118 Timo Flink cally polished formula, which is broken if πᾶσαν ψυχήν is chosen. Such a triplet is a common feature in the author’s Greek and fits well with the terse, picturesque and impassioned style of the author overall92. The chiastic structure supports the longer reading, because vv. 4 and 14-15 constitute a parallel pair, both reading ἀσεβεῖϛ if the longer reading is accepted. To conclude, the external evidence favours the longer reading while the internal evidence may go both ways. Thus, Jude 15 should read πάνταϛ τοὺϛ ἀσεβεῖϛ, which brings the text in harmony with the later Majority text and the Ethiopian tradition of 1 Enoch. Jude 18a This textual variation unit has to do with the second occurrence of ὅτι in the sentence ὅτι ἔλεγον ὑμῖν [ὅτι]. Should it be included (Wasserman) or excluded (ECM)? The editors of NA27 did not make a clear decision on this issue but placed the word in brackets. The ECM editors indicate by a bold dot that the inclusion of ὅτι2 is of equal value or that their exclude it without complete confidence93. The external evidence is as follows. omit ὅτι2 ὅτι2 ) B L* Ψ // – // 61 2344 // L:R Lcf P72 A C 81 88 307 326 431 436 442 453 808 1739 2200 // 18 33 35 323 621 623 630 665 915 1067 1409 1836 1837 1845 1852 1875 2374 // 5 6 93 254 468 1243 1292 1735 1846 1881 2186 2298 2805 2818 // pm L:V S:HPh Cyr Ephr The external evidence appears to be on the side of inclusion in terms of the amount of primary, secondary, tertiary and other witnesses. Both the “best” manuscript (codex 81) and the earliest witness (P72) support the inclusion, which is widespread, found in Egypt, Rome94, and Syria while the exclusion of ὅτι is found mainly in Egypt with some additional sup- port in Rome. Syriac witnesses do not necessarily support the inclusion, because their text may simply be in accordance with the Syriac usage to introduce a quotation and may therefore not give evidence as to the Greek J.D. Charles, “Literary Artifice in the Epistle of Jude”, ZNW 82 (1991) 111; Landon, 92 Text-Critical Study, 117-18. Clark, “Discourse Structure in Jude”, 137, has questioned whether four occurrences of the same word in a row is really a polished style but his re- marks do not change the trust of the argument in favour of the longer reading. ECM IV.4, 426. 93 Comfort notes that in Jude the text of P72 is more Western than Alexandrian. See 94 Comfort with Barrett, The Text of the Earliest New Testament Greek Manuscripts, 479.