HistoricalChristian.Faith

Deuteronomy 32:43 at Hebrews 1:6: “let all God’s angels worship him” — the Septuagint preserving an older Hebrew?

Type: supporting argument (off-timeline) · Assessment: sound — the catalog’s older-Hebrew tag on this row holds, and this is the honest counterweight to the gloss cases. When Hebrews 1:6 proves the Son’s superiority to the angels from “and let all the angels of God worship him” (καὶ προσκυνησάτωσαν αὐτῷ πάντες ἄγγελοι θεοῦ), it quotes a clause the Masoretic Deuteronomy 32:43 does not contain — but which stands in the Septuagint and is now attested in a Hebrew copy from Qumran (4QDeut^q). So unlike Isaiah 7:14, Psalm 40:6, or Amos 9:11–12 — where the Greek diverges from a stable Hebrew — here the Greek preserves an older Hebrew the Masoretic line dropped: it is Hebrew-over-Hebrew, not Greek-over-Hebrew. Two refinements keep the tag from being over-read: the wording narrows down the chain (Qumran “gods” → Greek “sons/angels of God” → Hebrews “angels of God”), so Qumran vindicates the line, not the exact christological word; and 4QDeut^q is a distinct textual form, not proof the Greek’s Vorlage was identical — the Septuagint preserves an older Hebrew, not the one.

See also: the contrasting species — a distinctively Septuagintal reading of a stable Hebrew — is the Septuagint’s interpretive gloss group (Isaiah 7:14 · Psalm 40:6 · Amos 9:11–12); the christologically-neutral twin of this species — the Greek and Qumran preserving an older Hebrew against the Masoretic Text — is worked at Goliath’s height (LXX and 4QSam^a “four cubits” against Masoretic “six”). The big-picture Septuagint-dependence thesis, and the “older Hebrew” caution this row is the sharpest instance of, are weighed on “did the New Testament authors quote the Septuagint?” (assessment sound). This page adds what those do not carry: the text-form adjudication of the Hebrews-1:6 catalog row itself.

The claim

“Rejoice, O you heavens, with Him, and let all the angels of God worship Him” and “Rejoice, O you nations, with His people, and let all the angels of God be strengthened in Him.” … by these words He declares that we, the nations, rejoice with His people — to wit, Abraham, and Isaac, and Jacob, and the prophets, and, in short, all of that people who are well-pleasing to God.”

— Justin Martyr, Dialogue with Trypho 130 (c. 155), newadvent.org.¹ Justin quotes Deuteronomy 32:43 in its full Septuagintal form — the summons to “the angels of God” to worship — two centuries before Jerome would translate the same verse “from the Hebrew” and find the whole clause missing. The interpretation this argument serves is the catalog’s classification of the Hebrews-1:6 row: that its proof-text is older-Hebrew — the New Testament follows the Greek at a point where the Greek does not diverge from an older Hebrew the Masoretes lost, but preserves it.

Who advances it

That Hebrews 1:6 draws on a text richer than the Masoretic Deuteronomy is not in dispute; the modern point is why. The text critics who edited the Qumran Deuteronomy — Patrick Skehan, who first published the fragment (1954), and Eugene Ulrich, who edited it in the Discoveries in the Judaean Desert series — and the standard handbooks (Emanuel Tov, Arie van der Kooij) hold that the Masoretic Text of Deut 32:43 is a shortened form and that the Septuagint, joined now by a Hebrew scroll, preserves the fuller, older song.² On the New-Testament side the same fact is a flat classification: Archer & Chirichigno list Hebrews 1:6 among the quotations following the Septuagint where it parts from the Hebrew, and the quotation tables (Kalvesmaki; the loci citati of the Nestle–Aland / UBS Greek New Testament) list it the same way; the Hebrews specialists (Gareth Lee Cockerill, Harold Attridge, Gert Steyn, Paul Ellingworth) go further, identifying the exact Greek form the author used.³ The ancient church simply used the line: Justin (above) reads it of the Gentiles’ calling; Athanasius makes “Let all the angels of God worship Him” an anti-Arian proof that the worshipped Son is no creature; and John Chrysostom expounds it as showing the Son greater than the angels “as much as a Master is than a slave.”⁴

The evidence for

The clause is absent from the Masoretic Hebrew and present in the Greek. The Masoretic Deut 32:43 is the shortest of the three witnesses — four cola, opening “Rejoice, O nations, with his people” and running straight to the vengeance and atonement. It has no summons to the heavens and the divine beings. The Septuagint doubles that invitation into two couplets: “Be glad, O heavens, with him, and let all the sons of God worship him (προσκυνησάτωσαν αὐτῷ πάντες υἱοὶ θεοῦ); be glad, O nations, with his people, and let all the angels of God be strengthened in him.” The clause Hebrews needs — divine beings commanded to worship — stands in the Greek and is missing from the Hebrew the Masoretes handed down.¹

A Hebrew scroll from Qumran carries the missing line. This is what lifts the row out of the gloss cases. The Song-of-Moses manuscript 4QDeut^q (4Q44), an excerpted copy of Deuteronomy 32 in Hasmonean script (c. 100 BC), preserves v. 43 with the very summons the Masoretic Text lacks: הרנינו שמים עמו והשתחוו לו כל אלהים — “Rejoice, O heavens, with him, and bow down to him, all gods (ʾĕlōhîm).” So the worship-summons the Septuagint carries is not a Greek invention: a Hebrew copy a thousand years older than the Masoretic codices carries it too. The three-way split is the signature of an older Hebrew the proto-Masoretic line later pruned.²

The mainstream reconstruction is that the Masoretic Text abbreviated. Tov treats Deut 32:43 as a textbook case: the versions preserve the longer, original text, and the Masoretic Text is the abridgment — most plausibly because a summons to “all the gods” to worship YHWH sat uneasily with later strict monotheism (the same theological pressure suspected at Deuteronomy 32:8, where the Masoretic “sons of Israel” displaced an older “sons of God” still read at Qumran and in the Greek). On that reconstruction the New Testament’s “Septuagintal” reading is not a preference for Greek over Hebrew at all: it is the Greek keeping a Hebrew line the synagogue’s text had already lost.⁵

Hebrews 1:6 quotes Deuteronomy 32:43, not Psalm 97:7. The one live rival source, Psalm 97:7 LXX (96:7), also pairs “worship” with “angels” — but it reads “all his angels” (οἱ ἄγγελοι αὐτοῦ), where Hebrews reads “angels of God” (ἄγγελοι θεοῦ). That “of God” is the fingerprint of Deut 32:43: it is absent from the Psalm and present in the Deuteronomy line. In the main Septuagint Deuteronomy text “worship” and “angels of God” sit in separate couplets; they come together in the first colon in the form carried by the Odes appended to the Greek Psalter (Ode 2), which reads exactly Hebrews’ words — so the author quotes Deut 32:43 in an Odes-type text-form. The majority of modern commentators (Cockerill, Attridge, Steyn, Ellingworth) so identify it.⁶

Adversarial assessment

What holds. The catalog’s classification is correct, and it is the sharpest example of its kind in the New Testament. Hebrews rests a christological proof — the angels worship the Son — on a clause the Masoretic Deuteronomy does not contain but the Septuagint does, and a Hebrew scroll a millennium older than the Masoretic codices carries that clause too. This is not the Greek reading a stable Hebrew its own way (the Isaiah 7:14 / Psalm 40:6 / Amos 9:12 species); it is the Greek preserving an older Hebrew the Masoretic line lost — the same phenomenon the Scrolls make christologically-neutral at Goliath’s height. Because the proof-text is genuinely older Hebrew, the ancient corruption-charge inverts: the reading Trypho’s heirs would call a Christian insertion is the one the Masoretes dropped.

What must be refined, so the tag is not over-read.

1. The word narrows as it travels; Qumran vindicates the line, not the exact christological word. The chain is not identical end to end. The Qumran Hebrew reads “all gods” (ʾĕlōhîm); the Septuagint renders that “sons of God” / “angels of God”; Hebrews quotes “angels of God.” So an interpretive narrowing — the divine ʾĕlōhîm read as created angeloi — happened on the Greek side (exactly as the Septuagint renders the “sons of God” of Job 1:6 as “angels”). What 4QDeut^q proves is that the colon summoning the heavenly beings to worship is old Hebrew, not that the specific word “angels” is. There are thus two layers here: the line’s existence (older-Hebrew, correctly tagged) and the word inside it (an ʾĕlōhîm→angeloi gloss, of a piece with the group next door). The tag reports the first; it must not be stretched to claim the second.⁵

2. The source-identification is not idle for the tag — and the same detail secures both. Whether Hebrews quotes Deut 32:43 or Ps 97:7 decides whether “older-Hebrew” is even the right label. Psalm 97:7’s Hebrew already reads “bow down to him, all gods” in the Masoretic Text — a stable Hebrew — so on the Psalm sourcing the only divergence would be the same ʾĕlōhîm→angeloi gloss, and the row would belong with the gloss group, not here. It is the Deuteronomy sourcing that makes “older-Hebrew” apposite, because it is the Deuteronomy line the Masoretes dropped and Qumran recovered. And the detail that fixes the source — Hebrews’ “of God” (θεοῦ), which the Psalm lacks — is the same detail that secures the classification. A minority (John Owen, and older expositors) read it from the Psalm; the θεοῦ tells against them.⁶

3. 4QDeut^q is a distinct form; the Septuagint preserves an older Hebrew, not demonstrably the one. The three witnesses do not line up in a single ancestry. 4QDeut^q has six cola and one worship-summons (to the “gods”); the extant Greek has eight, doubling the invitation — itself most likely a conflation of two Greek renderings, one of a 4QDeut^q-type Hebrew and one of a Masoretic-type Hebrew. So the Qumran scroll confirms that a fuller Hebrew existed and that its worship-line is genuinely old; it does not prove the Greek translator’s Vorlage was identical to 4QDeut^q. The honest statement is the modest one the tag already makes: here the Greek preserves an older Hebrew reading the Masoretic Text lost — not that Greek, Qumran, and Hebrews are three copies of one text.²

4. Here the New Testament proof does depend on the text-form — and that is the point, not an embarrassment. Unlike Psalm 40:6 or Amos 9:12, where the load-bearing clause survives in both text-forms, Hebrews’ angels-worship proof exists only in the fuller text: the Masoretic Deuteronomy carries no such line at all. But this is exactly why the row is the counterweight to a corruption-charge run in reverse. “The New Testament built its christology on a verse absent from the Hebrew” is a true sentence only if “the Hebrew” means the Masoretic Hebrew; against the older Hebrew now in hand, the clause is not absent but original, and it is the Masoretic line that is short. So the dependence is real and it is legitimate: the author of Hebrews quoted a Bible whose Deuteronomy still had its opening summons to the sons of God — the Bible the Greek-speaking church actually read, and, as it turns out, an older Hebrew besides.⁴

Assessment

Sound. The interpretation this argument serves — that Hebrews 1:6’s proof-text is older-Hebrew, the Septuagint preserving a worship-summons the Masoretic Deuteronomy 32:43 lost — holds, and it is arguably the clearest place in the New Testament where a pre-Masoretic Hebrew reading, recovered at Qumran, is put to explicit christological use. The clause is absent from the Masoretic Text, present in the Septuagint, and carried in Hebrew by 4QDeut^q (c. 100 BC); the mainstream verdict that the Masoretic Text abbreviated — the same pruning suspected at Deuteronomy 32:8 — means this is Hebrew-over-Hebrew, the honest counterweight to the gloss cases and to any “the New Testament invented it” charge. What the tag must be trusted to mean, and no more, is fixed by three refinements: (a) the word narrows as it travels — Qumran “gods” → Greek “sons/angels of God” → Hebrews “angels of God” — so the Scroll vindicates the line, not the exact christological word; (b) the source is Deut 32:43 in an Odes-type form (the θεοῦ tells against Psalm 97:7), and that identification is what makes “older-Hebrew” the right label rather than a gloss; and (c) 4QDeut^q is a distinct textual form and the extant Greek is itself conflate, so the Septuagint preserves an older Hebrew, not demonstrably the one behind the Greek. Kept to that, the row is a genuine, well-grounded instance of the Greek preserving an older Hebrew the Masoretes lost; stretched into “Greek, Qumran, and Hebrews are one identical text” or “the word ‘angels’ itself is the old Hebrew,” it overreaches. This row is the pole opposite the interpretive-gloss group: not the Septuagint reading a stable Hebrew, but the Septuagint keeping an older Hebrew — the finding the Dead Sea Scrolls turned from a guess into a fact.

Sources