100 Timo Flink that matter. Luke 9,35 is about the event on the mount of transfiguration, not about the testimony of John the Baptist as John 1,34. Why rewrite only this instance in the Fourth Gospel when other verses could as eas- ily be harmonized to the style of the Third Gospel? Thus, it does not seem probable that the scribe harmonized John 1,34 to Luke 9,35. This is supported by the fact that though some intentional changes occur in P75, the scribe did not systematically correct his text at all38. Therefore I submit that the scribe saw the singular reading υ ς κλεκτ ς in his exemplar, copied it, felt awkward about it, and corrected the text to a better attested reading. P106 as a Textual Witness There is no question that P66 and P75 are witnesses for the B-text but what about P106? Comfort states that it usually aligns with P66, P75, ) and B.39 Barbara Aland has listed some of the variant readings found in P106. What follows is a detailed comparison with other witnesses to find out its textual affinities or text-type40. I selected P66 P75 B C to present the B-text, ) b e to present the D-text, Ws Θ f1,13 to present the C-text and Byz to present the A-text. The following table contains all the variant read- ings against the selected manuscripts. It should be noted that since P106 contains only John 1,29-35a and 1,40b-1,46a with lacunae, this analysis gives only clues to its textual type41. In the following table the abbrevia- tion rell means “the rest of the studied manuscripts”. In case of the name Joseph in John 1,45 in P66, the in-process correction made by the original scribe is counted as P66*, not as P66c. 38 Colwell, “Method in Evaluating Scribal Habits”, 121; Royse, Scribal Habits in Early Greek New Testament Papyri, 538-39. 39 Comfort and Barrett, The Text of the Earliest New Testament Greek Manuscripts, 645. Comfort also argues that P106 concurs 8 out of 10 times with P66, P75 and Codex Sinaiti- cus. The same figure is 9 out of 10 times for Codex Vaticanus, making it the closest ally to P106. It is difficult to know what exactly did Comfort measure as he does not give any details except the numbers, but his results seem a bit misleading. In any case, I offer a more detailed analysis. See Comfort, Encountering the Manuscripts, 312. 40 B. Aland, “Kriterien zur Beurteilung kleinerer Papyrusfragmente des Neuen Testa- ments”, in A. Denaux (ed.), New Testament Textual Criticism and Exegesis. Festscrift J. Delobel (Leuven 2002) 1-13. The problem of establishing the text-type is related to the relatively short length of the text available. ) is Western in John 1-8 and Alexandrian elsewhere. See G.D. Fee, “Codex Sinaiticus 41 in the Gospel of John. A Contribution to Methodology in Establishing Textual Relation- ships”, NTS 15 (1968/69) 23-44.