Deborah, Jael, and the Oldest Song in the Bible

• Anchor Bible Dictionary — “Deborah,” “Barak,” “Jael,” “Sisera,” “Jabin,” “Hazor,” “Kenites,” “Kishon,” “Taanach,” “Megiddo,” “Meroz,” “Judges, Book of,” “Poetry, Hebrew.”
• Dictionary of the OT: Historical Books (IVP) — “Judges, Book of,” “Deborah,” “Women,” “Warfare,” “Chariots,” “Prophets and Prophecy,” “Poetry,” “Tribes of Israel.”
• NIDOTTE / TDOT — יָד (yāḏ, hand), שָׁפַט (šāpaṭ, judge), נָבִיא (nāḇîʾ, prophet), תִּפְאֶרֶת (tipʾeret, glory), יָתֵד (yātēḏ, tent peg), בָּרַךְ (bāraḵ, bless), אָרַר (ārar, curse), רֶחֶם (reḥem, womb).
• For the Greek — any dictionary article on the Septuagint of Judges should name the two texts. If it speaks of “the LXX of Judges” in the singular without qualification, treat it with care: this is the classic double-text book of the Greek Old Testament, alongside Esther, Tobit and Daniel.

• Barry G. Webb, The Book of Judges (NICOT); Daniel I. Block, Judges, Ruth (NAC); Trent C. Butler, Judges (WBC); Robert G. Boling, Judges (Anchor Bible); K. Lawson Younger Jr., Judges and Ruth (NIVAC).
• Susan Niditch, Judges (OTL); J. Clinton McCann, Judges (Interpretation); Tammi J. Schneider, Judges (Berit Olam).
• Natalio Fernández Marcos, Judges (Biblia Hebraica Quinta, 2011) — the current critical edition of the Hebrew, with a full account of the Greek witnesses.
• Philip E. Satterthwaite, “To the Reader of Judges,” in A New English Translation of the Septuagint — the clearest short statement in English of why there are two Greek texts and what each is.

The storm-theophany of 5:4–5 speaks a language the whole region spoke. In the Ugaritic Baal cycle, when Baal finally has his palace, the text says: “Baal op[ens] rifts in [the cloud]s. Ba[al gives] forth his holy voice… the mountains quake, A-tremble are… East and west, earth’s high places reel” (ANET 134–135, Ginsberg’s translation). Judges 5 uses the same furniture — the earth trembling, the clouds dropping water, the mountains quaking — and puts it to a use no Canaanite poem does. The difference is the direction of the march. Baal’s thunder announces a god who has just won a palace on his own mountain. Israel’s poem has the L' + SC[1:] + ' arriving from somewhere else — “when you went out from Seir, when you marched from the region of Edom” — to fight for a people who cannot fight for themselves. The shared vocabulary is not a problem to be explained away. It is how you say “God came” in the twelfth century BC, and the point is who came, and for whom.

Nine hundred chariots of iron is not a round number invented to impress. Centuries before this, Thut-mose III took Megiddo — the very ground of Judges 5:19, “at Taanach, by the waters of Megiddo” — and his scribes itemised the haul: “1 chariot worked with gold, with a body of gold, belonging to that enemy, [1] fine chariot worked with gold belonging to the Prince of [Megiddo]… and 892 chariots of his wretched army — total: 924” (Annals of Thut-mose III, ANET 237). A Canaanite coalition in that plain fielding a chariot corps in the high hundreds is exactly what the Egyptian records describe. Judges 4:3 is reporting a recognisable order of battle, and 1:19 had already told you Israel had no answer to it.

Dating and archaeology are contested, and this page names the positions without settling them. Three live questions touch these chapters. (1) Jabin and Hazor. Joshua 11 has Joshua kill a Jabin king of Hazor and burn the city; Judges 4 has a Jabin king of Canaan reigning in Hazor. Some read Jabin as a dynastic throne-name (as Abimelech functions in Genesis); some posit a rebuilt Hazor under a later ruler; some regard the two accounts as independent traditions of one conflict. Hazor’s excavated Late Bronze destruction is real and large, and its date and cause are argued about in print. (2) The chronology. The 480 years of 1 Kings 6:1 will not add up with the summed years of the judges unless some judgeships overlap — which many think they plainly do, since several are explicitly regional. (3) Iron I settlement. The highland villages of Iron I are the archaeological horizon in which these events are usually placed, and what they prove is disputed. We do not adjudicate any of these, and nothing in the Christian faith rests on the outcome. The Merneptah Stele (c. 1207 BC) is worth knowing as the first mention of “Israel” outside the Bible — a people, not a city, already in Canaan; and the frequently repeated equation of the Amarna letters’ ʿApiru with the Hebrews is a contested proposal, never an established result, and this page does not lean on it.

• The Baal cycle, II AB (the building of Baal’s house), Ugarit, c. 14th–13th c. BC (ANET 134–135) (a Canaanite myth — a witness, not Scripture) — “Baal op[ens] rifts in [the cloud]s… the mountains quake… earth’s high places reel.” The nearest thing we have to the idiom of Judges 5:4–5.
• Annals of Thut-mose III, the Megiddo campaign, fifteenth century BC (ANET 237) (an Egyptian royal record — a witness, not Scripture) — the captured-chariot inventory, “892 chariots of his wretched army — total: 924,” taken on the same plain the Song names.
• Merneptah Stele, c. 1207 BC (an Egyptian victory hymn — a witness, not Scripture) — the earliest extra-biblical naming of Israel, as a people in Canaan. (ANET prints it in the 370s, past the range of our local scan; the page number is therefore not given here.)
• Josephus, Antiquities 5.198–209 (= Whiston 5.5.1–4) (Second-Temple — a witness, not Scripture) — he keeps Deborah’s rebuke and sharpens it into direct speech: “Thou, O Barak, deliverest up meanly that authority which God hath given thee into the hand of a woman, and I do not reject it!” He turns the milk into an explanation — Jael “gave him sour milk, of which he drank so unmeasurably that he fell asleep” — and has her drive the nail “through his temples into the floor.” And then, tellingly, he omits the Song of chapter 5 entirely, closing instead with Barak commanding Israel forty years.
• Pseudo-Philo, Liber Antiquitatum Biblicarum 30–33 (first century AD — a witness, not Scripture) — a retelling that expands Judges enormously, gives Deborah a “new song” of her own (ch. 32) and a farewell speech in which she is called mother (ch. 33). At 31:7 Jael, who in Scripture never speaks before the act, prays: “Strengthen my arm this day, O Lord, for your sake and for your people’s sake, and for those who put their trust in you.” Note what has happened: the retelling supplies exactly the moral commentary the biblical narrator withholds. That is a witness to how early readers felt the silence — not a licence for us to fill it.
• b. Megillah 14a (rabbinic-era — a witness, not Scripture) — Deborah is counted among the seven prophetesses (“Sarah, Miriam, Deborah, Hannah, Abigail, Huldah, and Esther”), and the palm tree of 4:5 is explained practically: since men came to her for judgment, “she established for herself a place out in the open and visible to all.” The same folio reads “wife of Lappidoth” as “woman of flames,” from the wicks she made for the sanctuary — a homily on the name, not a translation of it.
• Sirach 46:11–12 (Second-Temple — a witness, not Scripture) — the judges remembered together, “may their memory be blessed,” under one general clause about those “whose hearts did not fall into idolatry.” No judge is named and none is examined. It is the earliest surviving example of the flattening this page is trying to avoid.

• Judges 9:53–54 — the second woman in the book to kill a commander, and this time the narrator makes the point out loud: Abimelech, dying of a millstone dropped by “a certain woman,” asks his armour-bearer to finish him “lest they say of me, ‘A woman killed him’” — after the narrator has already told you her aim was true. Deborah’s “into the hand of a woman” (4:9) is the first half of that pair (see Judges 9).
• 1 Samuel 12:9–11 — Samuel’s recital of the cycle names “Sisera, commander of the army of Hazor” and lists the deliverers God sent. It is the Old Testament’s own summary of the book, and it treats the whole period as a single argument about forgetting and being rescued anyway.
• Psalm 83:9–10 — “Do to them as you did to Midian, as to Sisera and Jabin at the river Kishon.” Centuries later this battle is still the standard example of a deliverance nobody engineered.
• Hebrews 11:32 — “Gideon, Barak, Samson, Jephthah.” The New Testament names among the faithful the man who said “if you will not go with me, I will not go” — and does not name Deborah. That is worth sitting with rather than tidying: the list is a list of people who acted in faith, not a ranking, and its most striking feature is how much it forgives.
• Exodus 15:20–21 — “Miriam the prophetess… took a tambourine in her hand” and sang the victory. Judges 5 is the second of these, and the resemblance is deliberate (see Exodus 15).

• Judges 2:11–19 — the formula, stated in the abstract, that every account in the book then varies: did evil → sold them → cried out → raised up a deliverer → the land had rest. Judges 4–5 keeps four of the five and drops the fourth: this is the first deliverance in the book of which it is never said that the Lord raised up the deliverer. Compare 3:9 (“the Lord raised up a deliverer… Othniel”) and 3:15 (“the Lord raised up for them a deliverer, Ehud”) with 4:4, where Deborah is simply already there (see Judges 1–2 and Judges 3).
• Deuteronomy 33:2 — “The Lord came from Sinai and dawned from Seir upon us; he shone forth from Mount Paran.” Judges 5:4–5 sings the same march from the same direction, and 5:5’s “this Sinai” is close enough to Moses’ blessing that the ESV footnotes an alternative rendering, “the One of Sinai” (see Deuteronomy 33). The same theophany recurs at Psalm 68:7–8 and Habakkuk 3:3–6: four poems, one memory.
• Exodus 14 and 15 — the pattern this passage is built on. A deliverance narrated in prose, then sung; water doing what an army could not; and a prophetess with an instrument leading the singing (see Exodus 14 and Exodus 15). Judges 5:21, “the torrent Kishon swept them away,” is the Song of the Sea in a Galilean wadi.
• Joshua 11:1–13 — Jabin king of Hazor, the northern coalition, and Hazor burned, “for Hazor formerly was the head of all those kingdoms.” Judges 4 has a Jabin king of Canaan reigning in Hazor again. Report both and do not silently harmonise them — the readings (throne-name, rebuilt city, parallel traditions) are laid out at Stage 2, and this page picks none (see Joshua 11).
• Judges 1:19 — “he could not drive out the inhabitants of the plain, because they had chariots of iron.” The problem of Judges 4:3 is set up one chapter into the book, and the Song’s answer is that the plain belongs to whoever controls the weather (see Judges 1).
• Genesis 35:8 — “And Deborah, Rebekah’s nurse, died, and she was buried under an oak below Bethel; so he called its name Allon-bacuth.” Judges 4:5 puts its Deborah under a named tree “between Ramah and Bethel.” The names and the neighbourhood coincide; Scripture draws no connection between the two women and neither will we — but it is a real echo, and readers notice it (see Genesis 35).
• Numbers 10:29–32; Judges 1:16; 4:11 — the Kenites, descended from Moses’ father-in-law’s people, living inside Israel without being Israel. Jael is a woman on a boundary in every sense, and the text tells you so before it tells you what she does.

• Barry G. Webb, The Book of Judges (NICOT, 2012) — the fullest recent evangelical commentary, and unusually careful about what the narrator does and does not say.
• Daniel I. Block, Judges, Ruth (NAC); Trent C. Butler, Judges (WBC); Robert G. Boling, Judges (Anchor Bible); K. Lawson Younger Jr., Judges and Ruth (NIVAC).
• Susan Niditch, Judges (OTL); J. Clinton McCann, Judges (Interpretation); Tammi J. Schneider, Judges (Berit Olam) — three quite different readings, all attentive to the women in these chapters.
• Robert Alter, The Art of Biblical Narrative and The Art of Biblical Poetry — for reticence and the gap in chapter 4, and for reading chapter 5’s parallelism as intensification rather than repetition. (Cited for the literary observation only; the theological frame here is the church’s, not his.)
• Frank Moore Cross and David Noel Freedman, Studies in Ancient Yahwistic Poetry — the standard entry point for the argument that Judges 5, Exodus 15 and Deuteronomy 33 preserve the oldest Hebrew verse we have.
• Michael D. Coogan, “A Structural and Literary Analysis of the Song of Deborah,” CBQ 40 (1978); Baruch Halpern, “The Resourceful Israelite Historian: The Song of Deborah and Israelite Historiography,” HTR 76 (1983) — on the poem’s shape, and on the relationship between the poem and the prose.
• Natalio Fernández Marcos, Judges (BHQ, 2011), and Philip E. Satterthwaite, “To the Reader of Judges” (NETS) — the two texts, the manuscript groups behind them, and the kaige question. Dominique Barthélemy, Les devanciers d’Aquila (1963), is the study that put kaige on the map.
• Robert Rezetko, “The Qumran Scrolls of the Book of Judges,” Journal of Hebrew Scriptures 13 (2013) — a full inventory of what Qumran actually preserves of this book.

• The Song’s Hebrew is genuinely hard, and pretending otherwise is not piety. Judges 5 is widely held to be among the oldest poetry in the Bible, and its language shows it: rare words, archaic forms, and several lines nobody can render with confidence. You do not have to take a scholar’s word for this — the ESV flags the difficulty three times in this one chapter: at 5:5 (“Or before the Lord, the One of Sinai”), at 5:10 (“The meaning of the Hebrew word is uncertain”), and at 5:11 (“Or archers; the meaning of the Hebrew word is uncertain”) — and at 5:14 it tells you outright that it has followed the Septuagint against the Hebrew. Then look at 5:8, 5:13–14 and 5:15–16 in the codex above and watch the two Greek translators produce, from the same consonants, sentences that are not about the same subject. Where the text is uncertain, this page says so.
• “Into the hand of a woman” — a rebuke, or simply a fact? Both are argued. On the first reading, 4:9 is a mild judgment on Barak’s condition: he asked for a prophetess as insurance and forfeited the credit. On the second, Deborah is announcing a divine method, not a penalty — God repeatedly wins by means the culture would not have picked (Ehud left-handed, Gideon with three hundred, a millstone dropped by a woman in ch 9). The text supports the first mildly and the second strongly, and it does not despise Barak either way: he goes, he wins, and Hebrews 11:32 puts him among the faithful. What it will not support is the sermon that makes Barak a coward and Deborah a reproach to him — the chapter never says either thing.
• The prose and the poem differ, and that is a feature. The prose has Sisera abandon the chariot and flee on foot; the poem has him fall at Jael’s feet in a fourfold cascade (5:27). The prose has him asleep; the poem’s picture is famously harder to reconstruct. Readers who require a single reconciled film reel will find the two accounts irritating. Readers who have ever heard a soldier’s report and then a nation’s song about the same day will not.
• The two Greek texts are the book’s signature problem, and they are not a threat to it. There was still no Göttingen critical edition of Greek Judges when NETS was prepared. Rahlfs printed A and B in parallel because he judged them separate recensions; Brooke–McLean printed a diplomatic text whose base is essentially Vaticanus. Most now think A and B descend from one archetype at different degrees of revision, and that the Old Greek is best sought in the Lucianic subgroup within the A tradition. What varies is the Greek, not the story: both texts have Deborah judging, Barak bargaining, Jael acting, the mother at the window and the land at rest. Where they change the sense of a line we print both, above.
• Qumran does not help here, and it is worth knowing why. The Judges scrolls are few and fragmentary: 1QJudg (1Q6) preserves parts of chapters 6 and 9 (and, less certainly, 8); 4QJudgᵃ (4Q49) preserves part of chapter 6 and is famous for lacking 6:7–10; 4QJudgᵇ (4Q50) preserves parts of chapters 19 and 21. None of them preserves any part of chapters 4–5. For the Song of Deborah we have the Masoretic Hebrew and the two Greek texts and nothing older, and a page that implied otherwise would be inventing evidence.
• Josephus is a witness to reception, not to the event. He keeps Deborah’s rebuke, explains the milk, and drops the Song entirely (Ant. 5.198–209). Pseudo-Philo goes the other way and gives Deborah a whole new song plus a farewell (LAB 32–33) and puts a prayer in Jael’s mouth (31:7). Both are doing what readers always do with a narrator who will not comment: they comment. Noting that is not a criticism of them; it is a warning to us.
• What this page deliberately does not do. It does not decide between the two Jabins. It does not date the battle. It does not turn Jael into a heroine or a villain. And it does not use the Song’s curse on Meroz as a stick for anyone alive: the verse is about a village that would not come to a battle in the twelfth century BC, and the only legitimate use of it is the one the Song itself makes — a question about whether you showed up.

• Read the report and the song. God gave this event to us twice on purpose. The prose tells you what happened; the poem tells you what it meant, who was missing, and what it cost the other side. Most of us were trained to read only one of those and to distrust the other. Both are Scripture.
• The formula is already breaking. Evil, sold, cried out, rest — but nobody is raised up. Deborah is simply already there. Small crack, early in the book; by the end there is no judge left at all, only a refrain about everyone doing what was right in his own eyes.
• Keep ‘approved,’ ‘reported,’ and ‘condemned’ in separate columns. The Song blesses Jael. The prose says nothing about her. The book condemns Meroz. Merge those and you will either turn Judges into a hero cycle or into a lecture, and it is neither.
• Where the text is uncertain, say so. Three ESV footnotes and two wildly different Greek translations in one chapter are not a crisis. They are what an ancient poem looks like when you are honest about it — and a faith that can only survive on smoothed edges is being trained to flinch.
• Meroz is the verse for churchgoers. The Song curses no Canaanite by name. It curses a village that stayed home — “because they did not come to the help of the Lord.” Reuben had a very thorough discussion about it first. That is the line most likely to be about us.
• And the poem ends by looking at the enemy’s mother. Before the last line of praise, four verses at a window with a woman waiting for a son who is not coming. Whatever else victory is in Scripture, it is never allowed to be simple.

Step in response: two things this week. First, read Judges 4 and Judges 5 aloud, back to back, in one sitting — it takes about eight minutes — and write down three things the poem tells you that the prose does not. Second, ask the Meroz question about one specific thing: where is help currently needed that you have had a great searching of heart about and not yet gone to?