Does the Bible say where demons came from?

No. The canonical text never narrates the origin of demons. The 'demons are disembodied spirits of dead Nephilim' doctrine that often fills this silence comes from 1 Enoch 15:8-12 — a pseudepigraphal work, not Scripture. The NT keeps demons (δαιμόνιον, G1140) and angels (ἄγγελος, G32) in separate vocabulary streams that never share a verse. The careful reader reports both what the text says and what it leaves unsaid.

The honest answer: the canonical text never says.

What the text does say

The Old Testament names spiritual entities sparsely. The Hebrew word שֵׁד (shed, H7700, "demon") occurs twice in the entire Hebrew Bible — both times as the recipient of illegitimate sacrifice (Deu 32:17; Psa 106:37). The text does not say where shedim come from. It does not give them a backstory. It reports what Israel did with them.

The wider OT vocabulary is equally spare: שָׂעִיר (saʿir, H8163, "goat-demon") in four desolation passages (Lev 17:7; 2Ch 11:15; Isa 13:21; 34:14); לִילִית (lilit, H3917) once at Isaiah 34:14; עֲזָאזֵל (ʿazaʼzel, H5799) four times in Leviticus 16. Roughly eleven verses across thirty-nine books. None of them narrates an origin.

The New Testament uses δαιμόνιον (daimonion, G1140) 63 times across 55 verses. Demons speak (Mrk 5:9), recognize Jesus (Mrk 1:24; Mat 8:29), possess (Mrk 5:1-13), petition (Luk 8:31), and are cast out (Mrk 1:34 and many others). At no point does the canon supply their origin story.

What the text says about a different population

The NT does describe rebel angels. Two passages name them:

εἰ γὰρ ὁ θεὸς ἀγγέλων ἁμαρτησάντων οὐκ ἐφείσατο, ἀλλὰ σειραῖς ζόφου ταρταρώσας παρέδωκεν εἰς κρίσιν τηρουμένους "For if God did not spare angels who sinned, but having cast them into Tartarus in chains of gloom, handed them over, kept for judgment" — 2 Peter 2:4 (TAGNT)

ἀγγέλους τε τοὺς μὴ τηρήσαντας τὴν ἑαυτῶν ἀρχὴν ἀλλ᾽ ἀπολιπόντας τὸ ἴδιον οἰκητήριον, εἰς κρίσιν μεγάλης ἡμέρας δεσμοῖς ἀϊδίοις ὑπὸ ζόφον τετήρηκεν "and angels who did not keep their own domain but abandoned their proper dwelling — for the judgment of the great day, in eternal chains under gloom, he has kept" — Jude 1:6 (TAGNT)

These are real texts about a real angelic rebellion. 2 Peter 2:5 places the event in immediate sequence with Noah and the flood — the antediluvian period. Jude 6 echoes Watchers vocabulary. The canon affirms a category of fallen, bound angels.

But notice what the canon does not do. It never calls those bound angels demons. It calls them ἄγγελοι. It does not say their offspring became demons. It does not say their disembodied spirits became demons. It does not describe a transformation from one category into the other. The two vocabulary streams remain separate even in the passages that speak most directly about angelic rebellion.

The empirical fact, run three different ways: δαιμόνιον (G1140) and ἄγγελος (G32) co-occur in 0 NT verses, 0 NT pericopes, and share 0 top-25 embedding-space neighbors. The text declines to weld them.

What 1 Enoch added

The popular doctrine fills the canonical silence with a Second Temple expansion. 1 Enoch 6-16 is a richly imagined narrative: 200 angels descend on Mount Hermon under the leader Semjaza, swear an oath, take human wives (Gen 6:1-4 read angelically), teach forbidden arts (Azazel teaches metallurgy and warfare), and beget the giants. When the giants die, their disembodied spirits become demons:

"And now, the giants, who are produced from the spirits and flesh, shall be called evil spirits upon the earth, and on the earth shall be their dwelling. Evil spirits have proceeded from their bodies; because they are born from men and from the holy Watchers is their beginning and primal origin." — 1 Enoch 15:8-9

This is the source of the doctrine. It is not in the canonical text. 1 Enoch is a pseudepigraphal work — written under a pseudonym (in this case, Enoch the seventh from Adam), not authored by him, dated by scholars to the Second Temple period (ca. 3rd century BC for the Watchers section). It is regarded as Scripture only by the Ethiopian Orthodox Church.

1 Enoch is a valuable historical witness. It shows what some Second Temple Jews believed about the origin of demons. It provides literary and conceptual context for some NT passages — the bound-angels vocabulary in 2 Peter 2:4 and Jude 6, the ζόφος ("gloom") and ταρταρόω ("cast into Tartarus") word-clusters that appear nowhere else in the NT, and Jude's direct quotation of 1 Enoch 1:9 in Jude 14-15. But "valuable historical witness" is not the same as "canonical authority." The two NT authors who engage 1 Enoch most closely (Peter and Jude) draw on its bound-angels imagery and its Lord-coming-with-myriads passage — they do not import its demon-origin doctrine.

What Tobit shows about the deuterocanonical handling

Even the deuterocanonical tradition that is most willing to narrate demons in story-form keeps the categories distinct. In Tobit (a Second Temple Jewish work, deuterocanonical in Catholic and Orthodox canons, not Protestant), the angel Raphael (Ῥαφαήλ, "God heals") is sent to bind the demon Asmodeus. The angel and the demon are different kinds of beings, given opposite roles in the narrative — the angel binds; the demon is bound. Tobit calls Asmodeus a πονηρὸν δαιμόνιον (Tob 3:8, "evil demon"); Raphael is ἄγγελος (Tob 5:4 etc.). The categorical distinction is preserved in narrative form. A book that wanted to teach "demons are fallen angels" would not write its central deliverance scene as an angel-binds-demon contest.

Why the silence matters

The NT authors knew the Watchers tradition. Peter and Jude both echo its vocabulary in their bound-angels passages. Jude quotes 1 Enoch 1:9 directly (Jud 14-15). They had every opportunity to import 1 Enoch 15's demon-origin doctrine if they had wanted to. They did not. The vocabulary they chose for the demons of the Synoptic exorcisms (δαιμόνιον, πνεῦμα ἀκάθαρτον, "unclean spirit") is consistently kept apart from the vocabulary they chose for the rebel angels (ἄγγελοι ἁμαρτήσαντες, "angels having sinned"). The discipline is visible across the corpus.

That selective non-importation is itself a form of canonical commentary. The NT authors did not treat 1 Enoch as Scripture — Tertullian thought they did; Augustine thought they did not, and the canonical tradition settled with Augustine. Jude's quotation of one verse from 1 Enoch is exegetical, not canonizing. Even at the Watchers' direct entrance into the NT (2 Pe 2:4 / Jud 6), the text affirms the angelic-rebellion category without affirming the further leap to demons-as-Nephilim-spirits.

What the careful reader reports

What the canon says about demons:

  • They exist as a category of unclean spirit (δαιμόνιον / πνεῦμα ἀκάθαρτον).
  • They speak, recognize Jesus, and possess.
  • They fear the abyss as a future destination (Luk 8:31).
  • They submit to the name of Jesus (Luk 10:17).
  • They will be tormented "before the time" (πρὸ καιροῦ, Mat 8:29).

What the canon does not say about demons:

  • It does not say they are fallen angels.
  • It does not say they are disembodied Nephilim spirits.
  • It does not name any individual demon (the Gospel encounters use no proper names for them; "Legion" in Mrk 5:9 is a number, not a name).
  • It does not narrate their origin.

The careful reader holds both columns. The canon's silence is not an invitation to fill the gap with extra-canonical material as if it were Scripture; it is a discipline to which the reader submits. We say what the text says, no more.

For the full lexical case — the three independent witnesses to demon-and-angel non-overlap, the chained-vs-roaming grammar, and the four distinct vocabulary streams the canon refuses to fuse — see Demons vs. Fallen Angels — What Does the Text Say?. For the OT's spare demonic vocabulary, see The Silence and the Storm. For the LXX-and-Second-Temple expansion that produced the NT word δαιμόνιον, see "Between the Testaments" — From Silence to Storm.

Related questions

Are demons fallen angels?

The New Testament does not say so. δαιμόνιον (G1140) and ἄγγελος (G32) never share a verse, never share a pericope, and share no top-25 embedding-space neighbors — three independent witnesses to a vocabulary the text refuses to fuse. Even the Tobit narrative, the closest deuterocanonical bridge between the two categories, keeps them in separate roles: the angel Raphael binds the demon Asmodeus.

Was Lucifer Satan's name?

No. 'Lucifer' is the Latin Vulgate's translation of the Hebrew word helel (הֵילֵל, H1966) — a hapax legomenon at Isaiah 14:12 meaning 'shining one' or 'morning star.' The Septuagint had already rendered it ἑωσφόρος ('light-bearer,' i.e., the planet Venus at dawn). The literary frame names the addressee outright: 'the king of Babylon' (Isa 14:4) and 'the man who shook the earth' (Isa 14:16). The Lucifer-equals-Satan identification is a post-canonical reading of a Latin translation choice.

What is Tartarus?

Tartarus is the custodial holding-place of rebel angels named once in the New Testament — 2 Peter 2:4 — by the verb ταρταρόω ('to cast into Tartarus,' G5020), a NT hapax. It is distinct from the abyss (ἄβυσσος, G12), which demons fear and where Satan is bound for the millennium, and from the lake of fire (λίμνη τοῦ πυρός, G3041 + G4442), which is the final convergent endpoint. Three distinct judgment-geographies, with zero shared NT verses among them.

Did Satan fall before creation, at the cross, or in the future?

The three texts most often cited as 'Satan's fall' do not describe a primordial pre-creation event. Luke 10:18 uses the imperfect ἐθεώρουν ('I was watching') to comment on the Seventy's exorcisms in real time. Revelation 12:7-12 dates the casting-down by ἄρτι ἐγένετο ('now is come') to Christ's enthronement. Isaiah 14:12 is a taunt-song explicitly addressed to 'the king of Babylon.' None of the three is a Genesis-prequel.