^{page 218 note 1} In Germany the solution is referred to as the Zweiquellentheorie, a title going back to the middle of the last century when the attempt was to find the relationship between the material common to the Synoptic Gospels: the English title allows for M and L as well as Mark and Q.I use the word source in as comprehensive a sense as possible, so as to include those who think of Q, for example, as a document, a block of oral tradition in firm or viscous form, or a collection of such traditions.

^{page 218 note 2} It may suffice to mention Farmer, W. R., The Synoptic Problem (New York, 1964)Google Scholar, Gaboury, A., La Structure des Évangiles Synoptiques (Leiden, 1970)CrossRefGoogle Scholar, Fuchs, A., Sprachliche Untersuchungen zu Matthäus und Lukas (Rome, 1971)Google Scholar, Schramm, T., Der Markus-Stoff bei Lukas (Cambridge, 1971)CrossRefGoogle Scholar, Boismard, M. E., Synopse des quatre Évangiles en français, Tome II, Commentaire (Paris, 1972)Google Scholar. All these alternative solutions have been criticized, in my view effectively, by Neirynck, F., especially in ‘La matière marcienne dans l'évangile de Luc’, L'Évangile de Luc (ed. Neirynck, F., Gembloux, 1973), pp. 157 ffGoogle Scholar., and in ‘Urmarcus Redivivus?’, L'Évangile selon Marc (ed. Sabbe, M., Gembloux, 1974), pp. 103 ffGoogle Scholar. Two further theories were adumbrated in this journal in 1975 by O'Neill, J. C. and Robinson, J. A. T.Google Scholar, neither of whom would claim to test the 4ST systematically.

^{page 218 note 3} Streeter's solutions to the minor agreements in The Four Gospels (London, 1924), pp. 295 ffGoogle Scholar., and Schmid's, in Matthäus und Lukas (Freiburg, 1930), pp. 35 ffGoogle Scholar., may be said to have settled the problem for forty years.

^{page 218 note 4} Both Sanday, , in Studies in the Synoptic Problem (Oxford, 1911), pp. 21 ffGoogle Scholar., and Hawkins, , in Horae Synopticae (2nd ed.Oxford, 1909), pp. 208 ffGoogle Scholar., found a number of the agreements too impressive to be explained away, and opted for a Deutero-Marcus solution. Four of Hawkins' twenty-one instances are included below.

^{page 218 note 5} Q is allowed by Streeter to overlap with Mark so as to cover any serious concentration of agreements, as in the baptism pericope; but once it is allowed to do this to any large extent, it begins to look like Matthew's Gospel. Cf. Sanders, E. P., ‘The overlaps of Mark and Q and the Synoptic Problem’, N.T.S. 19 (1973), 453 ff.Google Scholar

^{page 218 note 6} The theory of Schramm, see above, n. 2.

^{page 218 note 7} Fuchs returns to this theory, noting how often the agreements show a more developed form of the Marcan text.

^{page 218 note 8} So Boismard and Robinson, noting passages where Mark has a more developed form of the Matthaean text.

^{page 218 note 9} Neirynck, F., The Minor Agreements of Matthew and Luke against Mark (Leuven, 1974).Google Scholar

^{page 219 note 1} ML, p. 35.Google Scholar

^{page 219 note 2} Luke's versatility makes it an easy task to find explanations for his changes to Mark, including the agreements, by parallels from his writings. He has both a Septuagintal and a natural Greek style; he has a vocabulary considerably larger than the other evangelists; he has the disconcerting habit of varying his synonyms; and there is Acts besides. It is correspondingly harder for modern scholars to find points where he is writing uncharacteristically.

^{page 219 note 3} Attempts have been made to get rid of Q on grounds of philosophical logic, e.g. by Farrer, A. M., ‘On Dispensing with Q’, Studies in the Gospels (ed. Nineham, D. E., Oxford, 1954)Google Scholar, and Palmer, H., The Logic of Gospel Criticism (London, 1968)Google Scholar. It is not surprising that they have failed to carry conviction. Q is in possession of the field, and will not be removed without high explosives. To render it implausible, it is necessary to argue persuasively (1) that Luke knew Matthew in the triple tradition sections, (2) that a credible account can be given of Luke's order in the Q passages, and (3) that Matthew never develops a Q-saying found in Luke. The first is the purpose of this article; the second and third I have attempted in Midrash and Lection in Matthew (London, 1974)Google Scholar. Turner, N., in ‘The Minor Verbal Agreements of Mt. and Lk. against Mark’, Studia Evangelica (ed. Aland, K., Berlin, 1959), 1, 223 ffGoogle Scholar., uses a similar method to my own, but limits it to four controversial pericopes.

^{page 219 note 4} The Greek is cited from Aland, K., Synopsis Quattuor Evangeliorum (4th ed.Stuttgart, 1967).Google Scholar

^{page 220 note 1} Streeter, , FG, pp. 206 fGoogle Scholar., proposes to include Matt, . iv. 13Google Scholar /Luke, iv. 14 fGoogle Scholar. in Q, although Νααρά is the only word in common to justify the suggestion; and Matthew's form of Q would then have said that Jesus left Nazareth, Luke's that he came to it. But in fact virtually every other word in the accounts is either drawn from Mark or is characteristic of its author.

^{page 220 note 2} ML, pp. 85 f.Google Scholar

^{page 220 note 3} Cf. Streeter, , p. 206Google Scholar, ‘This (sc. the inclusion of the verse in Q) would also explain why in the Lucan version the story of the Rejection of Nazareth is inserted in this context – or rather it would justify the insertion.’

^{page 223 note 1} The distinction between νεανίας and νεανίσκος is less certain, but may correspond to our ‘young man’ and ‘youth’.

^{page 223 note 2} I am grateful to Prof. Neirynck for this point.

^{page 224 note 1} ML, p. 100.Google Scholar

^{page 225 note 1} Figures in this form throughout the article signify the number of uses in Matthew/Mark/Luke + Acts.

^{page 225 note 2} L'Évangile selon S. Matthieu (3rd ed.Paris, 1927), pp. cxx, cxxii.Google Scholar

^{page 225 note 3} The School of St Matthew (Uppsala, 1954), p. 51.Google Scholar

^{page 225 note 4} Das Evangelium des Markus (Göttingen, 1937), p. 11Google Scholar; cited by Stendahl.

^{page 225 note 5} The Use of the Old Testament in St Matthew's Gospel (Leiden, 1967), p. 11.Google Scholar

^{page 226 note 1} Neirynck, F., ‘Minor Agreements of Matthew-Luke in the Transfiguration Story’, Orientierung an Jesus (ed. Hoffmann, P., Freiburg, 1973)Google Scholar explains this as follows: ‘The fact that Luke prefers here the accusative…is linked with his representation of the cloud, not hanging over the disciples and over-shadowing them, but an involving reality in which they enter’. But was not the overshadowing of the Holy Ghost at i. 35 an even more involving reality into which Mary entered?

^{page 227 note 1} FG, pp. 325–8Google Scholar. The identical argument is reproduced by Schmid, , ML, pp. 157–9.Google Scholar

^{page 228 note 1} FG, p. 323.Google Scholar

^{page 229 note 1} Verleugnung, Verspottung und Verhör Jesu (Munich, 1969), pp. 54, 95 f.Google Scholar

^{page 229 note 2} Jeremias, J., The Eucharistic Words of Jesus (E. T. London, 1964), pp. 145 ffGoogle Scholar., cites the studies of Preuschen, E.—Pott, A., Tatians Diatessaron aus dem Arabischen übersetzt (Heidelberg, 1926)Google Scholar, and Hack, A. von, Marcion: Das Evangelium vom fremden Gott (Leipzig, 1921)Google Scholar, to show that ‘the pre-Marcionite Western text is characterized in its text of Luke by numerous assimilations to Matthew (and Mark)’. Of the eleven examples he gives of the concord D Marcion Tat^ned, six are of omissions by the Western text. Jeremias goes on to cite Western ‘major lacunae’ at Luke, v. 39Google Scholar, vii. 7a, 33, x. 41 f., xi. 35 f., xii. 19, 21, 39, xix. 25, xxi. 30, xxiv. 6, 12, 21, 36, 40, 50, 51, 52, which he argues to be omissions, and so to increase the likelihood that xxii. 19b –20 is an omission also. I have added arguments in this article that the same is true at Luke, iv. 15Google Scholar and vii. 27. We cannot ignore the possibility that the Western text of Luke was a text abbreviated by general tendency, like ⁴⁵, especially where parallels in either Matthew or Mark lacked the original material. This would be particularly credible if, as was believed in the second century, Mark was a Western Gospel.

^{page 231 note 1} FG, pp. 323 f.Google Scholar

^{page 231 note 2} Ibid.

^{page 232 note 1} Bauer, W. (-Arndt-Gingrich), A Greek-English Lexicon of the NT (4th ed.Cambridge, 1952)Google Scholar, ad voc.

^{page 232 note 2} J. T. S. XIV (1913)Google Scholar, 188 ff.

^{page 233 note 1} J. T. S. (1913), 538 ff., citation p. 546.Google Scholar

^{page 233 note 2} FG, p. 324.Google Scholar

^{page 233 note 3} J. T. S. 27 (1926), 179 ff., citation p. 181.Google Scholar

^{page 233 note 4} An Aramaic Approach to the Gospels and Acts (3rd ed.Oxford, 1967), pp. 136 ff.Google Scholar

^{page 234 note 1} Classically stated in Streeter, , FG, p. 183.Google Scholar

^{page 234 note 2} Pp. 241–442.