¹ tc There is a textual difficulty with this verb. The Greek form uses r and not d, giving the name Ra‘oul. There is even some variation in the Hebrew traditions, but BHS (following the Leningrad codex of a.d. 1008) has preferred the name Deuel.

¹ sn In ancient Judaism Ezra and Nehemiah were regarded as a single book with dual authorship. According to the Talmud, “Ezra wrote his book” (b. Bava Batra 15a). The Gemara then asks and answers, “And who finished it? Nehemiah the son of Hacaliah.” Accordingly, the two are joined in the Leningrad Codex (ca. A.D. 1008), the manuscript upon which modern printed editions of the Hebrew Bible (e.g., BHK and BHS) are based.

² tc The MT reads simply “the good,” but the Greek versions read “the good and the bad.” In contrast to the other four pairs in v. 2 (“the righteous and the wicked,” “those who sacrifice, and those who do not sacrifice,” “the good man…the sinner,” and “those who make vows…those who are afraid to make vows”), the MT has a triad in the second line: לַטּוֹב וְלַטָּהוֹר וְלַטָּמֵא (lattov vÿlattahor vÿlattame’, “the good, and the clean, and the unclean”). This reading in the Leningrad Codex (ca. a.d. 1008) – the basis of the BHS and BHK publications of the MT – is also supported by the Ben Asher text of the First Rabbinic Bible (“the Soncino Bible”) published in a.d. 1488-94. On the other hand, the Greek version in B (Aquila) has two pairs: τῷ ἀγαθῷ καὶ τῷ κακῷ, καὶ τῷ καθαρῷ καὶ τῷ ἀκαθάρτῳ (“the good and the bad, and the clean and the unclean”). Either Aquila inserted καὶ τῷ κακῷ (kai tw kakw, “and the bad”) to fill out a pair and to create six parallel pairs in v. 2, or Aquila reflects an early Hebrew textual tradition tradition of לַטּוֹב וְלַרָע (lattov vÿlara’, “the good and the bad”). Since Aquila is well known for his commitment to a literal – at times even a mechanically wooden – translation of the Hebrew, with no room for improvisation, it is more than likely that Aquila is reflecting an authentic Hebrew textual tradition. Aquila dates to a.d. 130, while the Leningrad Codex dates to a.d. 1008; therefore, the Vorlage of Aquila might have been the original Hebrew textual tradition, being much earlier than the MT of the Leningrad Codex. The alternate textual tradition of Aquila is also seen in the Syriac and Latin versions (but these are dependent upon the Greek = Aquila). On the other hand, the editors of BHK and BHS suggest that the presence of the anomalous לַטּוֹב was an addition to the Hebrew text, and should be deleted. They also suggest that the Greek pair τῷ ἀγαθῷ καὶ τῷ κακῷ (tw agaqw kai tw kakw, “the good and the bad”) does not reflect an alternate textual tradition, but that their Vorlage contained only לַטּוֹב: the Greek version intentionally added καὶ τῷ κακῷ (kai tw kakw, “and the bad”) to create a pair. The English versions are divided. Several follow the Greek: “the good and the bad, the clean and the unclean” (NEB, NAB, RSV, NRSV, NIV, Moffatt, NLT). Others follow the Hebrew: “the good and the clean and the unclean” (KJV, ASV, MLB, NJPS). None, however, delete “the good” (לַטּוֹב) as suggested by the BHK and BHS editors. If the shorter text were original, the addition of καὶ τῷ κακῷ would be intentional. If the longer text were original, the omission of וְלַרָע (“and the bad”) could have caused by unintentional homoioarcton (“similar beginning”) in the three-fold repetition of לט in וְלַרָע וְלַטָּהוֹר וְלַטָּמֵא לַטּוֹב (lattov vÿlara’ vÿlattahor vÿlattame’, “the good and the bad, the clean and the unclean”). The term וְלַרָע (“and the bad”) was accidentally omitted when a scribe skipped from the first occurrence of לט in לַטּוֹב to its second occurrence in the word וְלַטָּהוֹר (“the clean”).

¹ tc The textual problems in Hosea are virtually unparalleled in the OT. The Masoretic Text (MT), represented by the Leningrad Codex (c. a.d. 1008), which served as the basis for both BHK and BHS, and the Aleppo Codex (c. a.d. 952), are textually corrupt by all accounts and have a multitude of scribal errors. Many medieval Masoretic mss preserve textual variants that differ from the Leningrad and Aleppo Codices. The Qumran materials (4QXII^c,d,g) contain numerous textual variants that differ from the MT; unfortunately, these texts are quite fragmentary (frequently in the very place that an important textual problem appears). The textual tradition and translation quality of the LXX and the early Greek recensions (Aquila, Symmachus, Theodotion) is mixed; in some places they are inferior to the MT but in other places they preserve a better reading. The textual apparatus of BHK and BHS contains many proposed emendations based on the ancient versions (Greek, Syriac, Latin, Aramaic) that often appear to be superior readings than what is preserved in the MT. In numerous cases, the MT readings are so difficult morphologically, syntactically, and contextually that conservative conjectural emendations are necessary to make sense of the text. Most major English versions (e.g., KJV, ASV, RSV, NEB, NAB, NASB, NIV, TEV, NKJV, NJPS, NJB, NRSV, REB, NCV, CEV, NLT) adopt (either occasionally or frequently) textual variants reflected in the versions and occasionally adopt conservative conjectural emendations proposed in BHK and/or BHS. However, many of the textual problems in Hosea are so difficult that the English versions frequently are split among themselves. With this in mind, the present translation of Hosea must necessarily be viewed as only preliminary. Further work on the text and translation of Hosea is needed, not only in terms of the NET Bible but in Hosea studies in general. The text of Hosea should be better clarified when the Hebrew Old Testament Text Project completes work on the book of Hosea. For further study of textual problems in Hosea, see D. Barthélemy, ed., Preliminary and Interim Report on the Hebrew Old Testament Text Project, 5:228-71.