Jump To ECM Entry Online Commentary Introduction
GC: There is a coherent attestation with a small core of A-related witnesses for a, most notably 01, 03, and 81. The Kr witness 35 is at the top of a large and coherent strand of Byzantine witnesses including lectionaries. This shows an impressive continuity through all stages of the transmission. There is no other variant with equally strong support. The other part of the Byzantine tradition, including 18, the other Kr representative, supports c and is clearly derivative of a.
TP: As the context (cf. 11:30 and 13:1) clearly requires Barnabas and Saul to have returned from Jerusalem, a is lectio difficilior. All other variants can be explained as removing the difficulty by replacing εἰς with ἐξ or with ἀπό, or by building on this step (f-h).
Although it is clear that we have to accept a as initial text here, it remains a problem to interpret it. Hort (94) conjectured that τήν had to be read before εἰς, because it would be natural to take εἰς Ἰερουσαλήμ with ὑπέστρεψαν, not with πληρώσαντες τὴν διακονίαν, where it would make sense (“having fulfilled their ministry in Jerusalem”). This solution is already suggested by Wendt (277), although he does not insert τήν, but simply places a comma before εἰς. This, in fact, seems to reflect the understanding of the scribes who transmitted the difficult reading so faithfully until late Byzantine times.
 35* supports a. A corrector inserted εἰς ἀντιόχειαν, with the obvious intention to establish variant f. He forgot, however, to change εἰς before Ἰερουσαλήμ to ἀπό. A second corrector expunged the addition and consciously restored the difficult reading of the first hand.
 This holds true although the relationship of 35 to a potential ancestor within the a attestation (5 with ranking number 6) is relatively weak. The connectivity of this difficult variant is high, as is also shown by the correction back to the first hand text in 35 (see previous footnote).
 Haenchen (372), Ropes (114), Metzger (350-352) and Barrett (595-597) also favor this solution. Strutwolf (275-277) accepts this as a possibility, but in addition tries to make sense of taking εἰς Ἰερουσαλήμ with ὑπέστρεψαν by referring to 9:26-29. However, this would not alleviate the contextual tension. Pervo (316f.) deems a more forceful conjecture necessary (εἰς Ἀντιόχειαν ἀπὸ Ἰερουσαλήμ), but he does not explain how the other variants may have emerged from this perfectly acceptable reading.