-8.26) - almost certainly proven
                6 (s.e.)    Mk's lectionary sequence without 6.46-8.26 & Lk does not use Mk 6.45-8.26
             Mk earlier than Mt or Lk - basically proven (on basis of Mk-Mt, Mk-Lk ties)
                17 (e.)      Mk has survived as canonical gospel

        Matthew-Mark Relationship
            Possibility Mk dependent on Mt - basically excluded by next two items  
            |    14 (p.)    16 cases where Mk appears to shorten what Mt has fully
           or   24 (p.)   Mk's summaries lack Mt's idiomatic Semitic parallelism
            Mk not dependent on Mt - quite certainly proven
                 4 (e.)      Mk's Greek is colloquial, Mt's is not
                 5 (e.)      Coherent stories in Mk not always coherent in Matthew
                 6 (s.e.)   Mk 6.45-8.26 not lectionary sequence, but par. in Mt is possibly
                15 (e.)    Cruder Aramaisms in Mk than in same material in Mt
            Mt as dependent on Mk (& on Mk II in particular) - almost certainly proven
                4 (p.)      Mk's Greek is colloquial, Mt's is not
                5 (e.)      Coherent stories in Mk not always coherent in Mt
                6 (s.e.)   Mk 6.45-8.26 not lectionary sequence, but par. in Mt is possibly
                13 (p.)    Words in Mk occur in Mt & Mt or in Mt or in Lk
                22 (p.)    M + Mk - Q often makes good sense & is even more coherent
                26 (e.)    Pre-Matthaean 'catechism' includes solid block of Mk
            Mk & Mt closer in relationship than Mk & Lk - proven
                9 (e.)     Mk in blocks in Lk buty fused with M & Q in Mt
              16 (s.e.)  Mk's common expressions often in Mt, seldom in Lk
              25 (s.e.)  Aramaizing features in Mt & Mk, Hebraizing ones in Lk

        Mark-Luke Relationship
       or-Possibility Mk dependent on Lk - highly unlikely in view of next two items
        |--Lk dependent on Mk I (or possibly Ur-Mk) - almost certainly proven
        |     4 (p.)    Mk's Greek is colloquial, Lk's is not
        |     5 (e.)    Coherent stories in Mk not always coherent in Lk
        |     6 (e.)    Mk 6.45-8.26, not in Mk I, is missing from Lk
        |    13 (p.)   Words in Mk occur in Mt & Lk or Mt or Lk
        |    20 (e.)    In Lk Q + L mixed, but Mk primarily in large blocks
        & Mk not dependent on Lk - best explanation
              4 (e.)    Mk's Greek colloquial, Lk's is not
              5 (e.)    Coherent stories in Mk not always coherent in Lk
            14 (e.)    Of 16 cases where Mk appears to shorten, only 2 extended in Lk
            15 (e.)    Cruder Aramaisms in Mk than in Lk's parallels
            16 (e.)    Mk's most common expressions are infrequent in Lk 
            25 (e.)   Hebraizing features in Lk's Greek, but Aramaizing ones in Mk

        Matthew-Luke Relationship
        or-Possibility Mk dependent on Lk - excluded by proven independence of Mt & Lk
         |      18 (p.)    Much Q material (& also Mk material) identical in Mt & Lk
         |--Possibility Lk depends on Mt - excluded by proven independence of Mt & Lk
         |      18 (p.)    Much Q material (& also Mk material) identical in Mt & Lk
        or-Mt & Lk not directly dependent on each other - proven   
                2 (s.e.)   In sections common with Mk, Mt & Lk never agree in order vs Mk
                6 (e.)      Lk lacks Mk 6.45-8.26; Mt modifies it to fit lectionary sequence
               20 (e.)     Mk in Lk in blocks, Mk in Mt mixed with M & Q
               22 (p.)     M + Mk in Matthew often coherent; Q differs from M + Mk
               27 (s.e.)  Q material never at same point in Mk framework in Mt & Lk
               28 (s.e.)  Mt & Lk alternate in presenting primitive form of Q
            The above excludes the possibility that Mt makles midrash on Mk, Lk copies Mt (Goulder's hypothesis)
               24 (p.)     Mt has idiomatic Semitic parallelism where Mk lacks it

     Ur-Mark
            Mk dependent on Ur-Mark - fairly likely, even if not proven completely
                 7 (s.e.)    Mk 1.1-13 as shortening & modifying what Mt & Lk have fully
                 8 (e.)       Mt & Lk have some significant agreements against Mk's version
                10 (e.)      Mk differs from Mt & Lk in seeming to be abridgement (Torrey), with fuller version in Mt
                14 (p.)      16 times Mk appears to summarize what Mt gives fully
                24 (p.)      Mt has Semitic idiomatic parallelism where Mk lacks it
            Mt is directly dependent on Mk II, even if Mt draws on Ur-Mark - proven
                6 (s.e.)     Mk 6.45-8.26 not lectionary sequence, but par. in Mt is
            Lk dependent upon Ur-Mk - fairly likely, even if not proven completely
                7 (s.e.)     Additions to Mk 1.1-3 common to Mt & Lk come at same places
                8 (e.)        Mt & Lk have some significant verbal agreements against Mk
               34 (p.)       Narratives common to Mt & Lk not found in Mk

    Q
        Mt & Lk draw on Q (a lost source, mostly or entirely sayings) - proven
                18 (s.e)    Much verbally identical non-Mk material in Mt & Lk
                20 (e.)      Q materials used different ways in Mt & lk   
                21 (e.)      Mt & Lk distribute Q differently bit largely in the same order
                27 (s.e.)   Q material never in same relation to Mk framework in Mt & Lk
                28 (s.e.)   Mt & Lk alternate in presenting more primitive forms of Q
                29 (s.e.)   Sayings often occur twice in Mt & Lk, once in Mk, none non-Mk
        Mt draws on Q - already proven, this is data bearing on Mt only
                22 (e.)     Q theology not Mt's; Mt minus Q often makes sense
        The Q material was largely in a written document, now lost - proven
                18 (s.e.)  Vocabulary common to Mt & Lk in Q is over 50%
                19 (p.)     Mt & Lk diverge in Q at points contrary to their own usage
                21 (e.)     Mt & Lk have same order of Q despite different distribution 
                23 (s.e.)  Qaterial has characteristic vocabulary, style & theology
        Some Q material may have been oral, not written - likely
            |   18 (p/e.)  Rather slight agreement in some parts of Q material in Mt & Lk
      &/or 19 (p.)     Mt & Lk diverge in Q at points contrary to their own usage
        Some variations in Q material may be due to differing translations from Aramaic
            |   18 (p/e.)  Q sayings in Mt & Lk diverge in ways contrary to Mt & Lk
            |   34 (e.)      Examples of Aramaic that account for Mt's & Lk's differences
       &/or                    -- this may be The Logia compiled by Matthew - possible
        Q may have circulated in 2 versions (one to Mt, one to Lk) - quite possible
                18 (p/e.)  Q agreement is rather slight in some parts of Mt & Lk
                19 (p/e.)  Q sayings diverge in Mt & Lk in ways contrary to Mt & Lk
                28 (p.)     Mt & Lk alternate in having original form of Q

   Matthew & Q
        Matthew as starting to write before he took up Q as such - possible
                22 (p.)    If Q removed from Mt, remainder often coherent
        Matthew has pr-Matthaean unit drawn from Mk (or Ur-Mk) plus Q - almost certain
                26 (e.)    5.1-7.29; 22.15-24.35 chiastic lectionary unit of M+Mk+Q.

   Luke & Q
        Proto-Lk (Q + L) existed before Lk drew on Mk (Pierson Parker, Vimcent Taylor) - likely
            20 (e.)    Q + L mixed together; Mk used mainly in large blocks
                            [The likelihood of this is reinforced by the demonstration above that Mt & Mk are more closely related than Lk & Mk, and yet both Mt & Lk drew on Mk.  Did Lk use Mk to fill his lectionary &/or narrative gaps?]

III.    THE LIKELY SOLUTION TO THE PROBLEM
    A.    Summary of the results obtained in Section II:

                We can take as assured:
1.  Mt, Mk & Lk have written source(s) in common.
2.  Mk is the middle term between Mt & Lk, but Mk is not an 'agreed synopsis' of Mt & Lk.
3.  Mt is dependent on Mk II (which includes Mk 6.45-8.16).
4.  Lk is dependent on Mk I (or Ur-Mk)
5.  Mt & Lk independently use Q, a sayings source now lost, which was largely in written form.
 6.  Mk existed ion two stages, Mk I (which may have been Ur-Mk) without 6.45-8.26, and Mk II with it, the former used by Lk, the latter by Mt.
7.  Mt & Lk did not use each other.
8.  Mt & Mk are more closely related than Mk & Lk.
9.  The passion narrative was the first part of the Jesus-tradition to reach a (relatively) fixed form.

            We can take as very likely possibilities:
1.  Mt, Mk & Lk are all dependent on Ur-Mk.
3.  Some Q material may have been in oral form when Mt & Lk drew on it &/or Q may have reached Mt & Lk in two written versions.

            We can take it as quite possible that:
1.  Mt started to write before he took up Q in earnest (apart from Q material in Mt 5-7, 22.15-24.35), i.e., there was a Proto-Mt.
2.  Proto-Lk (L + Q) may have been draftyed before Lk took up Mk (or Ur-Mk), and this is somewhat more likely than Proto-Mt.
4.  Q may be The Logia compiled in Aramaic (not Hebrew) by Matthew (not 'Matthew' the evangelist).  I.e., this may well be what Papias' statement refers to.

               To the above we may add:
1.  Not only were the sayings and parables of Jesus gathered in Q, but there may well have been also one or more collections of miracle stories (aretalogies) drawn on by the evangelists.
2.  Since Mk is so complex, it is probably the first extant product of a process of attempts to write a gospel, rather than being itself the first gospel ever written.  (This probability is heightened by Lk 1.1., with its reference to many attempts to draw up a narrative of Jesus.)
3.. Papias' second century statement tha Mark was Peter's interpreter is probably an incorrect guess, since Mk shows too many signs of being well down the line of tyradition-transmission.  As A. Jülicher, Einleitugg in das NT [Intro. to the NT], 7th ed., 1931 (rev. with E. Fascher), has said, "Without the stimulus from Papias, we should scarecely have advertised Peter as the guarantor for material in the Markan narrative",.
4.  A separate passion tradition apart from Mk's may well have been drawn on by Lk, in the view of some scholars based on Lk's differences.

     B.    A Probable Solution to the Synoptic Problem

        Now we are ready to combine the foregoing into a solution.  Let our mottos be the famous maxim from Alfred North Whitehead:  "Seek simplicity - and mistrust it".
        The 'simple solution' to the Synoptic Problem sees merely Mk, Q, M & L as all the elements that are necessary for diagramming a solution, as has been done, for example, by A. M. Hunter in Introducing the New Testament (1946), p, 32.  This is the first diagram reproduced below.
        We have seen reason to think this diagram is an over-simplification, so let us take up this 'simplicity' as our basic pattern, and then we shall 'mistrust' it by adding modifications.  we shall do so by enlarging the rectangles and inserting into them the additional elements we have seen, plus adding some more things outside them which appear to pre-date them.

Elements which are 'assured' are underlined twice.
Elements which are 'probable' are underlined once.
Elements which are less assured are underlined with a dotted-line.

Relationships which are 'assured' are shown by a double-line arrow.
Relationships which are 'probable' are shown by a single-line arrow.
Relationships which are less assured are shown by a dotted-line arrow.
Relationships which are only possible are shown by a dotted-line arrow with a question mark in the middle.

If all the material in a rectangle is the source, the arrow is drawn from the boundary of the rectangle.
If an item within the rectangle is the source, the arrow is drawn from the item itself.
To keep the diagram as simple as possible, the arrows are merely drawn to the borders of a rectangle, not to items within it, usually.

The above could be made more complex, and the actual units and relationships may well be more complex.  Future work will likely change the probabilities of units and relationships, perhaps showing new ones.  But it seems unlikely that the 'simplicity' of Hunter's diagram will be upset, even if the 'mistrusting' changes with time.
        Note:  Further argument for Lk using Mk I (or Ur-Mk) and not Mk II:  Mk 6.45-8.26 concerns Gentile Mission conditions (see Mark 6.45-8.26 and the Gentile Mission), which Luke might have chosen to omit, since he covers the same issues in Acts, but this is unlikely since (1) he includes many Gentile motifs in Lk (e.), and (2) also includes Mk 4.35-5.20 (// Lk 8.22-39: Stilling of Storm & Gerasene demoniac), which are Gentile Mission items as shown by R. H. Lightfoot, History and Interpretation in the Gospels (1935), pp. 89-90 (e.).

POSTSCRIPT:  WHY BOTHER TO SOLVE THE SYNOPTIC PROBLEM?

        So we have solved the Synoptic Problem.  So what?  What differences, if any, does it really make for our understanding of the gospels?  The answer is plenty, and here are a few of them.

1.    It affects how we assess Mk 1.8: 'Holy Spirit' (vs Mt's & Lk's 'Holy Spirit/Wind and fire') and Mk 1.12-13 (Temptation).  If we accept that Mk is working from an earlier, fuller account (our Ur-Mk), then se see Mk as shortening and editing it.  Then Mk 1.8, with its shoft from the threshing (judgemental) image to simply 'Holy spirit' means that he is shifting John the Baptizer's remark to a reference to Christian baptism, and the association of the Holy Spirit with baptism.

2.    It affects greatly how we assess the difference between Mt and Mk regarding 'Messianic secret' and the meaning of 'faith' whether we think Mt depends on Mk or not.

3.    If we see Q as a source, then Mt's shift from 'Wisdom said' (in Lk) to 'I say' (in Mt) stands out and is significant for understanding what Christology Matthew is presenting.

4.    If we see Q as a source, then Q has a coherent outlook, etc. (Tödt, etc.) representing another independent line of early tradition than Mk, and it takes us back closer to the historical Jesus than does M. D. Goulder's thesis that Matthew made a poetic expansion of Mk.

5.    Just as it makes a difference whether a NT writer does not say that 'Jesus is God' or whether he says 'Jesus is not God' (the former is silence, but the latter is a straight denial), so it makes a difference whether Mt first has, 'Jesus said, "I am sending to you..."' and then Lk modifies it to: 'Jesus said, "Wisdom will send to them ..."' (this is M. D. Goulder's position).  If this were the case, then it would be proof that Luke was lowering the Christology he already found in Matthew by making a lesser claim for Jesus (namely, that Jesus is not God's Wisdom) than Matthew does.
       But if, as we have seen, Matthew and Luke independently of each other draw on Q, then Luke in all  probability simply presents the form of the saying as he found it (for it is part of Q's Christology), an Matthewuis the one who is shifting the Q-Christology to one which claims more for Jesus.

These are just a few examples, and they could be multiplied many times over, but they help to show how important it is for a real understanding of Matthew, Mark and Luke-Acts that we deal seriously with the Synoptic Problem.