¹ Rosché, Theodore R., The Words of Jesus and the Future of the “Q” Hypothesis, Journal of Biblical Literature. 79(1960), 210–20CrossRefGoogle Scholar; HONORS, A. M., A Statistical Study of the Synoptic Problem, Novum Teslamentum 10(1968), 95–147CrossRefGoogle Scholar. All pages cited below refer to these two articles.

² Nearly fifty years ago BULTMANN pointed out that the two-document hypothesis is basic to all further advance in gospel study; see his review of Fascher's, ErichDie formgeschichtliche Methode in Theologische Literaturzeitung 50 (1925), cols. 313-18Google Scholar. What he says there of Formgeschichte is certainly no less true of Redaktionsgescicte.

³ Listed in the order in which they occur in HUCK-ETZMANN: Lk. 3:1-2a, 15-16a; 6:12, 20, 22a; 5:12a; 7:3-6a; 5:17, 19; 9:3a; 6:5a, 12, 39a; 8:1-3; 10:23a; 13:18a, 22a; 8:40.

⁴ In the order in which they occur in HUCK-ETZMANN: Mtt. 5:1-2; 8:1; 9:1; 11:20; 13:31a, 33a; 9:18a.

⁵ In the order in which they occur in HucK-LiETZMANN: Lk. 3:5-6; 4:17-21, 23, 25-30; 6:21b; 12:57; 6:34-35; 11:36; 6:44b, 45c; 7:30; g:60b; 5:39; 10:1, 7-9; 12:52; 7:20-21, 29-30; 14:5; 6:8; 11:16, 18b, 21-22; 8:51-52a, 55-56, plus of course all entire sections (as, e.g., sect. 3) which are peculiar to Luke.

⁶ In the order in which they occur in HUCK-LTZMANN: Matt. 3:2, 4, 14-15; 4:13-16; 13:56, 58; 5:4-5, 7-, 14, 16, 21-24, 31, 8-39a, 41, 43, 45; 6:34; 7:12b, 15-17, 9-21b; 9:13a-15b; 10:5-8, 16b, 36; 11:14-15; 12:5-7, 11-12a; 15:14a; 12:23b, 32C, 36, 45b; 13:14-15; 9:21; 13:23; 14:2c; 16:17-19; 18:4, 10, 16-20, plus of course all entire sections (as, e.g., sect. 23) which are peculiar to Matthew.

⁷ In the order in which they occur in HUCK-LIETZMANN: Lk. 3:18; 4:15, 41; 9:36, 45; 14:35b.

⁸ In the order in which they occur in HUCK-LIETZMANN: Matt. 8:17; 13:26; 17:6-8.

⁹ As, e.g., Lk. 4:33-37.

¹⁰ As, e.g., Matt. 4:18-22.

¹¹ Thus, e.g., Matt. 8:14-15 is counted in sect. 13, but not in sect. 47.

¹² Matthew's παραλυτικόν is the equivalent of Luke's ἃνθρωπος δς ἣν παραλελνμένος but it is one word instead of four. Hence the difference in the number of ”common” words.

¹³ P. 97.

¹⁴ Ibid.

¹⁵ Proof of this fact can easily be obtained from the tables in Appendix B of Honoré's article (pp. 140-44). Matt.-Lk. ”agreements” (in his restricted sense) amount to 2489/8336 or 29.9 per cent, of Matthew and 2489/7884 or 31.6 per cent, of Luke in the Triple Tradition. In the Double Tradition, the figures are 749/4461 or 39.2 per cent, for Matthew and 1749/4476 or 39.1 per cent, for Luke. This is an average of 30.7 per cent, in the Triple Tradition and 39.1 per cent, in the Double Tradition. Expressed as a percentage of the Triple Tradition, the Double Tradition is, therefore, 39.1 — 30.7/30.7 or 27.4 per cent, more conservative than the Triple Tradition. If our tables are used, however, — and thus our definition of ”agreement,” — the Double Tradition is 71.0—56.0/56.0 or 26.8 per cent, more conservative than the Triple Tradition. (See Table Seven for these figures.) By either method, in other words, the Double Tradition is approximately 27 per cent, more conservative than the Triple Tradition.

¹⁶ Jerēmias, Joachim, Zur Hypothese einer schriftlichen Logienquelle, Zeitschriftfur die neutestamentliche Wissenschajt 29 (1930), 147–49Google Scholar.

¹⁷ Wrege, Hans-Theo, Die Ueberlieferungsgeschichte der Bergpredigt (Tübingen, 1968)Google Scholar. For a review of WREGE'S work, see CARLSTON, C. E. in the Journal of the American Academy of Religion 38(1970), 104–06CrossRefGoogle Scholar and GOULDER, M. D. in The Journal of Theological Studies, n.s. 20 (1969), 599–602CrossRefGoogle Scholar.

¹⁹ Appeal is often made to the extraordinary feats of memory in third-century Babylonian synagogues or modern Arab universities to explain the tenacity of oral tradition “in the Semitic world,” i.e., presumably, first-century Palestine. If this is legitimate, religionsgeschichtlick gesprochen, it is surely no less legitimate to cite the ancient Chinese proverb, “The strongest memory is weaker than the palest ink!”

²⁰ Robinson's paper, The Johannine Trajectory, now available only in mimeographed form, will be included in a forthcoming collection of essays by ROBINSON and HELMUT KOESTER, to be published in both English and German in 1970.

²¹ Ernest Colwell pleads guilty to the phrase (used about ten years ago before the Society of Biblical Literature) and adds that the tendency “has been demonstrated time and time again” (letter dated Jan. 1, 1970). KENNETH CLARK points out that Erasmus' notes “appear with greater frequency at the beginning of Codex 2, but then gradually diminish — as is so often the case with a διωρθωτ⋯ς”; see his The Erasmian Notes in Codex 2, Studia Evangelica (Texte und Untersuchungen, Band 73 [1959]), 750Google Scholar.

²² His figures for standard deviation and coefficients of variance in the Double Tradition (p. 145) do not suggest any pattern of the kind suggested here, and even his mean percentages do not fit it very well. Note that both the present study and HONORE'S article are based on the order of sections in HUCK-LIETZMANN, not the original order of Q. Figures somewhat more meaningful than those cited above could be obtained by rearranging sections (sayings?) in Lukan order (on the assumption that Luke more commonly preserves Q order than Matthew does), and these figures, on the basis of a few sample tests, do not reflect any pattern, so far as we can tell. In view of these many uncertainties, it would seem to be important t o avoid overgeneralization.

²³ See, e.g., the material in STREETER, , The Four Gospels (London, 1924). 273Google Scholar.

²⁴ The reconstruction in Manson's, T. W.The Sayings of Jesus (London, 1949), 6Google Scholar amounts to about 4000 words in Luke, which MANSON thinks preserves ”substantially all that Q ever contained” (ibid.).

²⁵ Above, note 15.