(1477285X - The Journal of Jewish Thought and Philosophy) Sefer Yesira and Early Islam - A Reappraisal

J. Jewish Thought and Philosophy, Vol. 3, pp.
1-30 © 1993
Reprints available directly from the publisher
Photocopying permitted by licence only
Sefer ye~ira and Early Islam: A

Reappraisal
Steven M. Wasserstrom
Reed College, Portland Oregon
A book impenetrable by contingency,

a mechanism of infinite purpose,
of infallible variations,
of revelations lying in wait,
of superimposed light
Borges
The short, cryptic Book of Creation, Sqer Yefira, (SY) entered] ewish academ-
ic life, if you will, with full tenure. From obscurity to the sustained close
attention of several major Jewish thinkers of the tenth century, this work
seemed to come out of nowhere. This apparently instant eminence, and
the absence of any historical background on this treatise, has reduced most
scholars searching for the origins of SY to look for isolated motifs in vari-
ous works of late antiquity. This largely fruitless procedure has resulted
neither in a substantial consensus nor in anything near a precise dating.*
• I thank William Brinner for his comments on an earlier, oral delivery of this paper, during a
session of the Annual Meeting of the American Academy of Religion in 1986. Thanks also to
Ronald Kiener, Michael Sells, Edwin Gerow, Sarah Stroumsa and Elliot Wolfson, who provided
fcedback on subsequent versions. All flaws herein are mine.
The status quaestionis is given by Nehemiah Allony in "The Time of Composition of Sefer
Yeq:ira," [Heb.] pp. mem-aleph to nun in Temirin 2 (1981) and Nicolas Sed, "Ie Sefer Ye~ira,
L'Edition critique, Ie texte primitif, la grammaire et la metaphysique," Revue des etudesjuives CXXXII
(1973) pp. 513-528. Also Vajda, "Recherches recentes sur l'esoterisme juif," Revue de I'histoire des reli·
gions 192 (1977) pp. 42-45; and Gershom Scholem, Origins of the Kabbalah (princeton, 1987) pp.
24-35. These surveys include a review of the available texts. Notice may also be taken of Sefer
Yef:{jrah. The Book of Creation, In Theory and Practice by Aryeh Kaplan (York Beach, Maine, 1990):
while not academic in orientation, this work does translate four versions of SY, and provides per-
haps the fullest available overview in English of the traditional commentaries on SY.
A review of the early philosophical commentaries now is provided by Raphael Jospe, "Early
Philosophical Commentaries on the Sefer Ye::<irah: Some Comments" Revue des etudesjuives CXLIX
(1990) 369-415.
Jorge Luis Borges, "A Vindication of the Cabala", Borges. A Reader, (eds.) E.R. Monegal, A. Reid
(New York, 1981) 22-24, at 24.
Downloaded from Brill.com08/02/2019 01:13:42PM

via Bar Ilan University
2 Steven M. Wasserstrom
In the present essay I shall reappraise the arguments that SY may have
been redacted in the form received by its first commentators sometime in
the ninth century. This hypothesis was set forth, most notably, by the
l
Islamicists Louis Massignon, Paul Kraus and Henry Corbin. Other schol-
ars who specialized in the early Islamicate period, especially Georges Vajda
and Nehemia Allony, further modified this theory.2 I will begin my discus-
sion with a review of the Kraus hypothesis. Next, I shall present circum-
stantial evidence demonstrating that in the early ninth century we can find
forms of cosmic semiotics and gnostic encyclopedism closely related to
those found in SY. In addition, I will try to show that this historical con-
text does satisfy adequately the form, content and function of this strange
text.
By a review of context, form, content and structure, then, I hope to do
more than present circumstantial evidence; to list yet another litany of
selected parallels; or to engage in one more comparative motif-analysis. In
so doing, however, I shall neither venture a close reading of SY as such,
nor of the early commentaries on it. I intend to be neither conclusive nor
comprehensive. My modest goal herein is (for what I believe to be the first
time) to review and reappraise the historical evidence pertaining to the
Islamicate setting for the earliest known versions of SY. This historical
period must be reconsidered in studies of SY, if we are to account for the
unique form, content, function of the extant redactions of SY.
Part I: The Kraus hypothesis
The Imrnot Alef/ Mem/ Shin

It was Louis Massignon who first noticed that the "Mother-letters" (Heb.
immot; Ara. ummahiit) of the third chapter of SY - alif, mem and shin -
] Louis Massignon, Salmdn Pdk et les premices spirituei/es de l'Islam iranien (Publications de la Societe des
Etudes Iraniennes et de I'art persan, # 7) Tours, 1934, p. 394, subsequently amplified in his "La
Philo sophie orientale d'Ibn SIna et son alphabet philosophique", Memorial Avicenne, Institute franfais
d'archeologie orientale IV (1952) 1-17, at p. 12. Kiaus followed Massignon's suggestion (which was
published later than the work of Kraus), and provided the most important parallels to date (see
below, esp. nn. 19-20). This notion also was fully approved by Henry Corbin: Avicenna and the
Visionary Recital (New York, 1960, repro Dallas, 1980) p. 276. Gershom Scholem, likewise, was
impressed with Kraus' findings: see Major Trends in Jewish Mysticism (New York, 1941, repr. 1972),
where he acknowledges that Kraus had adduced "some remarkable similarities between the Book of
Creation and early Islamic gnosticism ... " (p. 368 n. 128). But Scholem does not make the argu-
ment for priority made by Massignon and Corbin.
2 Georges Vajda, "Les lcttres et les sons de la langue arabe d'apri:s Abu I:Iatim al-RazI", Arabica 8
(1961): 113-130; "Le commentaire kairouanais sur Ie 'Livre de la Creation'," Revue des etudes juives
(105) 132-40; 107 (1946/47) 99-156; 110 (1949/50) 67-92; 112 (1953) 5-33; 119 (1961) 159-161;
Nehemia A1lony, "The Time of Composition" (see first note above).

Seier Yef}ira and Early Islam 3
would seem to be related to the ShI'I -gnostic triad of 'ain) mim and sin.3
To account for the discrepancy in the first term of these triads, Paul Kraus
insightfully suggested that the 'ain, standing for 'All, was meaningfully
transposed to an alef, representing Aaron:
We have here [in SY] an unsatisfactory explanation of a more ancient schema, the significance of
which the author does not reveal. Why, indeed, are there only these three elements, and why were
these letters chosen to the exclusion of others? Insofar as there exists a relationship between the
Sill'I series 'Aynl Mimi Sin and that of SY, the priority seems to go back to the Muslim system,
where it is rooted in the religious history of Islam. (my translation from Jabir, vol. II p. 267, n. 6)
Massignon and Corbin would subsequently use these apparent similari-

ties to assert the direct influence of early ShI'I jaJr on SY - an "influ-
ence", however, which has yet to be demonstrated.4
Because the triad AI Mis and the term designating it (immotl ummahat)
both would appear to present incontestable analogies with usages in SY, a
closer investigation of this curious parallel is in order. As Kraus suggested,
the matter of dating therefore is of some importance here, especially inso-
far as borrowing would seem to provide the only adequate explanation for
this similarity. Fortunately, the innovation of the AI Mis triad can be dated
fairly accurately: it must have originated in the early ShI'I -gnostic revolu-
tions of the mid-eighth century, and it must have preceded the rise of
Ism~nlIs at the end of the ninth century. Nor could SY have borrowed
this triad from the Ism~nlIs, since - as I hope to demonstrate below -
SY must pre-date the rise of the Ismii'IlI mission.
This triad must therefore originate in the eighth-ninth centuries. But was
it itself a borrowing from the more ancient SY? This ShI'I triad could not
have originated with SY, for the reasons just cited from Kraus. To expand
on Kraus' argument, several points must be amplified. First, the ostensible
explanations for the selection of these letters by an ancient Jewish author
on some hypothetical grammatical grounds may be too weak to be satisfac-
3 Jabir U, p. 267, (see n. 31 below) where he acknowledges Massignon's original suggestion. A

fully study on the critical cognates "ummahatlimmol' remains to be undertaken. When Vajda gath-
ered some sources which used the term ummahat, they were exclusively derived from the milieu
under discussion here (Rasa'il Ikhwan al-~afii', the early Isma'Ills, and the SilT al-Khal!qa): see "Un
opuscule ismaelien en transmission judeo-arabe (Risalat al:fawharayn)", Journal asiatique 246
(1957-1958) 459-466, at 465 n. 15. He also discussed the term in "Saadya Commentateur de Sefer
Ye~ira", p. 66. I would add that the term is also used in the Umm al-Kitab, a work deriving from the
eighth-century, which betrays Jewish and gnostic influences: see Pio Filippani-Ronconi, "The
Soteriological Cosmology of Central Asiatic Isma'Ilism" in Ismazl! Contributions to Islamic Culture (Ed.)
Seyyed Hossein Nasr (Teheran, 1977) 99-121, at 113, 115, 118, and E.F. Tijdens, Der myth%gisch-
gnoJtiJcheHinter;grund des 'Umm a/-Kitab', (Leiden, 1977) p. 269.
4 See n. 1 above. In support of Kraus' hypothesis, it might be pointed out that the 'ayn of 'AlI
was known to lexicographers as being capable of transposition to an a/if see Vajda, "Les lettres" (n.
1 above) p. 116, n. 4.

tory. For when one considers the far more intrinsically coherent reasons
for their selection on ShI'I-trinitarian grounds, the coherence of the
'Ali/Mul).ammad/Salman Pak triad simply appears to be far more cogent
than the strained, unprecedented phonological hypothesis eXplaining
Alephl Meml Shin as originating in different pronunciations. This evidential
and logical disparity weighs yet more heavily in favor of a ShI'I origination,
moreover, when one considers the milieu in which the work of Jabir origi-
nated, which I shall describe more fully below.
But more can be said at this point in favor of the theory of ShI'I origi-
nation of the triad AI Mis. For the historic problematic assumed in this
trinitarian abbreviation was, as Kraus put it, "rooted in the religious history
of Islam." That is, it designated the continuity of a certain form of ShI'I
authority. This implication is encoded by the succession of Mim
lMul).ammad] to Ain' [AlI] to Sin, or Salman Pak, a mythical hcro of the
ghuliit.5 fafr, letter-mysticism, as I shall show below, was an essential feature
in this ghuliit argument for self-legitimation. This self-legitimation by means
of letter-mysticism can be found in such eighth-century Shl'l works as the
cosmogony of Mughlra b. Sa'ld, or the cosmic semiotics of the Umm al-
Kitiib, a work which virtually apotheosizes Salman pak. In other words,
there was an intrinsic, even intimate connection between the historic
claims to authority made by the ghuliit, and the gnostic content of their
teachings, which specifically championed a kind of cosmic semiotics.
Another piece of evidence makes this argument clearer. Traditions ema-
nating from ghu/iit circles taught that 'Ali's two martyr-sons, I:Iasan and
I:Iusain, were joined by a third son, variously known as Mul).sin or
Mul).assin. These traditions furthermore specified that the names of these
three sons were given to 'AlI by the angel Gabriel, as translations of the
Hebrew names of the sons of Aaron, said to be Shabr, Shubayr and
Mushbir. This tradition, again, pushed the constitutional legitimation of
6
the "extremists" (ghu/iif), as they thereby depicted themselves holding the

historic third position in another triad formed on the basis of letter-mysti-
cism. For the purposes at hand - that of corroborating the relation
between the 'Ainl Mimi Shin triads - this tradition bears the following rel-
evant features: (1) The triadic permutation of a triliteral root; (2) The
extremist ShI'I legitimizing role of that triad; (3) the ostensible transmission
from Hebrew; (4) The divinely-ordained association between the lineages
of Aaron and 'Ali (which Kraus noted in explaining the relation bctween
the Arabic 'Ain and the Hebrew A/if).
5 "Louis Massignon, Salman Pak (see n. 1 above). The ghulat were the proto-Sh!'! "extremists",
"exaggerators", in the polemical characterization of their Sunni opponents.
6 See the sources cited in n. 54 above.

SeIer Ye$ira and Early Islam 5
It should be noted, in passing, that this ShI'I evidence can not be used
to identify specific sectarian groups by these three-letter abbreviations, as
Massignon repeatedly attempted to do. In reviewing the evidence, Halm
cautiously and correctly concluded that "Eine solche Systematisierung nach
einem einzigen Merkmal ist J edoch fur die Erkenntnis der sek-
tengeschichtlichen Zusammenhange unbrauchbar".7 In this regard, it also is
instructive to note the strain of the earliest known commentators, Dunash
ibn Tamlm and Saadia, in their attempts to account for the peculiar role of
the shin. Their problem was, of course, that in SY, shin is curiously associ-
ated with esh (fire), which does not begin with the letter shin, as might be
expected from the pattern of A/if/ Avir and Mem/Mayim.8
Nevertheless, while not to be taken as literal abbreviations of historical
groups, such triads are still significant for the emphasis they place on the
critical third term. Indeed, one might say that the ghu/iit innovation in ShI'I
authority and the SY innovations in Jewish cosmology are both, in a sense,
represented by the third letter of their respective occult triad. For the tri-
ads themselves would be meaningless without the essential recognition of
the role played by the third member, for which initiatic decipherment
alone renders the triad operative. Indeed, the very existence of the triliteral
occult code, in this sense, makes an eloquently mute historical argument
concerning itself: the mystery of the third element, in fact, is said by Jabir
to be either unpronounceable (!gmt), or even that of the "silent one"
(fiimi~.9 In other words, the secret identity of the author is appropriately
referred to, by means of indirection. The author, encrypting the text in the
historical present, is himself to be identified with the successfully "unpro-
nounceable", "silent" meaning of the third (= historically most recent and
still occultated) element in the triad.
There are other implications to the linguistic theory of unpronounceabil-
ity. The obscure term be/imah (not a neologism, but found in Job 26: 7, as
Sed noticed) reflects this concern, which finds its parallel in the alchemical
, Heinz Halm, Kosmologie und Heilslehre derfriihen [smii T!iya, p. 161, n. 94.
8 Canteins, op. cit. (n. 23, below) See also nn. 68-74 below.
9 It is not necessary to assume that the shin must refer to a personage parallel to Salman: after all,
the explanation which SY itself provides refers to a correspondence with the element esh (fire),
which does not begin with shin. As for the rclation between samech and sin/shin see Massignon,
Lexique, pp. 98- 101 ["Appendice: Tableau de l'alphabet 'philosophique' OAFR)"]on the letter sin:
"En arabe, Ie lettre samech, hebraique et syriaque . . . est tombee; elle est remplacee par sin . . ."
(p. 100). As for the 'UJ1nT, it should be noted that the early Isma'Ili Book of Revelation (Kitiib al-Kashj;
by Ja'far ibn Man~Ur al-YamanI (tenth century) associated the 12 ummahiit and the 7 tijmiyiit in the
context of a cosmological use of letters, which includes the image of 22 letters which are attached
to a circle, turning backwards and forwards, giving rise to the elements of all the worlds, and giving
rise to 231 gates (Halm, Kosmologie 58-50) these are indeed sttiking parallels with SY, as Halm prop-
erly observed. I hope to review the early Ism,fIli familiarity with SY on another occasion. For the
moment it is sufficient to note that these traditions circulated among the early Shi' a.

and subsequent Ismii'Ili teachings on the 'tij'mi. Thus, as Massignon noted,

Jabir ibn .f::Iayyanmakes this argument against the proponents of the Sin:
"the obscure consonant of the Sin is medial, latent, quiescent, absolutely
inert, in whatever state and in whatever position the Sin may be, it has,
therefore, only one genus, 'tij'mf (unpronounceable) . . ."10 To this may be
compared the anomalous character of the Shin in SY, as noted by the early
commentators.
The Kraus hypothesis, as he left it on his tragically early demise, was
provocative, but it had rather serious limitations. First, the crucial triads of
SY can not be eXplained exclusivelY by recourse to ShI'I "borrowings".
Perhaps most importantly in this regard, the triadic repetition of the
Hebrew Samechl Pehl Resh root, with which SY begins, may have been used
already by the payyetan Kallir, who has recently been dated to the lifetime
of prophet Mul;ammad.11 Still, the enthronement of these various triadic
theories in Jewish thought, of course, solely is the achievement of SY. And
the modernizing redactor of SY did seem to have been aware of contem-
poraneous (post-Islamic) linguistic (grammatical, phonological and semiotic)
theories concerning pronunciation, vowel-and-consonant theory, etc.
To summarize, there are two arguments of some strength which still
commend the position of Kraus. The first argument is that the reasons
given for the triad AI Mis are intrinsic and therefore strong in the Jabirian
system and in Shi'ism, while the reasons given for the triad AI MI Sh are
extrinsic and therefore comparatively weak in SY. Thus, Scholem could go
only so far in explaining the origins of SY's triad: "The choice of these
three consonants seems to reflect an ancient division related to the quantita-
tive force of articulation of the consonants, in explosives, aspirates, and
nasals [my emphasis, SMW]."12 Such an "ancient division", presumably
rooted in antiquity, has never been identified, so far as I know, though
some evidence for a kind of sacramental phonology does exist.
Nicomachus of Gerasa, for example, already associated certain isolated
sounds with the worship of theurgists:" . . . they invoke it symbolically
with hissing sounds and clucking, with inarticulate and foreign sounds."13
10 Salman Pak, (see n. 1 above, ET) p. 33. See also Corbin, "Le Livre du Glorieux de ]abir ibn
f:layyan," EranosJahrbuch XVIII (1950) p. 97; and Kitab al-Kashf(n. 35 above); and see the recent dis-
cussion by Yves Marquet, La philosophie des alchimistes et l'alchimie des philosophes (paris, 1988) 104-110.
In Origins (p. 28), Scholem still speculates that there may be a Greek original behind belimah.
11 See Aryeh Kaplan, Sefer YetiJrah. The Book of Creation in Theory and Practice, p. 348-349, n. 109,
for the considerable literature on this enticing citation. For the new dating, see now Ezra Fleischer,
"Solving the Qaliri Riddle," Tarbi[ LIV (1985) 383-429 [Hebrew].
12 Origins, p. 30, n. 49. Note that Scholem carefully expresses himself, "seems to reflect ... "
Moreover, Scholem did not take into account the evidence that the phonology of SY may represent
a stage of linguistic science post-dating the advent of Isalm.
13 Pearson, "Gnosticism as Platonism: with Special Reference to Marsanes (NBC 10, 1)", Harvard
TheologicalReview 77 (1984) 55- 72, at p. 69.

SeIer Yer,ira and Early Islam 7
Just such symbolic sounds were invoked in the language-speculation of the

ghuliit. For example, in the eighth-century, ShI'I-gnostic Kdiib al-Hqft wal-
A:rPla, we read this statement attributed to the Fifth Imam Ja 'far al-$adiq
(addressed to his disciple MufaQQalibn 'Umar):
o Mufa<Nal, the tongues were confounded at the time of Abraham. The Discourse (kalima) then
became Hebrew. And the fundaments of that discourse are four. For he added to the Discourse
these sounds: whistling, shouting, and percussive sounds added to the letters. For he who knows
their ways of connection and their method of separation understands the Discourse. . . . These are
the four modes of Discourse: what opens understanding is cooing; what is intrinsic to the mouth is
whistling; what scnds breath into the air derives from percussion; what comes forth from the gullet
is the expulsion of air through the open mouth. Understand that, if God wishes.14
Like the theory espoused by SY, this mystical phonetics is at once lucid
and cryptic. Subsequent ImamI and Isma'IlI ShI'I speculation picked up
analogous lines of sacramental phonology.15 All this must be compared
closely with the phonetic theory of the AI MI Sh triad in SY.
The second argument is in fact statistical. For the likelihood of these
systems coincidentallY hypostasizing the same three cognate letters - mutatis
mutandis - in the same order is simply out of the question, on arithmetical
grounds: there are (as calculated by Kaplan, p. 193) approximately one sex-
tillion possible permutations for the twenty-two letters of the Hebrew
alphabet.
Part II: Circumstantial evidence
Histon·cal context
Scholars agree that SY exhibits some knowledge of the Hellenistic occult
sciences. This agreement, however, has misled investigators to posit an
originating date sometime in late antiquity, usually said to be between the
third and sixth centuries. A reappraisal of the evidence, undertaken herein,
suggests that SY can equally well be argued to have emerged not out of
Hellenism itself, but rather out of the early medieval revival of Hellenism.
Even aside from the fact that SY contains no linguistic Grecisms, the latter
milieu provides a plethora of structurally-integrated parallels, the specifically
detailed likes of which cannot be discerned at anyone particular moment
in antiquity. In order to support the hypothesis of Kraus, in this section of
14Edited by 'Arif Tamir and 1. A. Khalife (Beirut, 1970) pp. 148-149.

15Abu I:1atim al-RazI, Kitab al-Zzna, analyzed by Vajda, "Les Lettres" (see n. 2 above); Ja'far ibn
Man~Ur al-YamanI, Kitab al-Kashf, (ed.) Mustafa Ghalib (Beirut, 1983) p. 92, for analogies to animal
sounds in a spiritual theory of pronunciation. See also Scholem (n. 11 above).

my paper I will gather contextual evidence which would support a dating

of SY to the ninth century.
My historical analysis of the context of SY is predicated on the follow-
ing premises, which shall be discussed in detail in the following pages.
First, it cannot date from any earlier than the end of the eighth century
because it demonstrably post-dates the extremist Sill'I revolutions which
marked the latter half of that century. Furthermore, it would seem to
emerge from the revival of Hellenistic intellectualism, which began to
flourish only in the first half of the ninth century. Nor can it be dated
much later than the second half of the ninth century, for it was repeatedly
commented on in the early tenth century. And, finally the pioneering
Isma'III missionaries of the early ninth century apparently knew some tra-
ditions from SY.
Having posited these chronological limit-conditions, I shall now describe
briefly the milieux in which, one could argue, SY originated. These milieux
are the general atmosphere of early 'Abbasid Islam, especially that of
Baghdad; the ShI'I context; and the heterodox Jewish context.
Al-Ma'mun and the 'Abbasid court

The second Islamic century ended with the reign of the controversial al-
Ma'mOO (813-833).16 It should be remembered that Baghdad was a new
city, having been founded in 762. In this booming metropolis, leaders of a
variety of religions interacted with what appears to have been an unusual
degree of frequency. AI-Ma'mOO, in fact, was known for his sponsorship,
variously, of ShI'I imams; of Hellenistic scholarship; and of certain selected
dhimmfs.17
Interconfessional intercourse was patronized at court. There are extant,
in fact, rather richly presented, if perhaps ultimately apocryphal accounts of
al-Ma'mOO's meeting with the leaders of the Jewish, Christian, $abian and
Zoroastrian communities. These reports are preserved in, among others,
1H
16"Al-Ma'moo," EP; D. Sourdel, "La politique religieuse du caliphe 'abbaside al-Ma'moo," Revue
des etudes isiamiques 30 (1962) 27-48.
17F. Gabrieli, "Al-Ma'mOO e gli 'Alidi," MOfl!,enidnd Texte und Forschunj!,enII/1 (Leipzig, 1929); F.E.
Peters articulates the conventional view that al-Ma' moo encouraged interconfessionalism: "In the
atmosphere of Ma'mOO's Baghdad, Thumama's liberal stand justified and excused the Christian,
Jewish and Zoroastrian beliefs of those other, non-Muslim residents of the Dar-aI-Islam", Allah's
Commonwealth, (New York, 1973) p. 187. On the other hand, see the dissent of Avraham Grossman,
"The Attitude of the Caliph al-Ma'mOO to the Jews" Zion XLIV (1978/1979) 94-110 [Hebrew].
18 Ibn Babuya, Kitiib ai-Tauhrd, (ed.) H. al-TiJ:1ranI(Teheran, 1387 H.H.) pp. 435-437, for a cosmo-
logical theory of language and its phonetics; for other versions, and a partial translation, see pp.
65-78 in David Thomas, "Two Muslim-Christian Debates from the Early Shi'ite Tradition" Journal
of Semitic Studies XXXIII (1988) 53-80. Thomas is appropriately skeptical concerning the historicity
of these reports. While we need not believe the details of these proceedings, accepting the existence
of such meetings still seems a reasonable position to hold.

SeIer Ye$ira and Early Islam 9
Ithna AsharI ShY'ytexts. There would seem to be little reason to doubt

that such meetings took place in some form. These Twelver reports depict
the Imam al-Ri<:laensconced beside the Caliph, just as comparatively reli-
able non-Twelver traditions also describe the sponsorship of al-Ri<:laby al-
Ma'miin.19Various sources also describe Hisham ibn al-J:Iakam at the court
of al-Ma'mun. Credited with the origination of the doctrine of the
Imamate, Hisham ibn al-J:Iakam moved from Kufa to Baghdad during the
reign of aI-RashId, and may have lived into the reign of al-Ma'mu.n.2o
Hisham was noted for a number of beliefs and activities which may be rel-
evant to the deep background of SY: he espoused an anthropomorphic
conception of the godhead; he taught that God occupies a point; he met
with Jews and Day~anites.21Hisham and al-Ri<:lawere two figures of central
importance in the foundation of the Imamate at the end of the second
Islamic century, but many other groupings of 'Alid loyalists remained
active in those years.
The sciences
Theological, philosophical, occult and linguistic sciences began to be culti-
vated intensively in the eighth century. AI-Ma'miin soon achieved lasting
fame for his establishment of the Bait a/-Hikmah ("House of Wisdom")
devoted to the study and the translation of the Greek sciences.22 From this
period of scientific efflorescence, the most important such works for con-
firming the original milieu of SY include: arithmological and Pythagorean
treatises; alchemical works attributed to Jabir b. Hayyan; the hermetic pam-
phlets attributed to Baunlis (Apollonius of Tyana), especially the Sirr a/-
Kha/iqa: and the encyclopedic Book if Treasures of Job of Edessa.23
19 "'All al-Rida", El'

20 Thomas, "Two Muslim-Christian Debates", pp. 54-60, translates a debate with Hisham ibn al-
l:lakam. Etan Kohlberg follows Madelung in asserting that Hisham was the founder of the doctrine
of the Imamate: "From Imamiyya to Ithna Ashariyya," Bulletin of the School of Oriental and African
Studies 39 (1976) 521-534, esp. at 521. The rich report by ShahrastanI has yet to be fully eXploited:
Livre des religions et des sectes I, (eds.) D. Gimaret, G. Monnot, (Louvain, 1986) 531-538, provides the
available literature.
21 For his view of the godhead, see ShahrastanI, cited in the preceding note. For his contacts with
Jews and Day~anites, see Ivanow, Ibn al-Qdda~. The Alleged Founder of Ismailism (Bombay, 1957) 85.
Pines, "Points of Similarity between the Exposition of the Doctrine of the Sefirot in the Sefer
Yezira and a Text of the Pseudo-Clementine Homilies," Proceedings of the Israel Academy of Sciences and
H;manites VII/3 (1989) 108-110, provides more evidence on the Daysanite connection.
22 See the vivid report in Ibn al-Nadlm, The Fihrist of al-Nadrm (2 vols., ed. and trans.) Bayard
Dodge (New York and London, 1970) II: 583-584, where MaOmlin is inspired to do so by a dream
of Aristotle.
23 Perhaps the best overview of this environment is now F.E. Peters, "Hermes and Hanan: The
Roots of Arabic-Islamic Occultism," Intellectual Studies on Islam. eds. M.M. Mazzaoui and Vera B.
Moreen (Salt Lake, 1990) 185-215: he does not mention SY.

Nicomachus of Gerasa's Introduction to Arithmetic (wr. ca. 100 c.e.), with

its Pythagorean emphasis on the perfection of the decad, and the greatness
of the triad, may well have been known to the redactor of SY, as Salo
Baron recognized.24 This ancient arithmology seems to have entered
Islamic civilization first under the administration of al-Ma'mlin. His gener-
al, Tahir ibn ai-Husain, hired one HabIb ibn BahrIz to translate this work
into Arabic. 25 Al-Kindi (d. ca. 850) was sympathetic to arithmology, but
had no direct dependence on it, though a student of his wrote a para-
phrase to Nicomachus' Introduction to Arithmatic.26 The Ikhwan al-Safa'
claimed closeness to Nicomachus.27 But the key figure in this regard is
probably Thabit ibn Qurra (836-901).28 Thabit translated the Introduction to
Arithmetic, wrote Hermetic works on magical images, and was known for
his mystical permutations of letters.29 And, finally, Thabit had at least one
Jewish student, and his mathematical work was known to Saadia Gaon.'o
24 Social and Religious History of the Jews, (New York and London 1958; repro 1971) vol. Vlll: 151.
See also Nicomachus, Introduction to Arithmetic (trans.) Martin Luther D'Oooge (New York, 1926). A
current overview on Nicomachus is provided by Dominic J. O'Meara, Pythagoras Revived Mathematits
and Philosophy in Late Antiquity (Oxford, 1989) 14-23.
25 Felix Klein-Franke, Iatromathematics in Islam (Hildesheim, Zurich, New York, 1984) 46, citing
Sezgin, Geschichte des Arabischen Schrijtums (Leiden, 197) vol. V, 164 ff. The most important carly
Islamicate mathematician, a1-KhwarizmI (d. ca. 844), was patronized by al-Ma'moo, worked with
both Greek and Hindu systems, and was familiar with the Jewish calendar (E.S. Kennedy, "Al-
KhwarizmI on the Jewish Calendar," Scripta Mathematica XXVlll ((1964) 55-59). Indeed, he some-
how knew the Mishnat ha-Middot: see Baron, A Social and Religious History of the Jews, vol. VIII, p. 350,
n. 10, where he surveys the relevant literature.
26 A. Ivry, AI-Kindi's Metaphysics (Albany, New York) 20-21.
27 For example, Rasa'il Ikhwan al-Safa' I, 24. See S.H, Nasr, Introduction to ,Islamic Cosmological
Doctrines (Cambridge 1964, repro Boulder 1978), p. 37; S. Hrentjes, "Die erstd Risala der Rasa'il
Ihwan al-~afa uber elementare Zahlentheorie: Ihr mathematischer Gehalt und ihr Beziehungen zu
spantantiken arithmetischen Schriften," Janus 71 (1984) 181-274; B.R. Goldstein, "A Treatise on
Number from a Tenth Century Arabic Source," Centaurus 10 (1964) 129-160. The consensus seems
to be stated succinctly by Alessandro Bausani: "The arithmetic of the Ikhwan seems chiefly inspired
by the work of Nicomachus of Gerasa." "Scientific Elements in Isma'IlI Thought: The Epistles of
the Brethren of Purity (IkhJPiin al-yafa)" in IsmaTlr Contributions to lrlamic Culture, (ed.) by Seyyed
Hossein Nasr (Teheran 1977) 127-141, at 124.
28 Tabit b. Qurra's arabische ubersetzung der arithmetike eisogoge des IVikomachos von Gerasa, (ed.) W.
Kutsch (Beirut, 1958).
29 The Astronomical Works of Thabit ibn Quraa, F.J. Carmody (Berkeley and L.A., 1960) 167-199;
!:lajjI KhalIfa, Kashf al-Zuniln (ed.) G. FlUgel (7 vols., Leipzig and London, 1835-1858) vol. 2,
"Jami' a", p. 604, on his transmission of jafr. He was not the only ~abian remembered for his lin-
guistic mysticism: see the Sabian dialogue at the court of al-Ma'moo, in Ibn Babuya (n. 5 above) p.
429 ff. See also D. Chwolson, Die Ssabier und die Ssabismus (2 vols. St. Petersburg, 1856) vol. I, 551;
L. Thorndike, History of Magic and Experimental Science (New York and London, 1923, repr. 1964) vol.
I, 664; and D. Pingree, "Some of the Sources of Ghayat al-!:lakIm," Journal of the Warbu1;J:'and
Courtold Institute 43 (1980) 1-15, at 4-5, esp. 5, n. 29.
30 Mas'UdI cites Yehuda ibn Josph ibn abi'l-Thanna of Raqqa as his student: see Kitab al-Tanbrh
wa 'I-lshriif (ed.) M.L de Goeje Leiden, 1893, p. 113; Pines, "A Tenth-Century Philosophical
Correspondence," Proceedings of the American Academy of Jewish Research XXIV (1955) 103-136, at 134
n. 106. Saadia apparently knew Nicomachus in the Thabit ibn Qurra translation: S. Gandz, "Saadia
Gaon as Mathematician," SaadiaAnniversity Volume (New York, 1943) 141-195.

Seier Yer;;iraand Early Islam 11
The great alchemist Jabit b. I:Iayyan also has been implicated in the con-
text of SY. While according to the ShI'I tradition a student of the mid-
eighth century Imam Ja 'far al-Sadiq, Jabir's vast alchemical corpus was
dated to the ninth and tenth century by its modern editor and master, Paul
Kraus.31 Kraus sets the parallels of SY with Jabir's language-speculations
against the background of the "extremist" ShY'ygnostics of 2nd-century
Islam. "However [Kraus suggests], these isolated elements have been inte-
grated in a rational system of great coherence and elaborated in the light
of the newly discovered Pythagorean science."32 Supporting this argument
with several pages of detailed parallels, Kraus asserts that the cosmic semi-
otics of the Jabir-tradition is shared by the SY-commentary of Saadia
Gaon. Specifically, Kraus suggests these parallels: (1) the centrality of the
permutation of word-roots; (2) the close similarity of terminology for this
permutation; (3) classification of letters and their articulation in the mouth
- unknown in Talmudic times, but pioneered by the early grammarians of
Arabic.J] That being noted, I would repeat that the present study is not
specifically concerned with SY-commentaries as such, though their study
may be relevant, if it is the case that SY and its earliest commentaries are
in fact contemporaneous.
The Secret of Creation (Sirr al-Khalrqa) is a pseudepigraphical hermetic work
said by the philosopher al-RazYto be written at the time of al-Ma'miln.34
The Secret of Creation, attributed to the ancient wonderworker, Apollonius of
Tyana (Balinlis), perhaps more than any other work with the exception of
the Jabirian corpus, can help us understand how SY came to be redacted
when and how it was. As their similar titles show, The Book of Creation and
the Secret of Creation are both treatises on the secrets of the formation of
the visible universe. Both are composed in a kind of hermetic-sounding,
31 Jiibir b. Ha.Iyiin (Memoires presentes a I'lnstitut d'Egypte vols. 44, 45 Cairo, 1943, 1945: Hereafter,
"Jabir"). A full update of the "Jabir problem" has now been made by Fuat Sezgin, Ceschichte des
Arabischen Schriftums vol. IV (Leiden, 1971) pp. 132-269. See now also Pierre Lory, A!chimie et mys-
en terre d'Islam (paris, 1989) [I thank Sarah Stroumsa for guiding me to this source].
Vol. II, p. 266 (my translation).
33 Ibid., pp. 266-269. Subsequently others have commented on the question of pronunciation. I
shall return to this point at greater length in my discussion of the grammatical sciences below. For
now, see Jabir, II, 187-188. It should be noted that Jabir occasionally expresses some indebtedness
to Jewish sources. One Jabirian treatise is titled ''Kitiibi alladhifassartu fihi al-Tawriit'~ that is, "The Book
in which I Explicate the Torah": See Jiibir, vol. I, p. 171; and see n. 50 below, for other contemporary
reliances on "Torah".
34 M. Plessner, "BalInlis," E1' vol. 1: 994-995 (see p. 994, for a Hebrew translation of this work).
We are fortunate to now have Ursula Weisser's study Das 'Such uber das Ceheimminis der Schopfung' von
Pseudo-Apollonios von Tyana (Berlin, New York, 1980) and her edition of the text, Buch iiber das
Ceheimnis der Schopfung und die Darstellung der Natur (Buch der Ursachen) (Allepo, 1979). See also the
important review of it by M. Ullmann, in Journal for the History oj Arabic Science 4 (1980) 90-94, 5
(1981) 121-126.

scientistic, apodictic tone. Both compnse SiX sections, sub-divided into

brief statements.
Pseudo-Apollonius begins by discussing the 24 letters of the alphabet
and the 4 "Causes."35 (It is interesting that Saadia's neopythagorean com-
mentary resorts to a related quaternity when attempting to make sense of
SY's three sefarim.36) The neopythagorean Pseudo-Apollonius, moreover,
addresses the centrality of the linguistic triad of root-permutation, ft'l, fii'ii,
and mqf'u1.37 From here Ps.-Apollonius addresses the hermetic topics of
numbers and elements (ummahiit), which are also topics of SY.
The philosophical influence through Christian hands was extensive: the
most important name in this regard is Job of Edessa, who (under al-
Ma'mOO) wrote his Book 0/ Treasures, a work which bears important con-
nections to Baiinlis, Jabir and Sy'38 This book betrays several significant
35See Weisser's edition p. 14; and her study, p. 75-76, and p. 157.
36See Kraus p. 268~269. and, more fully see Vajda's "Commentaire Kairouanais"; and Jean
Canteins, Phonemes et Archetypes (paris, 1972) 27-38.
l' See Weisser's discussion of these passages, Das "Buch ... , pp. 69, 157, 168, 176. It may be
noted that the "ancient explanations" which Judah ben Nissim ibn MaIka cites as a gloss on the
three sefarim of SY include .aq~ 'iiqil and ma 'qiiJ as well as 'ishq, 'iishiq and ma'shiiq. See Georges
Vajda, Judah ben Nisslm ibn Maika (paris, 1954) p. 25, n. 2. For a typically triadic mystical manipula-
tion of the f '-I root, see the early Ismii'Ili missionary, al-KirmanI, in Majmtla Rasa'il al-Kirmani (ed.)
Mustafa Ghalib (Beirut, 1983) p. 47-48.
38 job of Edessa's Book of Treasures, (trans.) A. Mingana, (Cambridge, 1935). It was of course the
parallels between Job, Jabir and Balinus which led to Kraus' great breakthroughs in dating these
works: see Jabir II, ["Jabir et Apollonius de Tyane"l, pp. 270-303, where this is brilliantly demon-
strated. For the fullest recent comparison of Job and BalInus, see Weisser, Das "Buch ... (cited in
n. 21 above) pp. 55-63. More generally, see also Bernhard Lewin, "Job d'Edesse et son Livre des
Tresors," Orientalia Suecana 6 (1957) 21-30. Hans Daiber observes, yet more recently ("Masa'il wa'l-
Adjwiba," EI', vol. 5: 637), that Job of Edessa and Balinus are both indebted to the same, ancient
(unidentified) Greek Problemata.
See also Ryding (below, n. 46) p. 119, for the observation that "The concept of mizan, or balance,
is a crucial one in both Jabirian alchemy and in the evolution of the Basra school of grammar."
(and n. 8 for the relevant literature). A rich exploration of this concept is found in Corbin's "The
Science of the Balance and the Correspondences between Worlds in Islamic Gnosis," in Temple and
Contemplation (translated by P. and L. Sherrard) (New York and London, 1986) 55-132. For a good
summation of the idea of the ]abirian balance, see jabir II, 187-188.
This important concept of SY has not been studied in this comparative context, so far as I know.
Willie Kraus sets his entire discussion of SY into the context of the origins of the mizan al-huru! (see
his remarks in jabir II, p. 266), he does not note the central importance of the concept "balance"
within SY itself. For the use of "balance" (makhria beintaim) see SY II: 1; III: 1, 3, 4, 5; VI: 3, 6, 7. It
should be noted that this usage repeatedly is associated with the three immot, alej/ mem/ shin, the com-
parative significance of which is explored below, nn. 68-75. At least one precise parallel concerning
SY's science of the balance is to be found in the alchemical Turba Philosophorum, which Plessner
dated to ca. 900. In the Turba, Empedocles states that Air mediates between Water and Fire: see
Martin Plessner, Vorsokratische Philosophie und GriechischeAlchemie, (ed.) Felix Klein-Franke (Wiesbaden,
1975) p. 53. Similarly, in SY VI: 3, we read, "Three immot, Alif/ Mem/ Shin, are Air, Water and Fire.
The Fire is above, the Water is below and the Spirit of Air sets them in balance (makhria beintaim)".
Finally, for the relation between the Turba and Sirr al-Khaliqa, see Ulrich Rudolph, Die Doxographie des
Pseudo-Ammonios (Stuttgart 1989) p. 211.

SeIer Ye!?iraand Early Islam 13
parallels with SY, such as those concerning the "character of the simple
elements" (Mingana's translation, p. 6); the physical and metaphysical
meaning of the number thirty-two (pp. 83-84); phonetics related to breath-
ing (p. 142); "the balance" (p. 229); the balance of the four elements (p.
235); and the irreducibility of the first elements (293-294).
In such an interconfessional milieu, the crossfertilization of religious and
linguistic studies was cultivated intensively, if not extensively. Nor were
these contacts marginalized as merely "occultist" contracts. Students of the
theological sciences, surprisingly, also contributed to the transmission of
cosmic semiotics as well. As an example, let me just mention Mu'tazilism.
Some Shl'l schools, such as the Zaydis, not only embraced Mu'tazilism as
the basis for their theology, but retained it long after Sunni Islam had
jettisoned it - just as the Karaites also tenaciously retained it. A
Mu'tazilite debate on the createdness of letters arose out of the larger con-
tention over the createdness of the text of the Qur'an. Thus, the jurist
Shafl'l is said to have declared: "Do not say that the letters were created in
time, for the first cause of the ruin of the Jews was in believing this."39
Goldziher suggested that this statement indicated an awareness of and a
response to Mu'tazilite Jews who believed that letters were created accom-
panying the act of creation.40 We know, moreover, that Jews were well-
known to have held for the createdness of the Word of God, for even a
Christian Mu'tazilite of the middle tenth century attacks them for this
belief.41While it is still not certain what conclusions can be drawn from
this evidence, it would appear that Mu'tazilite Jews held some such ideas
about the cosmological pre-creation of language, and that Mu'tazilite
Muslims and Christians were aware that they held those views.
Another Muslim Mu'tazilite, shortly after the Mu'tazilite Saadia, knew of
the Gaon and of his commentary on SY, and may have even known SY
itself. This author, Maqdisl, went so far as to assign a sect to Saadia,
asserting that this group, "the Fayyiliniyya, are followers of Abu Sa'ld al-
39 Goldziher, "Melanges judeo-arabes XXIV: La creation des lettres," Revue des etudes )uives 50
(1905) pp. 188-190 at p. 189.
40 Ibid. p. 190.
41 When Nemoy wrote his synoptic article on Karaite attitudes towards Christianity, he concluded
with the observation that the Karaites' extensive knowledge of Christianity must have been based
on direct contact. "No such contemporary Christian anti-Karaite works, so far as I know, have yet
been discovered, but this does not mean that they never will. It is devoutly hoped that a lucky find
will some day provide definite confirmation of this reasonable assumption." Happily, I made just
such a discovery: Severus b. al-Muqaffa', Jacobite Coptic bishop of al-Ashmunain, Arabic-language
Mu'tazilite, writing ca. 950, who criticizes the Rabbanites and Karaites for various heresies. He
explicitly criticizes the Karaites for believing in the Created Word: "The 'Anaiyya [a telling corrup-
tion for 'Ananiyya, a common eponym for Karaitel of Jews and Mu'tazilite Muslims assert that the
Word of God is created, is made in time and is subordinate," Sec Histoire des conciles (ed. and trans,)
L. Leroy, Patrologia Oriental is VI (1911) p. 535.

Fayyfunl. They interpret the Torah by a method of isolated letters, as do

the Batiniyya of Islam."42For our present purposes, it is especially interest-
ing to note that MaqdisI, a ncar-contemporary of Saadia, associates Saadia's
recently-published commentary on SY with the Batiniyya, that is, with eso-
teric ShI'Is.
One final science, that of linguistics itself, remains to be mentioned. It
should be remembered that virtually all the sciences were directly inspired
by the contemporaneous discoveries of the grammarians. As Massignon
stated the case: "It is useless to examine the works of Muslim mystics
unless one has studied very closely the mechanism of Arabic grammar, lex-
icography, morphology, syntax. These authors constantly connect the new
technical terms that they propose with their ordinary significance, in cur-
rent usage, ascertained by the grammarians ... "43Less explicitly, but no
less forcefully, such a case has been made for the similarly mystical utiliza-
tion of grammatical concepts in SY. From Abraham Epstein to Nicholas
Sed and Nehemiah Allony, scholars studying SY have observed that the
linguistic theory of SY is closely related to that pioneered by the Arabic
grammarians of the eighth and ninth centuries.44 This fact, if it is correct,
would suggest the the Jewish mystic who compiled SY, like his Muslim
mystic contemporaries, may have been inspired by the same sciences of
language.45
Aside from theories concerning pronunciation and categorization, the
grammatical sciences of language interpenetrated in other ways with the
occult sciences of language. Recent studies on the history of alchemy have
emphasized, for example, the rhetorical and linguistic dimensions of alche-
my.46 Most especially, the role of ambivalent or contradictory senses
42 Mu~ahhar ibn Tahir al-MaqdisI, Kitiib a!-Bad' wa a!-Ta 'rfkh (Livre de !a Creation et de !'Histoire) (ed.
and trans.) C. Huart (6 vots, Paris, 1899-1919) vol. 4, text, p. 34; translation p. 33. Kraus, in Jiibir II
p. 222, cites this work, and refers in a work written by J. Schwarcz [in Hungarian (Budapest, 1909)
non vidz] on the Jewish elements in MaqdisI's masterwork.
43 The Passion of A!-Ha!!a; (trans. Herbert Mason, Princeton, 1982) vol. III: 79
" Epstein, "Recherches sur Ie Sefer Yec;:ira," Revue des etudes juives xxviii (1894) 95-108, p. 101 "
. il semble presque que I'auteur du SY ait puise dans Sibawaihi."; Sed, "Le Serer Yec;:ira," esp. p.
527 (arguing that, despite some striking similarities to the eighth century Khalil's K a!- Ayn, SY's
grammar-theory may be yet more ancient); and Allony, "Time of the Composition of SY," passim
(eighth century). See also note 47 below on the Syriac Christian grammarians.
4) Jews paid Muslim teachers to learn grammar from them. We are told, for example, of a gram-
marian (al-MazinI, d. ca. 863) who refused to teach a Jew the Kitiib of Sihawaih: Yaqut, lrshiid a!-Arfb
(ed.) D.S. Margoliouth (London, 1923-1931) II, 382 (cited by Von Grunebaum, Is!am (London,
1955) p. 249).
46 Jacques Berque, "Hellenisme et alchimistes arabes," in L'ambiva!ence dans !a cu!ture arabe (ed.) J.P.
Charnay (paris, 1967) 111-115 [this collection is devoted to an exploration of the cultural implica-
tions of the theory of ambivalent words (ad'dad)]; Brian Vickers, "The Discrepancy between res and
verba in Greek Alchemy," Alchemy Revisited, (ed.) Z.R.W.M. von Martels (Leiden, NYC, Kyihenhaven,
Koln, 1990) 21-33; K.C. Ryding, "Alchemy and Linguistics: Connections in Early Islam," Ibid.,
117-120.

SeIer Yef!ira and Early Islam 15
(ad'dad) of the same word was a subject of linguists and of alchemists: this
concern may be compared with the seven fundamental oppositions
(temurot) listed in SY (4:3).
In addition to the context of the theological, philosophical and occult
sciences at the early 'Abbasid court, SY of course also must be understood
in its Christian and Jewish contexts. The Syriac Christian context never has
been analysed systematically in its relation to SY, though discrete parallels
have been adduced. This Syriac evidence pertains to, variously, matters
grammatica4 astrologica4 gnostic and philosophical. The grammatical evidence per-
tains to the theories of pronunciation, and of the categorization of kinds
of "double letters". Kahle and Pines thus regard the theory found in SY
necessarily to post-date the Syriac advances in this area, which date to the
time of the rise of Islam.47 In addition to grammarians, Christian astrologers
also were active in the late eighth-century 'Abbasid court.48 The gnostic con-
duit, finally, seemed to have served as a last (nominally) Christian channel
through which material may have come to the redactor of Sy'49 The philo-
sophical connection is discussed above, with regard to Job of Edessa.
Finally, a few words must be said concerning Jewish scientists at the
early 'Abbasid court. Jews and converted Jews were utilized at the
'Abbasid court for purposes of scholarship. For example, the biographer
of Mul,1ammad, Ibn Hisham (d. ca. 833) spent these years editing the
canonical 5rra (biography) authored by Ibn Isl,1aq,which in turn was gath-
ered from many sources, some of which were Jewish.50 One such Jewish
tradent was Al,1mad ibn 'Abd Allah b. Salam, a purported descendent of
one of the first Jewish converts, who was utilized to translate from Torah,
" P. Kahle, The Cairo Ceniza (2nd Ed., Oxford, 1959) 183-184; Pines, "Points of Similarity
between the Exposition of the Doctrine of the Sefirot in the Sefer Ye?ira and a Text of the
Pseudo-Clementine Homilies," Proceedingsof the Israel Academy of Sciencesand Humanities VII/3 (1989)
112-114. See also the important study of Sed, "Notes sur l'homeJie #34 de Narsai," Orient Syrien 10
(1965) 511-524; and Shlomo Morag, in Sefer Tur Sinai, 233-236 [in Hebrew].
18 Klein-Franke, Iatromathematics, (n. 12 above) p. 43, provides the literature on Theophilos of
Edessa (695-785).
49 The example which has received the most attention is that of the "Wedding Hymn" in the Acts
of Thomas. See the most recent study, by M. Marcovich, "The Wedding Hymn of Acta Thomae,"
Studies in Craeco-Roman Religions and Gnosticism, (Leiden, 1988) 156-173, esp. (on "The Thirty-Two")
170-172 [for more on this work see Lafargue, cited in the Appendix below]. Marcovich notes
approvingly that already Thilo in thc eighteenth century recognized that The Thirry-Two refer to
teeth. In this regard may now be added H. Koester History and Literature of the Early Church vol. II,
216. Significantly, Job of Edessa devotes several pages to the arithmology of thirty-two, with
emphasis on the teeth (pp. 82-84): "Further, the evolution of the number thirry-two is complete,
because it is divisible without change till it reaches uniry; and, similarly, when it is multiplied, it
keeps the same number till thirry-two; its quotient and its divisor are also the same". See also N.
Sed, "Le Memar samaritain, Le Serer Ye~Ira et les trente-deux sentiers de la Sagesse," Revue de l'his-
toire des religions 170 (1966) pp. 159-184.
50 Gordon D. Newby, The Making of the Last Prophet (Columbia, SC, 1989).

Gospels, and $abian (al-$iibiyyun al-Ibriihimiyya) sacred texts.51 And the Jews
Mash'aIlah ibn Athan (745-815), Shu 'aib (late eighth early ninth century),
Sahl ibn Bishr (first quarter of the ninth century) and Abul-Tayyib Sind ibn
'All (under Ma'mOO) worked as court astrologers, and also delved into
such occult sciences as the making of talismans.52
By way of summation to this point, it is clear from explicit evidence that
Muslims, Jews, Christians and pagans worked together in the burgeoning
sciences of the second and third Islamic centuries. Considering the evi-
dence of the officially patronized interpenetration of traditions at this time,
Jews thus must have been in contact with non-Jewish individuals and
groups who were interested in scientific matters, including the occult sci-
ences. Whatever the precise extent of this interconfessional activity, this
general context does provide a sensible setting for the emergence of SY. In
fact, no specific historical context in antiquity yet has been shown to carry
nearly so much detailed circumstantial evidence for the origination of SY
as this does.
The 5hz 'z- crucible

But perhaps the most important single context into which we must under-
stand the redaction of SY, and one fully complementary with the evidence
adduced above, is the ShY'ycrucible. I have already noted that MaqdisY, a
Muslim scholar writing in the mid-tenth century, associated Saadia's SY-
commentary with the Batiniyya, that is, with the Shi'a. Why? To begin
with, the cosmological mystique of language, which I term cosmic semi-
otics, was associated with this community from its inception. Among the
specific contents of the wa.fiyya, or "inheritance", which 'All was said to
have received from Mul)ammad, was jafr.53 Deemed the "Kabbala of
Sufism" by Goldziher, jaJr was a (variously defined) hidden teaching
vouchsafed to 'All and his lineage.54 jafr was said to have included such
things as the 73-letter name of God; the known name and the hidden
5] The Fihrist of a/-Nadrm (2 vols., ed. and trans.) Bayard Dodge (New York and London, 1970)
vol. I, pp. 41-42.
52 Mash' allah: See the useful overview of the sources in D.M. Dunlop, Arab Civilization to A.D.
1500 (London and Beirut, 1971) 215-216, Shu'aib: Ibn AbI U~aibi'a (Vol. 1., 130) cited in M.
Ullmann, Die Natur-und GeheimJvissenschaften (Leiden, 1972) p. 228, and in Klein-Franke,
Iatromathematics (cited above, n. 12), p. 59-60; for Sahl, see Sezgin, Geslhichte des Arabischen Schriftums
125-128, and David Pingree, "Classicial and Byzantine Astrology in Sassanian Persia" Dumbarton
Oaks Papers 43 (1989) 227-239, at 235; on Abul-Tayyib see Steinschneider, "An Introduction to the
Arabic Literature of the Jews", Jewish Quarterly Review, o.s. p. 109.

Seier Yef?iraand Early Islam 17
name of the Messianic Imam; and matters pertaining to the last days.55It
was also said to encompass previous revelations, explicitly including the
Torah.56 Its prestige was such that already in the second Islamic century
the ShI'Is could be derided as being those who are "deluded by the )ajr."57
J4r, along with the rest of the wa,rfyya, was claimed to be an inheritance
from the Banu Isra'Il, the Children of IsraeP Thus, these early Shn tradi-
tions do confirm (to some extent) anti-ShI'I accusations of a Judaic "inher-
53 See the important article by HajI KhalIfah, "Jami'a", in Kashf al-Zuniin, (ed. G. Fliigel) vol. II:
604); and T. Fahd "Djafr," EP p. 375-377. For some of the older studies see Casanova, Journal
Asiatique ser 9, t.ll: 156. See also the materials collected by Goldziher in Zeitschrift der Deutschen
Mor:genlandischenGesellschaftXXVI (1872): 782-785 (repr. in Gesammelte Schriften I led. by J.dc Somogyi,
6 vols., Hildesheim, 1967-1973] pp. 183-186), and Zeitschrift der Deutschen Mor:genlandischenGesellschaft
XLI (1887) pp. 123-125 (Gesammelte Schriften II pp. 284-286); St. Guyard in Notices et Extraits des
Mss. de la Biblio. Nationale t. xxii (1874) pp. 177-428, esp. pp. 292 ff.; Louis Massignon, Lexique, pp.
98-101; "La Philo sophie Orientale d'Ibn SIna et son alphabet philosophique," (see n. 1 above) pp.
1-17. For an extensive listing of Shi' jafr-works see Agha Buziirg al-Tihrani, al-Dhari'a Ita Tasanif al-
Shi'a, (Beirut, 1983) pp. 118 ff. On the wa[iyya, see Uri Rubin, "Prophets and Progenitors in the
Early Shi'a Tradition," Jerusalem Studies in Arabic and Islam I (1979) pp. 41-65.
That the primary characteristic of jafr is the permutation of letters (Hebrew [eru!; Arabic ta[rTf) -
of which practice SY is of course the fundamental Jewish classic - is emphasized by Kraus, "La
science du Jafr repose en premier lieu sur Ie principe de la permutation des lettres (Jabir II, 265, n.
8; and Corbin," Or, la science des lettres, Ie jafr, repose essentialement sur la permutation.": Histoire,
(n. 64 below) p. 205.
54 A. Altmann "Maimonides' Attitude to Jewish Mysticism," in Studies in Jewish Thought (ed.) Alfred
Jospe (Detroit, 1981) pp. 200-220 at p. 217 n. 29.
55 On the 73-letter name of God see the sources cited in Etan Kohlberg, "Some Sill I Views of
the Antediluvian World," Studia Islamica 52 (1980) pp. 41-66, at p. 43; A. Sachedina Islamic
Messianism (Albany, 1981) p. 190, the Imam possesses 72 of the 73 letters of the Greatest Name.
On the 72-1etter name in Jewish traditions see Joshua Trachtenberg, Jewish Magic and Superstition
(New York, 1st ed. 1939, repro 1977) pp. 94-96, 289.
On the known name and the hidden name of the Imam, see Goldziher, "Himmlische und
Irdische Namen,", pp. 157-162 (1922, Ges. Scr. V pp. 462-468) Ivanow, p. 7 n. 3; Kohlberg "From
Imamiyya to Ithna-' Ashariyya," Bulletin of the School of Oriental and African Studies 39 (1976) pp.
521-534 at p. 522 n. 9; Sachedina, Messianism, p. 50, and p. 207 n. 34; and further materials gathered
in Wasserstrom Species of Misbelief: A History of Muslim Heresiography of the Jews (PhD diss., University
of Toronto, 1985) p. 216 nn. 40,41.
On the eschatologicaljafr see Fahd, "Djafr," EI'; and A. Abel, "Changements politiques et littera-
ture eschatologique dans Ie monde musulman," Studia Islamica II (1954) pp. 23-43, esp. pp. 33-35.
Jafr, which includes the Torah (MajlisI, Bihar alAnwar, XXVI p. 18) will be revea!ed by the Mahdi
(See the tradition cited and discussed by Goldziher, "Litteraturgeschichte der SI'a," p. 55; and
Casanova, Journal Asiatique ser. 9, t. 11 (1898) p. 156).
56 MajlisI, Ibid., voL XXVI, p. 18; Rubin, "Prophets" (n. 47 above), p. 48, n. 40; and other sources
cited by Kister (n. 54 below) p. 232.
57 Goldziher, Introduction to Islamic Theologyand Law p. 190 n. 72, citing JaJ:ll~.
58 Rubin "Prophets," (n. 47 above) passim. Jafr is sometimes said to be made of cow-hide (e.g.
Sachedina, Messianism (n. 49 above) p. 22, from Kulaini). Could this reflect the fact that some of the
suhaf mentioned as included in the white jafr may have been made of skins. For the Jewish use of
skins for book-scrolls, see Menahem Haran, "Book-Scrolls at the Beginning of the Second Temple
Period: the Transition from Papyrus to Skins," Hebrew Union Colle.geAnnual LIV (1983) 111-123.

itance" for Shi'ism. These traditions concerning "inheritance" do not, of

course, confirm such anti-Shr'I calumnies as those alleging that the rene-
gade Jew 'Abdallah b. Saba' created Shi'ism as a mere put-up job.59 But
early ShI'I traditions themselves do claim a certain spiritual continuity with
Judaism.
Such a claim is unmistakably made concerning the establishment of the
'Alid line. Thus a widespread tradition has it that the two celebrated mar-
tyr-sons of All, MUDammad's only two grandsons, received their names,
I:Jasan and ~Iusayn, from the Angel Gabriel, as translations of the names
of the sons of Aaron. This in turn was an extension of the tradition,
60
accepted by both Sunnis and ShI'Is, which stated that 'AlI's status was to
MUDammad as Aaron's was to Moses.6] The Imamate, then, sought a kind
of Aaronid, "priestly" legitimation. Nor was that the only Judaic paradigm
upon which the Imams patterned their authority: other traditions sought to
establish a ShI'I parallel with the Davidic line as well.62 It should not be
entirely surprising, then, to find Sunni polemicists popularly accusing ShI'Is
os being "the Jews of our Community (Umma)."63
For present purposes, it is important to remember that this continuity,
legitimation and inheritance from Jews to ShI'Is was explicitly said to have
59 Israel Friedlaender, "'Abdallah ibn Saba', der Begrunder der SI'a, und sein judischer Ursprung,"
Zeitschrijt fiir Assyriologie 23 (1909) pp. 296-327; 24 (1910) pp. 278-292.
'" A rich selection of sources is provided by M.J. Kister in "f:1addithu 'an banI isra'ila wa-la
J:laraja," Israel Oriental Studies II (1972) pp. 215-239, at p. 223 n. 37. For various opinions as to the
meaning of these names, see Goldziher, "Himmlische," p. 159 [an Aramaic original]; Wilhelm Eilers
"Shiitische Wasserheilige," Die irlamische Welt Zu)ischen lvlittelalter und Neuzeit, (eds.) U. Haarmann, P.
Bachmann (Beirut, 1979) pp. 94-124, at p. 107 n. 42 [Aramaic original]; Ramzi Baalbaki, "Early
Arab Lexicographers and the Use of Semitic Languages," Berytus XXXI (1983) pp. 117-127, ar p.
126 [Hebrew original]. The encounter with Aaronid traditions was also associated with the Arabian
tribal context: MuJ:1ammad's wife Safiyya was said to be descended from Aaron (Goitein "Banu
Isra'I]" EF, vol. 1, citing Ibn Sa'd, viii, 85, line 27); the origins of the tribes of Na<;JIrand Qurayza
was said to be that they "were two brothers among the children of Harlin, the prophet." (Moshe
Gil, "The Origins of the Jews of Yathrib," Jemsalem Studies in Arabic and Islam IV (1984) p. 208 n.
16).
51 See also E. Kohlberg, "The Term Rafi~a in ImamI ShI'I Usage," Journal of the American Oriental
Society 99 (1979) pp. 677-679 at p. 679 n. 16.
(,2 For traditions associating the Exilarchatc with the Imamate, see Goldziher, "Renseignements de
source musulmane sur la dignite de Resch-Caluta," Revue des itudesjuives VIII (1884) pp. 133-136 [CS
II pp. 121-125]; W.J. Fischel '''The Resch Cal uta in Arabic Literature," [Heb.] The Magnes Anniversary
Book (eds.) 1'.1. Baer and others (Jerusalem, 1938) pp. 181-187. The many (mostly Sunni) texts per-
taining to the ra's jiiltit are largely derogatory, but the parallel of the Shi'i wa{iyya to the descent of
the Exilarchs, in the words of Stillman, "must have appealed to Shi'ite notions of 'Alid legitimacy."
Uews in Arab Lands, (Philadelphia, 1979) p. 176 n. 1).
63 See Wasserstrom, "The Delay of Maghrib: A Study in Comparative Polemics," Logos Islamikos
(eds.) R.M. Savory, D.A. Agius (Toronto, 1984) pp. 269-286, at p. 283 nn. 11-14, for this linkage
of Jews/Rafi~a in Sunni polemics. I have treated this issue more fully in "'ShI'Is are the Jews of our
Community': Israel Oriental Studies XIV [forthcoming].

Seier Yef?ira and Early Islam 19
included the secrets of language. These secrets were transmitted, it was

said, from 'All to his descendents. It is thus that we hear much more
about them in reports concerning the great fifth and sixth Imams, who
dominated 'Ali's line through the mid-eighth century.
The fifth Imam, Mul,1ammad al-Baqir, is assigned a key role in the prop-
64
agation of secrets of language. His "linguistic mystique" particularly con-
cerned aijad, the Arabic alphabet arranged in the order of the Hebrew and
Aramaic alphabets. This al!jad alphabetic system was to gain great prestige
in later mystical circles.65 Moreover, Mul,1ammad al-Baqir is usually assigned
the telling of the most influential early tale of abecedarian myth.66 This
mytheme is well-known from the apocryphal infancy gospels, whose form
is retained in Baqir's version: the child-Jesus goes to school for the first
time. The teacher tells him to recite "A," to which he responds "A." But
when the unsuspecting teacher then says "B" the wise child responds: Tell
me the meaning of "A" and I will tell you the meaning of "B"! The dumb-
founded teacher faints, whereupon the child recites a litany of mystic and
moralistic meanings, in the sequence of the aijad alphabet.67 In the gnostic
apocalypse known as the Umm al-Kitab, this myth erne is told of the five
year old child-Imam Mul,1ammadal-Baqir himself.68
64 For the role of this Imam in the transmission of Islamic gnosis generally, see Halm, Die
Islamische Gnosis (Zurich, Munich, 1982) "Die IIiiretiker um den Imam MuJ:lammad al-Baqir", pp.
84- 112.
65 "Abdjad" El' vol. 1 (G. Weil, G.S. Colin); For Avicenna's ah;ad alphabet, see Carra de Vaux,
Les Penseurs de l'Islam ((5 vols., Paris, 1921-1926) vol IV: 38-42; Seyyed Hossein Nasr, An introduction
to Islamic CosmologicalDoctrines (Cambridge, Mass., 1964, repro Boulder, 1978) pp. 209-210; Massignon
(n. 1 above). For abjad in thc Ikhwan al-~afa' scc Nasr, ibid, p. 50, Table 1. For abjad in Jabir's
usage, in Jabir II, 187 ff. In a brilliant footnote, Kraus showed that the order of the letters enunciat-
ed in the revelation of MuJ:lammad al-Baqir in the Persian-language Umm al-Kitab is based on abjad
ordering, and therefore reveals a more ancient source, which he identifies as the Arabic Infancy
Gospel See Jabir II, 263, n. 8.
For the Jcwish mystical use of ah;c;d, see the extensive trcatment in Elliot R. Wolfson, "Letter
Symbolism and Merkavah Imagery in the Zohar", Alei Shefer, Studies in the Literature of Jewish Thought
Presented to Rabbi Dr. Alexandre Safran (ed.) Moshe Hallamish (Bar-llan, 1990) 195*-236*, especially at
197*-214* (I thank Professor Wolfson for sharing this article with me). Abjad may have had a very
ancient magical use: see "'BGD - Magic Spell or Educational Exercise?" Alan R. Millward, in
Eretz-Israel18 (1985) 39*-42*.
66 The sources gathered by Galtier, Futuh al-Bahnasii. (Memoires publiis par les membres de l'Institut
franfais d'archiologieorientale du Caire) t. 22 (1909) p. 17.
67 The earliest version may be that of 'Umara ibn Wathima, (ed.) R.G. Khoury, in Les Jegendes
prophetiques dans I'Islam (Wiesbaden, 1978) p. 324 (two variants). For Shn versions, see, for example,
al-KulainI, A/-Kaf!, VI: 192; Al-MajlisT, Bihar al-Anwar, (120 vols, Teheran,) vol. XIV, p. 286. For
Sunni versions, see Abii Nu'aim, Hilyat al-Awliya' wa-Tabaqat aI-Aiftya', (10 vols., Cairo, 1932-1938)
vol. VII, 251-252; Tha'labi, Qisas al-Anbiya': (Cairo, 1951) 387; and the texts collected in BalawI, A/if
Ba' (2 vols., Cairo, 1287) vol. I, p. 76.
68 Edited by Ivanow in Der Is/am XXIII (1936) 1-132; Glossed by Ivanow, "Notes sur l'Umm al-
ICitab", Revue des etudes islamiques 6 (1933); translated partially in Halm, Is/amische Gnosis, (n. 58 above)
pp. 128-131.

In this mytheme (a version of which is also found in a work written in

the eighth-ninth centuries, in a Muslim country, the Alphabet of Ben Sira -
the same Ben Sira elsewhere said to have created a Golem by using SY),
the alphabet contains all meaning and is known primordially by the child-
prophet.69 These elements, the comprehensiveness and primordiality of lan-
guage, are central features of the cosmic semiology to which I have allud-
ed.
Even more than Mul)ammad al-Baqir, his son, the sixth Imam, ]a 'far al-
~adiq, is said to have been the formative theoretician of jajr in the form in
which it was to possess its influential mystique.70Not only are many jajr
teaching assigned to him personally, but he is also known to have been the
teacher and master of many noted figures of his day, several of whom
were to be celebrated for their knowledge of language mysteries.
Indeed, no less than four followers of Mul)ammad al-Baqir and his son
] a'far are also assigned salient roles in conveying myths of language. These
include a founder of the ZaydI ShI'I schism, and ]abir ibn I::Iayyan,the
"father" of Islamicate alchemy.71Among the followers who broke with
these Imams are several influential gnostic revolutionaries. One of these,
Mughlra b. Sa'ld, described a (recognizably gnostic) vision depicting the
body of the creator of the universe, upon whose members are seen the let-
ters of the abjad alphabet.72 The linkage of bodily members with letters, it
will be recalled, is also found in some versions of Sy'73
69 For some conjectures as to the provenance of this motif in Ben Sira see 1. Levi's "La nativite
de Ben Sira,"Revue des etudesjuives 29 (1894) pp. 197-205. For a useful new study and translation, see
Norman M. Bronznick (basing himself on the edition in Eisenstein's Otsar Midrashim, pp. 44-50) in
Fiction 7 (1983) [special issue Rabbinic Fantary] pp. 99-136, the mytheme being found at pp. 105-
109. For Ben Sira and the magical use of SY see Scholem On the Kabbalah and its Symbolism (New
York, 1978) pp. 178-179; M. Idel, Golem (Albany, 1990) s.v. "Ben Sira".
70 Corbin Histoire de la philosophie islamique (Paris, 1964) p. 187: "L'Imam Ja 'far est regarde
unanimement comme l'initiateur de la 'Science des lettres.''' In addition to being the source of many
language-myths, he is also said to be the author of long-influential mystical exegesis, the so-called
Tafsir JaJar. see Nwyia's "I.e TafsIr mystique attribue it Ga'far Sadiq," Melanges de l'Universite st.Joseph
XLIII (1968) pp. 182-230. It was probably during the lifetime of Ja 'far that the Jewish pseudo-mes-
siah Abii'lsa al-I~fahanI led his uprising. See Wasserstrom, "The 'Isawiyya Revisited," studia Islamica
LXXV (1992) pp. 57-80.
71 For the Zaydi HarM b. Sa'Id al-IjlI, see Ibn Khaldun, The Muqaddimah trans. Rosenthal (2nd
edition, Princeton, 1967) vol. II: 209; Brockelmann, Geschichte der Arabischen Literatur (Supplement I,
Leiden, 1937) 314; Ibn Hajar, Tahdhrb al-Tahdhrb (Hyderabad, 1907-1909) XI, p. 6; NawbakhtI,
(trans. M. Mashkur) Revue de l'histoire des Religions CLIV, p. 88 n. 3, for more sources.
72 Wasserstrom, "The Moving Finger Writes: Mughlra b. Sa'Id's Islamic Gnosis and the Myths of
its Rejection," History of Religions 25 (1985) pp. 1-29. See also Elliot R. Wolfson, "Anthropomorphic
Imagery and Letter Symbolism in the Zohar," in the "Proceedings of the Third International
Conference on the History of Jewish Mysticism," Jerusalem Studies in Jewish Thought VIII (1989)
147-181 [Hebrew].
73 See the sources given in Dan Cohn-Cherbok "The Alphabet in Mandaean and Jewish
Gnosticism," Religion 11 (1981) pp. 227-234, at 231.

Seier Ye$ira and Early Islam 21
Part III: The Jewish context
The problem of ]ewZ:shness

Even with all the foregoing evidence taken into consideration, it remains
the obvious case that SY is not a Sh1'I work but a Jewish one. That being
said, to characterize the Jewishness of SY as a "problem" may seem odd,
but it is to be remembered that this work carries no discernible traces of
explicitly Rabbinic ideology. This is a st1"~kingfact, often observed in mod-
ern scholarship, but one that has yet to be adequately explored. Indeed,
the oddity of this situation seems inescapable: a manifestly non-Rabbinic
work first commented on by the staunchest defender of Rabbinism, Saadia
Gaon, and soon thereafter by the learned Dunash ibn Tamim, who begins
his commentary with the pointed statement: "We have read some time ago
a book of rabbinic provenance . . ." (Arabic: qad qar 'ana kitaban Ii 'ahl mil-
latina al-rabbanryyin ... )."'74
Despite this statement by Dunash - which, as Vajda observes, in fact
may mean nothing more than that he came across the manuscript in non-
Karaite Kairouan - I would argue that the Jewish milieu into which we
must understand the ninth-century redaction of SY was indeed non-
Rabbinic. First of all, it should be remembered that the first centuries of
Islam engendered the most variegated "sectarian milieu" (to borrow the
phrase of Wansborough) in Jewish history after that of the Second Temple
period. Moreover, these first centuries of Islam produced an extensive
non-Rabbinic literature in Hebrew, which was cultivated throughout the
Jewish world.75The polemics of Hiwi al-Balkhi, the Alphabet if Ben Sira and
the rich Karaite literature, all products of this period, were also all openly
anti-Rabbinic in tone if not in intent. It is therefore certainly not incon-
ceivable that the neutrally non-Rabbinic SY emerges in the ninth century,
when it was possible (if not popular) to attack the proponents of the Oral
Torah in the very language of the Written Torah.
But this circumstantial and indirect evidence is also, of course, inconclu-
sive. A more promising avenue of analysis for identifying the J ewishness of
SY would investigate the specifically intra-Judaic innovations undertaken by
the author of SY. Scholem, the best-known proponent of this position,
argued that SY introduced "for the first time speculative reinterpretations
from the Merkabah."76 Following Scholem, Martin S. Cohen has wondered
74 Vajda, "Commentaire Kairouanais" RE} CVIl (1946- 47) 113.

75 Wasserstrom, "Who Were the Jewish Sectarians Under Early Islam?", in Jewish Sects, Political
Parties, and Religious Movements (ed.) Menahem Mar (Omaha, 1992) pp. 101-113.
76 Origins, p. 25.

"if the SY was not composed as a sort of great leap forward from the
unsophisticated meditative technique of the 5'hi ur Qomah and its sister texts
of hekhalot mystic literature."77However, A.P. Hayman now has studied the
relation between SY and hekhalot literature, and concluded that "it is very
difficult to show any direct connection indicating that our author was
deliberately developing and giving a new direction to the concepts of the
Ma 'aseh Merkava . . . [tJhere is no overlap in their essential concerns."78
This is the more persuasive position, all the more so considering that Peter
Schafer has now argued, on the basis of a systematic analysis of the
Genizah evidence, that the hekhalot literature is itself largely a post-Rabbinic
development. One could conclude that SY, like Bahir, is a medieval work
drawing on manifestly ancient sources.79
Still, even if my arguments for its historical context, generally identifiable
genre and specific textual counterparts should all prove successfully persua-
sive, we would still not know who wrote SY. Grunewald thus remains
accurate in his observation that SY "occupies a kind of spiritual isolation
that is positively unique in the history of Hebrew literature."8o But it is
nonetheless possible, I believe, to characterize SY's orientation to
Jewishness. For if SY is a kind of speculative great leap forward, then it is
oriented in a certain way to the past. It exhibits, in this regard, a certain
anxiety of influence insofar as it found itself unhappily in a posterior posi-
tion in history. To overcome its lateness, SY's redactor resorted to a vari-
ety of postures and attitudes: neologism) ahistorism) and an01'!Ymiry.
SY is famous for its neologisms, which have confounded researchers for
decades. The best known of these neologisms, "srjirot", continue to resist
definitive analysis.8 The Arabophone milieu into which I have placed the
!
redaction of SY, of course, was notable for its linguistic experimentation,
77 M.S. Cohen, The Shi'ur Qomah (Lanham, New York, London, 1983) p. 180.
78 A.P. Hayman, "Sefer Ye~ira and the Hekhalot Literature", Proceedings of the Fir.rt International
Conference on the History of Jewish Mysticism: Early Jewish Mysticism. (Jerusalem Studies in Jewish Thought VI,
1-2, 1987) 71-87 [English section].
79 Peter Schafer, "The Aim and Purpose of Early Jewish Mysticism" Hekhalot-Studien (Tiibingen,
1988) 277-295, at 293: " ... this literature is in fact a post-Rabbinic phenomenon"; Hayman's work
is predicated on the antiquity of SY.
80 Ithamar Gruenwald, "Some Critical Notes on the First Part of Sefer Ye?,ira," Revue des etudes
Juives 132 (1973): 477.
81 See also Pines, "Points", Appendix I, pp. 115-121. The classical discussion remains that of
Scholem: but see now the revisionary arguments of Moshe Idel, for example in "The Sefirot Above
the Sefirot", Tarbi{ LI (1982) 239-280 [Hebrew]. See also the recent work of Elliot R. Wolfson on
the question of the sefirot, especially in his "The Theosophy of Shabbetai Donnolo, With Special
Emphasis on the Doctine of Sefirot in Sefer Hakhmoni," Frank Talmadge Memorial Volume II, (ed.) Barry
Walfish (Haifa, 1992), 281-316. I thank Elliot Wolfson for sharing a version of this remarkable arti-
cle with me. My intention in the present study, as I noted at the outset, is not to analyse themes or
motifs of SY as such. As a complement to the present study, however, I hope elsewhere to analyse
the early Isma'IlI evidence which sheds light on the doctrine of sefirot in SY.

Seier Ye?ira and Early Islam 23
including a considerable amount of neologizing.82 At the least, neologizing

implies an authorial discomfort with the inadequacy of conventional idiom.
As Scholem puts it, " . . . the special charm of [SY] consists in the fre-
quently felicitous and in any event ever-vivid imagery and fullness of
meaning it lends to most of the concepts newly created in order to express
abstractions. The author finds concrete and appropriate designations that,
until then, Hebrew did not know how to render in adequate terms."83
This reaction to the inadequacy of existing modes of religious expression
is also discernible in the pronounced ahistorism of SY. SY's text exhibits no
development, no narrative, no clear sequence. Moreover, it is ahistorical in
another sense: it is reticent to address specified individuals among its read-
ership. Thus, in the effort to adopt a lapidary, oracular reserve, it mimics a
kind of (presumably scripturally) authoritative tone. The author of SY
thereby rejects not only the precisely temporal and personally vivid narra-
tive style of the Written Torah, but also the argumentative engagement, the
social cortesia responsibly obliged to persuade one's audience, typical of the
public voice of Oral Torah. In short, it is not submitting to the contempo-
raneous authority-structure of the Geonim.
SY - again, not surprisingly - was purposely anof!Jmous. Its author
found no Jewish sodality worthy of its public support, if we may judge by
its marked silence. Its anonymity would seem to reflect a propagandistic
technique, one that argues for the superiority of its own group precisely by
concealing its identity.
Anof!Jmiry, ahistorism and neologism seem to suggest that SY was not the
creation of a loyal Rabbinist, and that its function therefore must be found
outside the precise demand-system of Rabbinic Judaism. In the foregoing,
I have suggested a context where that alternative system may have been
located. On this hypothesis, SY functioned as a Jewish expression of the
well-known Shy'Y-influenced gnostic intellectualism of the Muslim world,
not that of a barely-documented secret Judaism of the Amoraic period.H4
But this point, if conclusive, only transposes the mysteriousness of its
emergence: why then did Saadia accept it? After all, the Geonim were
ambivalent about the raz traditions of Jewish antiquity.85One answer may
82 C.E. Bosworth, "Karramiyya," E1', vol. IV: 667-669, at 668, gives some examples from thl'
usage of this sect. .
83 Origins, p. 33. Note the apparent disparity between this statement and the one Scholcm puh
lished in the Encyclopedia Judaica ("Sefer Yqira", vol. 16: 785-786): "[SY] contains no hngulstlc torn)
which may not be ascribed to second or third century Hebrew." "
84 Joseph Dan, following Scholem, recognizes that SY appears fully discontlnuOUS \\'1lh IbhlHlllt'
tradition: " ... SY seems to be independent of the linguistic and cultural atmosphere of alll'll'lll
rabbinic texts." Three Types of Ancient Jewish Mysticism (Cincinnati, 1984) 20., ' ..
85 E.E. Hildesheimer, "Mystik und Agada im Urteile der Gaonim R. Schema 1I1ld R, Iial.
Festschrift fur Jacob Rosenheim (Frankfurt am Maim, 1931) 259-286.

be that Saadia found SY useful (unlike the Shi'ur Qomah or other more
mythological compositions) because, malgre lui, it was attuned to his intel-
lectual sensibility. Both its rational tone and its demythologizing effect
spoke to him, I suggest, because they both emerge from a common ratio-
nalizing historical moment. In this sense, I see SY as historically transition-
al, a fact which may account for its appeal among the early commentators.
Scholem seems to indicate this when he notes that "[p]erhaps one could
view the text as situated at the limits of esotericism, partly within it, but
partly beyond it."86
My argument with regard to the intra-Judaic innovations initiated by SY,
then, is that they do indeed constitute a kind of speculative great leap for-
ward, though not a leap beyond hekhalot literature. Rather, they derive from
late gnostic/intellectual circles, who were still active in the ninth century.
To cite an example, a converted Jew of the ninth century, Al;1mad ibn
Abdallah ibn Salam, translated both from the Bible and from Sabian texts,
which fact indicates that he was culturally conversant both with Jewish lit-
erature and with the Sabian form of gnostic literature. Yet more important-
ly, a considerable amount of (still-unexplored) evidence for Jewish gnosti-
cism of this period is provided by Ibn al-NadIm and by Theodor bar
I<:honai, which evidence clarifies the extent to which Jewish and gnostic
circles remained interactive in the first centuries of Islam.87 Finally, I have
shown elsewhere that these trajectories of what Scholem termed J ewish-
Gnosticism may corroborate Scholem's intuitions that the origins of .5ifer
ha-Bahir may be found in this milieu.88
Again, the non-Jewish milieu provides instructive parallels here. The
pagan gnostics themselves, the notorious Sabians most especially, nominally
"converted" in the eighth and ninth centuries, but not directly to SharI'a-
traditionalist Islam. Rather, they successfully became professional culture-
intermediaries, specializing in translating from Greek and in cultivating the
occult sciences.89 As mediating intellectuals, middlemen, they moved texts
86 Origins, p. 35. For Saadia's motivations for writing his commentary, see J ospe (cited in the first
note above); and Haggai Ben-Shammai, "Saadya's Goal in his Commentary on Seftr Ye:rira," in A
Straight Path. Studies in Medieval Philosophy and Culture. Essays in Honor of Arthur Hyman, (ed.) Ruth
Link-Salinger et al. (Washington D.C., 1988) 1-9. Ben-Shammai is correct to assert that Saadia's
intent is to demythologize (p. 9), though he does not suggest that this demythologizing is already
implicit in SY. This apparently paradoxical tendency was already ancient, as Adorno observes: "The
mythologizing equation of Ideas with numbers in Plato's last writings expresses the longing of all
demythologizing: Number becomcs the canon of the Enlightenment." Dialectic of Enlightenment,
Theodor W. Adorno and Max Horkheimer (New York, 1944, repro 1989) p. 7.
87 I have summarized this argument, with reference to the relevant literature, in "Who \'Vere the
Jewish Sectarians under Early Islam ?" (cited above, n. 75).
88 Ibid.
89 See Peters (cited in n. 10 above) Corbin speaks of the " ... secret intention which places the
~abian theosophers in a mediating position between positive Islam and the Ismaili exegesis of the
Brothers [= Ikhwan al-~aIaT': "Sabian Temple and Ismailism," Temple and Contemplation translated by
P. and L. Sherrard (New York and London, 1986) 132-183, at 162, n. 97.

N
(Jl

io
rt
(1)
en

(;J
o


(1477285X - The Journal of Jewish Thought and Philosophy) Sefer Yesira and Early Islam - A Reappraisal

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(1477285X - The Journal of Jewish Thought and Philosophy) Sefer Yesira and Early Islam - A Reappraisal

Uploaded by

Copyright:

Available Formats

J. Jewish Thought and Philosophy, Vol. 3, pp.

Sefer ye~ira and Early Islam: A

A book impenetrable by contingency,

Downloaded from Brill.com08/02/2019 01:13:42PM

Part I: The Kraus hypothesis

The Imrnot Alef/ Mem/ Shin

Downloaded from Brill.com08/02/2019 01:13:42PM

Massignon and Corbin would subsequently use these apparent similari-

3 Jabir U, p. 267, (see n. 31 below) where he acknowledges Massignon's original suggestion. A

Downloaded from Brill.com08/02/2019 01:13:42PM

the "extremists" (ghu/iif), as they thereby depicted themselves holding the

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

and subsequent Ismii'Ili teachings on the 'tij'mi. Thus, as Massignon noted,

Downloaded from Brill.com08/02/2019 01:13:42PM

Just such symbolic sounds were invoked in the language-speculation of the

Part II: Circumstantial evidence

14Edited by 'Arif Tamir and 1. A. Khalife (Beirut, 1970) pp. 148-149.

Downloaded from Brill.com08/02/2019 01:13:42PM

my paper I will gather contextual evidence which would support a dating

Al-Ma'mun and the 'Abbasid court

Downloaded from Brill.com08/02/2019 01:13:42PM

Ithna AsharI ShY'ytexts. There would seem to be little reason to doubt

19 "'All al-Rida", El'

Downloaded from Brill.com08/02/2019 01:13:42PM

Nicomachus of Gerasa's Introduction to Arithmetic (wr. ca. 100 c.e.), with

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

scientistic, apodictic tone. Both compnse SiX sections, sub-divided into

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

Fayyfunl. They interpret the Torah by a method of isolated letters, as do

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

The 5hz 'z- crucible

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

itance" for Shi'ism. These traditions concerning "inheritance" do not, of

Downloaded from Brill.com08/02/2019 01:13:42PM

included the secrets of language. These secrets were transmitted, it was

Downloaded from Brill.com08/02/2019 01:13:42PM

In this mytheme (a version of which is also found in a work written in

Downloaded from Brill.com08/02/2019 01:13:42PM

Part III: The Jewish context

The problem of ]ewZ:shness

74 Vajda, "Commentaire Kairouanais" RE} CVIl (1946- 47) 113.

Downloaded from Brill.com08/02/2019 01:13:42PM

redaction of SY, of course, was notable for its linguistic experimentation,

Downloaded from Brill.com08/02/2019 01:13:42PM

including a considerable amount of neologizing.82 At the least, neologizing

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

Downloaded from Brill.com08/02/2019 01:13:42PM

You might also like