^{Page 249 note 1} See Butler, B. C., The Originality of St Matthew (1951), pp. 62–71Google Scholar; Farmer, W. R., The Synoptic Problem (1964), pp. 63–7, 211–15Google Scholar; Palmer, N. H., ‘Lachmann's Argument’, N. T. S. XIII (1967), 368–78, especially 369, 377fGoogle Scholar.

^{Page 249 note 2} The argument from order which I am discussing is the one which depends upon the absence of agreement between Matthew and Luke against Mark. As Palmer has shown, there is another argument from order, one which attempts to show that Mark's order can be demonstrated to be the source of the order of Matthew and Luke at individual points by adducing reasons why Matthew and Luke would have altered Mark's order if they had it before them. This was Lachmann's argument; but as Palmer (p. 377) points out, ‘it may also be that we could find “reasons” telling the opposite way in each single case’. For a recent attempt to show that Mark has the earlier order at individual points, see Wood, H. G., ‘The Priority of Mark’, Exp. Times, LXV (1953–1954), 17–19Google Scholar. He argues, for example, that the connection of the conflict stories in Mark, II. I–iiiGoogle Scholar. 6 is better than it is in Matthew.

^{Page 249 note 3} Woods, F. H., ‘The Origin and Mutual Relations of the Synoptic Gospels’ (11 1886), Studia Biblica et Ecclesiastica, II, 60Google Scholar.

^{Page 250 note 1} Here Woods has a footnote listing the ‘doubtful exceptions’: Matthew, xii. 22–30Google Scholar = Luke, xi. 14–23Google Scholar; Matthew, xiii. 38–42Google Scholar = Luke, xi. 29–32Google Scholar; Matthew, xii. 43–5Google Scholar = Luke, xi. 24–6Google Scholar. Perhaps also Matthew, xii. 33–5Google Scholar = Luke, vi. 43–5Google Scholar.

^{Page 250 note 2} Op. cit. pp. 61–2 (italics his)Google Scholar.

^{Page 250 note 3} See, for example, Weisse, C. H., Die evangelische Geschichte (1838)Google Scholar, cited by Kümmel, W. G., Das neue Testament (1958), p. 184Google Scholar.

^{Page 250 note 4} Woods, op. cit. p. 63Google Scholar.

^{Page 250 note 5} Streeter, B. H., The Four Gospels ² (1930), p. 161Google Scholar. Streeter applies Woods's point three to proving Mark's direct priority to Matthew and Luke. Woods used the argument to show that Mark's range and order were identical with the range and order of the Ur-Gospel.

^{Page 251 note 1} Ibid. p. 183. Streeter's statement is curious, because actually Hawkins knew very well that there are several places at which Matthew and Luke agree in placing the same Q material at the same point in the Marcan outline. Hawkins gives a list of ten such places in Oxford Studies in the Synoplic Problem ( 1911), p. 102Google Scholar, and Horae Synopticae² ( 1909), p. 208Google Scholar. The text which Hawkins showed Streeter must have been based on full pericopes in Tischendorf's synopsis. The ten places which Hawkins discusses are all instances in which Matthew and Luke have the same Q material within a Marcan passage, and in which the entire passage (both Marcan and Q material) is considered by Tischendorf to be one pericope. The pericope division, of course, is artificial. A different division would permit at least ten clear exceptions to Streeter's statement and to Woods's point four. See section VI below.

^{Page 251 note 2} I use ‘Marcan’ to refer to a passage in Mark which is also in Matthew and/or Luke, without prejudice as to whether Matthew and Luke copied it directly from our Mark or not. Similarly, I use ‘Q’ only to indicate the tradition common to Matthew and Luke alone.

^{Page 252 note 1} See, for example, Sir Hawkins, John, Horae Synopticae ², p. 114 n. 3Google Scholar. Bultmann, in assessing the evidence for the two-document hypothesis, writes, ‘Das Hauptgewicht aber fällt auf die Reihenfolge…’, (Die Erforschung der synoptischen Evangelien³ 1960), p. 8Google Scholar. See further Farmer, op. cit. pp. 63 ffGoogle Scholar. The belief that Mark's style is that of dictation from an eye-witness also was very influential when the two-document hypothesis was first being formulated. The difference between the style of Mark and that of Matthew and Luke was regarded as ‘the difference which always exists between the spoken and the written language. Mark reads like a shorthand account of a story by an impromptu speaker…’ (Streeter, , op. cit. p. 163)Google Scholar. Few people today seem to hold this view, although still regarding Mark as earlier. The view that some of it, at least, represents dictation is still held in some quarters, however. See, for example, Vincent Taylor, The Formation of the Gospel Tradition² ( 1935), p. ixGoogle Scholar.

^{Page 252 note 2} See Streeter, , op. cit. pp. 293–331Google Scholar.

^{Page 253 note 1} On the significance of the sequence of events for research on the life of Jesus, see Schweitzer's chapters on Strauss and on the Marcan hypothesis in The Quest of the Historical Jesus. Lachmann's original discussion of the synoptic order was apparently motivated by interest in the ‘precise chronology of the history of Jesus’. He argued that since the three synoptics show a common order, except where one or the other alters the order for his own purpose, ‘no greater weight can be placed on the witness of three evangelists than if a single and indeed unknown author had testified’. The gospels reveal the order of the Ur-Gospel, which order then must be investigated for information about the chronology of Jesus’ ministry. See Palmer, op. cit. pp. 375 fGoogle Scholar.

^{Page 253 note 2} Tischendorf's synopsis was first published in 1851 and was revised in 1864. I have used the fifth edition of 1884. Huck's synopsis was first published in 1892. Sanday, , writing in 1911Google Scholar, noted that the Oxford Seminar used Tischendorf, but referred to other synopses, including Huck's (Oxford Studies in the Synoptic Problem, p. viii)Google Scholar. See also Hawkins's remark ibid. p. 102 n. I.

^{Page 254 note 1} Oxford Studies in the Synoptic Problem, pp. 29–30Google Scholar.

^{Page 256 note 1} Op. cit. p. 66Google Scholar.

^{Page 257 note 1} It would be easy to increase the number of items in this category by considering less significant phrases. See, for example, ‘why are you afraid’ (Matt, . viii. 26Google Scholar; Mark, iv. 40)Google Scholar and ‘and leaving him, they departed’ (Matt, . xxii. 22Google Scholar; Mark, xii. 12)Google Scholar.

^{Page 257 note 2} Op. cit. p. 94Google Scholar.

^{Page 257 note 3} Cf. Hawkins, , Oxford Studies, p. 50Google Scholar.

^{Page 258 note 1} Cf. the list given by Hawkins, , Oxford Studies, p. 102Google Scholar and Horae Synopticae ², p. 208Google Scholar.

^{Page 258 note 2} If Huck's arrangement is followed, this passage is an instance in which Matthew and Luke add the same separate discourse after a Marcan pericope. Huck divides the pericopes thus: (I) Matt, . xxiv. 23–5Google Scholar = Mark, xiii. 21–3Google Scholar = Luke, xvii. 21Google Scholar; (2) Matt, . xxiv. 26–8Google Scholar = Luke, xvii. 23–4, 37Google Scholar. He does not, however, count Luke, xvii. 21Google Scholar as a full parallel to Matt, . xxiv. 23–5Google Scholar and Mark, xiii. 21–3Google Scholar.

^{Page 259 note 1} Op. cit. p. 306Google Scholar.

^{Page 260 note 1} While these three sayings are usually assigned to Q, their Matthean and Lucan forms are not really different from their Marcan forms. Thus we might protest that they are improperly assigned to Q, apparently on the basis of context alone. This is not an adequate reason for attributing a passage which occurs in all three of our synoptics to that hypothetical document, as is well recognized. The sayings in question, however, are too short to admit much variation in any case, so I have left the passages in the list.

^{Page 260 note 2} Even so, these pericopes are instances in which Matthew and Luke agree in the placement of Q material.

^{Page 260 note 3} For arguments in favour of assigning the Parable of the Mustard Seed to Q, see Hawkins, , Oxford Studies, pp. 50–3Google Scholar and Harnack, , The Sayings of Jesus (1908), pp. 26 fGoogle Scholar. Both mention first the connection of the parable with the Parable of the Leaven and then the fact that Matthew agrees partially with Mark and partially with Luke. These points are taken as establishing the existence of different versions, one Marcan and one Q. The argument from context is irrelevant. The fact that Matthew and Mark have ‘shrubs’ and Matthew and Luke have ‘tree’ (to mention the most striking agreement in the second point) does not seem sufficient evidence for the existence of different versions which are conflated in Matthew. Or if it is taken as such, what will one make of the fact that Mark, iii. 8 agrees withGoogle ScholarLuke, vi. 17Google Scholar in reading ‘Tyre and Sidon’ and with Matt, . iv. 25Google Scholar in having ‘beyond the Jordan’?

^{Page 261 note 1} The parallels to Mark, i. 4Google Scholar, Matt, . iii. 2Google Scholar and Luke, iii. 3Google Scholar, are not placed in Q by either Harnack or Hawkins, but Streeter (op. cit. p. 291)Google Scholar does assign them to Q. I do not see that there is any objective basis for assigning these verses to Q, following Streeter's own criterion, and so have followed Harnack and Hawkins.

^{Page 261 note 2} Assuming that Matthew does not support Mark, vi. 1–6aGoogle Scholar. See IV. 3 above.

^{Page 261 note 3} Of course the suggestion that Luke used Matthew has often been made; but it has been countered by the argument from order: why do Luke and Matthew never agree in the placing of the Q material if Luke knew Matthew? (See Kümmel, W. G., Introduction to the New Testament (1965), p. 50.)Google Scholar I have just shown that they do so agree, and so perhaps one impediment to accepting Luke's use of Matthew is removed. The number of agreements in order between Luke and Matthew is too large to attribute to chance. For recent articles which accept Luke's use of Matthew, see Argyle, A. W., ‘Evidence for the View that St Luke used St Matthew's Gospel’, J.B.L. LXXXIII (1964), 390–6Google Scholar; Simpson, R. T., ‘The Major Agreements of Matthew and Luke against Mark’, N. T. S. XII (1966), 273–84Google Scholar. For a recent article which takes another view, see West, H. Philip Jr, ‘A Primitive Version of Luke in the Composition of Matthew’, N. T. S. XIV (1967), 75–95Google Scholar.