Scholarly consensus
Matthew, Mark and Luke tell the same story in the same order, in many places in the same words, and sometimes with the same rare Greek expressions and the same parenthetical asides (“let the reader understand”: Mark 13:14; Matthew 24:15). Independent accounts of the same events, written in different places by different people, do not do this. Some literary relationship exists between the three. The question of what it is, and which way it runs, is the synoptic problem, and it has been studied without a break since the eighteenth century.
The data
About ninety percent of Mark is in Matthew and about half of it is in Luke, usually in Mark’s order. Where Matthew and Luke both follow Mark, they agree with each other; where one departs from Mark’s order, the other almost always keeps it. A further two hundred and thirty or so verses, chiefly sayings (the Lord’s Prayer, the Beatitudes, the temptation story, John the Baptist’s preaching), are shared by Matthew and Luke but absent from Mark; this is called the double tradition. Each of Matthew and Luke then has material of its own: Matthew’s magi, Luke’s shepherds; Matthew’s sheep and goats, Luke’s prodigal son.
Markan priority
Karl Lachmann observed in 1835 that Matthew and Luke agree in order only where they agree with Mark, and the conclusion that Mark is the middle term, used by both, has held the field since Heinrich Julius Holtzmann set it out in 1863. The supporting arguments are cumulative. Mark’s Greek is rougher, and it is easier to see Matthew and Luke polishing it than Mark roughening theirs. Mark’s hard sayings are softened by the other two, not the reverse: Jesus “could do no deed of power” in Nazareth (Mark 6:5) becomes “did not do many” (Matthew 13:58); “Why do you call me good?” (Mark 10:18) becomes “Why do you ask me about what is good?” (Matthew 19:17); Mark’s Jesus does not know the day or the hour (13:32), and some manuscripts of Matthew drop the words “nor the Son”. Matthew and Luke each add what Mark lacks, a birth story and resurrection appearances, and add different ones. An abbreviator who had Matthew and Luke in front of him and left out the Lord’s Prayer and the resurrection appearances is harder to imagine than two editors who supplied them.
Q
If Matthew and Luke each used Mark, the double tradition needs a source too, since neither seems to have known the other’s special material or the other’s order for the shared sayings. Christian Hermann Weisse proposed in 1838 that they drew on a lost collection of sayings; it has been called Q, from German Quelle, since the 1890s. The Gospel of Thomas, found at Nag Hammadi in 1945, showed that a gospel consisting only of sayings was a real literary form (see the Gospel of Thomas). No copy of Q survives, and its wording is reconstructed from the places where Matthew and Luke agree. B. H. Streeter’s four-source hypothesis of 1924 (Mark, Q, M and L) remains the standard textbook account.
Q is reconstructed by comparing Matthew and Luke, and is numbered by Luke’s verses. In the reconstruction in The Critical Edition of Q, which is the base of Mark M. Mattison’s public-domain translation at gospels.net, it holds the preaching of John the Baptist and the temptation (Q 3-4), the beatitudes and the love of enemies (Q 6), the centurion’s servant (Q 7), the instructions for the mission (Q 10), the Lord’s Prayer and “ask, seek, knock” (Q 11), the woes on the Pharisees (Q 11), the sayings on anxiety (Q 12), the mustard seed and the yeast (Q 13), the great dinner (Q 14), the lost sheep (Q 15) and the coming of the Son of Humanity (Q 17). It is almost all sayings, with little story. The Gospel of Thomas is also a collection of sayings, and about half of its sayings have parallels in the gospels, which is one reason Thomas and Q are often read together.
Minority view
A minority of scholars reject Q. The Farrer hypothesis, developed by Austin Farrer, Michael Goulder and Mark Goodacre, holds that Luke used Matthew as well as Mark, which accounts for the double tradition without a lost document and explains the “minor agreements” where Matthew and Luke agree against Mark in Markan passages. The older Griesbach or Two-Gospel hypothesis, revived by William Farmer, holds that Matthew was first, that Luke used Matthew, and that Mark abbreviated both. A few scholars still defend the independence of the three, but almost none who work on the texts themselves.
What every solution has in common
On any of these solutions, the synoptic gospels are not three independent eyewitness reports that happen to agree. They are edited works in a chain of dependence, each writer copying, correcting, rearranging and supplementing the one before. Where they agree, the agreement counts once, not three times, because it is copying. Where they disagree, the disagreement is a window on what the second writer thought the first had got wrong. The “fourfold witness” of Christian apologetics is, on the gospels’ own evidence, a single tradition in several editions.
Why it matters here
The Islamic Dilemma assumes that the gospels are a fixed and unified body of testimony that the Qur’an must either accept whole or reject whole. The synoptic problem shows the opposite. The gospels were revised in the act of being written, by their own authors, within a generation. A revelation given to ʿĪsā could lie within such books; it could not be identical with them. That is what the measure on this website is for. See the measure and The Islamic Dilemma, answered.
Sources
- Mark Goodacre, The Synoptic Problem: A Way Through the Maze.
- B. H. Streeter, The Four Gospels: A Study of Origins (1924).
- Bart Ehrman, The New Testament: A Historical Introduction to the Early Christian Writings, chapters on the synoptic problem.
- E. P. Sanders and Margaret Davies, Studying the Synoptic Gospels.
- James M. Robinson, Paul Hoffmann and John S. Kloppenborg (eds.), The Critical Edition of Q (Minneapolis: Fortress Press, 2000), and Mark M. Mattison’s public-domain translation of Q at gospels.net/quelle.