The Style of Galatians 511 in 5,19-21 and 22 he gives lists of vices and virtues. Elsewhere Paul has used connectives in Galatians. One instance of repetition for the sake of clarity can be seen in repetition of the conjunction e[peita in Galatians 1,18.21; 2,1. Galatians does not make use of the type of dependent constructions Demetrius seems to find incompatible with the plain style, i.e., indirect discourse. In general it does use natural word order. However, the subject of the clause in 2,7-10 is not mentioned until v. 9, and the direct object precedes the subject in 3,8a.15.17; 5,1; 6,4.17. We have already seen that Galatians does not generally use long periods; neither does it use long clauses. Galatians does not completely avoid hiatus between long vowels and diphthongs as Demetrius recommends. In the first 7 verses of the letter I count four instances of hiatus between diphthongs, one instance of hiatus between a diphthong and a long vowel, and one instance of hiatus between long vowels. While Galatians does include many tropes and figures, in general they are not conspicuous. However, there are some exceptions. Most obvious are the paronomasia in 5,2-3 (wjfelhvsei – ojfeilevth") and 5,7-8.10 (peivqesqai – peismonhv – pevpoiqa). The latter includes one of Paul’s seven new coinages in Galatians. In these respects too, Paul may write an Asian plain style. One passage that is particularly vivid is Gal 1,12-2,14. Throughout the passage we find an abundance of precise, circumstantial detail concerning times, places and people. The verbs in this passage are generally past tense (one exception is 1,20). The passage gradually reveals the course of Paul’s life; in particular 2,6-10 gradually reveals the outcome of Paul’s meeting with James, Cephas and John. All of these things contribute to vividness. The anacoluthon in 2,4-5 might be seen as an instance of omitting something for the listeners to work out for themselves for the sake of persuasiveness. Or it might be seen as faulty composition in that a clause is abruptly broken off, perhaps because the thought is unpleasant. It is a little uncertain how the listeners were expected to complete the anacoluthon. Perhaps they were expected to supply something like “we were urged to do this, i.e., circumcise Titus” as the completion of the sentence (37). The anacoluthon in 2,6 does not require (37) Betz (Galatians, 90) and Longenecker (Galatians, 50) suggest that readers were expected to supply “Now this happened” as the beginning of the sentence in vv. 4-5. It seems more likely that the end of the sentence is missing than that the beginning is missing whether the anacoluthon is intentional or not.