Introduction
1. See further below, pp. 16–21.
2. Pekah’s twenty-year reign (2 Kgs 15:27) may include a period of rule only over Gilead (Thiele 1994: 53); cf. Dearman 2010: 23–24. Reference to Israel and Ephraim alongside Judah (5:5) may reflect division in the northern kingdom.
3. Wolff (1974: 111, 124) links it with Hoshea’s assassination of Pekah. See comments below.
4. The northern kingdom is frequently referred to as ‘Ephraim’, the name of its largest tribe.
5. Alt relates 5:8 – 6:6 to a Judean incursion into Israel during the Syro-Ephraimite war. See also Mays 1969: 86–88; Wolff 1974: 111–113; M. E. W. Thompson 1982: 63–78; Stuart 1987: 101; Hubbard 1989: 118–121; Macintosh 1997: 194–198. Arnold (1989: 447–460) sees, instead, a reference to the Syro-Ephraimite advance towards Jerusalem. Dearman (2010: 179–181) also suggests a Syro-Ephraimite background but does not specify the direction of attack. See also Good 1966a; Garrett 1997: 148–149; Sweeney 2000: 60–65; Ben Zvi 2005: 140–142; Bos 2013: 85–87; Moon 2018: 108–109.
6. Though this may not have been his only, or primary, motive (Ehrlich 1996: 93; Na’aman 2005: 30–31).
7. Tiglath-Pileser records the death of Pekah and his installation of Hoshea as king (COS 2.288, 291); cf. 2 Kgs 17:3.
8. King So of Egypt (2 Kgs 17:4) may refer to Pharaoh Osorkon (Kitchen 2003: 15–16); though see J. Day 1992a: 289–301.
9. On the fall of Samaria, see Na’aman 1990; 1993; Becking 1992; Younger 1998: 201–227; 1999; Bergland 2011–12; Park 2012. Cf. Dearman 2010: 27–29.
10. ‘In the year that King Uzziah died’ (Isa. 6:1); see Thiele 1994: 12. This is often understood as a reference to Isaiah’s call.
11. There are frequent references to ‘Baal’ (e.g. Judg. 2:13; 6:25–32; 1 Kgs 16:31–32; 18:19; 2 Kgs 17:16; 21:3; Jer. 2:8; 7:9; 11:17; 12:16; Zeph. 1:4) and ‘the Baals’ (e.g. Judg. 2:11; 3:7; 8:33; 10:6; 1 Kgs 18:18; 2 Chr. 24:7; 28:2; Jer. 2:23; 9:14), as well as specific references to ‘Baal of Peor’ (Num. 25:3, 5; Deut. 4:3; Ps. 106:28; cf. Hos. 9:10), ‘Baal-Berith’ (Judg. 8:33; 9:4) and ‘Baal-Zebub’ (2 Kgs 1:2–6, 16).
12. The request for Anat to ‘pour well-being out into the earth’ is a frequent refrain in Baal and the Sea.
13. Hess (2007: 14–15) suggests that Baal may refer to ‘the chief male spirit of a region’; cf. M. S. Smith 2002: 79.
14. Associated with hieros gamos was the view that women were required to have sexual relations with a stranger, usually before marriage, as part of a ritual ensuring fertility (Wolff 1974: 14–15, 86–87), though this was almost certainly not part of Israel’s worship; cf. Westenholz 1989: 261; Macintosh 1997: 123–125; Abma 1999: 14–15; Marsman 2003: 497–498; Kelle 2005: 132–135.
15. Some suggest that qĕdēšâ does not necessarily suggest sexual activity; see Westenholz 1989: 248; Keefe 1995: 81 n. 6; Miller 2000: 205–206; Marsman 2003: 497–498; Ringgren 2003: 542–543; Kelle 2005: 123–132; Yee 2012: 209; Lim and Castelo 2015: 123.
16. Tamar is described as a zônâ and a qĕdēšâ (Gen. 38:15, 21–22), and the terms appear together in Deut. 23:17–18 where both seem related to prostitution (see Craigie 1976: 301; van der Toorn 1989: 203).
17. Some see little or no formal connection to the cult (van der Toorn 1989: 203; Gruber 1995: 176–177), though van der Toorn suggests that qĕdēšîm may have engaged in prostitution as a source of temple funds.
18. It could also mean ‘my husband’, though that does not fit in this context.
19. 2 Sam. 2:8 (cf. 1 Chr. 8:33); 1 Chr. 8:34 (cf. 2 Sam. 4:4); 1 Chr. 14:7 (cf. 2 Sam. 5:16).
20. Baal and the Sea describes Lotan as ‘the fleeing serpent . . . the twisting serpent, the close coiling one with seven heads’ (COS 1.265). In Isa. 27:1, Leviathan is referred to as ‘the fleeing serpent . . . the twisting serpent’ (nrsv). Ps. 74:14 further refers to Leviathan’s several heads. See Emerton 1982; J. Day 1985: 4–5; M. S. Smith 2002: 86; Anderson 2015: 93.
21. Baba Batra notes that Hosea’s ‘scroll is so small that if copied on its own it might get lost’ (b. B. Bat. 14b).
22. Sirach 49:10 refers to ‘the Twelve Prophets’. Baba Batra’s list of canonical texts also includes ‘the twelve prophets’ (b. B. Bat. 14b).
23. E.g. House 1990; Nogalski 1993a; 1993b; Nogalski and Sweeney 2000; Redditt and Schart 2003; Kakkanattu 2006: 181–183; Albertz, Nogalski and Wöhrle 2013; see also Lim and Castelo 2015: 27–30; Wenzel 2018. Seitz (2007: 189–247) argues that interpretation needs to take into account the individual books’ historical context and place within the canon.
24. E.g. Ben Zvi 1996: 125–156; 2005: 6–7; Petersen 2000: 3–10; Hadjiev 2010: 325–338; 2020: 90–108.
25. E.g. Joel 1:15; 2:1; Amos 5:18; Obad. 15; Zeph. 1:7, 14; Mal. 4:5; see also Mic. 2:4; 4:6; 5:10–15; Hag. 2:21–23; Zech. 2:10–12; 12:1–9.
26. There is a suggestion that the picture in Hos. 11:1–4 may point instead to God as mother (Schüngel-Straumann 1995), though this seems unlikely; see on 11:4 below.
27. Examples include ‘the Lord will roar from Zion and thunder from Jerusalem’ (Joel 3:16; Amos 1:2), being ‘cast’ (šālak) into the ‘depths’ (mĕṣûlâ) of the ‘sea’ (yām) (Jon. 2:3; Mic. 7:19) and the use of Exod. 34:6–7 in Joel 2:13; Jon. 4:2; cf. Mic. 7:18–19; Nah. 1:3.
28. The mt and lxx order the books differently. Hosea is first in both and the last six are in the same order. This does not affect House’s overall structure.
29. The same expression ‘the word of the Lord came to’ (dĕbar yhwh ʾšer hāyāh ʾel) introduces Hosea, Micah and Zephaniah (though also appears in Joel 1:1). Amos and Hosea both refer to northern and southern kings (Uzziah and Jeroboam). See also Lim and Castelo 2015: 43–44.
30. Albertz (2003: 250) suggests that Hosea served as ‘the model that shaped the concept and the structure of the Book of the Four’.
31. Andersen and Freedman 1980: 75; J. Day 1986a: 1–12; Dearman 2010: 39; Robson 2012: 48–49; see also Nogalski 1993a: 278–279; 1993b: 274–275; Albertz 2003; Radine 2013: 287–302; Vielhauer 2013: 55–75.
32. E.g. 2 Chr. 31:5; 32:28; Neh. 5:11; 10:39; 13:5; Jer. 31:12; Joel 2:19; Hag. 1:11; cf. Gruber 2017: 128.
33. Yee suggests a double redaction: during the reign of Josiah and again during the Babylonian exile.
34. This is most commonly associated with the so-called ‘Deuteronomistic History’, comprising the historical books Joshua–2 Kings. See further, Routledge 2016: 253–257; see also Campbell and O’Brien 2000: 1–37.
35. Some also date the idea of God’s relationship with Israel as a covenant from this period. See, however, J. Day 1986a; Nicholson 1986: 179–189; Routledge 2008a: 160–163; 2016: 120–123; Ansberry and Hwang 2013: 77.
36. Most recent scholars date Deuteronomy in the late seventh century bc. However, see Craigie 1976: 22–24; McConville 1984; 2002: 21–38; Wenham 1985a; 1985b. See also Vang 2011.
37. It has also been suggested that, while sharing Deuteronomic ideas, Hosea’s conclusions are not the same as those of the ‘Deuteronomistic history’ (Albertz 2003: 247–251; Vielhauer 2013: 60). Wilson (1980: 227–228) links Hosea with an Ephraimite tradition with affinities with Deuteronomy and Jeremiah.
38. Wolff suggests that the book was put into its final form by Deuteronomistic redactors.
39. This expression does not occur in other prophetic books.
40. See the discussion of ḥesed below, pp. 31–32.
41. On šûb, see also below, p. 24. Lalleman (Lalleman-de Winkel 2000: 91–115; Lalleman 2013: 37–40) notes that for Jeremiah šûb relates to a return to the land, while Hosea focuses primarily on the restoration of the relationship. This reflects their respective historical contexts.
42. The verb yādaʿ occurs several times through Hosea. It may refer, as here, to the knowledge of Yahweh. It is also linked with Israel’s failure to acknowledge the source of their blessings (2:8; 11:3) or their own weakness (7:9). The related term daʿat, also referring to the knowledge of Yahweh, occurs in 4:1, 6; 6:6. See Fretheim 1996.
43. There are also questions about the Deuteronomistic editing of Jeremiah. See Routledge 2016: 294–295; see further, McConville 1993: 12–23; Lalleman 2013: 28–37.
44. The reference to Judah in 11:12 is interpreted both positively (nrsv, esv) and negatively (niv). See discussion in the commentary.
45. Wolff (1974: xxxi) suggests that the positive view of Judah arose in Hosea’s later ministry, and the threat of judgment on Judah is part of the Judean redaction; see also Emmerson 1984: 88–95. Childs (1979: 379–380, 381) considers that the message of judgment on Judah came first, and the more positive view emerged later.
46. However, Gruber (2017: 268–269), following Ginsberg, argues that in all of the texts which parallel Ephraim with Judah (5:5, 12, 13, 14; 6:4; 10:11), ‘Judah’ replaces an original ‘Israel’. Ephraim–Israel is the usual word pairing and has been changed by a Judean redactor after the fall of Samaria.
47. Emmerson (1984: 68–74, 77) takes references to Judah in 5:10–14 as original, though sees 6:4 as part of a Judean redaction. Most commentators, though, interpret 6:4 in its present context; cf. Mays 1969: 96–97; Wolff 1974: 119; Macintosh 1997: 228–229; Lim and Castelo 2015: 134.
48. Some regard this as original (Wolff 1974: 24–29; Emmerson 1984: 95–98); others take it to reflect Hosea’s thought, but come short of direct attribution (e.g. Mays 1969: 31; Macintosh 1997: 33–35).
49. It occurs in later texts where both terms refer to the surviving southern kingdom (Isa. 65:9; Jer. 5:20); see, however, Hos. 10:11; Mic. 1:5.
50. Emmerson (1984: 64–65) notes that the change is unnecessary, though still regards the evidence for it as conclusive.
51. For a defence of Hosean authorship of references to Judah, see Moon 2018: 13–18.
52. See the discussion below, p. 175.
53. Ben Zvi (2005: 6–16; 2006: 42–45) attributes the book to literati living in Yehud (Judah) during the Persian period. In his view Hosea exists only within the book but has been integrated into a reconstruction of Israel’s past. See also Bos 2013; Trotter 2001. Lim and Castelo (2015: 32) suggest ‘significant redactional additions’ to the original material during this period.
54. Yee (1987: 112–115) suggests a much more intrusive editorial role.
55. See the further discussion below, p. 22.
56. It is widely held that Hosea used a northern regional dialect (Seow 1992a: 292; Yoo 1999; Lim and Castelo 2015: 31; Kim 2018: 55–63); though see Andersen and Freedman 1980: 67–68; Macintosh 1997: liv. For a survey, see Kelle 2010: 317–321.
57. Kim (2018) gives a detailed defence of the mt. For commentary on the lxx text, see Glenny 2013.
58. The noun rîb, translated charge (niv) or ‘indictment’ (nrsv), occurs twice (4:1; 12:2); the associated verb appears in two pairs. In 2:2 it is translated ‘to rebuke’ (niv), ‘to plead with’ (nrsv); and in 4:4, ‘to bring a charge’ (niv), ‘to contend’ (nrsv). See further, Bracke 1996; Ringgren 2004. On covenant lawsuit, see Huffmon 1959; Nielsen 1978; VanGemeren 1990: 108, 400–402; Davidson 2010.
59. See above, p. 15 n. 42.
60. E.g. 1:4–5; 2:9–13; 4:3, 10; 7:12–13; 8:13–14; 9:7–9, 15–17; 10:14–15; 11:5–7; 12:2; 13:7–9, 15–16.
61. As noted above (pp. 20–21), each of the book’s three main sections ends with a message of hope.
62. See also discussion of ḥesed (below, pp. 31–32) and allusions to the exodus (below, pp. 33–34).
63. This formula comprises variations of ‘I will take you as my people, and I will be your God’ (Exod. 6:7); e.g. Lev. 26:12; Jer. 11:4; 24:7; 31:1, 33; 32:38; Ezek. 11:20; 37:23, 27; Zech. 13:9. See Rendtorff 1998.
64. Another suggestion (Davies 1993: 87–92; Moughtin-Mumby 2008: 215–224) is that Hosea became involved with a prostitute and their illicit relationship reflects Israel’s prostitution to Baal. However, the expression in 1:2, where lāqaḥ (‘to take’), lĕ (‘to’) and ʾiššâ (‘wife, woman’) appear together, usually relates to legitimate marriage (e.g. Gen. 24:3–4; 27:46; 28:2; Exod. 6:20, 23, 25; Deut. 21:11; Judg. 3:6). See also Kruger 1983: 107; Fensham 1984: 72; Els 1996: 814; Seebass 1997: 19; Routledge 2018: 32.
65. For a recent overview of the debate, see Routledge 2018; see also Rowley 1956; Macintosh 1997: 113–126; Kelle 2009: 179; Dearman 2010: 80–88.
66. See above, pp. 7–8.
67. Moughtin-Mumby (2008: 95–96) argues that Israel’s idyllic past is contrasted with her current unfaithfulness; see also J. A. Thompson 1980: 163; Sherwood 2004: 207–208; Lalleman 2013: 76. Fox (1973) argues that ḥesed is not shown by people to God and, in Jer. 2:2, must refer to God’s love for Israel. However, see Routledge 1995: 193–195. Ezek. 23 suggests that Israel’s promiscuity began in Egypt, though this is different from the picture in Hosea and Jeremiah and has a different rhetorical purpose.
68. See also van der Woude 1982: 44–45; Dorn 2000; Sweeney 2000: 15; Lim and Castelo 2015: 80.
69. Moon (2015; 2018: 70–71) argues that the primary rhetorical factor here, and in 1:2, is the disgrace attached to the man who marries such a woman. Yahweh willingly accepts that disgrace because of his continuing commitment to his people. See also Yee 2003: 46–48. The emphasis here, though, is on the sin and the shame of the unfaithful wife (2:5; 2:10; cf. 2:3). See Hadjiev 2012; 2016. For further discussion of honour–shame, see Bechtel 1991; Olyan 1996.
70. E.g. Setel 1985; Weems 1989; 1995; van Dijk-Hemmes 1996; Brenner 1996a; 1996b; 1997; Baumann 2003; Yee 2003; 2012; J. M. O’Brien 2008. See also Castelo (Lim and Castelo 2015: 229–233). For a response, particularly to Brenner and van Dijk-Hemmes, see Sloane 2008; cf. Garrett 1997: 124–133.
71. Other texts discussed in this context include Jer. 2 – 4; Ezek. 16; 23.
72. Setel (1985: 95) suggests that some ‘pornographic’ texts should not be described as ‘God’s Word’ in a public setting; Baumann (2003: 104) argues for the need to ‘confront the marriage imagery and its problems’ and ‘seek counter images’.
73. See e.g. Routledge 2008a: 102–103, 108–112.
74. It occurs proportionately more times only in Psalms and 2 Samuel.
75. For further discussion of ḥesed, see Routledge 1995; see also Glueck 1967; Sakenfeld 1978; Zobel 1986; Clark 1993; Britt 2003.
76. Clark’s analysis indicates that in around two-thirds of its occurrences, ḥesed is shown by God to human beings (1993: 49). Ḥesed is also linked with the relationship between king and subject (2 Sam. 3:8; Prov. 20:28; Isa. 16:5), between host and guest (Gen. 19:19; 21:23; Josh. 2:12) and between powerful and weaker members of society (Ps. 109:16).
77. Ḥesed may also be shown in the context of friendship (2 Sam. 3:8; Job 6:14), between relatives (Gen. 24:49) and between members of the covenant community (Mic. 6:8; Zech. 7:9).
78. Compassion (raḥămîm), here, may suggest divine activity, though may indicate an important aspect of the people’s renewed relationship with one another.
79. See above, p. 26 n. 63.
Chapter 1
1. evv render these other names ‘Hoshea’, though the Hebrew term (hôšēaʿ) is the same. Two (Num. 13:8; 1 Chr. 27:20) are from Ephraim, furthering speculation that Hosea was an Ephraimite. See Wolff 1974: 5.
2. It is unlikely that Beeri is a place in Ephraim.
3. See above, pp. 1–5.
4. See above, pp. 16–19.
5. This is the more widely accepted translation; see niv, nrsv, esv and many recent commentaries.
6. See above, p. 26 n. 64.
7. See above, pp. 27–28.
8. Gomer is sometimes linked with gāmar (‘to complete, bring to an end, fail’). Diblaim (diblāyim) has been taken as the dual form of dĕbēlâ, indicating a pressed-together fruit cake, and, based on the reference to raisin cakes (3:1), which are assumed to be part of Baal worship, to suggest Gomer’s involvement in the Baal cult (Keil and Delitzsch 1980: 38–39). See also Pressler 1992.
9. See above, pp. 23–24.
10. The names of Gomer’s second and third children relate to specific Hebrew terms and in the first instance are transliterated. Subsequent references take them as proper names (cf. nrsv).
11. See above, pp. 16–19.
12. See above, p. 26 n. 63.
13. This is also a major emphasis in Isa. 1 – 39; see e.g. Isa. 22:8b–11; 30:1–2; 31:1; cf. 7:9; 25:9; 28:16; 30:15. See also Johnston 2009: 104–121.
14. See Macintosh 1997: 33–35.
15. See above, p. 33.
16. Wolff (1974: 26) notes Israel’s relative insignificance on the world stage at this time.
17. The terms for children are different, though bĕnê zĕnûnîm (‘children of whoredom’, nrsv) is used in 2:4.
18. Andersen and Freedman (1980: 203) take it as the desert, where Israel was first described as God’s children (cf. Deut. 14:1). Mays (1969: 32) and Kelle (2005: 111–112) suggest Jezreel. For Moon (2018: 46) it is the land where they have been unfaithful but where they will be given an honourable name.
19. The expression in 1:10, ʾēl ḥāy, also occurs in Josh. 3:10; Pss 42:2; 84:2. Expressions using ĕlōhîm instead of ʾēl occur in Deut. 5:26; 1 Sam. 17:26, 36; 2 Kgs 19:4, 16; Jer. 10:10; 23:36; Dan. 6:20, 26. A related expression, ‘as the Lord lives’ (ḥay yhwh), appears in Hos. 4:15.
20. Aster (2012: 43–44) understands Jezreel geographically. By analogy with ‘the day of Midian’ (Isa. 9:4), the day of Jezreel points to the destruction of Jezreel and the self-reliance it represents, opening the way for restoration. The rhetorical link between Jezreel and Israel, though, makes it more likely that this is a positive expression analogous to the ‘day of the Lord’ (cf. 7:5).
21. The hope of salvation also immediately follows the threat of judgment in 11:5–11.
22. On the relationship between the church and Israel as the people of God, see Routledge 2013.
23. See above, p. 22.
24. Dearman regards the situation as unclear, though notes the absence of legal elements that might be expected to be present in divorce.
25. Though see above, pp. 7–8.
26. For discussion of what some see as the sexually abusive content of the verse, see above, pp. 29–30.
27. See above, p. 7.
28. On šûb, see above, p. 24.
29. On yādaʿ, see above, p. 15 n. 42.
30. See above, p. 13.
31. The New Moon festival is noted in Num. 29:6; 1 Sam. 20:24; Ezek. 46:6.
32. Similar lists relating to legitimate festivals occur in 1 Chr. 23:31; 2 Chr. 8:13; 31:3; Neh. 10:33; Ezek. 45:17. Yearly festivals translates ḥag, which usually refers to Israel’s pilgrim festivals: Passover/Unleavened Bread, Weeks and Tabernacles (see e.g. Exod. 23:14–17; Deut. 16:1–17).
33. See above, pp. 9–10.
34. The term translated pay (ʾetnâ) occurs only here. It is generally viewed as a variation of ʾetnan, which refers to a prostitute’s pay in Deut. 23:18; Ezek. 16:31; Hos. 9:1.
35. Ruining vines and fig trees symbolizes divine judgment in Ps. 105:33; Jer. 5:17; Joel 1:7; cf. Hab. 3:17.
36. See above, pp. 33–34.
37. See above, pp. 10–11.
38. On the use of hinnēh in biblical narrative, see Berlin 1994: 62–63; Routledge 2016: 166.
39. Similar expressions occur in Gen. 34:3; 50:21; Judg. 19:3; Ruth 2:13; 2 Chr. 30:22; Isa. 40:2.
40. ‘Speak to her heart’, corresponding to pātâ, may also include the sense of persuasion.
41. The covenant formula ‘you will be my people, and I will be your God’ (Ezek. 36:28) (see above, p. 26 n. 63) indicates that Ezekiel also anticipates a new covenant relationship (cf. Hos. 2:18).
42. ‘Achan’ is probably original, though appears as ‘Achar’ in 1 Chr. 2:7 and the lxx of Josh. 7, making the link with ‘trouble’ clearer. For further discussion see Hess 1994.
43. See above, pp. 33–34.
44. See above, pp. 9–10.
45. See above, p. 26.
46. This may be indicated by the absence of reference to sea creatures, which would not be a threat. Where the emphasis is on creation (4:3), the fish in the sea are included.
47. The niv does not include the repeated ‘to me’.
48. 2 Sam. 3:14 uses the same expression ‘betrothed [ʾēraśtî] to me [lî] for [b-]’ as in Hos. 2:19–20.
49. The terms ṣedeq and ṣĕdāqâ seem interchangeable; thus, ṣĕdāqâ appears in 2 Sam. 22:21, 25, while ṣedeq appears in the parallel passages in Ps. 18:20, 24. See Reimer 1996: 746.
50. See above, pp. 31–32.
51. For discussion of the key terms related to ḥesed, see Clark 1993: 108.
52. The Hebrew term here is ʾĕmet. This is closely related to ʾĕmûnâ and appears in similar contexts. See Moberley 1996.
53. See n. 49 above.
54. See above, p. 15 n. 42.
55. See above, pp. 5–6.
56. The term ʿānâ might refer to an answer to prayer (e.g. 1 Sam. 7:9; 8:18; 1 Kgs 18:24–26, 37; Isa. 58:9; Jon. 2:2) or a response to need (e.g. Isa. 41:17). Both are appropriate here.
57. Jezreel (yizrĕʿeʾl) and Israel (yiśrāʾēl) look and sound similar (cf. 1:4–5).
58. See above, p. 26 n. 63.
59. See above, pp. 28–29.
60. Moon, for example, speculates that Gomer might have died by this time (2018: 70).
61. Ben Zvi dates the whole book in the Persian period.
62. See above, p. 18.
63. The term ʾiššâ means, literally, ‘(a) woman’, though it may be translated ‘wife’. The niv your wife implies a reference to Gomer.
64. The mt verb is passive. Following the lxx, Wolff (1974: 56) reads it as active, ‘who loves’. This better fits the parallel with Israel’s love for raisin cakes; cf. Mays 1969: 54. See also Andersen and Freedman 1980: 296.
65. For rēaʿ as ‘lover’, see Song 5:16; Jer. 3:1.
66. Elsewhere, ʾăšîšâ on its own is translated ‘raisin cakes’ (2 Sam. 6:19; Isa. 16:7) or ‘raisins’ (Song 2:5). Here, the link with grapes (ʿēnāb) is specific.
67. The niv reflects this by adding sacred (another example of paraphrase rather than translation). 2 Sam. 6:19 may suggest a connection with Yahweh worship. In the context of Baal worship, there may be a comparison with ‘cakes’ (kawwānîm) offered to the ‘Queen of Heaven’ (Jer. 7:18) (Andersen and Freedman 1980: 298).
68. This translates the noun ʾăhābâ (once) and the verb ʾāhab (three times).
69. The form is unusual, and Macintosh (1997: 99–100) links it instead with ‘to recognize’, in the sense of recognizing a transfer of ownership. See also Gruber 2017: 170.
70. Kings and princes are mentioned together negatively in 7:3–7; 8:4, 10; 13:10–11; kings are further indicted in 5:1; 10:3–4. Israel’s sacrifices are condemned in 4:19; 6:6; 8:13; 9:4, and sacred stones in 10:1–2. Ephods and household gods are not mentioned elsewhere in the book.
71. On tĕrāpîm, see Motyer and Selman 1980.
72. Cf. 2:7. See also above, p. 24.
73. The verb here is pāḥad. The niv also translates ḥārad (11:10–11) and rĕtēt (13:1) in the same way.
Chapter 2
1. For discussion of the book’s structure, see above, pp. 20–21.
2. E.g. 1 Kgs 22:19; 2 Kgs 7:1; Isa. 1:10; 28:14; 66:5; Jer. 2:4; 7:2; 19:3; 21:11; 29:20; 42:15; Ezek. 13:2; 36:1.
3. See above, p. 22.
4. The niv adds you, maintaining the direct address to the people.
5. See comments on 2:19–20. Faithfulness (ʾĕmet) and acknowledgment (daʿat) are closely related to the equivalent terms used there. On ḥesed, see above, pp. 31–32.
6. E.g. Exod. 19:24; 2 Sam. 5:20; 6:8; Neh. 2:13; Job 16:14; Ps. 89:40; Eccl. 10:8; Isa. 5:5. The related noun refers to a ‘robber’ or other violent person (e.g. Ps. 17:4; Jer. 7:11; Ezek. 7:22; 18:10; Dan. 11:14).
7. E.g. Gen. 28:14; 30:30; Exod. 1:12; 1 Chr. 4:38; Job 1:10; Isa. 54:3.
8. Literally, ‘bloodshed touches bloodshed’, suggesting either that one act of bloodshed leads to another or that one act of bloodshed rapidly follows another.
9. Cf. Ezek. 18:10. Andersen and Freedman (1980: 338) link this with human sacrifice (see also Hubbard 1989: 98). Though primarily connected with Molech (e.g. Lev. 18:21; 2 Kgs 23:10; Jer. 32:35), human sacrifice is also linked with Baal (e.g. Jer. 19:5; 32:35; cf. 7:31), and Molech may have been included in the general category of ‘Baals’. On child sacrifice, see J. Day 2000: 210–212; Lange 2007. There is, though, no direct polemic against human sacrifice elsewhere in Hosea (see on 5:2; 8:13; 9:13; 13:2).
10. The expression tōhû wabōhû (‘formless and empty’) appears only in Jer. 4:23 and Gen. 1:2.
11. Hosea’s view of prophets seems generally positive (see on 6:5; 9:7–8; 12:10, 13). He would, though, have challenged ‘official’, state-sponsored prophets who opposed true prophets (1 Kgs 22:5–28; Jer. 14:13–14; 23:16–22) and whom he saw as part of Israel’s problem. Priests and prophets appear to have conspired against Jeremiah and are frequently condemned together (Jer. 2:8, 26; 5:31; 6:13; 14:18; 23:11).
12. See above, p. 15 n. 42.
13. See above, p. 24.
14. This may follow a later scribal correction, perhaps based on Ps. 106:20; Jer. 2:11.
15. The parallel with ʿāwôn (‘iniquity’, nrsv, esv) suggests the former, though the double meaning allows the wordplay.
16. See above, p. 24.
17. This has been linked to the futility of fertility rites associated with Baal worship, though, as noted above (pp. 7–8), it is unclear how much of that was part of the Baal cult.
18. Mays (1969: 72) suggests that zĕnût (‘harlotry’) both ends v. 10 and begins v. 11, but a copyist has omitted the repeated word.
19. E.g. Deut. 4:2; 5:1, 10; 6:2–3; 7:11–12; 8:1–2; 11:1; 12:1; 27:1.
20. Understanding here translates lēb (‘heart’), which in the OT is the seat of will and reason.
21. Mays (1969: 74) refers to ‘the intoxication of the mind brought on by desire for the gifts of the Baals’.
22. See above, pp. 27–29.
23. On Asherah, see J. Day 1986b: 385–408; 1992b; 2000: 43–67.
24. Num. 33:52; 1 Kgs 13:33; 14:23; 22:43; 2 Kgs 12:3; 14:4; 17:9–11; cf. 2 Kgs 18:4.
25. Oak (ʾallôn) suggests strength (Isa. 6:13; Ezek. 27:6; Amos 2:9; Zech. 11:2). Terebinth (ʾēlâ) is a large tree (Isa. 6:13) with leafy shade (Ezek. 6:13; cf. Judg. 6:11, 19; 2 Sam. 18:9). It is usually translated ‘oak’, except where it appears with ʾallôn (Isa. 6:13; Hos. 4:13). Poplar (libneh) occurs only here and in Gen. 30:37.
26. Some relate this to the activities of the people as a whole (Wolff 1974: 86–87; Macintosh 1997: 160–161), maybe as part of fertility rites (see above, pp. 7–8).
27. On references to Judah, see above, pp. 16–19.
28. The term ʾāwen may mean ‘deceit, nothing’; in 6:8; 12:11 it is better translated ‘sin, evil’.
29. The related expression ‘living God’ appears in 1:10.
30. This may compare with words from the Baal myth: ‘For Mighty Baʿlu is alive’ (COS 1.271); see Mays 1969: 78.
31. See above, pp. 23–24.
32. The term also appears in 8:4; 10:6; 13:2; 14:8.
33. The infinitive alongside the finite verb, zānâ (‘to practise prostitution’), expresses intensity.
34. [They] dearly love translates ʾāhăbû hēbû. The first term is from ʾāhab (to love), and hēbû is probably a dialectical variant or corruption of the same verb, with the repetition expressing intensity (as in the first part of the verse); see Wolff 1974: 73; Dearman 2010: 156; Moon 2018: 91. The mt reads ‘their shields’ (māgēn), which may refer metaphorically to protectors and so ‘rulers’. Some, with the lxx, read ‘pride, insolence’: ‘they love shame more than their pride’ (cf. Stuart 1987: 71; Dearman 2010: 156–157).
35. Reading šaḥăṭâ (‘slaughter’) as šaḥăt (‘pit’) and śēṭîm (‘rebels’) as šiṭṭîm (‘Shittim’).
36. On yādâ see above, p. 15 n. 42.
37. Moon’s suggestion that Israel, as a divorced wife who has become attached to another, may not return (2018: 101–102) seems unlikely; see on 2:2, 6–7. On šûb, see above, p. 24.
38. See above, p. 3 n. 2.
39. The sequence seek . . . [but] not find also occurs in relation to Israel’s ‘lovers’ (2:7).
40. For suggested emendations see Wolff 1974: 95; Stuart 1987: 88; Macintosh 1997: 189.
41. See also Hubbard 1989: 116. Ben Zvi (2005: 134) suggests the possibility of multiple meanings.
42. See above, p. 3 n. 5.
43. See above, pp. 16–19.
44. Beth Aven may be a derogatory reference to Bethel (see on 4:15) or a town to the east of Bethel (Josh. 7:2; 18:12).
45. On Benjamin, see Schunk 1992.
46. For further discussion, see above, p. 3 n. 5.
47. Good’s suggestion that this has cultic rather than military associations (1966a: 282–283) seems less likely.
48. The same expression (ʾaḥăreykā binyāmîn) occurs in the song of Deborah and Barak (Judg. 5:14) and may have become familiar as a battle cry.
49. If this related to Judah’s internal injustices (Ben Zvi 2005: 140), like would be unnecessary.
50. It is unlikely that this refers to the relatively limited encroachment by Judah.
51. Macintosh (1997: 207) prefers ‘emaciating disease’.
52. See above, pp. 2, 4.
53. This requires, too, that the -î is read as part of the particular Hebrew construction and not translated.
54. The term ‘great king’ is also attested in an Akkadian inscription from Sefire (COS 2.214).
55. Lion (šaḥal) and great lion (or ‘young lion’, nrsv, esv; kĕpîr) are synonymous terms.
56. Macintosh (1997: 216–219) notes the main options.
57. Some take the first acknowledge with v. 2 (e.g. Stuart 1987: 98; Dearman 2010: 191).
58. On šûb, see above, p. 24; on yādaʿ, see above, p. 15 n. 42.
59. Similarity between ‘dawn’ (šaḥar) and ‘earnestly seek’ (šāḥar, 5:15) suggests another link with the previous section.
60. Examples of x/x+1 parallelism include Ps. 62:11; Prov. 30:15b; Amos 1 – 2. Barré (1978: 129–135) argues that hāyâ and qûm are also a recognized parallel pairing and interpretation should focus on the whole expression.
61. J. Day (2000: 118–220) suggests that this reapplies imagery relating to Baal’s dying and rising; cf. Levenson 2006: 206–207. Here, though, the people die and rise, emphasizing that Yahweh, not Baal, is the source of his people’s life.
62. Barré (1978: 139) notes that a prognosis of recovery within three days, probably indicating a short time, is common in Akkadian medical, and other ANE, texts.
63. The removal of threats or a reversal from death to life is indicated in Gen. 22:4; 40:12–13; 42:18; Josh. 2:16; 1 Sam. 30:12; 2 Kgs 20:5; Jon. 1:17; see Russell 2008; see also Dempster 2014. Moon (2018: 115) suggests a link with Yahweh’s appearing on the third day on Mount Sinai (Exod. 19:11, 16).
64. See further, Dempster 2014. Moon (2018: 118–122) notes that several Jewish writings take 6:2 to suggest general resurrection.
65. For an overview of typology see Routledge 2008a: 43–47; see further, Baker 2010: 169–189.
66. Some suggest that elements of the typological understanding were known to the original writers; cf. Beale 2012. See also on 11:1.
67. See the discussion of ḥesed above, pp. 31–32.
68. Like Leviathan (see above, p. 9), Rahab appears in the context of God’s victory over the primaeval sea (Job 26:12; Ps. 89:9–10). See J. Day 1985: 6, 38–40; 1992d.
69. Another possibility, ‘like a man’ (cf. lxx), seems unlikely.
70. V. 8 is generally translated Gilead is a city of evildoers, though Andersen and Freedman read, instead, ‘in Gilead is the city [i.e. Adam] of evildoers’.
71. Evil translates ʾāwen (see on 4:15).
72. E.g. Deut. 30:3; Jer. 29:14; 31:23; 32:44; Ezek. 39:25; Joel 3:1; Amos 9:14.
73. The verb appears in the first expression; in the second, it is assumed.
74. See above, p. 15 n. 42.
75. Pātah also occurs with lēb in 2:14, though in a more positive context.
76. The niv translation, When I hear them flocking together, I will catch them, maintains the imagery of the fowler but is unlikely.
77. ‘To their assembly’ is emended by Andersen and Freedman (1980: 469) to ‘of their treaty’, and by Wolff (1974: 107) to ‘of their wickedness’, but these have little textual support.
78. Following the lxx, and reading gādad (‘to cut, slash’) for gārar (‘to gather’).
79. The niv has added appealing to their gods.
80. Here the noun rāʿâ and the adjective rāʿ mean the same.
81. See above, p. 24.
82. Further suggestions include ‘not to me’ (Wolff 1974: 108); ‘Not on High’, as a parody of ‘God on High’ (Garrett 1997: 173); ‘not what is above’ (Macintosh 1997: 284–285; cf. esv).
83. Andersen and Freedman (1980) further divide 8:1–14 into two sections (8:1–8, 9–14).
84. Garrett (1997: 181) likens the vulture, as something unclean, to the priesthood. Emmerson (1975: 704) emends nešer to naśśār, which, based on an Arabic term, she translates ‘herald’: ‘set the trumpet to your lips like a herald’. She then gives a cultic interpretation, linking pursuit by an enemy to covenant curses rather than a specific military threat.
85. Israel in the verse is awkward. It may be the subject of ‘to cry out’ (niv; cf. nrsv), or taken in apposition to ‘we’: ‘we, Israel, know you’ (nrsv, esv; cf. Emmerson 1975: 702; Macintosh 1997: 294; Moon 2018: 144). Andersen and Freedman (1980: 481, 490) suggest ‘God of Israel’. It is not in the lxx and may be a dittograph from v. 3, or a gloss (Dearman 2010: 216).
86. See also above, p. 15 n. 42.
87. It is less likely that there was such a calf in Samaria (see Dearman 2010: 224–226) or that the calf was moved there (Gruber 2017: 348). Cook (2004: 111) suggests that the reference to Samaria links the calf-cult to the Israelite king.
88. For discussion of possible translations, see Macintosh 1997: 308–310.
89. See above, pp. 2, 4. Irvine (1995) sets 8:8–10 at the end of Hoshea’s reign.
90. See above, p. 23.
91. The term translated lovers is ʾahab (‘love gift’); Macintosh (1997: 316) translates it ‘love affair’. It is probably better to repoint as ʾāhab (‘lovers’); cf. 2:5, 7, 10, 12.
92. The lxx, ‘they will cease to anoint kings and princes’ (cf. 3:4), reading māšaḥ (‘to anoint’) for maśśāʾ (‘burden’) (see Stuart 1987: 127), seems less likely.
93. For alternatives, see Andersen and Freedman 1980: 508–509; Macintosh 1997: 823–825.
94. Andersen and Freedman’s suggestion, they sacrifice ‘my loved ones’, indicating child sacrifice (1980: 501), is less likely (see on 4:2).
95. On šûb, see above, p. 24.
96. Some suggest that this is original to Hosea but placed here by a redactor (Wolff 1974: 146; Emmerson 1984: 74–77; Macintosh 1997: 332–333).
97. Garrett (1997: 188) notes that palaces were often heavily fortified.
98. Other passages which accuse the people of forgetting their Maker (e.g. Ps. 95:6–7; Isa. 17:7–10; 51:13) use the term primarily to assure them of his ability to redeem them.
99. The opening expression in the mt is, literally, ‘do not rejoice to rejoicing’. Following the lxx, ‘to’ is often emended to repeat ‘do not’ (cf. niv, nrsv, esv). Alternatively, the repetition may indicate exuberant rejoicing (Macintosh 1997: 337).
100. Garrett’s suggestion (1997: 189–190) that this refers to a failed harvest prefiguring judgment seems less likely.
101. This may suggest having sex on threshing-floors (Andersen and Freedman 1980: 523; Hubbard 1989: 156), which may have become significant within the fertility cult.
102. Andersen and Freedman (1980: 526) suggest that the term translated mourners may indicate a pagan deity and that this is bread offered to idols.
103. The term used here, nepeš, may be translated ‘throat, neck’, hence ‘hunger’ (nrsv, esv); more usually it refers to a ‘living being’ (cf. Gen. 2:7), and ‘their nepeš’ could mean ‘their lives’, i.e. ‘to keep them alive’, or simply themselves (niv).
104. Macintosh (1997: 343) sees this as part of a later redaction.
105. Num. 11:25–29; 24:2–3; 1 Sam. 10:6; 19:20; 2 Sam. 23:2; Neh. 9:30; Ezek. 2:2; 11:5; Zech. 7:12.
106. Isa. 20:1–6; Jer. 13:1–11; 19:10–13; 27:1–7; Ezek. 4; 5:1–4; 12:1–16.
107. The same term (ʿāwōn) occurs in 4:8; 5:5; 7:1; 8:13; 9:9; 10:10; 12:8; 13:12; 14:1–2.
108. The statement in Judg. 19:30 comes after the events at Gibeah, but before the reprisals against Benjamin (Judg. 20), and it is unlikely that those reprisals are its primary focus (contra Arnold 1989: 452–454). However, the allusion to the whole incident raises the issue of internecine conflict (Andersen and Freedman 1980: 534).
109. See above, p. 14.
110. See above, p. 123.
111. See above, pp. 27–28; cf. Eidevall 1996: 149–150.
112. Following the lxx, some suggest a parallel with the second part of the verse: ‘Ephraim . . . has made his sons a hunter’s prey’ (Mays 1969: 131; cf. Wolff 1974: 160), though this is unnecessary.
113. See above, p. 23.
114. The term here usually refers to breaking the neck of an animal (Exod. 13:13; Deut. 21:4; Isa. 66:3).
115. See above, p. 121 n. 87.
116. ‘Inhabitant’ is singular but is better understood collectively (niv: The people who live in Samaria). Some see ‘Resident/Inhabitant of Samaria’ as a title for the calf (Andersen and Freedman 1980: 556; Hubbard 1989: 174).
117. See above, p. 14.
118. Kābôd (‘glory’) and gālâ (‘to depart’) occur in both passages.
119. See above, p. 4.
120. A similar term in Joel 1:7 refers to a strip of wood; see also lxx; Mays 1969: 138; Hubbard 1989: 175; Garrett 1997: 211–212. ‘Wrath’ may be linked to ‘boiling’ or ‘bubbling’, hence the translation ‘foam’; see also Macintosh 1997: 406–407; Ben Zvi 2005: 213; Dearman 2010: 259; Moon 2018: 167.
121. The same term (ʾāwen) refers to ‘evil, sin’ in 6:8; 12:11.
122. Most see ʿalwâ (mt) as a copyist’s error and read instead ʿawlâ (‘injustice, wickedness’) (cf. 10:13).
123. See above, pp. 16–19.
124. The terms ṣedeq and ṣĕdāqâ appear interchangeable; see above, p. 63 n. 49.
125. Planted (niv) translates ḥāraš, ‘to plough’ (nrsv); cf. v. 11. Rešaʿ is contrasted with righteousness in Job 35:8; Pss 45:7; 125:3; Ezek. 33:12.
126. Kāḥaš refers to political treachery (7:3) and lies (11:12). The verbal form is included in the general indictment of the nation (4:2) and refers to the failure of the crops that the people depend on (9:2). A related term, kāzāb, appears in 7:13; 12:1.
127. See above, pp. 3, 4.
128. The lxx reads ‘house of Israel’ instead of Bethel, though there is no need to emend the mt.
129. For a comparison between Israel’s adoption and adoption in the ANE, see Melnyk 1993.
130. LaSor (1978) takes 11:1 as sensus plenior, by which the fuller significance of the text is implicit in the prophecy; Garrett (1997: 220–222) takes a typological approach but suggests that Hosea recognized that his words had further significance; see also Beale 2012. A fully typological approach, though, makes finding a link with Matt. 2:15 from Hosea’s side unnecessary.
131. Suggestions that Hosea may have had access to a different version of the Decalogue seem speculative.
132. There may be wordplay between Ephraim (ʾeprayim) and rāpāʾ (Dearman 2010: 282).
133. Comparison with Isa. 30:28, which refers to putting a bridle ‘on the jaws’ (Dearman 2010: 283), is not convincing: a bridle is put on the jaws, whereas a yoke is not. Moon’s suggestion that lifting the yoke enables the animal to feed (2018: 184) still does not explain the yoke being ‘on the jaws’.
134. Schüngel-Straumann (1995) amends lĕḥî (‘jaw, cheek’) to read ‘breast’ and argues that the imagery is of God as mother. That, though, seems unlikely, as does the suggestion that ‘I am God and not man’ (v. 9) indicates non-masculine behaviour. See also Wacker (2012).
135. See above, p. 24.
136. Aroused (niv) translates kāmar (‘to agitate, grow hot’; cf. nrsv, esv).
137. See above, p. 24.
138. The same expression, ‘Holy One . . . among you’, occurs in the context of salvation in Isa. 12:6.
139. Literally, ‘the sea’ (i.e. the Mediterranean) and so the west.
Chapter 3
1. Rāʿâ relates to shepherding. With flock or shepherd as subject, it means, respectively, ‘to graze’ (cf. niv) or ‘to pasture, tend, shepherd’ (cf. nrsv).
2. See the discussion above, pp. 18–19.
3. ‘God ruled, and (the angel) prevailed’ (Yee 1987: 231–234; cf. Wolff 1974: 206; Macintosh 1997: 483–484) is less probable, as is the view that ‘angel’ is a later addition.
4. For further discussion, see Seow 1992b.
5. See above, p. 24.
6. See above, pp. 31–32.
7. The term ʾôn can mean ‘wealth’ or ‘strength’.
8. The noun yĕgîaʿ suggests something acquired through labour (cf. Gen. 31:42).
9. This is a literal translation. The niv is a paraphrase, though conveys the correct sense.
10. Visions are important in prophetic revelation (1 Sam. 3:1; Lam. 2:9; Ezek. 1:1; Zech. 1:8; cf. Isa. 1:1; Obad. 1:1; Mic. 1:1; Nah. 1:1).
11. Dāmâ may mean ‘to destroy’ (nrsv; cf. 4:5–6; 10:7, 15), though here it means ‘to compare’ and hence ‘to speak in parables’ (niv).
12. It has this sense in 1QH 4.33.
13. It is less likely that this relates to the idols; however, see Macintosh 1997: 525–526.
14. See above, p. 15 n. 42.
15. The terms appear together in 4:8; 8:13; 9:9 (see also 12:8); ʿāwōn also occurs in 5:5; 7:1; 9:9; 10:10; 12:8; 14:1–2; and ḥaṭṭāʾt in 10:8 (cf. 13:2).
16. Nōḥam is from the same root as niḥûmîm (cf. 11:8).
17. Plagues (deber) can also mean ‘thorn’; destruction (qeṭeb) may be translated ‘sting’.
18. Resurrection imagery may derive from Baal’s dying and rising (see on 6:2).
19. A link with pereʾ (‘wild donkey’; cf. 8:9) (Andersen and Freedman 1980: 625; Macintosh 1997: 550) is less likely.
20. Kruger (1988b) also views 14:1–8 as a later redaction.
21. E.g. stumble (cf. 4:5; 5:5; 14:1), ‘rebel’ (cf. 7:13; 8:1), ‘know’ (yādaʿ; cf. p. 15 n. 42), ways (cf. 4:9; 10:13). Israel is also described as being without understanding (4:14) and wisdom (13:13). See Andersen and Freedman 1980: 647–648; Seow 1982; Dearman 2010: 345–346.
22. On šûb, see above, p. 24.
23. See above, p. 171 n. 15.
24. Receive us graciously (niv) is unlikely.
25. Horses are specifically noted in relation to Judah in 1:7.
26. See above, p. 13.
27. Cf. n. 35 below.
28. The mt here reads ‘like Lebanon’, though the reference to ‘roots’ may suggest trees. Though Lebanon was noted for its cedars, the niv rendering is too specific.
29. The niv again understands ‘cedars’, though this is unnecessary.
30. Again (niv, nrsv) translates šûb, taken as an auxiliary verb. It may also suggest a return to Yahweh and/or the land (cf. esv; Stuart 1987: 211), and the double sense may be intentional.
31. Following the lxx, some read ‘What more has Ephraim to do with idols?’ (nrsv mg.; Stuart 1987: 211).
32. See above, pp. 9–10.
33. The niv reads the verbs as future (cf. Macintosh 1997: 576), though a present sense seems preferable (nrsv, esv). Some read ʾšr (‘to bless’) instead of šûr (Mays 1969: 184; Stuart 1987: 211).
34. Continuing the wordplay between Ephraim (ʾeprayim) and pĕrî (‘fruit’); cf. 9:16.
35. Commentators note Wellhausen’s conjecture that ‘he answers’ (ʿānîtî) and ‘he watches over’ (ʾăšûrenû) refer to the fertility deities Anat and Asherah, suggesting that Yahweh provides what the people previously sought from them. However, reference to them here for the first time seems unlikely. See Hubbard 1989: 232–233; Dearman 2010: 343; Moon 2018: 221.
36. Cf. n. 21 above.
37. E.g. wise (ḥākām), ‘upright’ (ṣaddîq), ‘to understand’ (bîn), to walk in ways (derek) that are right (yāšār); see Hubbard 1989: 233–234.