³ tn Grk “a Jewish man” (ἀνδρὶ ᾿Ιουδαίῳ, andri Ioudaiw).

² tc ‡ Most mss have either ἀνθρώπῳ (anqrwpw, “for a man” [so א² C D W Θ 087 Ë^1,13 33 Ï latt]) or ἀνδρί (andri, “for a husband” [1424^c pc]) before the infinitive ἀπολῦσαι (apolusai, “to divorce”). The latter reading is an assimilation to the parallel in Mark; the former reading may have been motivated by the clarification needed (especially to give the following αὐτοῦ [autou, “his”] an antecedent). But a few significant mss (א* B L Γ 579 [700] 1424* pc) have neither noun. As the harder reading, it seems to best explain the rise of the others. NA²⁷, however, reads ἀνθρώπῳ here.

¹ tc Grk “the brother.” Later witnesses (א² D² Ï) have ἀνδρί (andri, “husband”) here, apparently in conscious emulation of the earlier mention of ἀνήρ (ajnhr) in the verse. However, the earliest and best witnesses (Ì⁴⁶ א* A B C D* F G P Ψ 33 1739 al co) are decisively in favor of ἀδελφῷ (adelfw, “brother”), a word that because of the close association with “wife” here may have seemed inappropriate to many scribes. It is also for reasons of English style that “her husband” is used in the translation.