To appear in Exchange and Transmission across Cultural Boundaries: Philosophy, Mysticism

and Science in the Mediterranean (Proceeding of a Workshop in Memory of Prof. Shlomo Pines,
the Institute for Advanced Studies, Jerusalem; 28 February - 2 March 2005), ed. H. Ben-
Shammai, S. Shaked and S. Stroumsa (Jerusalem: The Israel Academy of Sciences and
Humanities, forthcoming).
 THE BOOK OF WATCHERS IN THE QUR<ÅN

Patricia Crone

Sura 2:102 informs us that the Jews follow the magic taught by demons (shayåã•n) and
"such things as were sent down to the two angels at Babylon, Hår¥t and Mår¥t", but that
the latter two will not teach anyone without warning them first: "we are just a temptation
(fitna), do not disbelieve", they will say. In explanation of this passage the exegetes tell a
gripping story about a woman named Anåh•d or Zuhra (i.e. Venus) or BŸdukht (daughter
of God) who tricked two amorous angels into telling her the great name of God, or some
other magic formula, that the angels had used to ascend to heaven, whereupon she rose to
heaven herself; there she became the star Venus, while the two angels were left behind on
earth and punished by being hung upside down in a well in Babylon, where people would
come to them for magic knowledge. The exegetes add that angels had descended to earth
with God's permission to act as judges because they were upset by the terrible behaviour
of human beings, believing that they could do better than humans even if they had to
contend with passions; God decided to put them to the test and provided them with
passions for purposes of the experiment; and the woman proved them wrong by inducing
them to drink wine, kill and practise idolatry before making them reveal their secret
formula to her.1
It is well known that both the Qur<ånic passage and the gripping story are
developments of the account of the fallen angels in the so-called Book of Watchers, a
work attributed to the antediluvian figure Enoch (great-grandfather of Noah). It is
perhaps not so well known that this book is reflected more than once in the Qur<ån. A
further example has recently been identified by John Reeves.2 In what follows I propose
another two and discuss the identity of >Uzayr, who should perhaps be seen as a fifth
example. The interest of all four or five examples lies in the light they throw on the
religious milieu in which the Qur<ån arose and the relationship of the Qur<ån (and indeed
the exegetical tradition) with an old debate in the Near East about how sin had come into
the world.

I should like to thank Annette Reed for giving me a copy of her thesis (now a book), which taught me an
enormous amount and inspired this article. I am also grateful to her, as well as to Michael Cook, Behnam
Sadeghi, Shaul Shaked, Adam Silverstein and, last but not least, Joseph Witztum for helpful comments on
diverse versions of the article.
1
Numerous accounts are given al-$abar•, Jåmi> al-bayån, Beirut 1988, i, 452ff, ad 2:102, translated along
with other versions in E. Littmann, 'Hår¥t und Mår¥t', in Festschrift Friedrich Carl Andreas, Leipzig 1916,
70-87; cf. also L. Jung, Fallen Angels in Jewish, Christian and Mohammedan Literature, Philadelphia
1926, 126ff; Encyclopaedia Iranica, s.v. 'Hår¥t and Mår¥t' (Shahbazi); add al-Kalb•, cited in early Persian
tafs•r (Ab¥ Bakr >At•q N•shåb¥r•, Tafsir-i S¥råbåd•, ed. S. S•rjån•, Tehran 1381 [2002], i, 105f; Lahore
Tafs•r, facsimile edition and Russian translation by F. I. Abdullaeva, Tolkovanije korana (Lakhorskhij
tafsir), Moscow 2001, 42f = 79ff); and also in the long version attributed to Ab¥ Ja>far, i.e. Mu˙ammad al-
Båqir, in al-Qumm•, Tafs•r, Beirut 1991, i, 65-67. Some versions omit the motif regarding God's name or
other magic formula.
2
See below, note 76.
 2

The fallen angels.

Genesis, 6:2-4, informs us that in the antediluvian past there were giants (nephilim) on
earth and that at that time "sons of God" consorted with "daughters of men", siring
mighty heroes (gibborim); thereafter the wickedness of man led to the flood. This passage
has its roots in an ancient Near Eastern myth about rebellion in the pantheon, and it was
to generate a vast number of narratives itself. In fact, it is one of those cases where we
can follow the history of a couple of motifs and their endless transmogrifications from the
dawn of history until today, seeing them meander like a huge river with a mass of
constantly shifting arms and canals over the literary landscape of western Eurasia. It is
quite an awe-inspiring sight.3 Here, however, we need to zoom in on the rivulet
constituted by the Book of Watchers.
The Book of Watchers is the first of five or more separate works which together
make up the Ethiopic Book of Enoch (known as 1 Enoch).4 It is based on the assumption
that the Biblical "sons of God" are angels, the normal understanding of the term in
antiquity,5 and it casts the angels as sinners. They are guilty of transgressing both sexual
and epistemic boundaries, for not only do they take human wives, as Genesis says, they
also teach them sorcery and other illicit sciences, and the outcome is disastrous. They
have giant offspring (the gibborim and nephilim, identified), and it is these giants who
wreak all the havoc on earth that causes God to send the flood, without the damage ever
being fully repaired; for though the rebellious angels are bound and jailed while their
giant offspring are killed by obedient angels sent against them, their wicked activities
continue due to evil spirits that have issued from them. The message of the book is that
super-human forces rather than human beings are responsible for the existence of evil on
earth; God sent the flood to rid the earth of the dreadful giants, not to punish humans,
who suffer as innocent victims of superior powers.6
Aramaic fragments of this book, apparently dating from the second century BC
and suggesting that the book itself goes back to the third century BC, have been found at
Qumran, along with fragments of related works such as the Book of Giants.7 Substantial
Greek portions are also extant, partly in an Egyptian papyrus dating from perhaps the
fifth or sixth century CE and partly in extracts by the ninth-century Byzantine author

3
Nobody has tried to draw a picture of the entire river, but for a fine account of the already much ramified
section from the ancient Near East to Augustine, see N. Forsyth, The Old Enemy: Satan and the Combat
Myth, Princeton 1987. For the ancient Near Eastern roots, see also P. D. Hanson, 'Rebellion in Heaven,
Azazel, and Euhemeristic Heroes in 1 Enoch 6-11', Journal of Biblical Literature 96, 1977, 195-233.
4
1 Enoch stands for the Enoch book in Ethiopic, 2 Enoch for that in old Church Slavonic, 3 Enoch for that
in Hebrew (more properly called Sefer Hekhalot). Watchers are a certain category of angels. For editions
and translations, see below, notes 7-9.
5
Cf. P. S. Alexander, 'The Targumim and Early Exegesis of "Sons of God" in Genesis 6', Journal of
Jewish Studies 23, 1972, 60f.
6
Forsyth, Satan and the Combat Myth, 167, 169f; cf. also J. J. Collins, 'The Origin of Evil in Apocalyptic
Literature', in his Seers, Sibyls and Sages in Hellenistic-Roman Judaism, Leiden 1997, 292-99; M. Delcor,
'Le mythe de la chute des anges et de l'origine des géants comme explication du mal dans le monde dans
l'apocalyptique juive', Revue de l'Histoire des Religions 190, 1976, 3-53; and the interesting discussion in
S. Schwartz, Imperialism and Jewish Society, Princeton 2001, 77ff.
7
J. T. Milik (ed. and tr.), The Books of Enoch: Aramaic Fragments of Qumrân Cave 4, Oxford 1976; L.
Stuckenbruck (ed. and tr.), The Book of Giants from Qumran, Tübingen 1997.
 3

George Syncellus.8 The Ethiopic version, which preserves the book in full, is a translation
made on the basis of a Greek original between the fourth and the sixth century CE.9
There are also numerous references to, and retellings of, the book in Jewish, Christian,
pagan, Manichaean, and other Gnostic literature.10
It used to be thought that the Enoch literature and other pseudepigraphic works
originated in sectarian or socially marginal circles, but this is no longer the prevailing
view; the Book of Watchers seems to have been regarded as authoritative by many Jews
down to the second century CE. Then the rabbis turned against it, however. The second-
century rabbi Simeon b. Yohai cursed all those who explained the "sons of God" as
angels: in his view they were sons of judges. Some said they were called "sons of God"
because they lived long and easy lives.11 The Aramaic targums duly translated "sons of
God" as "sons of judges" (Neophyti) or "sons of nobles" (Onqelos and Pseudo-Jonathan;
similarly the Samaritan targum), while Symmachus, a Jewish or Jewish Christian
translator of the Bible into Greek active in the late second or early third century, opted for
"sons of the powerful".12 Thereafter they were explained as human beings of one kind or
another by all the main commentators on the Bible, Qaraites included.13 The protagonists
of the story appear here and there in rabbinic literature, but with little trace, before the

8
M. Black (ed.), Apocalypsis Henochi Graece, Leiden 1970. The Egyptian manuscript (Codex
Panopolitanus) is sometimes dated to the eighth century or later.
9
M. A. Knipp (ed. and tr.), The Ethiopic Book of Enoch in the Light of the Aramaic Dead Sea Fragments,
Oxford 1978. There are later editions and translations too. The Book of Watchers covers chapters 1-36 and
has been used in this article in the translations of E. Isaac in J. H. Charlesworth (ed.), The Old Testament
Pseudepigraphica, vol. i, New York 1983, 5-89; and G. W. E. Nickelsburg and J. C. VanderKam, 1 Enoch:
a New Translation, Minneapolis 2004.
10
Cf. Jung, Fallen Angels; B. J. Bamberger, Fallen Angels, Philadelphia 1952; F. Dexinger, Sturz der
Göttersöhne oder Engel vor der Sintflut?, Vienna 1966; J. C. VanderKam, '1 Enoch, Enochic Motifs and
Enoch in Early Christian Literature', in J. C. VanderKam and W. Adler (eds.), The Jewish Apocalyptic
Heritage in Early Christianity, Minneapolis 1996; A. Y. Reed, Fallen Angels and the History of Judaism
and Christianity: the Reception of Enochic Literature, Cambridge 2005. G. A. G. Stroumsa, Another Seed:
Studies in Gnostic Mythology, Leiden 1984, esp. chs. 2 and 8; J. C. Reeves, Heralds of that Good Realm:
Syro-Mesopotamian Gnosis and Jewish Traditions, Leiden 1996, 183-206; id., Jewish Lore in Manichaean
Cosmogony: Studies in the Book of the Giants Traditions, Cincinnati 1992; id., 'Jewish Pseudepigrapha in
Manichaean Literature: the Influence of the Enochic Library', in id. (ed.), Tracing the Threads: Studies in
the Vitality of Jewish Pseudepigrapha, Atlanta 1994. For what appears to be reflections of the Watcher
story among the Mandaeans, see J. J. Buckley, The Mandaeans, Oxford 2002, 8. For the pagans, see below,
notes 34, 38, 40f.
11
Genesis Rabba, 26, 5; discussed, inter alia, by Bamberger, Fallen Angels, 91; Alexander, 'Targumim and
Early Exegesis', 61f; Reed, Fallen Angels, 208ff. The idea that they were sons of judges has a long and
fascinating history of its own, rooted in Psalms 82.
12
Alexander, 'Targumim and Early Exegesis', 64, 70f; Reed, Fallen Angels, 213ff; J. Fossum, 'The Angel
of the Lord in Samaritanism', Journal of Jewish Studies 46, 2001, 53n. Ps.-Jonathan also has a passage in
which they are angels, cf. below, note 14.
13
Bamberger, Fallen Angels, 149ff. Sa>adiya Gaon duly has ban¥ 'l-ashråf (Saadiyah Ben Joseph al-
Fayy¥m•, Version arabe du pentateuque, Paris 1893, 12), and elsewhere dismisses the idea of angels
mating as a monstrous invention (id., Iyov: >im targum e-ferush Se>adiya ben Yosef, ed. Y. D. Qafa˙,
Jerusalem 1972f, 27f (ch. i, §6) = The Book of Theodicy, tr. L. E. Goodman, New Haven and London 1988,
155f.
 4

rise of Islam, of their angelic descent or sexual misconduct, let alone their illicit
teaching.14
The Christians stuck to the story of the fallen angels for another century or two,
impressed by its capacity to account for the prevalence of pagan cults and all the sins
with which idolatry was held to go in tandem. The angels and their demonic offspring
had enslaved mankind by teaching them murder, war, adultery, magic and other terrible
things, not least worship of themselves in the guise of pagan deities, Justin Martyr (d.
165) explained, developing 1 Enoch 15, 8 and 19. The angels had taught humans
astrology, magic, metallurgy, cosmetics, and idolatry, as Clement of Alexandria (d. c.
215), Tertullian (d. after 220) and Lactantius (d. c. 320) said.15 Philosophically inclined
Christians such as Origen (d. c. 255) interpreted the story allegorically,16 but even he held
the angels to have taught humans astrology.17 That they and their demonic offspring were
responsible for idolatry and diverse forms of illicit knowledge came to be a generally
accepted Christian view. It was also as false gods and demons that the angels passed into
Manichaean literature.18
By the third century, however, the Christians too had begun to turn against the
story, and the Enoch literature in general.19 The "sons of God" were not angels, it was
now said, but rather righteous men, more precisely descendants of Seth who had been
seduced by lascivious women descended from Cain in the period between the expulsion
from Paradise and the flood. First encountered in Julius Africanus (fl. c. 200, a Syrian
despite his name), this version of the story was to prevail in Greek and Syriac literature,
and indeed in Catholic and Protestant interpretation up to modern times.20 In short, the
responsibility of evil was shifted from superior powers to humans themselves. In line
with this, the origin of evil increasingly came to be located at the beginning of human
history, in the disobedience and expulsion of the devil and his hosts from heaven on the
one hand and the sin and expulsion of Adam and Eve from Paradise on the other, as

14
Cf. Babylonian Talmud, Niddah 61a; Yoma 67b; Deuteronomy Rabbah 11,10 (guilty of sexual
misconduct, but no mention of their illicit teaching); Targum Ps.-Jonathan (the nephilim were "Šem˙azai
and Azael, these fell from heaven", taken by Alexander to antedate the suppression of the angelic
interpretation, 'Targumim and Exegesis', 70f; followed by A. Y. Reed, 'From Asael and Šemi˙azah to
Uzzah, Azzah, and Azael: 3 Enoch 5 (§§7-8) and Jewish Reception-History of 1 Enoch', Jewish Studies
Quarterly 8, 2001, 123n; differently Reed, Fallen Angels, 213f, where it is a later insertion). They reappear
as angels in Sefer Hekhalot (3 Enoch), 5, also assigned to a late date by Reed ('From Asael and Šemi˙azah',
132ff; Fallen Angels, 256f).
15
VanderKam, '1 Enoch, Enochic Motifs', 44-54, 68-70, 85.
16
Origen, Contra Celsum (tr. H. Chadwick, Cambridge 1953), 5, 55, based on Philo, On the Giants, 2, cf.
L. R. Wickham, 'The Sons of God and the Daughters of Men: Genesis VI 2 in Early Christian Exegesis', in
J. Barr and others, Language and Meaning. Studies in Hebrew Language and Biblical Exegesis, Leiden
1974, 142ff;VanderKam, '1 Enoch, Enochic Motifs', 54-59, 81f.
17
Origen, Philocalia (tr. G. Lewis, Edinburgh 1911), 23:6.
18
Cf. the Psalms of Thomas IV (210.1 and 210:10), in T. Säve-Söderberg, Studies in the Coptic
Manichaean Psalm-Book, Uppsala 1949, 127.
19
Bamberger, Fallen Angels, 74ff; Forsyth, Satan and the Combat Myth, 349ff; Wickham, 'Sons of God
and the Daughters of Men', 135-47; VanderKam, '1 Enoch, Enochic Motifs', 59f, 100f; Reed, Fallen
Angels, 194ff.
20
Bamberger, Fallen Angels, 78ff; Dexinger, Sturz, 106ff; A. F. J. Klijn, Seth in Jewish, Christian and
Gnostic Literature, Leiden 1977, 61ff; Africanus in VanderKam, '1 Enoch, Enochic Motifs', 80f; Brock
(tr.), St Ephrem the Syrian: Hymns on Paradise, Crestwood NY 1990, I, 11 and the note thereto (pp. 81,
189).
 5

opposed to in the voluntary descent of the angels from heaven in the period before the
flood. For all that, the Book of Watchers continued to be read by Greek and Syriac
Christians. Jacob of Edessa (d. 708) even defended the Enoch book of which it formed
part, convinced of its antediluvian origins; and unlike their Byzantine counterparts,
Syriac chroniclers rarely use disclaimers about the reliability of the Jewish
pseudepigrapha when they cite them.21
One reason why the Jews turned against the story is that angels, both pure and
fallen, were getting out of control. There is a fair amount of evidence for veneration or
actual worship of angels among the Jews of the first centuries CE,22 sometimes involving
the angels in general and sometimes a principal angel cast as mediator between God and
mankind; speculation about such an intermediary had probably contributed to the rise of
Christianity, and devotion to angels and/or an angelic viceregent (notably in the form of
Metatron) continued in Judaism after the first two centuries, too.23 Gnostics also operated
with the notion of an intermediary, but they postulated that he was evil and filled the
void between the all too distant God and mankind with demonic beings, convinced, like
Pseudo-Enoch, that human beings were victim of superior forces of evil beyond their
control. Gnostic myths abound in sexual union by seduction or rape between evil archons
and primordial humans, now by direct retelling of the Watcher story,24 now by distant
echo of it (or independently), and the result is sometimes called abortions (Hebr.
nephalim, an alternative understanding of the nephilim of Gen. 6:1).25 A similar filling up
of divine space was underway in the "underworld" of Platonism, as Dillon calls the
confluence of Gnostic, Hermetic and Chaldaean thought characteristic of late antique
paganism.26 That humanity is at the mercy of unfathomable forces of the universe is also
the key conviction behind late antique magic, devoted to the control of such forces by
manipulation of the angels in charge of them.27
The rabbis reacted to these developments by both rejecting the story of the fallen
angels and belittling the angels as a class.28 (A similar reaction is observable on the part

21
Cf. W. Adler, 'Jacob of Edessa and the Jewish Pseudepigrapha in Syriac Chronography' in J. C. Reeves
(ed.), Tracing the Threads. Studies in the Vitality of Jewish Pseudepigrapha, Atlanta 1994, 145; S. Brock,
'A Fragment of Enoch in Syriac', Journal of Theologial Studies 19, 1968, 626-31.
22
How far the angels were actually worshipped is controversial, cf. L. T. Stuckenbruck, 'The Angelic
Refusal of Worship: the Tradition and its Function in the Apocalypse of John', Society of Biblical
Literature Seminar Papers 1994, 695f; id., Angel Veneration and Christology, Tübingen 1995; differently
M. Simon, 'Remarques sur l'angélolâtrie juive au début de l'ère chrétienne', Académie des Inscriptions &
Belles-Lettres, Comptes Rendus 1971, Javier-Mars, Paris 1971. It is with reference to excessive regard for
angels that Alexander explains the change in Judaism ('Targumim and Early Exegesis', 68f).
23
A. Segal, Two Powers in Heaven, Leiden 1977; N. Deutsch, Guardians of the Gate. Angelic Vice
Regency in Late Antiquity. Leiden 1999; D. Boyarin, Border Lines: the Partition of Judaeo-Christianity,
Philadelphia 2004, chs. 5-6; for Metatron, see below, notes 96f.
24
Stroumsa, Another Seed, 32 (note 54), 33, 35-7; Reeves, Jewish Lore in Manichaean Cosmogony, 75f,
81.
25
Stroumsa, Another Seed, chs. 2, 8 (casting the Watcher story as the key to Gnosticism altogether);
Reeves, Jewish Lore in Manichaean Cosmogony, 71f.
26
R. Majercik, The Chaldean Oracles: Text, Translation and Commentary, Leiden 1989, 8f; cf. J. Dillon,
The Middle Platonists, London 1977, 384.
27
H. D. Betz, The Greek Magical Papyri in Translation, Chicago and London 1986, xlvii; cf. also S.
Shaked, 'Popular Religion in Sasanian Babylonia', Jerusalem Studies in Arabic and Islam 21, 1997, 104.
28
Noted by Bamberger, Fallen Angels 92.
 6

of the Samaritans.)29 The rabbis did not have an answer to the problem of evil, which was
acute thanks to the political disasters of the first and second centuries and the
Christianisation of the Roman empire that had followed them. Rather, they coped with it
by closing the door on the world outside and crediting evil, like everything else, to God
while at the same time making Him so homely and familiar a figure that everything He
did seemed bearable. It was not the case that there were "two powers in heaven", let alone
"many ruling powers in heaven"; rather, God and Israel formed a tightknit family,
whatever the ups and downs. Personal morality was what mattered, not the uncontrollable
developments in the world of gentiles: evil was now dealt with primarily as an "evil
inclination" (yetzer ha-ra>) in the human heart.30
The Christians, meanwhile, had split the godhead into three and allowed for an
almost autonomous realm of evil, setting them well on the way in the Gnostic direction;
but they saved their Biblical concept of the deity by casting God as the ultimate creator of
the evil realm, while at the same time absolving him of responsibility for it by recourse to
the concept of free will. The devil they placed in charge of the evil realm had a long
history entwined with that of the fallen angels, who accompanied him in the transfer of
the decisive fall from the period before the flood to the beginning of human history.31
Pseudo-Enoch's explanation of the flood was discarded.
To both Jews and Christians, eliminating Enoch's fallen angels was all the more
desirable in that the idea of angels copulating with women had come to feel offensive.32
Angels were superior beings. Besides, they had no passions, and even if they did, they
had no bodies. How could they desire the corporeal, let alone cause human reproduction?
The whole story was perverse and absurd, Cyril of Alexandria (d. 444) insisted, stressing
that the "sons of God" mentioned in Genesis were not to be understood as angels. "Is it
not probable that many will be discouraged by this and choose sensuality.....if we believe
that even the very angels fell subject to passion?", he asked.33
Meanwhile, the story of the "sons of God" had been taken up by pagans. Some
adduced it in polemics against Christianity, arguing, as did Celsus (c. 170), that it showed
Jesus not to have been the only angel to have come, or, as did Julian the Apostate (d.
363), that it proved Moses to have believed in many gods without knowing anything
about Jesus. 34 (To late antique pagans, gods and angels were interchangeable, and Christ
was commonly envisaged as an angel in early Christianity. 35) Others read the story as an
account of the origins of the occult sciences. It was after all to the fallen angels that

29
Fossum, 'Angel of the Lord', 52f.
30
Bamberger, Fallen Angels, esp. 49, 57, 95, 101f; E. E. Urbach, The Sages, Jerusalem 1975, ch. 15.
31
For all this, see Forsyth, Satan and the Combat Myth, 222ff and passim. The Watcher story began to be
connected with that about Adam and Eve already in the first century BC.
32
That they cannot sin is affirmed already by the Jew in Justin Martyr's Dialogue with Trypho, 79,1.
33
Wickham, 'Sons of God and Daughters of Men', 135-38. Wickham's further argument regarding the role
of Christology is opaque to me.
34
Celsus in Origen, Contra Celsum (tr. H. Chadwick, Cambridge 1953), 5, 52; Julian, Against the
Galilees, 290B-E, in W. C. Wright (ed. and tr.), The Works of the Emperor Julian, Cambridge Mass. and
London 1913-23, iii, 400-3.
35
For a classic work, see J. Barbel, Christos Angelos, Bonn 1941; for a more recent one, see C. A.
Gieschen, Angelomorphic Christianity, Leiden 1998. Like Jewish angel veneration/worship, it is best
attested in the first two centuries CE, but surfaces thereafter too (cf. S. J. Shoemaker, Ancient Traditions of
the Virgin Mary's Dormition and Assumption, Oxford 2002, ch. 3 and Appendix A, translating an Ethiopic
work preserving a fifth-century Syrian narrative).
 7

astrologers, alchemists, soothsayers, diviners, magicians and their likes were believed to
owe their knowledge, as they themselves were well aware. A Syriac, i.e. pagan or
Christian, incantation bowl against "all the evil magical arts" (presumably written by
someone who saw his own magic as beneficent) refers to "angels that reveal the
mysteries of their lord".36 Zosimus (c. 300), a Hermetic alchemist from Panopolis, where
the book of Enoch was still read in the fifth or sixth century, 37 says that the holy
scriptures mentions angels who descended from heaven and mated with women, teaching
them "all the arts of nature", and that they were punished for this since these arts were
bad arts "of no benefit for the soul"; he adds that Hermes, too, talked about these events
in his Physika, and that indeed they are mentioned in almost every exoteric or esoteric
treatise; and he is clearly pleased by the Biblical and Hermetic agreement on the angelic
origin of the arts of nature, however lacking in benefit for the soul they might be.38
Zosimus' claim that the story had gone into esoteric works accords with a remark
by Jacob of Edessa (d. 708), according to which Athanasius of Alexandria (d. 373) had
proscribed the Enoch literature because heretics had incorporated it in their secret
books.39 There are in fact traces of 1 Enoch in the Hermetic literature,40 and among the
Hermetic works that Zosimus may have in mind is a small treatise known as Isis the
prophetess to her son Horus. In this work, Isis tells of how one of the angels (or,
according to one version, a prophet or angel) caught sight of her and wanted to make love
with her, which she refused to do unless he would tell her about the preparation of gold
and silver first; the angel said that this was beyond his ability, but sent another angel,
Amnael, who also wanted to make love with her and who was eventually persuaded to
part with his secrets. (Whether he got his payment is not stated.)41
As de Ménasce observes, in Isis' account the Watcher story seems to have fused
with an Indian myth regarding two ashvins, twin Vedic gods who roam about in the
world of mortals and try to seduce a married woman, but merely succeed in rejuvenating
her old husband instead.42 In this story as in Enoch, the events lead to the appearance of
temptations on earth, here by the distribution of intoxicating substances in drinks,
women, gambling and sports; and this story, too, would have appealed to the practitioners
of the arts of nature, for the ashvins were physicians to the celestials and practitioners of
the healing arts. The hypothesis that their adventures had fused with those of the fallen
angels would account for three otherwise puzzling features of Isis' version: there are only

36
J. Teixidor, 'The Syriac Incantation Bowls in the Iraq Museum', Sumer 18, 1962, 53.
37
Cf. the date of Codex Panopolitanus (above, note 8).
38
Zosimus quoted by Syncellus in R. P. Festugière, La Révélation d'Hermès Trismégiste, i, Paris 1944,
256; in J. Lindsay, The Origins of Alchemy in Graeco-Roman Egypt, London 1970, 179; in Stroumsa,
Another Seed, 139f (whose translation I have followed); in VanderKam, '1Enoch, Enoch Motifs', 83f. For
Zosimus' position, see D. Stolzenberg, 'Unpropitious Tinctures: Alchemy, Astrology & Gnosis according to
Zosimus of Panopolis', Archives Internationales d'Histoire des Sciences 49, 1999, 28 (I owe this reference
to Albert de Jong).
39
Adler, 'Jacob of Edessa', 145.
40
Cf. M. Philonenko, 'Une Allusion de l'Asclepius au livre d'Hénoch', in J. Neusner (ed.), Christianity,
Judaism and Other Greco-Roman Cults: Studies for Morton Smith at Sixy, Leiden 1975, ii, 11-63.
41
Festugière, Révélation d'Hermès Trismégiste, i, 256-60 (two versions); Lindsay, 194f (amalgamated); cf.
also M. Idel, 'The Origin of Alchemy according to Zosimos and a Hebrew Parallel', Revue des Etudes
Juives 145, 1986, 117-24.
42
P. J. de Ménasce, 'Une légende indo-iranienne dans l'angélologie judéo-musulmane: à propos de Hår¥t et
Mår¥t', Etudes Asiatiques 1, 1947, 10; Mahåbhårata, III (Vana Parva), 123-5.
 8

two angels in it, as opposed to the two hundred in the Book of Watchers;43 it only
features one particular woman, as opposed to women in general; and the woman now
obtains her secret knowledge by not mating with the celestial beings. It is presumably the
same quasi-Indian version which is reflected in rabbinic allusions to the story in which
the angels form a pair.44 It is in any case this version which lies behind the Gnostic story
of a woman called Norea, Noria, Noraia, Horea, Orea, Nuraita or the like, who resisted
the attempts of the wicked archons to seduce her, to be rescued by Eleleth, a holy angel
who revealed the truth to her.45 (This seems to be the only form in which the Gnostics
took a positive interest in the teaching of the Watchers, which they normally condemn, in
so far as they mention it at all.)46 It is also in this form that it was familiar to Muslim
exegetes: two angels court one woman, who tricks them into parting with their secrets
without mating with them. Though the rabbis and the churchmen had not succeeded in
killing the story, they had clearly managed to relegate it to the fringes of respectable
society. It was now mainly among pagans, Gnostics and devotees of the occult that it
flourished, outside the mainstream communities or, if within them, in the somewhat
marginal circles of alchemists, diviners and magicians.47 It must have been from such
circles that it passed to the Qur<ån.

The fallen angels in the Qur<ån.

In the Qur<ånic passage on Hår¥t and Mår¥t as in the Isis story, the angels form a pair,
but they are not guilty of any sexual sins; they merely teach people magic. So too do the
demons (indeed, the passage can be taken to say that the demons teach magic which they
have learnt from the angels), but it is only the demons who render themselves guilty of

43
1 Enoch, 6. There are only 60 or 70 of them in Celsus (Contra Celsum, 5, 52).
44
Babylonian Talmud, Yoma 67b (Azael and Uzza), Deuteronomy Rabba (11:10: Azah and Azael), and a
probably late insertion in Targum Ps.-Jonathan (Azael and Šem˙azai, cf. above, note 14); similarly the
much later 'Midrash on Šem˙azai and Azael' (ed., tr. and discussed on the basis of four versions in Milik,
Books of Enoch, 321-39; also discussed in Reed, Fallen Angels, 258ff), and a late midrash on the virgin
Istahar who is turned into a star as a reward for her resistance to sin (S. Liebermann, 'After Life in Early
Rabbinic Literature', Harry Austryn Wolfson Jubilee Volume, English section, ii, Jerusalem 1965, 497).
45
B. A. Pearson, 'The Figure of Norea in Gnostic Literature', Proceedings of the International Colloquium
on Gnosticism, Stockholm 1973, Stockholm and Leiden 1977, 143-52; Stroumsa, Another Seed, 53-55,
140f.
46
The Watchers revealed the arts in the world and the mysteries of heaven to men, teaching them all they
had seen in heaven, hell and on earth, according to Mani (Kephalaia, 92:24-31, in Reeves, Jewish Lore in
Manichaean Cosmogony, 81); they taught magic, idolatry, and bloodshed, according to 'On the Origins of
the World' (The Nag Hammadi Library in English, ed. J. M. Robinson, Leiden 1988, II, 5, 123). Most
Gnostic works retell the story of the Watchers in recognizable form without saying anything about their
teaching, see the 'Valentinian Exposition' (op. cit., XI, 2, 38); the 'Apocryphon of John' (op. cit., II, 1, 29);
Agapius' summary of Awdi's doctrine in Agapius, Kitåb al->unwån (ed. and tr. A. Vasiliev in Patrologia
Orientalis VII, 1911, 564; cf. H.-C. Puech, En quête de la Gnose, Paris 1972, 275f); and the survey in
VanderKam, '1 Enoch, Enochic Motifs', 70-76.
47
Just how marginal (or disreputable) these circles were, from the point of view of churchmen and rabbis, I
do not know. But recourse to magicians and soothsayers was condemned at a synod of 576 (A. V.
Williams, 'Zoroastrians and Christians in Sasanian Iran', Bulletin of the John Rylands University Library of
Manchester 78, 1996, 43; cf. 45 for similar condemnations on the Zoroastrian side), and though the rabbis
made some concessions, their attitude comes across as basically hostile too (cf. Urbach, The Sages, 97ff;
B. Kern-Ulmer, 'The Depiction of Magic in Rabbinic Texts: the Rabbinic and the Greek Concept of Magic',
Journal for the Study of Judaism 27, 1996, 289-303).
 9

kufr thereby. The angels are cast as agents of God: they warn their customers not to
render themselves guilty of kufr, explaining that their own function is to test people's
faith (innamå na˙nu fitna); and what they teach is something sent down to them. Why
angels should be teachers of magic, and how they came to be in Babylon, we are not told;
nor are we told how, if at all, they relate to the demons. There is no suggestion that the
latter are their giant sons; and though the false angels/gods of the mushrik¥n are
sometimes explained as demons (jinn) in the Qur<ån,48 it is nowhere suggested there
either that they are the offspring of angels.
The problem that preoccupies the Qur<ån in the passage on Hår¥t and Mårut is
that some People of the Book (i.e. Jews or Christians) prefer magic to the truth. In the
preceding verse it complains that a party of the People of the Book react to the fact that a
messenger has come to them from God by throwing the book behind their backs (2:101);
they prefer to follow that which the demons related to Solomon, i.e. magic.49 Solomon
was not an unbeliever (i.e. even though he used magic),50 but the People of the Book
clearly are, for they disregard the advice of Hår¥t and Mår¥t not to turn infidel by using
their services. From the two angels they learn "that with which to split up a man and his
wife" (just as in the Book of Watchers the angels teach the daughters of men "to make
hate-inducing charms", 1 Enoch, 9:7), and thereby they forfeit their share in the hereafter
(2:102). We clearly find ourselves right in the middle of Jewish magic, a well attested
phenomenon, and one in which speculation about Solomon is well known to have played
a role.51 Famed in antiquity, it is represented in the Greek magical papyri from Egypt,
dating from the second century BC to the fifth century CE (though these texts are
generally pagan),52 in Aramaic amulets mainly from Palestine, in incantation bowls from
Sasanian Iraq,53 in the Hekhalot literature, reflecting the period c. 200-800 CE,54 in
manuals for sorcerers reflecting the period from late antiquity onwards,55 and in the
Geniza.56 In Mesopotamia and Iran the great majority of incantation bowls were made by
Jews, often for clients bearing Iranian names, suggesting that magic was regarded as

48
Q. 6:100; 34:40f; 37:158.
49
According to many exegetes, the demons did not relate things to Solomon, but rather against him. Talå
>alayhi normally means "he related/recited to somebody", but the verse is problematic because it has them
relate things >alå mulk Sulaymån rather than >alå 'l-malik Sulaymån. Some exegetes tried to solve the
problem by understanding the >alå as adversarial (against the kingship of Solomon) ; others read it as
chronological (at the time of Solomon's kingdom). Since none of these constructions really click, it seems
more likely that mulk is in need of emendation.
50
Or, in the understanding of the exegetes, even though the demons maligned him by calling him a
magician rather than a prophet.
51
Cf. P. Torijano, Solomon the Esoteric King: from King to Magus, Leiden 2002, esp. 119ff, 192ff.
52
Betz, Greek Magical Papyri, xli, xlv.
53
See (for example) the introduction to J. Naveh and S. Shaked, Amulets and Magic Bowls: Aramaic
Incantations from Late Antiquity, Leiden 1985.
54
Cf. P. Schäfer, 'Merkavah Mysticism and Magic', in Gershom Scholem's Major Trends in Jewish
Mysticism 50 Years After, ed. P. Schäfer and J. Dan, Tübingen 1993; id., 'Jewish Magic in Late Antiquity
and Early Middle Ages', Journal of Jewish Studies, 41, 1990, 76-79.
55
Cf. M. Gaster (ed. and tr.), The Sword of Moses: an Ancient Book of Magic, London 1896; reprinted in
his Studies and Texts, New York 1928, i (translation) and iii (text); M. A. Morgan (tr.), Sepher Ha-Razim:
the Book of Mysteries, Chico CA 1983; Torijano, Solomon, 198ff, with further references.
56
P. Schäfer and S. Shaked (eds. and trs.), Magische Texte aus der Kairoer Geniza, Tübingen 1994.
 10

something of a Jewish specialisation there;57 and it must have been from a region within
the Iranian sphere of influence that the story passed to the Qur<ån, for Hår¥t and Mår¥t
are Haurvatåt and Ameretåt, two of the the Zoroastrian divine beings known as amesha
spentas 58 and it is in Båbil that the Qur<ån places them.
De Ménasce conjectured that it was via the Manichaeans that the angels passed
into the Qur<ån, on the grounds that it was probably Mani who gave the angels Iranian
names, just as he (or his disciples) renamed the giants.59 But such evidence as we have
does not support him. In Genesis the angels and giants are anonymous; in the Book of
Watchers they have acquired names, and here the leading angel is called Shemi˙azah
while the one who reveals the divine secrets is called Asael (>s<l), also rendered Azael
(>z<l) and Azazel (>z<zl – a name which conflated him with the devil).60 But while the
fragments of Mani's Book of Giants do indeed iranianise the names of two giants as Såm
and Nar•mån, they just adapt Šem˙azai as Šahm•zåd;61 and though we do not know how
Azael's name was rendered, the fragments operate with two hundred Watchers (now
demons) rather than two.62 It is more likely to have been via magic that the angels were
renamed, given that it is in the context of magic that the Qur<ån mentions them. Artat
Amurtat (presumably from Haurvatåt and Ameretåt) figures among the nomina barbara
in magical texts from Iraq.63

Tafs•r.
It is a striking fact that although the Qur<ån gives the angels Iranian names and says very
little about them, the exegetes effortlessly recognized them as the fallen angels from the
Watcher story. The Iraqi exegete al-Kalb• (d. 763) even knew their pre-Iranian names. In
his version there are three angels, as also in the Sefer Hekhalot (3 Enoch), perhaps by
development of the three (or four) angels who observe the misbehaviour of the giants
from heaven in the Book of the Watchers.64 Sefer Hekhalot called them Uza, Aza and
Azael (>wz<, >z< and >z<l) and the like, with Azael as the stablest element.65 An Aramaic
incanation bowl which also operates with three angels on the theme of >z (and which
invokes pagan, Jewish and Christian divinities alike) calls them Azael, Azael, and
Az(a)ziel (>z<l, >z<l, >zzy<l), where the second Azael should perhaps be understood as
57
Naveh and Shaked, Amulets and Magic Bowls, 18.
58
De Ménasce, 'Une légende indo-iranienne', 10f.
59
De Ménasce, 'Une légende indo-iranienne', 16f.
60
Cf. the useful survey in M. Black (tr.), The Book of Enoch or 1 Enoch, Leiden 1985, 121; below, notes
128-32. The form Azazel arose by identification of the fallen angel and the demonic figure to whom a sin-
laden scapegoat was sent on the Day of Atonement (Leviticus 16).
61
W. B. Henning, 'The Book of Giants', Bulletin of the School of Oriental and African Studies 9, 1943-6,
60; W. Sundermann, 'Ein Weiteres Fragment aus Manis Gigantenbuch', in his Manichaica Iranica, ed. C.
Reck and others, ii, Rome 2001, 496 (of the original pagination).
62
Henning, 'Book of Giants', 68f, 71; W. Sundermann, 'Mani's "Book of the Giants" and the Jewish Books
of Enoch', in his Manichaica Iranica, ii, 42 (of the original pagination).
63
Shaked, 'Popular Religion in Sasanian Babylonia', 110, 113.
64
1 Enoch 9; cf. Jubilees (tr. O. S. Wintermute in J. H. Charlesworth (ed.), Old Testament Pseudepigrapha,
ii, New York 1985), 7:21-26; compare below, note 79, and the text that goes with it.
65
Reed, 'From Asael and Šemi˙azah', 122 and n64. P. Alexander transliterates the names as >Uzzah,
>Azzah and >Aza<el in his translation of 3 Enoch, 4 and 5 (in J. H. Charlesworth (ed,), The Old Testament
Pseudepigrapha, vol. i, New York 1983, 258, 260).
 11

Azzael, Uzael, or the like.66 Other bowls which only have two >z names call them Aza
and Azael (>z<, >z<l).67 Al-Kalb• calls them >Azå,>Azåyå, and >Azaz•l (>z<, >z<y<, >zzyl). One
of them failed to make the descent, he says, while the other two bore the additional names
of Hår¥t and Mår¥t.68 Al-Kalb• and other exegetes also knew the events to be associated
with Enoch (now called Idr•s).69
As the exegetes tell the story, however, it is not about angelic revolt or the origin
of sin. Rather, it is about how tough it is to be a human being: even the angels lost control
of themselves when they experienced the enormous surge of sexual passion. Cyril of
Alexandria's warnings notwithstanding, this plainly did not serve to encourage
immorality, but on the contrary to warn against smug self-confidence; and if the angels
came out badly in the story, it only went to show that they had no reason to feel superior.
The exegetes thus linked the story with the theme, familiar from rabbinic literature, of
rivalry between angels and human beings: the story put the angels in their place.70
Though the exegetes went out of their way to stress that angels were not normally
endowed with passions, some continued to find the story offensive and construed Hår¥t
and Mår¥t as human beings by reading malikayn ("two kings") for malakayn ("two
angels") at Q. 2:102, or they read the verse as saying that magic was not sent down to the
two angels.71 (The story of the sons of Seth seduced by the daughters of Cain was well
known, but not as an alternative to the Watcher story, which had changed too much by
now for the two to be interchangeable.)72 Rationalists such as the Mu>tazilite Muãahhar
al-Maqdis• (wr. c. 355/966) objected to it in much the same terms as Cyril of Alexandria:
how could spirits without bodies have passions or make love?73 For all that, the story
retained its popularity. Indeed, it returned to the world of high culture, and not only for
66
D. Levene, A Corpus of Magic Bowls: Incantation Texts in Jewish Aramaic from Late Antiquity, London
2003, no. M163:18 (§5). The clients want their opponent punished in the same way that >z<l w->z<l w->zzy<l,
who transgressed the command of their lord, were pressed under the mountain with their faces downwards
by angels sent against them.
67
Cf. below, notes 116f. There are also manuscripts of Sefer Hekhalot which only has two >z names, again
>z</>zh and >z<l (cf. Reed, 'From 'From Asael and Šemi˙azah', 122).
68
Ab¥ Bakr >At•q, Tafsir-i S¥råbåd• , i, 105; Abdullaeva, Lakhorskhij tafsir, 42f = 79ff.
69
The angels descend to earth in the time of Idr•s ($abar•, Jåmi> al-bayån, i, 458, citing al-Rab•>; al-
Måward•, Tafs•r, ed. Kh. M. KhiËr, Kuwait 1982, i, 142; Qazw•n• in Jung, Fallen Angels, 130); as in the
Book of Watchers (13:4), they ask him to intercede for them (Kalb• in Abdullaeva, Lakhorskhij tafsir, 44 =
81).
70
Cf. P. Schäfer, Rivalität zwischen Engeln und Menschen, Berlin 1975; J. C. Reeves, 'Some Explorations
of the Intertwining of Bible and Qur<ån', in J. C. Reeves (ed.), Bible and Qur<ån: Essays in Scriptural
Intertextuality, Atlanta 2003, 52ff. Cf. also S. Stroumsa, '"What is man": Psalms 8:4-5 in Jewish, Christian
and Muslim Exegesis in Arabic', Henoch 14, 1992.
71
Cf. Jung, Fallen Angels, 126f, 128f, 135, 139. The reading malikayn is reported for al-Óasan b. >Al• and
Ibn >Abbås (Ibn Khålawayh, al-Mukhtaßar f• shawådhdh al-Qur<ån, ed. G. Bergsträsser, Paris 2005 (first
published Leipzig 1934), 8), al-Îa˙˙åk b. Muza˙im (Tafs•r al-Îa˙˙åk, assembled and edited by M. Sh. A.
al-Zåwayt•, Cairo, 1999, 161, no. 68), Sa>•d b. Jubayr, al-Zuhr• and others (>A.-L. al-Khaã•b, Mu>jam al-
qirå<åt, i, Damascus 2000, 164 (drawn to my attention by Joseph Witztum).
72
Cf. al-$abar•, Ta<r•kh al-rusul wa'l-mul¥k, ed. M. J. de Goeje and others, Leiden 1879-1901, i (tr. F.
Rosenthal, The History of al-$abar•, i, Albany 1989), 168ff; al-Ya>q¥b•, Ta<r•kh, ed. M. Th. Houtsma,
Leiden 1883, i, 5ff; Jung, Fallen Angels, 124ff.
73
Muãahhar b. $åhir al-Maqdis•, Kitåb al-Bad< wa'l-ta<r•kh, ed. and tr. C. Huart, Paris 1899-1919, iii, 15
(Jung, Fallen Angels, 128). Similarly Saadiya Gaon (above, note 13).
 12

Muslims, but also for Jews, who liked the Muslim version of the story even though their
own exegetes continued to identify the Biblical sons of God as human beings.74 The story
proved extremely longlived, too. In 1915 an Indian Muslim in Germany wrote a summary
of the story of Hår¥t and Mår¥t in Persian for the Orientalist Littmann: it now involved
two women to match the two angels, the women had become singers and dancers, and
both ended up as planets; that apart, the story was much as it had been told by al-Kalb•
and his likes over a thousand years earlier.75

Other echoes.
As Reeves notes, there seems to be a second reflection of the Book of Watchers in Q.
2:30, on God's creation of Adam.76 In the Qur<ån as in the Jewish and Christian literature
of the time, it is at the dawn of human history that sin comes into the world, thanks to an
arrogant angel (related to the same ancient Near Eastern myth of rebellion in the
pantheon as that behind Gen. 6:2-4) who is expelled from heaven and proceeds to seduce
the daughters of men in the form of Eve. There had been much interaction between the
Biblical story of Adam and Eve and that about the sons of God and daughters of men as
developed by later authors, and motifs originally associated with the flood had been
transferred to the time of the creation.77 In the Qur<ånic account of the arrogant figure,
here known as Ibl•s, the fact that he is envisaged now as an angel (Q. 7:11; 15:28-31;
17:61; 38:73) and now as a demon (min al-jinn, Q. 18:50) is presumably an Enochic
legacy. But it is in the account of God's creation of Adam that we encounter a more direct
reflection of the Book of Watchers.

In Q. 2:30 God tells the angels that he intends to create Adam, which is also what
he does in rabbinic accounts (as not in Christian ones);78 and here as there, the angels
object to God's plan. Both versions, in other words, pick up the theme of rivalry between
angels and humans. In the rabbinic accounts the angels sometimes object to God's plan on
the ground that mortals are useless and weak (cf. Psalms, 8:5; 144:3-4);79 at other times
they object that he is all falsehood and discord,80 or they enquire what his deeds will be
like,81 or God tells them that he will be righteous without telling them that he will be
wicked too.82 In the Qur<ån they object to the creation of a being who "will do corruption

74
Cf. Bamberger, Fallen Angels, 113ff and 119ff; Reed, Fallen Angels, ch. 7; Jewish Encyclopedia, s.v.
'fall of angels'.
75
Littmann, 'Hår¥t und Mår¥t', 70ff. Hår¥t and Mår¥t also passed into English literature, both serious and
lighthearted; in Sir Rider Haggard's Ivory Child, published in 1916, they are African magicians who are
announced by the butler as "Mr. Hare-root and Mr. Mare-root" (Encyclopaedia of Iran, s.v. 'Hår¥t and
Mår¥t'; I owe this reference to Mohsen Ashtiany).
76
Reeves, 'Some Explorations', 52ff.
77
See the references given above, 3, 31.
78
Origen is exceptional in holding the same opinion on the Christian side (R. McL. Wilson, 'The Early
History of the Exegesis of Gen. 1.26', Studia Patristica, i, 1957, 420). The Christians typically seeing God
as addressing Christ, the Logos.
79
Thus Reeves, 'Some Explorations'. 53.
80
Bereshit Rabba 8, 5, where the angels are divided over the question.
81
Babylonian Talmud, Sanhedrin 38b, where God responds by destroying them until he gets the answer he
wants.
82
Bereshit Rabba, 8, 4.
 13

in the earth and shed blood", and God overrules them, saying that he knows what they do
not. Here as there God knows something that he is not telling the angels, but here it is the
angels who know about the wicked men of the future, and there is a new stress on
bloodshed. As Reeves says, this would appear to reflect Genesis, 6:11-13, on the
generation of the flood who "corrupted the earth" and "engaged in violence". In the
retelling of that passage in the Book of the Watchers three or four angels observing the
earth from heaven see "much blood being shed upon the earth, and all the oppression
being wrought upon the earth".83 The events leading to the flood having been transferred
to the beginning of human history, the angelic objections are now made, not on the basis
of observation of the earth in the time of Enoch, but rather by way of foresight at the time
of Adam's creation, and the angels no longer comment on the terrible behaviour of giants
in the pre-flood generation, but rather on that of all too fallible humans in general.
("Corruption in the earth" is denounced in many other passages, too.) In the Qur<ån as in
the rabbinic literature, however, God overrules the angels, putting them in their place: for
all their faults, humans have a special place with Him.
In the exegetical literature the transfer of the events from the flood generation to
human pre-history had other repercussions. The Book of Watchers presents the righteous
angels as descending to the earth to fight the giant offspring of their fallen colleagues. In
the exegetical and historical literature of the Muslims, the giants have become an angelic
tribe of spirits (jinn) who lived on earth, where the angel and/or spirit Ibl•s, the future
Devil, was sent to serve as their judge, or he was the ruler of heaven and earth at the time
until he grew arrogant and disobeyed; or the jinn became infidels and caused corruption
in the earth, whereupon Ibl•s was sent against them with an army of angels, which caused
him to become arrogant and rebel.84 These events are sometimes used to explain Adam's
status as khal•fa (deputy or successor): Adam succeeded or replaced those angels or
spirits on earth, we are told. The implicit message is that Adam's title did not mean
"deputy of God on earth", or in other words that the caliphs could not invoke Qur<ånic
support when they styled themselves deputies of God, as they did from >Uthmån
onwards. One part of the story thus came to be associated with a wholly new set of
burning problems, once again in connection with political changes.85
There could be a third reflection, or more precisely development, of a theme from
the Watcher story in the Qur<ånic stress on the fact that the angels only descend at the
command of God, or with His permission, of God. In Q. 19:64 unnamed speakers,
83
1 Enoch 9; cf. Jubilees, 7:21-26, both adduced by Reeves, 'Some Explorations', 53f. There are three
angels in Isaac's translation (Michael, Surafel and Gabriel), four in Nickelsburg and Vanderkam's (Michael,
Sariel, Raphael and Gabriel). Reeves' alternative suggestion that Q. 2:30 alludes to Cain's murder of Abel is
less persuasive. Cf. also his 'Sefer >Uzza wa->Aza(z)el: Exploring Early Jewish Mythologies of Evil', an
account of a monograph in process, at
http://www.religiousstudies.uncc.edu/jcreeves/sefer_uzza_waazazel.htm.
84
For an accessible survey, see $abar•, Ta<r•kh, 79-85; for many others, in which the fallen angels
(complete with the name >Azåz•l) are easily recognized, see M. J. Kister, 'Legends in tafs•r and ˙ad•th
Literature: the Creation of Ådam and Related Stories', in A. Rippin, Approaches to the History of the
Interpretation of the Qur<ån, Oxford 1988, 88ff.
85
Cf. Kister, ' 'Legends in tafs•r and ˙ad•th Literature', 85f; W. al-QåË•, 'The Term "khal•fa" in early
exegetical literature', Die Welt des Islams 28, 1988, esp. 399f, 410f; P. Crone and M. Hinds, God's Caliph.
Religious Authority in the First Centuries of Islam, Cambridge 1986, 6-19; cf. also P. Crone, Medieval
Islamic Political Thought (American title God's Rule), Edinburgh 2004, 40f.
 14

generally assumed to be the angels, declare that, "We do not descend (natanazzalu)
except at the command of your (sg.) Lord", adding that everything belongs to Him and
that He does not forget, with the next verse spelling out the implications: "so worship
(sg.) Him", the Lord of heaven and earth and everything between them. In 97:4 the night
of qadr is identified as the night in which the angels and the spirit descend (tanazzalu)
with God's permission.86 Neither passage has anything to do with the Enoch story (19:64f
comes not long after the mention of Enoch's elevation to an exalted place, 19:56f, but it
does not seem to be connected with it, or indeed with anything that precedes or follows
it). The insistence that the angels only descend with God's permission, or at His
command, is nonetheless striking, especially in 97:4, where there does not seem to be any
reason to stress it. (The exegetes claim that the obscure 19:64 was a response to
Mu˙ammad's impatience at a time when Gabriel had long stayed away.) In 65:12 it is
simply God's command that descends (yatanazzalu), without reference to the angels or
the spirit who serve as its bearers: in all three passages we are reminded that the only
power in the universe is God. The same point is also made in polemics against the alleged
angels of the polytheists: whether they are genuine angels falsely worshipped, demons or
empty names, they have no power (e.g. 7:191-3; 21:42f; 36:23, 74f); it is only with God's
permission that they can act (53:26). Rebellious angels who descended from heaven of
their own accord were problematic from this point of view. But attempts to cast the
Watchers as obedient until their encounter with human females had been made well
before the rise of Islam. In Jubilees (c. 150 BCE) they come down "to teach the sons of
man, and perform judgement and righteousness on earth".87 In the Christian Jewish
Pseudo-Clementine Homilies (c. 300 CE) we are expressly told that they asked for
permission to descend, here as in tafs•r because they are upset by human ingratitude to
God and wish to convict and punish the guilty, that is to act as judges.88 And in Sefer
Hekhalot the angels who descend no longer seem to include the wicked ones at all: it is
the ministering angels who come down from heaven, they do so to execute God's will on
earth, and their descent is placed in the quasi-Paradisical period after the fall familiar
from the Christian story of the sons of Seth and daughters of Cain; that those who taught
mankind sorcery were also angels by origin is left unstated.89 In line with this, the Qur<ån
nowhere says that Hår¥t and Mår¥t had defied God. Since their status as angels is
accepted, they are presented as obedient, even as teachers of sorcery (2:102). What is
stressed here is that they cannot harm anyone "except with God's permission". But the
repeated reminders in other contexts that angels do not descend without being ordered or
permitted to do by God are likely to be inspired by familiarity with claims to the contrary.
Finally, in the Qur<ån, 33:33, the following words are addressed to the wives of
the Prophet: "stay in your houses and do not make a display, like that of the first Jahiliyya
(al-jåhiliyya al-¥lå)". The reference here would seem to be to the women who had been
taught to beautify themselves by the angels (1 Enoch 8:1), an innovation which was

86
My thanks to Joseph Witztum for drawing this passage to my attention.
87
Jubilees, 4:15.
88
Clement of Alexandria (attrib.), The Homilies (tr. A. Roberts and J. Donaldson, Ante-Nicene Christian
Library, xvii, Edinburgh 1870), VIII, 12; the entire passage on the angels is translated in VanderKam, '1
Enoch, Enochic Motifs', 76-78. On the work, see A. Y. Reed, '"Jewish Christianity" after the "Parting of the
Ways"', in A. H. Becker and A. Y. Reed (eds.), The Ways that Never Parted, Tübingen 2003.
89
Sefer Hekhalot, 5.
 15

singled out for particular reprehension by Tertullian in number of writings: the angels
were responsible for the means of "womanly ostentation" such as jewelry and eye-make
up; it was on account of the angels that they had to be veiled, he claimed.90 The Book of
Watchers, as also Jubilees and Christian works, sees the flood as the first global disaster
to overtake mankind, prefiguring the last judgement.91 In the Epistle of Enoch, another
part of 1 Enoch, the flood is explicitly called "the first end".92 That the first end should
have been preceded by the first period of ignorance/barbarism is a natural inference, and
the Christians seem to have made it too, though the one example I have come across
lacks the eschatological perspective.93 Some exegetes duly assign the first Jahiliyya
mentioned in the Qur<ån to the period before the flood, either between Adam and Noah or
between Idr•s and Noah, explaining it with reference to a story of the people of the
mountain versus those of the plains (i.e. a version of the Christian story in which the sons
of God and the daughters of men are replaced with Sethians and Cainites).94
The overall impression conveyed by these references is that the Watcher story
formed part of the general background against which the Qur<ån was revealed. The story
clearly did not come directly from the Book of Watchers. The fact that there are only two
fallen angels in the Qur<ån, that they bear Iranian names and are located in Båbil, that the
angelic comments on human misbehaviour are set in the time of Adam rather than that of
Enoch, that they are associated with the theme of rivalry between angels and humans, and
that Enoch's own time had apparently come to be known as the first Jahiliyya, all this
goes to show that the material had long circulated, in written and/or oral form, in circles
which continued to revere the Enoch tradition, but which also participated in
developments among mainstream Jews and Christians. Three of the four Qur<ånic
allusions come in Medinese suras (2:30, 102; 33:33); the fourth, which is not so much an
allusion as further thought about angelic descent, comes in Meccan ones (19:64; 97:4).

>Uzayr.
The possibly fifth example appears in the 9:30, another Medinese sura, in which the Jews
are famously accused of regarding a certain >Uzayr as the son of God:
"The Jews say, >Uzayr is the son of God; the Christians say, the Mas•˙ is the son
of God. That is what they say with their mouths, imitating the unbelievers before
them. God curse them, how deluded they are".
In the Qur<ån, a son or daughter of God is always an allegedly divine being, usually
Christ or the gods/angels worshipped by the pagans, so the charge that is being levelled
against the offending parties here is deification of created beings. Even when the Jews
and Christians are accused of calling themselves sons of God, the implicit charge is of
deification: the retort includes the point that the Jews and Chritians are just human beings

90
VanderKam, '1 Enoch, Enochic Motifs', 51, 68; cf. also 66, on Clement of Alexandria.
91
J. C. VanderKam, 'The Righteousness of Noah', in G. W. E. Nickelsburg (ed.), Ideal Figures in Ancient
Judaism, Chicago 1980, 25f.
92
1 Enoch, 93,4.
93
Epiphanius identifies the first sect as Barbarism (barbarismos), which lasted from Adam to Noah,
marked by Adam's fall, Cain's fratricide, and the introduction of sorcery, witchcraft, licentiousness,
adultery and iniquity in the time of Jared, Enoch's father (Panarion, book I, ed. H. Kroll, Leipzig 1915; tr.
F. Williams, Leiden 1987, Proem, I, 3, 1; Anacephalaeosis 1, 1; sect 1).
94
Cf. $abar•, Jåmi>, xii, 4f. For the Sethians and Cainites, see the references above, note 20.
 16

created by Him (bashar mimman khalaqa, 5:18). Of course, the Jews and Christians did
not deify themselves, nor is the author of the accusation likely to have thought that they
did: he is simply being polemical. But here at least we know where he got his polemical
ammunition from (ultimately Deut. 14:1; Psalms 82:6; John 1:12). The same cannot be
said in the case of >Uzayr.
The exegetes almost unanimously identify >Uzayr as Ezra,95 and modern
commentators usually follow suit; but the Jews did not call Ezra the son of God, let alone
deify him, as they themselves repeatedly pointed out. This was well known to the
exegetes, who responded by postulating that a small group of Jews had worshipped Ezra
as the son of God in Medina at the Prophet's time, or that just one man had done so, or
alternatively that all of them had done so in the past, when Ezra had restored the Torah to
them, but that they had since stopped, or that they were still doing it somewhere else: the
Jewish denial of the charges could thus be discounted. 96 Similar suggestions have been
made by modern scholars too.97 But the Jews were surely right to remain unpersuaded.
Polemics are not of course a genre conducive to accuracy, but polemical charges
do need a sting in order to hurt, and it is hard to see where it was in this particular case.
The passage proved more of a liability to the Muslims themselves than it did to the Jews.
It thus seems unlikely that Ezra was meant, but what is the alternative?
One of the more interesting suggestions is by Newby. According to him, >Uzayr is
indeed Ezra, but only in name; in substance he is Enoch, with whom, in Newby's view,
he had come to be identified thanks to the fact that both of them were scribes who had
been translated directly to heaven instead of dying; Enoch in his turn was identified with
the angel Metatron, who was regarded as "the lesser YHWH" in circles cultivating
merkaba mysticism; and as Metatron he was chief of the angels who were known as
"sons of God" (and whose appellation might somehow have rubbed off on him?); the
term "son of God" could in any case be applied to any righteous men. "It is easy, then, to
imagine that among the Jews of the Óijâz who were apparently involved in the mystical
speculations associated with the merkâbâh, Ezra, because of the traditions of his
translation, because of his piety, and particularly because he was equated with Enoch as
the Scribe of God, could be termed one of the Bene Elohîm".98
This is a bit complicated. If I have understood Newby correctly, he sees the
Qur<ån as taking issue with Jewish speculation that "perhaps -- God forbid -- there are
two powers in heaven", as disapproving rabbis put it.99 The second power was a principal
angel, who was envisaged in some circles as Metatron, the angel who is a transfigured
version of Enoch in Sefer Hekhalot;100 and the relevant part of this work may be pre-

95
For an exception, see below, note 104.
96
Cf. M. Ayoub, '>Uzayr in the Qur<an and Muslim Tradition' in W. M. Brinner and S. D. Ricks (eds.),
Studies in Islamic and Judaic Traditions, Atlanta 1986, 11ff; H. Lazarus-Yafeh, Intertwined Worlds:
Medieval Islam and Bible Criticism, Princeton 1992, 51ff; J. Dammen McAuliffe (ed.), Encyclopaedia of
the Qur<ån, Leiden 2001-6, s.v. 'Ezra' (Abu-Rabi>).
97
Cf. below, notes 102f. Cf. also V. Comerro, 'Esdras est-il le fils de Dieu?', Arabica 52, 2005, 166f, 170-
72, where other modern suggestions are noted.
98
G. D. Newby, The History of the Jews of Arabia from ancient Times to their Eclipse under Islam,
Columbia SC 1988, 59-61.
99
See the references given above, note 23.
100
3 Enoch, 3-16; cf. Deutsch, Guardians of the Gate, 35f, 53.
 17

Islamic.101 So far, so good, but as the complications suggest, the evidence does not quite
fit. The Qur<ån does not speak of Metatron or Enoch or even Idr•s (the name under which
Enoch usually figures in the Islamic tradition), but rather of >Uzayr -- and getting Ezra
into position as the second power in heaven is hard work. One can try with reference to
IV Ezra (= 2 Esdras, 3-14, in the Apocrypha), 14:9, where God promises Ezra that "you
shall be taken up from among men, and henceforth you shall live with my Son and with
those who are like you, until the times are ended".102 It sounds as if Ezra is being
promised angelification similar to Enoch's here, and the work probably did originally end
with his assumption to heaven.103 For this reason IV Ezra figures in the attempts of
several scholars to solve the problem of >Uzayr, sometimes along with the Greek
Apocalypse of Ezra.104 But it is the messiah, not Ezra, who is here called the son of
God.105 Scholars who focus on the Ezra literature accordingly have to postulate either that
there was a hitherto unknown Jewish sect which elevated Ezra to divine sonship106 or else
that Mu˙ammad simply got his information wrong.107 That Mu˙ammad got something
wrong is also required by the hypothesis that in substance >Uzayr is Malachi.108 Newby
wisely refrains from going down that road.
101
Alexander placed the material about Enoch and Metatron (chs. 3-16) in the period c. 450-850 in an early
article (P. S. Alexander, 'The Historical Setting of the Hebrew Book of Enoch', Journal of Jewish Studies ,
28, 1977, 159, 164); but he has since dated the entire work in its present forn to probably the sixth/seventh
century CE ('From Son of Adam to Second God', in M. E. Stone and T. A. Bergren, Biblical figures outside
the Bible, Harrisburg, PA, 1997, 104f); cf. also J. Dan, The Ancient Jewish Mysticism, Tel Aviv 1975, 122,
124 (c. sixth century CE). For further references, see Reed, 'From Asael and Šemi˙azah', 108, note 10.
102
Cf. B. M. Metzger (tr.), 'The Fourth Book of Ezra', in J. H. Charlesworth (ed.), The Old Testament
Pseudepigrapha, i, New York 1983, with an introduction.
103
Cf. R. A. Kraft, '"Ezra" Materials in Judaism and Christianity', in Aufstieg und Niedergang der
Römischen Welt: Geschichte und Kultur Roms im Spiegel der Neueren Forschung (ed H. Temporini and W.
Haase), II (Principat), 19.1, Berlin 1979, 129. The text casually refers to "before he was taken up" (8:19);
Ezra is told to divest himself of his human nature (14:14), and the work concludes with his assumption in
eastern versions, including the Syriac (given in the margin in the New Revised Standard Version of the
Bible).
104
Cf. M. E. Stone, 'Greek Apocalypse of Ezra', in J. H. Charlesworth (ed.), The Old Testament
Pseudepigrapha, i, New York 1983. Here Ezra actually gets to heaven, but apparently only on a temporary
basis, since the text ends with his death and burial (7:14f).
105
The son also appears at IV Ezra, 7:29, on "my son the Messiah".
106
J. Horovitz, Koranische Untersuchungen, Berlin and Leipzig 1926, 127f; H. Speyer, Die Biblischen
Erzählungen im Quran, Gräfenhainischen n.d. (preface dated 1931), 413. Even if such a sect had existed, it
is hard to believe that it should have come to Mu˙ammad's attention without being known to the Jews
themselves. For comparable suggestions without reference to the Ezra literature, see H. Z. Hirschberg,
'Ezra', in Encyclopaedia Judaica (a Yemeni sect postulated by Ibn Óazm); J. Walker, 'Who is >Uzair?', The
Muslim World 19, 1929, 303-6 (the Samaritans made up the charge).
107
Cf. D. Künstlinger, '>Uzair is der Sohn Allåhs', Orientalistische Literaturzeitung 35, 1932, 381-3,
suggesting that Mu˙ammad mistook the name of the book for the name of the son of God mentioned in it.
For another hypothesis requiring him to mistake something, cf. Ginzberg in the following note.
108
This theory was advocated already in medieval times. Al-Biqå>• (d. 885/1480), NaΩm al-durar f• tanåsub
al-åyåt wa 'l-suwar, Hyderabad 1969-84, viii, 439, cites the Jewish convert al-Samaw<al al-Maghrib• for the
view that >Uzayr is not Ezra in the sense of the restorer of the Hebrew Bible. Al-Mahghrib•'s own view was
that >Uzayr, whom he calls "al->Uzayr", was Eleazar (cf. his If˙åm al-yah¥d, ed. M. >A. al-Sharqåw•, Cairo
1986, 152), but al-Biqå>• says that >Uzayr is the prophet Malachi (who is identified with Ezra in the
Talmud (Megilla 15a) and elsewhere). The same suggestion was made, apparently independently, by L.
Ginzberg, Legend of the Jews, Philadelphia 1909-38, vi, 432. Since Malachi (which means "my angel") is
 18

Metatron and Enoch were not called sons of God either, however. This is why
Newby claims both that Metatron was chief of the fallen angels known as "sons of God"
and that any righteous man could be thus called. But the first claim appears to be a
straightforward mistake. To the extent that Metatron was Enoch, he was indeed
associated with the fallen angels known as sons of God, but he was not their chief and he
did not have any connection with them as Metatron; nor is there any evidence that the
label "son of God" was ever transferred to him. As for the second claim, it was back in
Graeco-Roman times any righteous man could be known as son of God.109 The rabbis are
said to have extended the sonship to every Israelite, or indeed every human being, at
some point, but this does not actually solve the problem; for what the Jews are being
accused of in Q. 9:30 is not calling all Israelites or human beings sons of God (à la Q.
5:18), but rather of using the expression of one specific figure; and if it had been possible
to speak of Enoch/Ezra as a son of God in circles cultivating merkaba mysticism, why
were the Jews so unanimously puzzled by the Qur<ånic charge?
Given that the Jews simply did not recognize the sin they were accused of, it
seems more likely that the identification of >Uzayr as Ezra is mistaken. This possibility
was rejected by both Horovitz and Künstlinger, but their own suggestions work no better
than Newby's.110 Where do we go from there?
Newby must be right that we are up against something to do with angel worship.
This could admittedly be questioned on the grounds that 9:30 presents the Jewish view of
>Uzayr and the Christian view of Christ as parallel errors, suggesting that both parties
were guilty of deifying human beings. But the parallelism lies in the fact that both are
deifying created beings. Worship of Christ and angels is also put on a par in 4:172: "The
Messiah, Jesus son of Mary does not disdain being a servant of God, nor do the angels
who are drawn near (al-malå<ika al-muqarrab¥n)". Like Christ, the angels wrongly
deified by the pagans were actually "righteous servants", or this is one view of them in
the Qur<ån (Q. 21:26; 43:19; elsewhere they are demons or empty names); that the one
was a human being and the others angels was immaterial, and so it would have been in
the case of the Jews. If we persist in the search for a human being called "son of God" by
the Jews, we are unlikely to get beyond the conclusion that Mu˙ammad simply got
something wrong. If we are prepared to consider the possibility that he knew what he was
talking about, the only way in which Jews could be plausibly be accused of polytheism
was with reference to their "logos theology", as Boyarin calls it, or in other words their
veneration of a divine power, personified as an angel, as an intermediary between God
and mankind.111 "To this logos, His archangel, the Father of all has given the special

nowhere called a son of God, Ginzberg asks whether Mu˙ammad confused "messenger (i.e. angel) of God"
with "son of God". (I owe almost all of this to Joseph Witztum.)
109
The Jewish Encyclopaedia, New York [1964], s.v. 'son of God', with reference to Wisdom of Solomon,
2:13, 16, 18; Ecclesiasticus, 4:10; cf. also Philo, On the Confusion of Tongues, 145; id., Questions and
Answers on Genesis, I, 92; Justin Martyr, Apology, 1, 22, 2, where Jesus would deserve to be called son of
God for his wisdom alone even if he were wholly human. Cf. also G. Delling, 'Die Bezeichnung "Söhne
Gottes" in der jüdischen Literatur der hellenistic-römischen Zeit', in J. Jervell and W. A. Meeks (eds.),
God's Christ and His People: Studies in Honour of Nils Alstrup Dahl, Oslo 1977, 18-28.
110
Horovitz, Koranische Untersuchungen, 167, note to p. 127; Künstlinger, '>Uzair is der Sohn Allåhs',
382; cf. above, notes 102f, for their suggestions.
111
See the reference given above, note 23.
 19

prerogative to stand on the border and separate the creature from the creator", as Philo (d.
50 CE) expressed it, happily referring to the logos as "the second God" and "son of
God".112 The Christians duly took Philo to have been a Christian avant la lettre, but it is
now recognized that such ideas were widespread in Judaism, especially in the first two
centuries CE, but apparently much later too. In Sefer Hekhalot, a composite work which
may date in its present form to the sixth/seventh century,113 it is Enoch-Metatron who is
the second God, or more precisely "the lesser God", as Newby observes. The trouble is
that no form of the intermediary is known ever to have been called anything like Ezra.
Some eighty years ago, however, Casanova proposed that >Uzayr is a misreading
of Azael.114 He made the suggestion in a brief communication without developing the
thesis, and perhaps for this reason it fell flat: in effect, he simply substituted one strange
name for another, with perfunctory reference to rabbinic sources. Even in our present
state of knowledge it has to be said that the thesis has its problems. But as Wasserstrom
says, it deserves to be revived,115 if only for a proper hearing.
If >Uzayr is a misreading (or mishearing) of Azael, the force of the passage would
be that the angelic intermediary venerated by the Jews was actually an evil figure, a
demon trying to mislead them. This works well in that it is also one of the reactions to the
pagan worship of angels/deities in the Qur<ån: on the day of judgement God will ask the
genuine angels, "Was it these who worshipped you?", and the angels will reply, "Glory
be to you. You are our friend (wal•), not these. Rather, they worshipped the demons
(jinn)" (34:40f). Or again, "they have made the demons (jinn) partners of God, though He
created them, and falsely credit Him with sons and daughters, with knowing anything
about it" (6:100).116 The charge is all the more plausible in that Azael was an ambivalent
figure. Though he was widely known as a fallen angel, a Greek amethyst lists Ichtys,
identified as Christ by the chi-ro sign, with Raphael, Renel, and Uriel on one side and
Michael, Gabriel, and Azael on the other;117 and in the Aramaic magical texts he is
sometimes a fallen angel118 sometimes an anti-demonic power along with the archangels
Michael and Raphael. One amulet goes so far as to include him among "the holy angels
who stand in front of the throne of the Great God";119 one bowl text invokes Aza and
Azael (>z< w->z<l) as well as Metatron against the demons;120 another, directed against

112
Philo, Who is the Heir of Divine Things, 205; Questions and Answers on Genesis, ii, 62 (second God);
On Husbandry, 51 (firstborn son).
113
Cf. above, note 94.
114
P. Casanova, 'Idrªs et >Ouzaïr', Journal Asiatique 205, 1924, 356-60 (opting for the form >Uzi<el).
115
S. M. Wasserstrom, Between Muslim and Jew. The Problem of Symbiosis under Early Islam, Princeton
1995, 183.
116
Compare also 37:158: "They have set up a genealogical relationship (nasab) between Him and the
jinn". Contrast 21:26 and 43:19: "They say, 'al-Ra˙mån has produced children', but they are servants raised
to honour (>ibåd mukram¥n); "they have made the angels, who are servants of the Ra˙mån, females".
117
J. Daniélou, The Theology of Jewish Christianity, London 1964, i, 122, citing F. Cabrol and H. Leclercq,
Dictionnaire d'Archéologie chrétienne et de liturgie, i, Paris 1924, s.v. 'anges', col. 2088f (where no date is
offered).
118
Thus in the reference given above, note 66.
119
Naveh and Shaked, Amulets and Magic Bowls, 68f, amulet 7:3-5; compare 1:1-3; Magic Spells and
Formulae, 62f, amulet 19:17-23; cf. also Milik, Books of Enoch, 131.
120
C. H. Gordon, 'Aramaic Incantation Bowls', Orientalia 10, 1941, pp. 279f (Ashmolean, no. 1932,620),
transliterated as >Azza and >Azza<el.
 20

Jewish, Arab, Persian, Indian, Greek and Roman sorceries, and of disputed date, invokes
God, "who sent Aza, Azael and Metatron (>z< w->z<l w-myããrwn), the great prince of His
throne".121 Here Metatron takes the place of the third member of the >z trio, normally
Azael or Azazel. One would dearly like to know what the Qur<ån has in mind when it
holds the Jews and Christians who credit God with a son to be imitating "the unbelivers
before you" (alladh•na kafar¥ min qablu): is this another reference to the "first Jahiliyya",
now singled out for the polytheism rather than the female immodesty that prevailed at the
time? There does not seem to be any way of telling.
If Casanova's emendation is accepted, sura 9:30 would reflect much the same
environment as the passage on Hårut and Mår¥t, and carry much the same message: "the
Jews say that Azael is a son of God", i.e. the Jews prefer a figure associated with magic
to the messenger who has been sent to them. The next verse tells us why: they, i.e. the
Jews and the Christians, "have taken their learned men (a˙bår) and monks (ruhbån) as
their lords (arbåban) apart from God, and al-Mas•˙ son of Mary" (9:31). This can hardly
be a restatement of the claim that they deify >Uzayr and Christ, for Christ does not belong
in the category of ruhbån, and he is mentioned again as a separate object of deification. It
is equally implausible that the Jews should be accused of deifying an >Uzayr from among
their a˙bår and ruhbån, while the Christians deify Christ, for ruhbån are always Christian
figures in the Qur<ån. Most probably, then, this verse attacks the authorities to which the
Jews and Christians owe their horrendous beliefs: they elevate their own authorities to the
position of God by following them in defiance of God, as represented by the Messenger.
That they deify al-Mas•˙ appears to be mere repetition, possibly because it was the only
genuine charge, and one that the Messenger often took issue with. "They were
commanded only to worship one God", the verse continues, presumably meaning in the
past and now also through the Messenger whom they ignore. Both groups hope to
extinguish the light of God with the enormities they utter with their mouths, the
Messenger says (9:32), using a phrase elsewhere associated with those who dismissed
Jesus as a magician (61:6, 8). In short, here as in 2:102 the key issue seems to be the
Messenger's own authority.
On this reading, the charge against the Jews would not reflect ignorance or
misunderstanding of a Jewish belief, but rather anger and the polemical exaggerations
that this tends to induce. The observer knew very well that the Jews did not really
worship an angel, righteous or fallen, as the son of God: had he genuinely believed that
they did, he would have argued against it as frequently as he did against the sonship of
Jesus, not just on a single occasion. He is claiming that idolatry is what the beliefs of his
stubborn opponents really amount to. How literally did he intend the charge? It could be
argued that all he resented was the expression "son of God", having heard Azael
described as such on some occasion: metaphorically meant or otherwise, it was wrong to

121
C. H. Gordon, 'Aramaic Magical Bowls in the Istanbul and Baghdad Museums', Archiv Orientáln* 6,
1934, 328-30 (Baghdad Museum bowl no. 6519:11), transliterated as >Azzå and >AzzâŸl. Gordon places this
bowl after the Arab conquests with reference to the "Sh•>ite" sorceries in line 9; but Shaked reads "Arab"
sorceries and regards it as probably pre-Islamic (S. Shaked, 'Jews, Christians and Pagans in the Aramaic
Incantation Bowls of the Sasanian Period', in A. Destro and M. Pesce (eds.), Religions and Cultures,
Binghamton 2001, 72f).
 21

say such things.122 If this is all he meant, his reaction would be comparable to that of a
third/fourth-century Palestinian rabbi, who was offended by the passage in Daniel in
which Nebuchadnezzar observes in amazement that the three youths in the fiery furnace
are unharmed and that they have been joined by a fourth "like the son of God", i.e. an
angel (Dan. 3:25): the rabbi claimed that an angel came down and slapped
Nebuchadnezzar on the mouth for presuming that God had a son.123 What he meant was
presumably no more than that one should not say anything conducive to the blurring of
the boundaries between Jews and Christians (though he too could have been worried by
Jewish "logos theology").
That the Qur<ånic observer was only bothered by words is at first sight suggested
by the fact that he dismisses the claims regarding >Uzayr and Christ as "(just) something
they say with their mouths" (dhålika qawluhum bi-afwåhihim)" (9:30), a phrase elsewhere
used of comparing one's wife to one's mother for purposes of repudiation and calling
somebody else's son one's own by way of adoption (33:4). But the parallel passage is not
in fact about words alone. What 33:4 rejects is the opponents' belief that the words create
or reflect something real, whereas the truth is that they are only words (similarly 18:4f;
24:15; 61:1-8). Similarly, the truth about the angels/gods deified by the pagans is that
they are just names (when they are not angels falsely worshipped or mere demons, 7:70f;
53:23). It is only when opponents are held to say with their mouths what is not in their
hearts, or to have even worse thoughts in their hearts than in their mouths, that a
distinction between mere words and actual beliefs is postulated (3:118, 167; 5:41; 9:8;
48:11f). In the passage on >Uzayr and al-Mas•˙ the opponents do not simply use an
offensive expression; they go so far as to reject the Messenger for the sake of the belief
expressed by it. This is why they are accused of deifying their leaders too, and also why
they are told that God has sent the Messenger with guidance and the religion of truth to
make it prevail even though the mushrik¥n may dislike it (9:33). The intertextual
reference to the Jews who dismissed Jesus as a magician (in its turn conjuring up the
polytheists who dismissed the Messengers as such) also shows the issue to be divine
authority, this time along lines similar to the passage on Hår¥t and Mår¥t: the Jews go for
the wrong leaders, preferring the very magic that they wrongly impute to the genuine
Messenger. In 5:17-8, where only the Christian belief in a son of God is mentioned, both
the Jews and Christians are once more accused of deification, not of their leaders, but
rather of themselves: the concept of shirk extends to any form of authority other than that
of God represented by Mu˙ammad, as Comerro observes.124 Here, too, the passage
culminates in the claim that "Our messenger has come to you" (5:19).
In short, 9:30 is directed, not just against the offensive expression "son of God",
but also against actual beliefs held by Jews and Christians and the leaders under whom
they upheld them. As far as the Jews are concerned, it has to be said if the reference is
indeed to Azael, the charge was a brilliant polemical move, for it was one to which the
Jews could only reply, "yes, but.....": yes, the fallen angels were described as sons of God
in the Bible and the Book of Watchers; yes, Azael was one of them; and yes, it was to
him that humans owed their knowledge of magic, in which he was often called upon; but

122
This possibility was suggested to me by Behnam Sadeghi.
123
Alexander, 'Targumim and Early Exegesis', 61f, with reference to Jerusalem Talmud, Shabbat, 6, ad fin.
124
Comerro, 'Esdras', 170.
 22

the implication that he was worshipped in the same way that the Christians worshipped
Christ was not true at all. What with so complicated a defence, the audience will have
inferred that there was something to the charge. It merely so happened that a scribe was a
bit too fast when he copied the name of the demonic figure, or alternatively that the name
had come to sound too much like Ezra's, so that far from persuading later readers that the
Jews had sold their souls to the devil, the verse persuaded the Jews and Christians that
Mu˙ammad was an ignoramus: his scripture was full of errors, the Jews did not regard
>Uzayr as the son of God, as the Christians said in polemics of their own.125
There are no problems with Casanova's theory on the linguistic front. In Arabic,
Azael (>Azå<Ÿl) would be written >Az(å<)•l. The main difference between that (>zyl) and
>Uzayr (>zyr) in early Arabic script would be the size of the final letter:126 if the change
took place in written transmission and the copyist was transcribing from Arabic to
Arabic, all we need to put things right would thus be to postulate a minor scribal mistake.
But the change could also have been effected orally, given that the shift from l to r is
common in Semitic languages:127 in that case all we need to postulate is that Azael was
pronounced as something like Ozael/Ozaer. We do in fact find the forms Uzael (>wz<l),
Uziel (>wzy<l), and related forms, both on undatable magic amulets from the Syria region
and in texts on incantation bowls from Iraq.128
But one could also suggest other ways of achieving the same result. For example,
an angel by the name of >Azriel (>zry<l) figures on amulets and magic bowls.129 He has no
independent existence, and is nowhere identified as the son of God, but as yet another
bearer of a name on the theme of >z, he could easily have been treated as another
manifestation or associate of Azael, as in fact he seems to be in the trio Azael, Azriel, and
Ariel (all righteous).130 The diminutive would presumably be >Uzayr<•l, but the<•l would
have been rejected on the grounds that God had nothing to do with him, leaving the
contemptuous >Uzayr. Or, as Comerro suggests, the name Azariah (>zryh) could be
lurking in the background. It was borne by one of the three youths who were thrown into
the fiery furnace and saved by the angel described by the awed Nebuchadnessar as "like
the son of God".131 As it happens, it was also the pseudonym adopted by the archangel

125
al-Jå˙iΩ, 'al-Radd >alå 'l-naßårå', in his Raså<il, ed. >A.-S. M. Hår¥n, iii, Cairo 1979, 303f, 333f, 343, cf.
also 346f.
126
Cf. B. Gruendler, The Development of the Arabic Scripts, Atlanta 1993, 59, 95.
127
Cf. E. Lipiński, Semitic Languages: Outline of a Comparative Grammar, Leuven 1997, 135 (my thanks
to Adam Raziel for arguing the case for oral transmission and mentioning this work ).
128
H. Gordon, 'Aramaic Incantation Bowls', Orientalia 10, 1941, 123 (no. 5:7); J. B. Segal (ed. and tr.),
Catalogue of the Aramaic and Mandaic Incantation Bowls in the British Museum, with a contribution by E.
C. D. Hunter, London 2000, nos. 040A:6, 048A:36, 109M:10 (my thanks to Dr Hunter for drawing this
work to my attention); Naveh and Shaked, Amulets and Magic Bowls, 40f, (amulet 1:1); cf. 218f (magic
book from Islamic times, 2, 5); J. Naveh and S. Shaked, Magic Spells and Formulae: Aramaic Incantations
from Late Antiquity, Jerusalem 1993, 62f, 66 (Palestinian amulet 19:23).
129
J. A. Montgomery, Aramaic Incantation Texts from Nippur, Philadelphia 1913, 154 = 156, no. 8:14;
Gordon, 'Aramaic Incantation Bowls', 123 (no. 5:7); Naveh and Shaked, Amulets and Magic Bowls, 40f
(amulets 1: 13). Cf. also E. A. W. Budge, Amulets and Superstitions, London and New York 1930, 277f.
My thanks to Shaul Shaked for noting the possible relevance of this angel.
130
Naveh and Shaked, Amulets and Magic Bowls, 68f (amulet 7:3).
131
Dan. 1:6f; 3:20-25; Comerro, 'Esdras', 172f, with exegetical traditions which actually identify >Uzayr as
one of Daniel's companions.
 23

Raphael in Tobit, in which Raphael teaches Tobias to make medicine and drive away a
demon, which Raphael then binds (Tobit, 8:11). Curing by binding demons is what most
of the magic bowls of Iraq were designed to achieve. They often invoke Raphael. Given
the premium on invoking as many powers as possible, the chances are that they invoked
him under the name of Azariah too, thereby causing him to join the list of angels/sons of
God whose names are variants on the theme of >z. In fact, it could have been Azariah-
Raphael who generated the above-mentioned Azriel.
Very few of the angels and demons invoked in magic texts had stable
personalities, and strange-sounding names proliferated, but to disapproving observers
such as the rabbis, the names on the theme of >z invariably conjured up fallen angels, or
even the devil himself. The chances are that the same was true for the Messenger. In
whatever form he may have heard the name of the offending son of God, he will have
understood him as a rebellious angel and used him to unmask the wayward beliefs of the
Jews: what they so stubbornly rejected him for was a demon.
If this is accepted, one would assume the transmissionto have taken place via Iraq,
as in the case of Hår¥t and Mår¥t. Of course, >Uzayr could have entered via a different
channel, but the Ethiopic Book of Watchers uses the forms Asael and, more commonly,
Azazel,132 making Ethiopia unlikely as a source for our verse. The form Azael is attested
in both Syria and Iraq. In Greek it is attested in the fifth/sixth-century Codex
Panopolitanus and the ninth-century Syncellus, in both of which it is the only form used
(AzaŸl); in the Christian(-Jewish?) amethyst, and in two magical texts from Egypt, one
dating from the fourth-century, the other from the sixth or seventh.133 In Aramaic it
appears on a Palestinian amulet,134 in rabbinic sources,135 and on incantation bowls.136
The related forms Uzael, Uziel, and Azriel also appear on both sides.137 We know from
Justinian that there were Jews on the Byzantine side who denied that "the angels exist as
God's work and creation", i.e. they held the angels to be uncreated (and thus divine), but
whether they venerated a principal angel is not said.138 It is in the Babylonian Sefer
Hekhalot that the identification of Enoch/Metatron as the lesser God is attested, in circles
associated with magic, just as it is here that the magical texts associate Azael with
Metatron.139 It is also on an Iraqi bowl text that we encounter the actual expression "sons
of God" in conjunction with (but not clearly identified as) Azael, Azael, and Azaziel.140
One would thus assume Iraq to have been the source.
132
Black, Book of Enoch, 121.
133
Above, note 111; K. Preisendanz (ed. and tr.), Papyri Graecae Magicae, Leipzig 1928-31, ii, nos. xxxvi,
174 (also Aziel); xlv, 3.
134
Above, note 113.
135
Cf. above, note 44.
136
Cf. above, notes 66, 114-15.
137
Cf. above, note 122.
138
Justinian (legislating in 553), Novella 146, peri Hebraiøn, ed. and tr. A. Linder, The Jews in Imperial
Legislation, Detroit and Jerusalem 1987, 406f = 409. Practically all the voluminous literature on this
novella is about Justinian's regulation of the language to be used in the synagogue. That he also legislates
against Jews who deny the resurrection and last judgement, as well as the createdness of the angels, seems
to have passed virtually unnoticed.
139
See the references given above, notes 54, 114-5.
140
Thus the curse text mentioned above, note 66. In this text (which mixes Jewish, Christian and pagan
elements), the "lower foundation" of the universe is occupied by seven "sons of God" (bny <lhy), who keep
the universe together with seven powerful words (Levene, Corpus, M163:9, and commentary). By origin,
 24

Problem.
So far, Casanova's hypothesis seems to work wonders, but it raises one intractable
problem. Why did the exegetes not recognize Azael behind his new name? The nearest
we get to it is Muqåtil's claim that >Uzayr was described in the Pentateuch (tawråt), by
which, of course, he could simply have meant somewhere in the Hebrew Bible.141 It is all
the odder in that the exegetes effortlessly recognized the fallen angels behind their
Iranian names of Hår¥t and Mår¥t.
It has to be stressed that the exegetical reaction to >Uzayr is peculiar even if we
discard Casanova's hypothesis, for given that the early exegetes knew very well that Jews
did not deify Ezra, one would have expected them at the very least to discuss the person
intended before settling on the identification of Ezra and >Uzayr. Instead, they
unhesitatingly agreed on what must in fact be a mistake. Apparently, they had grown up
with the idea that Ezra was intended. As early as 170/786 certain Sa>d pronounces
Mu˙ammad, Jesus and >Uzayr along with all created beings, to be marb¥b¥n, servants of
God, in an inscription.142 If the exegetes had grown up with their understanding, one
wonders if Mu˙ammad had as well. "God does not command you to take the angels and
prophets as lords", we are told in Q. 3:80: did >Uzayr belonged in the category of
prophets rather than that of angels to the speaker of those words? If he did, we are back
where we started: how could the author of the Qur<ån claim that the Jews called Ezra the
son of God?
Either the Messenger's understanding is peculiar or else that of the exegetes is: in
effect, Casanova's hypothesis merely shifts the problem from the one to the other. It has
to be said, however, that this is not the only occasion on which the exegetes settled
without discussion or disagreement on what to a modern scholar looks like an obvious
mistake; there is another example in Q. 24:33, and here there can be no doubt that they
did so in departure from authorial intentions.143 Of course, exegetes are everywhere in the
habit of disregarding authorial intentions: they make of the revelation what they need.
But they do not usually do so with complete lack of hesitation or disagreement, least of
all when they know their interpretation to be problematic. This suggests that the problem
of >Uzayr lies in the early history of Muslim exegesis rather than in the understanding of
the Prophet, or in other words that Casanova had the better hunch. Until that can be
shown, however, the verdict on his theory must be "not proven". That the accusation in

they are presumably a new version of the fallen angels, but whether the magician sees them as such is not
clear. He proceeds to speak of the "sons of glory" (line 13) and thereafter about the trio Azael, Azael, and
Azaziel, pressed under a mountain with their faces downwards (line 18). Levene thinks that the "sons of
glory" may be identical with the "sons of God" (comm. to line 13), but does not say whether he thinks the
same could be true of the trio.
141
Comerro, 'Esdras', 168, citing Muqåtil b. Sulaymån, Tafs•r, ad 112:1.
142
Y. D. Nevo, Z. Cohen and D. Heftman (eds. and trs.), Ancient Arabic Inscriptions from the Negev, i,
Midreshet Ben Gurion, Negev, 1993, 54 (no. ST 640(34), cf. plate 34. (My thanks to Haggai Ben Shammai
for this reference.)
143
Cf. P. Crone, 'Two Legal Problems bearing on the Date of the Qur<ån', Jerusalem Studies in Arabic and
Islam 18, 1994 (reprinted in ead., From Kavåd to al-Ghazål•, Aldershot 2005), 3ff, on the kitåb understood
as a manumission document, though the context dictates that it is a marriage document.
 25

9:30 refers to Jewish veneration of an angelic viceregent remains the most plausible
solution; how this being acquired the name of >Uzayr remains unclear.

Conclusion.
Though Azael's presence in 9:30 remains conjectural, the four other echoes of the Book
of Watchers have at least done something to relate the Qur<ån to a well-documented
context on the fringes of the Arab world in late antiquity. Relating the Qur<ånic material
to earlier traditions could be said to be one of the most pressing needs for historians of
the rise of Islam.144 This is now coming to be generally recognized, after a long hiatus in
which origins tracing had a bad name. One can see why it was rejected. Back in the late
nineteenth and early twentieth century there was a tendency for Western scholars to
envisage Mu˙ammad as picking up bit and pieces of religious lore from his Jewish,
Christian and diverse other neighbours without much understanding of what they meant,
in order to mix them all up and then use them in altogether different contexts in Mecca
and Medina. Much work was done on where he had picked up his bits and pieces, but
given the magpie model, it did not seem illuminating: what did it matter whether this
came from Christianity and that from Judaism, or somewhere else if it had all been
denatured in the process of transmission? More recent scholars not unnaturally found the
interest of the ideas to lie less in their origin than in their meaning in the new contexts to
which Mu˙ammad applied them. Origins-tracing never seemed to further our
understanding of anything, merely to harp on the theme of the "parasitic dependence" of
Islam on earlier religions, as Reeves says.145
But religious (and all other) ideas do grow out of earlier ideas, by tiny incremental
changes; even revolutionary changes are achieved by very small steps; and though the
older literature never showed these steps, merely a haphazard collection of information
and mistakes, as if Islam had arisen by misunderstanding, there was more than prejudice
to the picture it presented. The Orientalists were reacting to the fact that it was, and
remains, extremely difficult to overcome the sense that Islam arose in a world apart. The
tribal societies evoked in pre-Islamic poetry, the ayyåm, Ibn Hishåm or al-Wåqid• are so
utterly different from the Near East described in Greek, Syriac, Aramaic, Coptic or
Iranian works that one automatically classifies ideas which can be shown to have
originated in the non-Arabian Near East as "foreign elements", or in other words as
features appearing out of their normal context, so that they have to be explained by
mechanisms such as traders accidentally picking up this and that on their journeys.
What we see in the Qur<ånic treatment of the fallen angels in the four (possibly
five) passages examined here, however, is not the impressions of a passer-by who had
picked up some ancient story without much sense of what it meant to his informants.
What we see is the story in the context to which it had come to belong by late antique
times, complete with the magic practices it was held to explain and the angry sense of
being outflanked by disreputable people that the situation induced in the observer.
Wherever or whenever the encounter(s) took place, the observer is engaging with the
144
Cf. F. E. Peters, 'The quest of the Historical Muhammad', International Journal of Middle East Studies
23, 1991, 292f.
145
Cf. Reeves in his introduction to his Bible and Qur<ån, ix.
 26

tradition as it looked at his time, not simply plundering it, let alone getting things wrong.
Looking back, we can follow the tradition he grappled with till it disappears in the dawn
of history; looking forward, we can see what it came to mean to his many followers
thereafter down until today. Islam here grows by imperceptible steps, however drastic the
observer's reaction, out of the environment that was before it, creating a new one as it
does so. It would be enormously illuminating if we could see the entire book in this way.