113 Reconsidering the Text of Jude 5, 13, 15 and 18 Jude 13 This textual variation unit has to do with whether the text should read ἐπαφρίζοντα (the NA27, the ECM) or ἀπαφρίζοντα (Wasserman) followed by τὰς ἑαυτῶν αἰσχύνας66. There are no bold dots in the ECM, so the editors believed their choice is beyond reasonable doubt. There are no text-critical umlauts for this verse in codex B either. The external evidence is as follows. ἐπαφρίζοντα ἀπαφρίζοντα P72 C 81 326 431 442 1739 2200 // 33 323 ) A B L Ψ 88 307 436 453 808 // 18 35 621 623 630 665 1837 1852 // 5 61* 93 254 915 1067 1409 1836 1845 1875 2374 // 6 61c 468 1846 2186 2818 // pm BasRt 1243 1292 1735 1881 2298 2344 2805 // pc Ephr Pall Phot PsOec ThdSt67 CnOec68 Isid The external evidence is quite evenly divided. Though both the earliest witness (P72) and the “best” witness (codex 81) support ἀπαφρίζοντα (“to skim”, “to scum”), the rival reading ἐπαφρίζοντα (“foam”) is also attested by early witnesses () A B Ephr). Both readings go back to the second century exemplars, ἀπαφρίζοντα via P72 and ἐπαφρίζοντα via codex B. Geographical distribution is also evenly divided as both readings are found in Egypt and Syria. It is possible that this is a case of phonetic confusion but that does not tell which way it went. The P72 has such ir- regularities, though in none α is taken for ε (unless this is the only excep- tion), though vice versa has taken place once in 2 Pet. 3,1469. The external evidence is rather inconclusive, giving only a tiny edge for ἀπαφρίζοντα on the basis of the P72 81 pair. However, G. Mink argues that ἐπαφρίζοντα has a perfect genealogical coherence while ἀπαφρίζοντα emerged repeat- edly by coincidence, even if an average connectivity between genealogical states is assumed, which Mink doubts70. Thus, the decision needs to be made on internal grounds. A suggested third variant reading metafri/zonta presumably supported by 429 in 66 Kubo, P72 and Codex Vaticanus, 87; Landon, A Text-Critical Study, 109, is incorrect as the minuscule reads ἐπαφρίζοντα. This is noted by Wasserman. Theodorus Studites, Parva Catechesis 25.28, reads ἐπαφρίζοντα τὰς ἑαυτῶν 67 αἰσχύνας. Concilia Oecumenica, Concilium Lateranense a. 649 celebratum 6.406.25, reads 68 ἀπαφρίζοντα τὰς ἑαυτῶν αἰσχύνας. Albin, Judasbrevet, 611; J.R. Royse, Scribal Habits in Early Greek New Testament 69 Papyri (Ann Arbor, MI 1981) 473-74. G. Mink, private communications with the author, July 26, 2007. I remain sceptical 70 about this, because too many early witnesses have been lost. It is not certain that the now genealogically unrelated witnesses really were so in the past.