[1] In N.T.S. 24 (1978), pp. 218–34Google Scholar; his final example (no. 12 below) is also examined in more detail in his appended article ‘Mark xvi.1–8 and Parallels’, N.T.S. 24 (1978), pp. 235–40Google Scholar. In a subsequent article, ‘Farrer on Q’, Theology 83 (1980), pp. 190–5CrossRefGoogle Scholar, Goulder refers to his examples and says: ‘It will be interesting to see whether it is possible to make a convincing reply to them.’ (p. 195). The present article is thus an attempt to take up this challenge. I would also like to thank Dr Goulder for discussing his views with me in private correspondence and for encouraging the production of a reply to his original article.

[2] Page references in parentheses refer to Goulder's N. T.S. articles.

[3] Both classic ways of explaining minor agreements on the two-document hypothesis.

[4] Cf. Goulder's own observations about this as a possibility (p. 234) in his discussion of the argument, often used to support the Q hypothesis, that Luke's version is sometimes more original than Matthew's. Goulder in fact believes that Luke never has a more primitive version (cf. his earlier full-length study Midrash and Lection in Matthew (London, 1974)Google Scholar, but a discussion of this is impossible within the confines of a single article. See, however, my final paragraph here.

[5] Holtzmann, H. J., Lehrbuch der historisch-kritischen Einleitung in das neue Testament (Freiburg, 1885)Google Scholar; Morgenthaler, R., Statistische Synapse (Zürich, 1971), esp. pp. 300 ff.Google Scholar

[6] The abbreviations MattR and LkR are used here for Matthean redaction and Lukan redaction respectively.

[7] Strecker, G., Der Weg der Gerechtigkeit (Göttingen, 1962), p. 64Google Scholar; cf. too Matt. 16. 4; 21. 17; see also Prabhu, G. M. Soares, The Formula Quotations in the Infancy Narrative of Matthew (Rome, 1976), p. 129 f.Google Scholar

[8] These are the readings of the 26th edition of the Nestle-Aland text. There are textual variants at all the instances, but virtually no support for reading Ναδαρά in Matt. 2. 23; 21. 11. There are variants at Matt. 4. 13 and Lk. 4. 16, but Ναδαρά is the harder reading and is more likely to have been changed by later scribes. For a full discussion of the variant readings, see Soares, Prabhu, op. cit., p. 130 f.Google Scholar; also Chilton, B. D., God in Strength. Jesus' Announcement of the Kingdom (Freistadt, 1979), pp. 105, 128 f., 311–13.Google Scholar

[9] For the unusual spelling as implying the existence of a Q tradition here, see Streeter, B. H., The Four Gospels (London, 1924), p. 206 f.Google Scholar; Schweizer, E., The Good News according to Matthew (London, 1976), p. 67Google Scholar; Schürmann, H., Das Lukasevangelium I (Freiburg, 1969), p. 227 f.Google Scholar; Soares, Prabhu, op. cit., pp. 130–2Google Scholar, among others. For a full bibliography for all the minor agreements, giving references for the various explanations which have been proposed for each agreement, see Neirynck, F., The Minor Agreements of Matthew and Luke against Mark (Louvain, 1974).Google Scholar

[10] Kilpatrick, G. D., The Origins of the Gospel according to St. Matthew (Oxford, 1946), p. 110 f.Google Scholar; cf. Matt. 4. 23; 9; 35; 10.17; 12. 9; 13. 54; 23. 34.

[11] See Kümmel, W. G., Introduction to the New Testament (London, 1975), p. 59.Google Scholar

[12] Kümmel, , op. cit., p. 59.Google Scholar For the importance of miracles in leading people to faith in Luke's writings, see Achtemeier, P. J., ‘The Lucan Perspective on the Miracles of Jesus’, J.B.L. 94 (1975), pp. 547–62, esp. pp. 553 ff.Google Scholar Strictly this need not demand a change in order, since there is a miracle within the call story itself. But the new positioning, with the story coming after the miracles in Capernaum, gives added weight to the events to which Peter is responding.

[13] See, for example, Streeter, , op. cit., p. 299.Google Scholar

[14] Goulder says that ‘these may well have been more like beds than mattresses’ (p. 222), but does not indicate how he knows this.

[15] Cf. Cadbury, H. J., ‘Four Features of Lucan Style’, in Keck, L. E., Martyn, J. L. (eds.), Studies in Luke-Acts (London, 1968), pp. 87–102, esp. pp. 91 ff.Google Scholar

[16] Cf. Schmid, J., Matthäus und Lukas (Freiburg, 1930), p. 100.Google Scholar

[17] The apocalyptic discourse in Lk. 21 is addressed to all the disciples, not to a chosen group of four, or three, or two.

[18] Schürmann, , op. cit., p. 316 f. n. 37. See Lk. 9. 28; 22. 8; Ac. 3. 1 ff.; 4. 13 ff.; 8. 14.Google Scholar

[19] Schürmann, , op. cit., pp. 316 n. 35, 318Google Scholar; but see the detailed counter arguments of Neirynck, ‘The Argument from Order and St. Luke's Transpositions’, in Minor Agreements, pp. 315–8.Google Scholar

[20] Neirynck, ‘Urmarcus Redivivus?’, E.T.L. 34 (1974), pp. 103–45, on p. 11730Google Scholar, points out that, on the two-document hypothesis, Matthew only uses γέγραπται in Q passages (4. 4, 6, 7, 10; 11. 10) and in passages derived from Mark (21. 13; 26. 24, 31). This gives a reasonably consistent account of Matthew's varied formulae introducing OT citations.

[21] Cf. Taylor, V., The Gospel according to St. Mark (London, 1952), p. 153Google Scholar, who also cites earlier authors; Strecker, , op., cit., p. 63 n. 1Google Scholar; Manson, T. W., The Sayings of Jesus (London, 1949), p. 69Google Scholar, even goes so far as to say that ‘it makes nonsense’ here. However, Marxsen, W., Der Evangelist Markus (Göttingen, 1959), esp. p. 21Google Scholar, has shown that the quotation fits well into the whole structure of the paragraph, showing John as the new Elijah (cf. v. 6). See too Suhl, A., Die Funktion der alttestamentlichen Zitate und Anspielungen im Markusevangelium (Gütersloh, 1965), pp. 135–8.Google Scholar

[22] Cf. Schweizer, E., The Good News according to Mark (London, 1975), p. 28 f.Google Scholar Both Marxsen and Suhl ascribe the quotation to MkR. Their analyses show how the citation functions within the whole pericope, and that it is an integral part of the structure. But this does not necessarily say anything about the ultimate origin of the individual parts which go to make up the whole.

[23] Schürmann, , op. cit., p. 410 n. 19Google Scholar; Jeremias, J., Die Sprache des Lukasevangeliums (Göttingen, 1980), p. 161 f.CrossRefGoogle Scholar

[24] It is also not to be found in the LXX: there, θεραπεύω in the sense to ‘to heal’ is never used with disease or sickness as the direct object.

[25] Cf. Held, H. J., ‘Matthew as Interpreter of the Miracle Stories’, in Bornkamm, G., Barth, G., Held, H. J., Tradition and Interpretation in Matthew (London, 1963), p. 250 f.Google Scholar

[26] With θεραπεύω, see Matt. 4. 24; 8. 16; 10. 8; 12. 15, 22; 14. 14; 15. 30; 17. 16, 18; 19. 2; 21. 14;with ίάομαι, see Matt. 8.8, 13; 13. 15; 15. 28.

[27] See Schürmann, , op. cit., p. 504Google Scholar; Schramm, T., Der Markus-Stoff bei Lukas (Cambridge, 1971), pp. 26 ff.CrossRefGoogle Scholar

[28] This is denied by Hoffmann, P., Studien zur Theologie der Logienquelle (Münster, 1971), p. 245Google Scholar, on the grounds that the phrase is too common. But he does not note the unusual syntax.

[29] The word occurs only in these three Matthean passages in the NT.

[30] See Schürmann, , op. cit., p. 561Google Scholar for this meaning; also Plummer, A., The Gospel according to St. Luke (Edinburgh, 1898, p. 252Google Scholar: ‘a luminous cloud cannot overshadow; but it may veil’. Cf. too the remarks of Neirynck, , ‘Minor Agreements of Matthew-Luke in the Transfiguration Story’, in Hoffmann, P. (ed.), Orientierung an Jesus (FS for J. Schmid, Freiburg, 1973), pp. 253–66, on p. 264Google Scholar, who says that the cloud here is an ‘involving reality’. Goulder cites Neirynck here, and asks if the overshadowing of the Holy Spirit in Lk. 1. 35 was not also an involving reality. However, this is more metaphorical than a literal ‘envelopping’ by a cloud.

[31] τότε, λέγοντες, έμπτύειν είς, οί δέ = ‘others’, sarcastic use of ό Χριστος.

[32] E.g. in his account of the death of John the Baptist, and the question of the rich young man (Matt. 14. 3 ff.; 19. 16 ff.). See Styler, G. M., ‘The Priority of Mark’, in Moule, C. F. D., The Birth of the New Testament (London, 1962), p. 228 f.Google Scholar

[33] Streeter, , op. cit., pp. 325–8Google Scholar; cf. too Schmid, , op. cit., p. 159Google Scholar; McLoughlin, S., ‘Les accords mineurs Mt-Le contre Mc et le problème synoptique’, E.T.L. 43 (1967), pp. 17–40, on pp. 31–5Google Scholar. Further literature in Neirynck, Minor Agreements, ad loc.

[34] See Gundry, R. H., ‘I Q Isaiah a 50,6 and Mark 14,65’, R.Q. 2 (1960), pp. 559–67, on p. 563 f.Google Scholar

[35] For the shorter text, see too Taylor, , op. cit., p. 570 f.Google Scholar; Schneider, G., Verleugnung, Verspottung und Verhör Jesu (Munich, 1969), p. 39Google Scholar; Catchpole, D. R., The Trial of Jesus (Leiden, 1971), p. 175.Google Scholar

[36] So Streeter, , op. cit., p. 327.Google Scholar Goulder says that this is ‘unexampled in the NT’ (p. 228).

[37] Juel, D., Messiah and Temple. The Trial of Jesus in the Gospel of Mark (Missoula, 1977), p. 77 f.Google Scholar

[38] It may be that Jesus is here being asked to prove his messianic claims by showing an ability to act without seeing or hearing, and so satisfying the conditions implied in Isa. 11. 3; see Catchpole, , op. cit., p. 175 f.Google Scholar, with reference to the similar test applied to bar Cochba (b.Sanh.93b). Lk. 22. 63 f. may thus be part of a pre-Lukan source which Luke, then misunderstands in his summary verse 65 (cf. Catchpole, , op. cit., p. 183Google Scholar; for v. 65 as LkR, see too Schneider, , op. cit., p. 103 f.).Google Scholar

[39] 0171 a b e etc. See Streeter, , op. cit., p. 323Google Scholar; Schneider, , op. cit., pp. 54, 95 f.Google Scholar; Catchpole, , op. cit., p. 169.Google Scholar

[40] McLoughlin, , op. cit., p. 29Google Scholar; cf. Taylor, V., Behind the Third Gospel (Oxford, 1926), p. 50.Google Scholar

[41] See Catchpole, , op. cit., pp. 160 ff.Google Scholar

[42] Taylor, , Mark, p. 601, citing Abbott.Google Scholar

[43] It may also be implied by Matthew: contra Goulder, Matthew may be intending to describe events happening on the Saturday evening (cf. Black, M., An Aramaic Approach to the Gospels and Acts (Oxford, 1967), pp. 136 ff.Google Scholar; Schweizer, , Matthew, , p. 523).Google Scholar This would then involve giving όψέ its normal meaning. The fact that Matthew has totally rewritten Mark here means that one cannot assume that he is simply paraphrasing Mark; thus one may not simply transfer Mark's meaning to Matthew's text.

[44] Cf. Schmid, , op. cit., p. 165 f.Google Scholar

[45] Cf. Suggs, M. J., Wisdom, Christology and Law in Matthew's Gospel (Cambridge, Mass., 1970), passim.CrossRefGoogle Scholar I have examined this text in more detail in my, The Revival of the Griesbach Hypothesis (Cambridge, 1983), pp. 157 ff.Google Scholar