Icons of Power

THE MAGIC IN HISTORY SERIES

FORBIDDEN RITES
A Necromancer’s Manual of the Fifteenth Century
Richard Kieckhefer

CONJURING SPIRITS
Texts and Traditions of Medieval Ritual Magic
Claire Fanger

RITUAL MAGIC
Elizabeth M. Butler

THE FORTUNES OF FAUST

Elizabeth M. Butler

THE BATHHOUSE AT MIDNIGHT

An Historical Survey of Magic and Divination in Russia
W. F. Ryan

SPIRITUAL AND DEMONIC MAGIC

From Ficino to Campanella
D. P. Walker

ICONS OF POWER
Ritual Practices in Late Antiquity
Naomi Janowitz

The Magic in History series explores the role magic

and the occult have played in European culture, religion,
science, and politics. Titles in the series will bring the
resources of cultural, literary, and social history to bear on the
history of the magic arts, and will contribute towards an
understanding of why the theory and practice of magic have
elicited fascination at every level of European society.
Volumes will include both editions of important texts and
significant new research in the field.
 M a g i c I n H i s t o r y

Icons of Power
Ritual Practices in Late Antiquity

Naomi Janowitz

THE PENNSYLVANIA STATE UNIVERSITY PRESS

university park, pennsylvania
 Library of Congress Cataloging-in-Publication Data

Janowitz, Naomi.
Icons of power : ritual practices in late antiquity / Naomi Janowitz.
p. cm. — (Magic in history)
Includes bibliographical references (p. ) and index.
ISBN 0-271-02147-0
1. Magic, Jewish—History—To 1500. 2. Ritual—Rome. 3. Magic, Ancient.
I. Title. II. Series.
BF1622 .J45 2002
291.3'8'093—dc21 2002001255

Copyright © 2002 The Pennsylvania State University

All rights reserved
Printed in the United States of America
Published by The Pennsylvania State University Press,
University Park, PA 16802-1003

It is the policy of The Pennsylvania State University Press to use acid-free paper for the
first printing of all clothbound books. Publications on uncoated stock satisfy the minimum
requirements of American National Standard for Information Sciences—Permanence of
Paper for Printed Library Materials, ANSI Z39.48–1992.
For Andy
 Contents
Acknowledgments ix
Introduction xi
Abbreviations xxvii

1 Late Antique Theories of Efficacy 1

2 The Divine Name as Effective Language 19

3 Thinking with the Divine Name:

Theories of Language in Christian Exegesis 33

4 The Meaning of Letters:

From Divine Name to Cosmic Sounds 45

5 Using Names, Letters, and Praise:

The Language of Ascent 63

6 Combining Words and Deeds:

Angelic Imprecations in the Book of Secrets 85

7 Transformation by Deed Alone:

The Case of Alchemy 109

Concluding Note 123

Select Bibliography 129
Primary Source Index 145
General Index 153
 Acknowledgments
For general intellectual guidance, comments on individual chapters, and
answers to specific technical questions, I want to thank Michael Silver-
stein, Richard Parmentier, Gregory Shaw, David Olster, Cale Johnson,
Sara Rappe, and Rebecca Lesses. Editorial help came from Priscilla
Stuckey, Devorah Schoenfeld, Jonathan Cale Harris, and Abby Limmer.
Peggy Hoover at Penn State Press edited the manuscript with great care.
Thanks are due to the series editor Richard Kieckhefer and to Peter J. Pot-
ter, editor-in-chief at Penn State Press. Chapter 2 was previously published
in slightly different form in Reflexive Language: Reported Speech and
Metapragmatics, edited by John Lucy, whose his meticulous work im-
proved the chapter greatly. The chapter was also delivered as a paper at
the Center for Psychosocial Studies in Chicago, Illinois, and the comments
from the participants were very valuable. History of Religions published
a version of Chapter 3. The research was supported by grants from the
University of California–Davis.
 Introduction
During the first centuries of the Common Era, Jews, Christians, and ad-
herents of Greco-Roman religions engaged in a stunning and bewildering
variety of rituals, and practitioners from these traditions deployed all
manner of wonders. Individuals traveled through the heavens to see flam-
ing chariots and hear angels give praise. Imprecations were used to sum-
mon angelic assistants to a host of tasks, including procuring lovers and
evading enemies. Practitioners used knowledge of cosmic secrets and di-
vine names in their rites and could also show their power by transforming
metals, thus revealing the hidden processes of nature.
All the rituals depended on making an alliance with superhuman pow-
ers. Once manifested on earth, divine power could be used for everything
from settling an old score to changing the very nature of earthly existence.
Small tasks and major transformations were all enacted in the context of
a finely calibrated cosmos where human speech and action were inextri-
cably bound up with divine speech and action. Human and divine become
at points inseparable. For many Jews, Christians, and Greco-Roman be-
lievers, the limits of human access to the supernatural were the limits of
the human imagination.
Some of these rituals presumed a vast cosmos with dizzying layers of
heavens full of entourages of angels—the higher and farther away from
earth and matter, the better. In this multiheavened cosmology, a trip up-
ward was a means for traversing the cosmos and gaining access to the
upper regions where the deity dwelt. Escape was the goal, escape from all
the bonds of earthly existence and life in the body, including fate, decay,
and death.1 The gap between heaven and earth was not just spatial.
Heavenly existence was entirely different, and infinitely superior, to
earthly existence.
1. This cosmology characterized by J. Z. Smith as “utopian” is described by Nilsson
1948:96–11. The contrasting “locative” view is described in the next paragraph.
xii Introduction

For others, the world was far from the uppermost heavens, but it was
still possible to find the highest spirit in the lowest level of existence. Na-
ture held as much lure as the skies, but alignment, not escape, was the
key.2 Because the highest heavenly elements were available to humans in
the lowest earthly elements, it was possible to work one’s way upward
from any point on the continuum. Ascent permitted the individual to join
the upper realm, where he could participate in the heavenly world of the
heavens along with its divine occupants.
This spectrum of beliefs produced a rich array of rituals, rituals that are
evidence of the tremendous diversity of religious ideas at the turn of the
millennium and in the centuries after. The rituals selected for this study
stem primarily from Jewish texts, but they all have Christian and Greco-
Roman parallels (first century b.c.e. to mid-fourth century c.e). They
were chosen both because they have been marginalized in modern schol-
arship and because they present a rich variety of ritual strategies, and they
include ascents through the heavens, alchemical pursuits, and acts aimed
at gaining special help with a variety of life issues, such as problems with
love and personal advance. These are found in a variety of Hebrew and,
secondarily, Greek texts, all of which are introduced chapter by chapter in
the present volume.
The ritual texts, and the more speculative texts that develop the philo-
sophical underpinning for rituals, do not present themselves as being mag-
ical, suspect, or in any way marginal vis-à-vis their religious traditions,
but this does not mean that the rituals were without controversy. There
are issues of secrecy and strict limitations of both knowledge and practice
to select individuals and circles. The practitioners of these rites were no
doubt enmeshed in debates about who was permitted to engage in them
and who was not, and what the use of divine power implied. For exam-
ple, in rabbinic circles we find evidence of hostile attacks about who was
and was not permitted to make claims about what they had seen in the
heavenly world.3 Modern scholarship often presents the ancient debates
about access to power as being a question of who engaged in “magic,” but

2. This cosmology, characterized as “locative,” represents a vision of the world that had
been traditional since the time of the Ancient Near Eastern religions. See J. Z. Smith 1978:
1–13, 13–42, 147–51, 16–66, 169–71, 185–89, 291–94, 38–39.
 Introduction xiii

this stance simply repeats the ancient prejudices without analyzing them
and is thus rejected in this study.
While the thrust of this study is directed at elucidating the Jewish ritu-
als, every chapter includes comparable rituals from other religious tradi-
tions (Christian and Greco-Roman). The reasons for this are many. First,
the Jewish texts and the rituals they describe are often difficult, if not im-
possible, to date. For all these rituals, we depend on textual traditions that
are fraught with problems. The Christian and Greco-Roman material is
thus crucial in dating the Jewish ideas. Second, and in some ways more
important, one theme of this study is that we can understand these rituals
only when we look at them from a broader than usual base. These rituals
reflect, so the argument goes, common ideas about efficacy. These ideas
were shared across the boundaries of what we tend to think of as inde-
pendent, and essentially distinct, religious traditions.
While these rituals may seem exotic to the reader, we shall see that
they were based on the same notions of how rituals work as more fa-
miliar rites. The power of the divine name drives every Jewish blessing,
and the Christian Eucharist is dependent on the transforming capability
of words.
The ritual texts have been the source of much modern controversy,
which in part reflects ancient debates about who is permitted to engage in
these practices. The rituals were understood to be effective, but not neces-
sarily permissible for everyone (as Chapter 1 explains). Ancient competi-
tors—that is, other practitioners than those who preserved and copied the
texts for us—sometimes labeled these rites “magic.” To those who prac-
ticed them, all such actions were in the image of the deity.
This book seeks to rescue these rituals that modern scholarship, per-
petuating some Late Antique biases, continues to denigrate, or mis-
characterize, as “magic.” The very term “magic,” the source of so much
recent controversy, is only one small piece of a complex map of ideas
about how and why some rituals work and how and why others are less
effective, less suitable, or wrong. Many modern scholarly studies of Late
Antique rituals fall into the “Yes, but” category, first critiquing the term

3. See the discussion in Chapter 5.

xiv Introduction

“magic,” then using it anyway.4 This Introduction briefly reviews how the
current categories used for classifying these texts (“magic” and “mysti-
cism”) have limited our engagement with them, and then introduces the
methods used in this study.
It is beyond the scope of this book, and seems redundant to me, to re-
hearse the demise of classical twentieth-century definitions of “magic.”5
This study builds on the growing consensus that such labels as “magic”
are inseparable from their pejorative use in the past. Modern writers can-
not use the term as if it represented more than a polemical distinction
about who was permitted to indulge in what activities.6
Nevertheless, some scholars of Late Antique Jewish rituals continue to
use the label “magic,” just as, for example, Ludwig Blau did in his 1914
Das altjüdischen Zauberwesen. Blau included as “magic” any references
to the evil eye, powerful words, use of the divine name, amulets,
mezuzahs, and exorcisms.7 Other scholars have not missed the trenchant
criticism of once-standard definitions of magic and are eager to embrace

4. H. S. Versnel accepts many of the critiques of traditional notions of magic but balks at
totally abandoning substantive definitions (1991). He relies in the end on a “common sense”
definition of magic, which includes the use of compulsive means in activities such as cursing.
However, labeling a means “compulsive” is often subjective. Equally problematic is the idea
of keeping certain types of unpleasant behavior such as cursing out of standard definitions
of religion. C. R. Philips explicitly endorses a sociological definition of magic for the Late
Antique world, but then explains that the lack of a more coherent definition was due to poly-
theism and its “receptive” nature (1986). A proper definition of religion and hence of magic,
he argues, was only to be worked out after the Theodosian Code. Yet the internecine charges
of magic between Protestants and Catholics belie a simple and consistent Christian defini-
tion of magic. Even Susan Garret in her laudable critique falls back on older definitions in
such statements as “Although there were widespread prohibitions against it, magic flour-
ished in the Greco-Roman world” (1989).
5. For an extremely lucid summary and critique, see Tambiah 1990.
6. For one recent analysis of how the term functions as a polemic, see Neusner
1989:61–81. See also Janowitz 2001:13–26.
7. This model is followed closely by Peter Schäfer, who includes as magic the use of seals,
“magic” speech, amulets, and “magical” acts (1990). He does not explain what magical
speech and acts are, but uses the terms liberally and makes it clear that they are by definition
nonrabbinic. Describing one text discussed below (Chapter 5), he writes: “We are dealing
with nothing less than a primitive hocus-pocus or abracadabra” (1992:165). Likewise, he
writes about the heavenly journey and the adjuration of angels that “in both cases the means
of achieving this is magic. The worldview that informs these texts is deeply magical. The au-
thors of the hekhalot literature believed in the power of magic and attempted to integrate
magic into Judaism” (1988a). For a similar usage, see Kern-Ulmer (1996).
 Introduction xv

new definitions. Hence, modifications have been made during the past
decades to Blau’s schema, primarily by reclassifying some of the items,
such as the use of mezuzahs, as religion. The central theme left in the rap-
idly shrinking Blau model is the use of the divine name, which appears to
be the last-gasp substantive definition of magic.8 For this reason, attention
is directed in this study toward the ideologies of the divine name, and an
attempt is made to find a more precise way to describe its role.
In this study I also eschew the label “mysticism” for the texts, including
the hekhalot (palace) texts analyzed in Chapter 5. A consistent definition
of mysticism has never been formulated. In addition, the term sets up an
implicit hierarchy of religious expression, much of which comes from
Christian writers.9 Mysticism means, for example, spiritual as opposed to
legalistic modes of religious expression. The classical mystical worldview
is usually connected with a specific cosmology, such as the Hellenistic one
with its deus abscondus (absent god) in the seventh heaven or beyond. The
search for communion or union with this far-off god becomes the core of
mystical pursuit. Some religions, because they have other cosmologies with
different structures for their interactions with divinity, end up being by def-
inition “nonmystical” and thus lower forms of religious expression. For
example, the religious expressions found in parts of the Hebrew Scriptures
have suffered from this characterization, and even more so have “primi-
tive” religions. Rather crudely put, the deity is too “close” to have exten-
sive concern with the struggle for mystical communion. Communion with
a far-off deity ends up being classified as a more spiritual experience than,
for example, meeting a god at a tent or sitting down to a meal with a god.

8. For one example, among many, of magic consisting primarily of divine names, see Elior
1993:11. Schäfer distinguishes between an ascent to heaven based on ecstasy, and another
that uses magical and theurgic means. This distinction rests entirely on calling the use of di-
vine names “magic” (1992:154).
9. In his history of the term “mysticism,” Boyer unwittingly shows the Christian bias in
the shift from translating the Greek term mystikos not as “mysterious” but as “mystical”
only when the term appears in the writings of church fathers. While the Jewish use is evalu-
ated more highly than the pagan use, it still falls short and does not represent “mysticism”
per se. Boyer explains that “with Philo it was merely a way of poeticizing about technical ex-
positions of the most abstruse problems, with Clement and Origen it was generally em-
ployed for all that touched upon what was considered the most difficult theological
problems presented by Christianity” (1980:45).
xvi Introduction

Yet another important theoretical problem with the term “mysticism”

is the vast range of religious experiences which it conflates. Mysticism as
a scholarly category grew in popularity in part because it appeared to of-
fer an easy solution to the vexing problem of comparing religions. R. C.
Zaehner found reassurance in his observation that “comparisons between
mystical writings of quite divergent religions are at least comparisons be-
tween like and like” (1960:2). If orthodox theological doctrines look too
particularistic for comparison, the experience of the sacred in its most
generalized form is more easily compared. Everyone was presumed to
experience the same ultimate reality.
These presumptions are now widely doubted. Steven Katz (1978,
1988) points out that there is no such thing as an unmediated religious
experience. Stripping away the language in hopes of finding the pure
experience it encodes cannot be done. To posit a phenomenological real-
ity, “mysticism” presumes that the experiences “behind” the language
used to describe it are of the same god, or, more ecumenically, the same
Reality. However, as Hans Penner notes, “We must remember that all we
have for understanding mysticism is language, not experience” (1983:91).
Huston Smith (1990) recounts a touching episode in which a visitor to
America deposited money in a bank and then asked to have the very same
dollar returned. Smith intimates that, unlike the misguided native, we cor-
rectly know that all dollars really are the same. This is a stunning demon-
stration of how positing a phenomenological “sameness” results in the
loss of the meaning of a cultural artifact. Money does not function in
every society as it does in America.10
Some have tried to shrink the theoretical claims entailed in the term
“mysticism” by shifting the term from a noun to an adjective, emphasiz-
ing the “mystical” dimension of Judaism. Penner has critiqued this stance,
arguing that the notion of a mystical aspect to a religious tradition still
distorts the evidence (1983). Penner’s particular example comes from
Hinduism. The isolation of ascetic practices as the core of Hindu mysti-
cism distorts the traditions because it separates asceticism from issues of
caste. Asceticism and caste are two sides of the same coin. Placing Hindu

10. For an illuminating exchange between Katz and Smith, see Katz 1988.
 Introduction xvii

ascetic practices next to ascetic practices from Christianity, for example,

ignores the fact that choosing to be an ascetic is a choice to step out of the
caste system. Without this context, we cannot understand what choosing
to become an ascetic meant within the religious system. He concludes:
“The mystical illusion is the result of an abstraction that distorts the se-
mantic or structural field of a religious system. As such it is a false cate-
gory, unreal, regardless of whether it is taken as the universal essence of
religion or as a particular feature of a religious system” (1983:96).
From yet another point of view, Proudfoot points out that the term
“mysticism” carries with it modes of evaluation. The term, therefore,
comes with built-in explanations, as does, for example, the word “mira-
cle.” Thus the term cannot be used and explained in “naturalistic terms”
(1985:137). Similarly, mystical experience is, “a phrase that includes
among the rules for its proper application an explanatory commitment—
namely, the judgment that whatever physiological or mental states are be-
ing identified as mystical could not be accounted for in naturalistic terms”
(1985:139).
At the same time, the mystical experience is considered “authoritative,
revelatory, and . . . support for the teachings of the tradition within which
it is identified and interpreted” (1985:153). Scholars then have the odd
task of choosing which religious experiences will be sanctioned as “mys-
tical” and which other, nonmystical, experiences must be explained with-
out recourse to any notion that they are grounded in some reality.
Finally, classifying a text as “mysticism” puts an emphasis on the ex-
periential dimension of the texts. This is the most inherently elusive aspect
of the material. because we have so little information about the experi-
ences of those who used the texts. Despite this lack of information, ques-
tions of experience have dominated much recent debate.
The need for “mystical” Judaism derives from a once-dominant picture
of Late Antique Judaism as legalistic, insular, and abstracted from the gen-
eral historical context. Such scholars as Gershom Scholem argued the no-
tion of Jewish mysticism in order to make room for modes of Jewish
religious expression neglected by the previous generations of scholars.
Ironically, this strategy now functions to keep the texts Scholem champi-
oned at the periphery of Jewish expression.
xviii Introduction

Modern Theories of Ritual Efficacy

The issue of the perceived efficacy of Late Antique rituals has two dimen-
sions, and these supply the twin goals of the study as well as the grounds
for discussion. The first goal is to illuminate how ancient practitioners
understood their rituals to work. The second goal is to demonstrate that
the semiotic vocabulary developed by modern anthropologists and
philosophers of language offers us precise tools for analyzing the Late An-
tique ritual texts. These twin goals will necessitate shifting between some
dense Late Antique texts and some equally dense modern terminology;
each will be explained as carefully as possible.
The first goal implies many questions. Did the practitioners have theo-
ries to explain how and why their rituals worked? How were the various
components of the ritual (the words said, the actions taken) understood to
bring about the transformation of a person or a metal? Uncovering Late
Antique ideas about rituals is not difficult; the challenge is to analyze these
notions instead of simply adopting them. Ancient texts themselves some-
times express directly the ideologies that informed these rituals, as we
shall see in the case of the Jewish exegetical texts discussed in Chapter 2.
One claim of this study is that we need better terminology for analyz-
ing rituals than the Late Antique terms we have tended to recycle, which
have been laden with Late Antique implications and modern polemics.11
This drive for more precise terminology leads us perforce to the work of
linguists and anthropologists as a source of unbiased tools for analysis.
From the vast recent scholarship on ritual, this study draws on the work
of S. J. Tambiah, Michael Silverstein, and Richard Parmentier, anthropol-
ogists who have developed the semiotic analysis of ritual.12 Semiotics
refers to the use of signs, a sign being something that stands for something
else. Rituals are highly structured uses of signs that have complex rela-
tionships with their contexts of use. That is, they can “do things” to the
world around them based on socially conceived models of efficacy. To
pour oil on someone, to put on a special robe, to recite a formula, to point

11. For example, to state that pagans or polytheists engaged in magic, as so much mod-
ern scholarship does, is simply to repeat a Christian and/or Jewish polemical stance.
12. For introductions to these scholars, see in particular Tambiah 1985:17–59, 60–86,
123–66; Silverstein 1978 and 1993; and Parmentier 1994:128–34.
 Introduction xix

a scepter at someone—all these actions change the contexts in which they

occur. In his article “A Performative Approach to Ritual,” S. J. Tambiah
in calls attention to the formality, stereotypy, condensation, and redun-
dancy of ritual (1979:119). These aspects of ritual are often considered
potentially negative, in contrast to, for example, spontaneous or unique
events. Indeed, rituals are, to use Rappaport’s wording, “among the most
perfectly recurrent social events” (1992:14). Rituals can seem so rigid and
devoid of meaning that they are simply declared meaningless. Oddly
enough, the very aspects of ritual that appear so backwards turn out to of-
fer a key to how rituals work. Formality does not always imply lack of co-
herence; the repetitive structure of poetry is a key to its ability to create
new meanings. The ability of signs to affect the contexts in which they are
used is dependent on the structures in which they are employed.
Tambiah argues, for example, that the repetitive patterning of rituals,
which he calls “ritual involution,” is a key to their function (1985:
123–66). The repetitious structures of a rite construct a model that is a
copy of every prior rite leading back to the original model.13 Each individ-
ual Eucharist reminds us, for example, of the original event on which it is
patterned. It must resemble the original event closely enough for it to de-
rive implications from that original event. As Parmentier explains, ritual
actions “are not just conventional, they are so conventionalized that
they highlight or call attention to the rules, that is, to the pattern, model,
or semiotic type that the ritual actions instantiates (is an instance of)”
(1994:133).
In religious rituals the original model includes the actions and words of
a god or some type of supernatural figure. Each rite is one instance of a
general type of socially conceived ritual action in which divine power is
active. To the extent that rituals are based on culturally specific patterns,
such putting into play of divine powers may be mystifying to outsiders, as
well as insiders.
The repetitious structures of rituals not only point toward past events
but also help create the settings for the rite. The language, and other signs,
used in the rites construct the context for the successful completion of the

13. Thus even a “happening” must recreate the specific form of chaos based on the orig-
inal “happening.”
xx Introduction

rite. The formal structure of ritual is in part an articulation of the “cul-

turally-derived aspects of the cosmologies” (Tambiah 1985:142)—that is,
the words spoken, and the objects used in rituals, construct for the partic-
ipants the context they need for the rite to be successfully completed. We
need to find ways of describing the context-creating ability of words and
of objects, or, in the terms of this study, of all the various types of signs
(semiotics) employed in rites.
On the issue of language, to the outsider ritual language often looks ab-
surd because it appears to violate rules of semantics. It is highly repeti-
tious or even nonsensical. Modern scholars of religion have tended to
look askance, for example, at Late Antique ideas about divine names that
diverge from our own “words are names for objects” (referential) model
of language.14 Late Antique ideas about the divine name, and other odd-
sounding ideas about language we will encounter, are not oddities that
defy any mode of analysis. Everyone who uses language is aware that, at
times, words seem to “do things.” The most well-known and influential
formulation of this concept is J. L. Austin’s How to Do Things with
Words (1962). Among his most famous examples is the formula “I now
pronounce you man and wife,” which effects the marriage. Austin’s study
has been used by numerous scholars of ritual, including Tambiah. Austin
distinguishes between several types of verbal acts, including illocutionary
and perlocutionary acts; the latter are acts where the act is achieved by the
utterance itself. His notion of “felicities” raises the question of the rela-
tionship between various formulas and the context of use and attempts to
make systematic observations about these relationships.
This study does not use the particular terminology of Austin, but in-
stead locates his ideas within a more general theory of the multifunction-
ality of language.15 This theory about the different possible functions of
language supplements the more familiar function of words as referential
(semantics, how words refer to objects) with the less familiar notion of
pragmatic functions. Pragmatic functions relate language to the contexts
of use, much as Austin was attempting to do, but without his emphasis on
certain verbal forms. Thus, pragmatic speech refers to speech that affects

14. See Silverstein 1976 for a discussion of the referential drive in linguistic theory.
15. For Austin’s limitations, see Silverstein 1978.
 Introduction xxi

the context in which it is spoken—that is, speech that has consequences in

the real world. These linguistic functions lead us to the heart of ritual, be-
cause ritual formulas (and other linguistic units) influence the context in
which they are spoken.16
These functions of language are not unique to ritual. S. J. Tambiah
named his article the punning “Magical Power of Words” in order to
point out that this property belongs to words in general. One place to be-
gin thinking about the functions of language is with legal language. In cer-
tain social contexts, for example, words are understood to effect
contracts. These functions of language are culturally specific. They neces-
sitate in turn a meta-system—that is, a set of rules that clarify and delin-
eate how and when words work (function in this way). Much of the
training in law school has to do with getting students to move beyond the
“semantic-interpretive orientation” they learned in previous schooling
(Mertz 1996). The students practice for example, oral exchange and de-
bate, thereby learning the rules for the special language uses associated
with law. The students need to grasp such notions as precedent and pro-
cedural history. All these are meta-level concepts that determine what can
be done with the words of law. As Elizabeth Mertz explains, “A legal
reading of case law focuses on the metapragmatics of the text, in which
lies the key to its authority” (1996:236).
So too the function of political oratory is often not simply what the po-
litical speech is “about.” As one example among many, Parmentier ana-
lyzes for us the political speech that a high-ranking titleholder in Belauan
society gave during a crisis in the village. In this speech the titleholder
sought to bring about a resolution to the crisis not so much by talking
about the resolution as by enacting it in his very performance (1994:96).
Repeatedly in law, politics, and religious rituals, we are reminded that
semantics is not the be-all-and-end-all of language. The general notion
that words “stand for” objects—which is the basis of semantics—will not
suffice. The patterns, or models, of rituals are replete with intricate and
bewildering uses of words and other signs that are less familiar modes of
representation. We must have ways of describing the rules by which

16. For a detailed discussion of the many distinct functions of language, see Silverstein
1993.
xxii Introduction

words and objects “stand for” other things in these rituals: how an image
“stands for” the person against whom a rite is directed, how a name
“stands for” a divine presence, how incense “stands for” a cosmic ex-
change. If we are going to do anything other than guess at the meaning of
these rituals we need a way to articulate exactly how these “standings
for” work and how they differ among themselves.
This study uses the sign system developed by C. S. Peirce to delineate
modes of representation (1940).17 For those dissatisfied with the vague
concepts of “symbol” found so often in the analysis of religion, Peirce of-
fers a detailed analysis of how signs stand for objects.18
A sign does not simply “stand for” an object in the familiar way in
which, for example, a name stands for an object. Instead, for a sign to
have meaning it must include a triadic relationship between the object it-
self, a sign that represents that object, and an interpretant—that is, the
“translation, explanation, meaning or conceptualization of the sign-ob-
ject relation in a subsequent sign representing the same object” (Parmen-
tier 1994:5). Each sign “stands for” its object in different ways, depending
on the particular relationship between the three dimensions of the sign
(object, sign, and interpretant). The specific terms Peirce uses for the three
distinct ways of “standing for” are “icon,” “index,” and “symbol.”19
Peirce defines the first type of sign, “icon,” as “a sign which would pos-
sess the character which renders it significant, even though the object had
no existence; such as a lead pencil streak in representing a geometrical
line” (1940:14). If we think about the example of the line, its form is im-
portant, indeed crucial, to its function. Alter the form, and the meaning is
altered.
Icons are not arbitrary in the way that we generally think of words as
collections of sounds arbitrarily chosen to represent an object. All icons

17. Peirce’s succinct essay “Logic as Semiotic: The Theory of Signs” is a good summary
of his theory (1940:98–119). For a clear introduction to Peirce, see Parmentier 1994:3–44.
See also Sheriff 1989 for a short introduction and some examples of application.
18. See, for example, Asad’s critique (1993:3) of Geertz’s concept of symbol as being in-
adequately defined in Geertz’s influential definition of religion. Asad contrasts Geertz’s
vagueness with the precision of Peircean terminology.
19. The term “symbol” is thus preserved for a specific kind of signifying relationship.
Here Peirce believed he was returning to a definition of “symbol” that was closer to its an-
cient usage based on the Greek root “to throw together.”
 Introduction xxiii

have formal resemblances to the entities they represent. Maps are iconic;
if they were totally arbitrary they would not be much help in finding one’s
way. International signs try to be as iconic as possible so that people from
a variety of cultures can understand them. Hence the iconic representa-
tions of men and women on bathroom doors.
In the world of law, one example of an icon is a personal signature. In
the American legal system it is necessary to establish one’s signature. The
standard way this is done is by signing what is known as a signature guar-
antee—a paper that states that a certain signature is in fact one’s signa-
ture. Once that icon, the signature, is established it is a legally binding
representation of that person.
Deities are also represented in rituals by means of culturally specific
notions of, as it were, signature guarantees. These icon signs are espe-
cially important because they establish divine presences in rituals. Any rit-
ual in which a deity is believed to be present will have some form of
iconic representation of that deity—that is, the deity is not simply referred
to in a ritual, but is physically present in some sign with formal links to
the deity.
Peirce’s second type of sign is an index that is linked to what it stands
for by a “pointing” relationship, such as smoke to fire, or a weather vane
to wind. For this type of representation, a sign must be in spatial-tempo-
ral contiguity with that which it represents. Someone yelling “Come
here!” can be understood only when “here” is made clear by the context.
Establishing “pointing to” relationships is important in rituals, because
it links all the components that are part of the ritual setting—including,
for example, human participants and divinities. The “here” in which a
particular ritual takes place is established by the very ritual itself, where
the setting is created, whether it is a room in which a deity is believed to
live or one of the highest heavens.
The symbol, the third type of sign, has an arbitrary relationship with
that which it represents, and lacks both the formal and spatial relations of
the other signs. American stop signs are arbitrarily chosen to represent the
concept of stopping at a cross street. Given this arbitrary relationship, the
form of the symbol is not motivated by its sign in any way. These signs are
based entirely on social convention and thus vary from culture to culture.
xxiv Introduction

Words, for example, are based on social convention.20 Symbols are likely
to have complex and shifting clusters of meaning that are difficult to un-
derstand in cross-cultural contexts.21
While this terminology may seem cumbersome at first, many of the rit-
uals discussed below are predicated on the distinct semiotic function of
certain signs. It is impossible to understand this function of the divine
name, and to compare it with other possible ways in which words and
signs can function, without the very specific semiotic vocabulary of icon,
index, and symbol. The larger iconic structures of ritual texts also turn
out to be crucial.
Abstracted from their contexts, many of these uses of signs will look
like nonsensical, misdirected actions (hocus-pocus), just as legal texts
sound like gobbledy-gook to the uninitiated. Consider how much more so
this will be the case for religious rituals, where many of the forces in-
volved in the rituals are themselves represented by signs because of their
very nature (invisible gods and so on). Without these icons and indices,
rituals would be empty, because the divine powers are accessible to hu-
mans to the extent that their signs are manifest or point to their presence.
Anytime a deity acts it will be via the semiotic signs (wine, name, angel)
that “stand for” the deity.
This book uses these tools to investigate how rituals were considered
effective modes of social action in the Late Antique period, however they
may appear to us now. What captured the imagination of Late Antique
practitioners was the perceived efficacy of the rituals. Much of the con-
temporary debate about these rituals (real versus literary, magical versus
mystical versus religious) falls by the wayside when we are able to articu-
late the perceived efficacy of the rituals in their historical contexts.
This study describes both how ancient practitioners believed rituals
worked and how modern semiotic theory can illuminate the rituals. Some-
times the connections between the worlds of theory and of practice in the
Late Antique evidence are tenuous or difficult to discern, and often we

20. An interesting exception is onomatopoetic words, which are believed to be formally

linked to what they represent.
21. When a meaning is easily located, it may be because the symbol is in effect “dead.”
See Parmentier 1994:47–69.
 Introduction xxv

lack clear theoretical texts (there are more extant Christian and Greco-
Roman theoretical texts than Jewish). Even in these cases it is still possi-
ble to connect rituals to the historical context and its ideas of efficacy, and
thereby avoid relying on the sweeping generalizations that have domi-
nated contemporary discussions of magic and ritual. How was it that rit-
uals are able to reach the lofty goals outlined at the beginning of this
Introduction? Only an analysis of the pragmatics of Late Antique rituals
will help us answer this question.
 Abbreviations
Ber Marcellin Berthelot and C. Ruelle, Collection des anciens
alchimiste grecs. 1888. Reprint, London: Holland Press, 1963.
PG J. P. Migne, Patrologia graeca. Parisiis: Migne, 1857–66.
PGM K. Preisendanz, ed., Papyri graecae magicae (Greek Magical
Papyri). Stuttgart: Teubner, 1928–31.
PL J. P. Migne, Patrologia latina. Parisiis: Migne, 1844–91.
 1
Late Antique Theories
of Efficacy
Rituals work by faith, truth, and love
—Proclus Commentary on Alcibiades 357.12

P
aradoxically, rituals that claim to reveal divinity on earth can look
to outsiders as if their purpose is to manipulate that same divine
power. The difference is in the eye of the beholder. Similarly, distin-
guishing between the work of an angel and a daimon, between the work
of good forces and evil forces is a subtle, if not impossible, task. (I am us-
ing the word “daimon” and not “demon” to emphasize that in Late An-
tiquity these beings were not always evil.) Even if everyone agrees, for
example, that someone was healed, that does not preclude variant expla-
nations of exactly what happened. Hostile outsiders and internal oppo-
nents are apt to levy charges against practices and rituals that look
suspicious to them, even if they admit that the rituals work.
In late antiquity, the word “magic” (Greek mageºa, Latin magia) was
widely used in intergroup and intragroup polemics. For the Persians, their
word for “priest,” magos, did not have any of the negative connotations
it accrued in Greek culture. Beginning in the sixth century b.c.e., however,
the term was associated in Greek literature with beggars and wizards.1 this
usage became standard despite a few references that more closely fol-
lowed its historical origin, such as in Cicero.
It is not surprising that Christians denounced Jewish rituals as magic2

1. For the development of the Greek and Latin usages, see Graf 1997:2–6 and Janowitz
2001:9–13.
2. Jewish practices that were considered magic include fasting, dietary laws, and Sabbath
observance. For Justin, all Jewish rituals are magic and identical to pagan rites, with their
“magical fumigations and incantations” (Dialogue with Trypho 85.3). John Chrysostom’s de-
nunciations claimed, among other things, that daimons live in synagogues (Homilies Against
the Judaizers 1.6). The Council of Laodicaea prohibited “Judaizing,” which was understood
as the adoption of magical practices, including the use of phylacteries (Canon 35–37).
2 Icons of Power

while Jews denounced Jesus as a magician.3 Representatives of both

groups considered pagan rituals to be magic, and this equation helped de-
termine that many of the rituals currently labeled “magic” are simply pa-
gan rituals.4 No doubt pagans believed that both Christianity and Judaism
were permeated with magic, as the small amount of extant evidence
demonstrates.5 Opponents within a community were often tarred with the
label “magician” in attempts to marginalize them, much as the term
“heretic” was used.6
The stakes in these debates were high; magic was associated with dan-
gerous and barbaric practices, real and imagined, including human sacri-
fice and cannibalism.7 Pliny the Elder matter-of-factly stated that magical
rites involve killing and eating men (Natural History 30.13). In addition
to the negative associations, the practice of magic was a criminal offense
under Roman law (Iulius Paulus Excerpts 5.23.14–18). To be accused of
practicing magic had all the drama of modern charges of high treason and
could have lethal results.
Much of this charge and countercharge can be correctly characterized

3. Many of the Jewish texts were censored, and thus it is difficult to reconstruct references
to Jesus. See the discussion in M. Smith 1978:46–55.
4. For the standard equation of paganism with magic and suspect practices of all types
found in the writings of the Christian apologists, see Thee 1984:328. For Justin, for exam-
ple, becoming a Christian was a turn away from magic (First Apology 14.2). See Pseudo-
Clementine Homilies 5.4–7; Clement Exhortation 1 and The Teacher 3.4. Simon
1986:361–62.
5. For Christian rituals that looked magical to pagans, see, for example, Suetonius
Nero 16.2 and the comments of Celsus preserved by Origen in Against Celsus. For the
common pagan equation of Judaism and magic, see, for example, Juvenal’s denunciation
of Jewish oracles (Satire 6.546), Lucian of Samosata’s reference to Jewish incantations
(The Gout Tragedy, 171–73), Celsus’ statement that Jews are addicted to magic (Against
Celsus 1.26), Posidonius’ claim that Jews were sorcerers (Strabo Geography 16.2.43), and
Pliny’s statement that magic comes from the Jews (Natural History 30.11). In a modern
continuation of this stereotype in what is otherwise a very sophisticated discussion, Simon
appears to accept the idea that in the Late Antique world Jews were magicians and in-
fected early Christianity with magic. He writes, “Though it is true that Judaism, by means
of its theodicy and its morality, was able to seduce the noblest of the pagans, it also exer-
cised a more murky influence on the masses by its reputed ability to ward off the Powers”
(1986:341).
6. See the accusations of magic made against Christian rivals, such as Marcus, Simon, and
Menander, in the writings of Irenaeus (see Against the Heresies 1.13.1).
7. The connection between “magoi” and human sacrifices goes back to Herodotus His-
tories 7.114.
 Late Antique Theories of Efficacy 3

as “your ritual is magic, mine is religion” (Neusner 1989). But that does
not exhaust Late Antique period debates about magic. A full discussion of
the use of the term “magic” in late antiquity is beyond the scope of this
chapter.8 The most important point that needs to be made here has to do
with efficacy: magic was believed to be effective. No doubt there were
some individuals who rejected any supernatural interventions in daily life,
but such people were rare. For most Late Antique writers the term usually
implied potentially effective but nonetheless inappropriate uses of divine
powers. Either the wrong powers were put to work (evil demons) or the
wrong people were engaged in using the powers (outsiders, women, and
the like).
The other half of this highly polemical debate, which will occupy this
chapter, was the attempt by some Greco-Roman thinkers both to defend
specific rituals and explain the use of rituals in general. While these issues
were of great importance to diverse Late Antique thinkers, the discussions
among the Neoplatonists are particularly lucid and well developed. These
thinkers wrote commentaries on Plato’s dialogues and other writings dur-
ing the second through fifth centuries c.e. The explanation of ritual that
interests us specifically is the development of theories of “theurgy,” a term
that is parallel to theology (words about divinity) but that has the empha-
sis on effective action.
E. R. Dodds developed the standard conception, still widely cited, for
the meaning of the term “theurgy” (1947). Dodds used two closely con-
nected approaches to define it, first defining it as religious ends reached
through magical means and then delineating it as a specific set of rituals
(animation of statues, divination by means of mediums). In the last few
decades the inadequacy of Dodds’ schema of theurgy, which was based on
a rigid evolutionary model of religion and magic, has been repeatedly em-
phasized. Dodds did not invent this evolutionary schema; his contribution
was to characterize Plotinus as a philosopher representing the top of the
hierarchy, and theurgy, sharing methods with magic, as the lowest form of
religion. Dodds had a decidedly antiritual stance in general, setting Ploti-
nus sharply against an irrational environment enmeshed in ridiculous rit-

8. See Janowitz 2001:9–26.

4 Icons of Power

uals.9 Dodds’s definition has unraveled as scholars realized that the list
was too short and began adding additional elements. Georg Luck’s recon-
struction of the theurgic ritual now includes periods of silence, references
to a “warming fire,” manipulation of material objects and writing down
of things, use of tools, and drugs or mind-altering substances (Luck
1989:192–93).10 It is to Luck’s credit that he has moved theurgy from the
fringe of pagan practice to its center, thereby making imperative a recon-
sideration of its relationship to traditional paganism. But his definition is
so broad that it delineates no clear set of rituals.
As the review below demonstrates, the rituals associated with theurgy
by ancient writers were for the most part quite old.11 That is, the term re-
ferred not so much to a new set of “magical” techniques but rather to a
hodgepodge of some older rituals (for instance, animating statues, prayer,
and animal sacrifice) and some newer rituals (for instance, ascents
through the heavens). What was new was that these rites were granted the
philosophical imprimatur of “theurgy.” Our argument is the opposite of
Luck’s recent statement, “The need of pagan believers to enter into direct
contact with their gods led to the development of a certain technique or a
set of techniques codified during the reign of Marcus Aurelius, it seems,
and given the name “theurgy’” (1989:185). On the contrary, the main im-
port of calling a rite “theurgy” is the new interpretations of ritual efficacy
it articulated. These new interpretations explain how the specific tech-
niques, and rituals in general, work.
The term “theurgy” was not only aimed at skeptical outsiders, nor is
that necessarily its origin. In addition to deflecting charges of “magic,”
discussions of theurgy explain rituals in light of shifting attitudes toward
specific rituals and the resulting need to constantly update them. Explain-
ing theurgy leads us deep into issues of ritual efficacy and helps paint a
background against which to consider the rituals discussed in the chap-
ters below.

9. Gregory Shaw has the most insightful critique of the antiritual bias of much scholarly
work (1985:9). Athanassiadi’s recent discussion is corrective of Dodds but is only for
Iamblichus. The issue of a general definition remains.
10. As the category “magic” shrinks, the category “theurgy” blossoms.
11. Dodds acknowledges this in his highly influential article, though he still presents
theurgy as a new type of quasi-magical expression (1947:63).
 Late Antique Theories of Efficacy 5

Late Antique Concepts of Theurgy

As reported by his friend Iamblichus (250–325 c.e.), the Neoplatonic

thinker Porphyry (232–305 c.e.) raised questions about the philosophical
basis of ritual. “How is it that rituals work?” Porphyry pondered. He had
been trained by Plotinus, who believed that many components of tradi-
tional religion were not theoretically grounded. Porphyry was daring to
ask why one would bother to pray, or do any other ritual.
In response, Iamblichus argued that it was impossible to proceed to
the goal of unity with divinity without ritual actions.12 He chided Por-
phyry for trying to solve every problem with dialectics (On the Mysteries
10.1–7; 285.10–293.13). Those who only talk and employ dialectics will
never achieve the contact with divinity they all sought. “It is not thought
that links the theurgist to gods; else what should hinder the theoretical
philosopher from enjoying theurgic union with them? This case is not so.
Theurgic union is attained only by the perfective operation of ineffable
acts, correctly performed acts which are beyond all understanding, and
by the power of the unutterable symbols which are intelligible only to the
gods” (2.11; 96.13–97.2). It would be difficult to improve on this state-
ment as a contemporary explanation of ritual efficacy. Rituals,
Iamblichus explains, do not always make sense to humans, even to insid-
ers, but they do make sense to the gods. The gods send down instructions
to complete certain actions that look strange to humans but that “per-
fect” humans. These acts provide what the dialecticians lack in their in-
vestigation of divinity.
The comment of Iamblichus that some aspects of ritual may be incom-
prehensible on the surface reappears in early uses of the term “theurgy.”
In what may be the earliest extant use of the word, Nicomachus of Gerasa
(fl. c. 50–150 c.e.), who wrote a number of technical treatises, including
“Introduction to Arithmetic,” “Introduction to Geometry,” and a manual
on harmony and was a seminal figure in philosophical circles, uses the
term “theurgists” in an explanation of why seemingly nonsensical vowel

12. The most important recent work on Iamblichus and theurgy is Shaw 1995. In partic-
ular, Shaw demonstrates how Iamblichus’ theory of the descent of the soul meshed with his
ideas about theurgy.
6 Icons of Power

sounds are uttered in rituals.13 Vowels are imitations of planet sounds, he

explains, and thus are effective in rituals. He sets up an analogy between
soul : body and vowel : consonant, arguing that unpronounceable vowels
must be “clothed” in consonants in order to be used.14 Once given mate-
riality, they can be employed as powerful forces in ritual.15 Nicomachus is
arguing that vowels are a special kind of sign. This is only one instance of
complex Neoplatonic philosophical discussions that insist that certain as-
pects of language are existentially connected with cosmic reality.16 In
Nicomachus’ terms the vowels function “symbolically.” In the Peircian
terms of this study, the vowels function iconically, because each sound is
a formal equivalent to the sound of a planet. Uttering the sounds brings
the power of that force directly into the ritual.
The specific practices of these “theurgists” are not elaborated here.
Whatever these rituals entailed, Iamblichus claimed they were “perfective
of divine things.” Surely for Nicomachus, as for Iamblichus, these are the
most important rituals, though they are also likely to be misunderstood.
Other early uses of the term “theurgy” are associated with a father-
and-son team of Julians who lived during the reign of Marcus Aurelius
(161–68 c.e.).17 The elder Julian is called a philosopher in the Suda entry
under his name (Suda 1.433, s.v. Julianus).18 The younger is remembered
as the author of esoteric tractates including Theurgika/Ueoyrgika (1.434).
Anecdotes about the father and son depict them as champions of flashy
rituals. “They say about him that once, when the Romans were exhausted
by thirst, he suddenly caused dark clouds to gather and discharge torren-
tial rain with incessant thunder and lightning; and that Julian worked this
by some wisdom” (1.433). These actions were apt to be considered magic
in the eyes of their opponents. Our picture of their practices is dependent

13. Dodds 1961:271 n. 31 rejects the term as a later addition, but his argument is not
convincing.
14. Nicomachus on sounds is also discussed below in Chapter 4.
15. For further discussion of sounds in rituals, see Chapter 5.
16. See Stephen Gersh 1978 and the comments in Chapter 4 on Neoplatonic theories
about letters.
17. For recent reviews of the dating problems and the evidence about the father and son,
see E. des Places (1984) and G. Fowden (1987). H. D. Saffrey (1981) is skeptical of much of
the biographical data, contra H. Lewy (1978).
18. The Suda is a historical encyclopedia derived from more ancient scholarship, com-
piled at the end of the tenth century.
 Late Antique Theories of Efficacy 7

on these writers, many of whom had no great sympathy for the Julians.
For these writers, “theurgy” is indistinguishable from “magic.”
The rituals attributed to the father-son team were not new or unique to
the Julians.19 They were embraced by their followers not so much for their
novelty as for their presumed success with manifesting divinity on earth.
The elder Julian was said to be able to make Plato’s soul speak through a
human medium, “conjoining” his son “with the gods and with Plato’s
soul.”20 H. D. Saffrey, who rejects most of the biographical data, accepts
the idea that the father used his son as a medium to receive oracles from
Plato (1981:218–19). The father could bring about an appearance of the
god Chronos (Proclus Commentary on the Timaeus 3.20.22), separate his
body from his soul (Proclus Commentary on the Republic 2.123.9–1),
and heal.21 They were renowned for a rain miracle, which in some versions
included use of a human mask that cast thunderbolts at the enemy.22 Fow-
den’s careful analysis of the thunderbolt episode demonstrates that the
story may originally have been attributed to another famous pagan figure
and only later attributed to Julian by his followers (Fowden 1987).
The text most closely associated with the Julians is the patchy collection
of enigmatic verses known as the Chaldean Oracles.23 This text, recon-
structed first by Kroll and later by des Places, is still the source of much con-
troversy. Johnston concludes, “Despite the murkiness that surrounds the
Juliani, the Oracles emerged during the mid to late second century (1990:3).
The extant fragments from the Chaldean Oracles do not include the
word “theurgy,” but one fragment explains, “Theurgists do not fall into
the herd which is subject to destiny” (frag. 153). Engaging in theurgy
thus elevates one above the common people, and even cosmic forces.
The fragments urge the reader to engage in various rituals. One fragment
calls specifically for a combination of “action (‘rgon) with sacred word

19. See note 11.

20. Psellus The Golden Chain, ed. Sathas 1875:217; French translation in Leveque
1959:79. Julian also was said to have prayed that his son have the soul of an archangel.
21. In a medieval Christian legend, Julian competes with two other healers, Apollonius and
Apuleius, to end a plague in Rome (Anastasius of Sinai Quaestiones [PG 89, col. 252ab]).
22. The mask version is preserved in Psellus Minor Writings 1.446.8–447.14 K-D; see des
Places 1971:221–22.
23. See Lewy 1978 and the additions by des Places 1983. For a recent critical study, see
Majercik 1989.
8 Icons of Power

(logoq)” (110).24 An inventory of the specific techniques mentioned in the

Oracles nets a rich list. We find mention of an invocation or calling on the
gods (7), which may or may not be related to the standard technique of
reciting divine names (150).25 The writings also describe the consecration
of a statue of Hecate (224), use of a ritual wheel called a strøfaloq
(206, 208),26 breathing techniques (123, 124), periods of silence (132),
and sprinkling of water (133). Fragment 135 mentions a process of initi-
ation. The status of animal sacrifice in the Oracles is debated; the only
“sacrifice” directly discussed is that of a stone (149).27 In a suspect frag-
ment, Hecate appears in response to prayer (222).28
It seems unlikely that these references to ritual actions were all part of
one specific “theurgic” ritual;29 instead, they were probably part of dis-
tinct rituals, many of which, again, were traditional Greco-Roman reli-
gious practices. The officiant is referred to as “‡ereÂq/priest” (frag. 133)
and the participant as “m¥sthq/initiate” (132), both traditional terms.30
The traditional practices likely were supplemented by some new rituals
that reflected Late Antique cosmological concerns, primarily rituals of as-
cent (“nodoq) of the soul to the heavenly realm.31
The author of the Oracles distinguished his favorite rituals from other
rituals he had rejected. The latter rituals included the use of birds and en-

24. For a discussion of shifting combination of words and actions in rituals, see the Con-
cluding Note.
25. Several other fragments appear to be about divine names, including fragment 87,
which describes a holy name leaping into the world, and fragments 76 and 77.
26. Majercik translates the term as “magical wheel,” but the word “magic” does not oc-
cur in the text.
27. This appears to be some form of reinterpreted sacrifice. Lewy (1978:289) cites Psel-
lus’ statement that the “Chaldeans” offered up animals (Minor Writings 1.446.23). See also
Lewy’s Excursus 8 (1978:487–89).
28. For epiphanies of Hecate, see Johnston 1990:111–33.
29. Lewy reconstructed them as one ritual he considered a form of magic, as noted in the
title of his chapter “The Magical Ritual of the Chaldeans” (1978:227–57). The nine ele-
ments of the ritual he outlined include, among others, “conjunction” with the gods, suppli-
cation, conjuring Hecate, consecrating a statue to Hecate, using a noise-making top, and
divination.
30. Johnston outlines parallels between the language of the Oracles and terms from the
mystery religions, though no doubt these terms appeared in a number of religious settings
(1997:176–77).
31. Argued by Lewy 1978 and subsequently by Johnston 1997. She places an ascent rit-
ual at both the core of the Oracles and thus her notion of “theurgy.” For a broader discus-
sion of ascent, see Chapter 5.
 Late Antique Theories of Efficacy 9

trails for divination (frag. 107), which are called “supports of a deceptive
trade (®mporic∂q Ωpåthq sthrºgmata).” The brief reference to “decep-
tive” ritual is a tantalizing hint about modes of categorizing rituals. “De-
ceptive” may mean that the rituals appear to work but are at some level
fraudulent.
The Oracles represent not so much a new set of techniques as a
grounding for a set of older techniques. Though abstruse, they present a
theological system for the rituals that is replete with Platonic imagery (see
Johnston 1997:169). Just as theology designates the very highest level of
discourse about the gods, in the Oracles “theurgists” engage in a group of
highly valued, traditional rituals with divine backing.
The first explicit definition of theurgy occurs in the writings of Por-
phyry, whom we encountered briefly above worrying about the theoreti-
cal basis of ritual. Porphyry could not settle on a philosophy that
sufficiently explained the workings of rituals. This was true even for ritu-
als aimed at achieving their most cherished goal of divine contact. In an
excerpt preserved in Augustine, Porphyry states that by means of certain
“theurgic consecrations (teletai), the spiritual element of the soul is put
into a proper condition, capable of welcoming spirits and angels, and of
seeing the gods” (Augustine City of God 10.9). “Consecration” or “initi-
ation” (telev) was a traditional term for religious rituals.32 The impor-
tance of theurgic consecrations is that they can attain a very lofty goal,
which in Porphyry’s case was to condition the soul to see the gods.
Porphyry’s concept of ritual efficacy is carefully articulated and re-
stricted. As with the Oracles, not all rituals are viewed as equal. Augus-
tine preserves some of Porphyry’s cautions and hesitations: “In fact he
[Porphyry] says that the rational soul (or, as he prefers, the intellectual
soul) can escape into its own sphere, even without any purification of the
spiritual element by means of ‘theurgic art,’ and further, that the purifica-
tion of the spiritual part by theurgy does not go so far as to assure its at-
tainment of immortality and eternity” (City of God 10.9).
Ritual activity is not necessary for Porphyry’s highest goals, nor is the
success of even the best rituals ensured. Porphyry’s theory does not mean

32. The term is connected, for example, with the mysteries of Demeter (Herodotus His-
tories 2.171).
10 Icons of Power

that theurgic consecration can reach only modified goals while some other
type of ritual can reach higher ones. It is a statement about the limitations
of ritual in general. And theurgic acts of purification are not unique in
their ability to help the soul. Unfortunately, however, we cannot tell from
Porphyry’s brief statement preserved in Augustine what else helps the soul
escape to its sphere.
The highest goal, liberation of the soul, could be sought by means of
rituals to the extent that there is a theoretical explanation of the efficacy
of these rituals. The very best rituals, theurgy, can be of some help to in-
dividuals, but at the same time rituals with no conceptual basis are mere
magic.33 The potential of even “theurgy” was open to doubt. Until the ef-
ficacy of rituals is explained, all rituals look like magic. And for Porphyry
it is folly to try to put magical constraints on the gods (Eusebius Prepara-
tion for the Gospels 5.10.199).
Iamblichus, who tried to answer Porphyry’s doubts, did not necessarily
plan to develop a philosophical defense of ritual, but he ended up devel-
oping one anyway.34 Iamblichus’ real concern, he tried to explain to his di-
alogue partner, was to illuminate the relationship of the human soul to the
cosmos. Porphyry believed that the soul shunned any actions that look to
the world and try instead to remain always in the upper cosmic realm. The
soul could save itself without having to “do” anything, without having to
engage in any type of ritual action. In Iamblichus’ eyes, this was a frontal
attack on traditional Platonic thought and threatened to lead people away
from the gods.
For Iamblichus, moving toward the gods was the purpose of traditional
practices, including even animal sacrifice (On the Mysteries 5.11;
215.3–7). He is particularly clear that these rituals do not work by com-
pelling the gods.35 The person performing the rituals must have the correct
“inner disposition,”36 a necessary condition, but not a sufficient one, for it
does not legitimate or guarantee every type of ritual. He separated effec-

33. Johnston writes that Porphyry “also apparently recognized a level of theurgy con-
cerned not at all with spiritual salvation but with more worldly, immediate goals” (1990:79).
Perhaps it is more clear to state that according to Porphyry rituals can be directed toward a
variety of goals, including some that may look surprisingly mundane to us.
34. As has been so elegantly argued by Gregory Shaw (1995).
35. See On the Mysteries 2.11; 96.13—97.9.
36. See A. Smith’s discussion (1974:99), commented on by Shaw (1985:25).
 Late Antique Theories of Efficacy 11

tive ritual (theurgy) from false ritual (magic) based on an inner attitude to-
ward the gods or spiritual disposition. Iamblichus’ defense of this position
foreshadows a mode of classifying true versus false ritual still used in reli-
gious polemics today.
Iamblichus revealed his philosophical snobbery in his rejection of
“sympathy.” According to this Late Antique cliché, divine processes exist
on earth in hidden forms despite the apparent gulf between the deity and
the material world. Earthly “symbols” can be manipulated so that they re-
veal divinity. Sympathy does exist, he admits, but it does not explain
everything (On the Mysteries 5.7–9; 207.7–210.14). According to
Iamblichus, divination by sympathy is “entirely different” from real
theurgy (6.4; 244.13–245.10).37
Rituals Iamblichus rejects are labeled, as expected, “magic,” for these
cannot by definition have any spiritual or pious aspect (On the Mysteries
3.26; 161.10–16). In particular, he denounced the animation of statues as
fraudulent and called makers of images “magicians” (3.25; 160.15).38 The
study of animal entrails and observation of the stars are mere human sci-
ences (3.15; 135–136.10), while theurgy is not.
Proclus (b. 410/12) continued Iamblichus’ positive evaluation of some
rituals and attempted to explain the theoretical value of the rituals. His
definition of “theurgy” is noteworthy in its total lack of reference to spe-
cific rituals. Theurgy is “a power higher than all human wisdom embrac-
ing the blessings of divination, the purifying powers of initiation and in a
word all the operations of divine possession” (Platonic Theology
1.26.63).39 Human wisdom cannot reach as high as divine power. Hu-
mans who enact rituals copy the gods, because gods “illuminate by ac-
tions and not by words” (4.26.220).
Proclus’ promotion of ritual, expressed here, is supplemented and ex-

37. See Shaw 1985:204–5 and 233.

38. Dodds points out a secondhand report that Iamblichus believed statues contained di-
vine presences (1947:64). Porphyry wrote a treatise titled About Statues that included sym-
bolic interpretations of the statues of deities, the objects they hold, their stances, and so on
(Bidez 1964: appendix 1). It is difficult to discern from the fragments whether he claimed
that divine powers are present in the statues. Eunapius recounts how Maximus made a
statue of a goddess smile, and the varied views of this type of activity (Lives of the Philoso-
phers 475). Animation of statues is also associated with the Chaldeans by Psellus Letter 187
(see Lewy 1978:496).
39. See also Platonic Theology 1.25.113.
12 Icons of Power

panded on in his various treatises. His treatise Elements of Theology,

which supplies, in the words of Dodds, “the philosophical basis for
theurgy” (1963:276)40 is an abstract text and does not refer to specific rit-
uals at all.41 That is, the propositions explain the theoretical principles on
which all successful rituals rest and do not support any particular ritual.
Proposition 145, for example, states: “The distinctive character of any di-
vine order travels through all the derivative existents and bestows itself
upon all the inferior kinds” (1963:129). Even the most “inferior” level of
existence has a distinctive character that connects it with the highest level,
meaning that it is possible to work with the inferior level and affect the
higher levels.
Rituals attributed to Proclus by his biographer, many of them familiar
by now, include weather manipulation, evocation of a vision of Hecate,
and participation in a ritual of divine communion called “s¥stasiq”
(conjunction).42 Statues can give revelations; gods make them speak the
future (On the Timaeus 3.155.18–25).43 A meeting of soul and divinity
takes place in prayer, which no doubt explains Proclus’ own practice of
praying three times a day (Marinus Life of Proclus 22). And the liturgy
probably included recitation of divine names, because “it is these names
which theurgy expresses in imitating them by exclamations, sometimes
distinct, sometimes indistinct” (Commentary on the Cratylus 71, p.
31).44 Theurgists are taught the names of gods by the divine powers.45
Ascent was also among Proclus’ ritual practices; at the age of forty-two,
he proclaimed that his soul had arrived in the heavens, which resound
with immortality (Marinus Life of Proclus 28). Rituals he rejected are la-
beled “magic.”46 It is clear that for Proclus theurgic rituals are not com-

40. Shaw (1985:11) points to the importance of proposition 57 for understanding the
philosophical basis for ritual.
41. As Dodds noted (1963:xvii).
42. Marinus Life of Proclus 28 (Oracles, frag. 28). See H. Lewy 1978:228–38.
43. See note 38 above.
44. Explaining this quotation, Dillon writes: “Theurgy teaches us how to represent the
structure of the symbola in the physical world by means of inarticulate utterances”
(1985:29). All nonsense is not theurgy, however, and not all theurgy uses nonsense.
45. Commentary on the Timaeus 1.274.16; see also On the Cratylus 72.8 and Platonic
Theology 1.29.124.
46. Commentary on the Republic 1.255.19 and 1.29.14.
 Late Antique Theories of Efficacy 13

pulsive or manipulative magic, but represent the best of traditional ritual

practices.47
Some Christians adopted similar usage and employed the term
“theurgy” to refer to divine action or the most effective and powerful rit-
uals. In an early liturgy, God is said to display “theurgic power (tÎ
ueoyrgikÎ soy dynåmei)” in miracles.48 Theurgic power is the opposite
of “magic” because it is divine. Dionysius the Areopagite, a fifth-century
Christian writer we will meet again, was particularly enamored of the
term; “theurgy” appears forty-seven times in his corpus and follows the
Neoplatonic usage very closely.49 The most significant ritual for him was,
of course, the Eucharist, and this he calls theurgy. “Theurgics,” he ex-
plained, are the consummation of theology (Ecclesiastical Hierarchy
3.5.432B), which is necessary because it is impossible to rise up toward
the deity without material symbols (Celestial Hierarchy 1.3.121CD).
For other Christians, attacking “theurgy” was a way of attacking the
heart of “pagan” beliefs and practice. For example, Augustine’s strategy is
very revealing. He attacked Porphyry’s distinction between effective ritu-
als (theurgy) and fraudulent rituals (magic). Brushing aside any distinc-
tion between these two categories, Augustine makes his own strategic
hierarchy in which all pagan practices ranked at the bottom. For Augus-
tine, theurgy is the invention of lying demons (City of God 10.10).
Porphyry’s rituals were no better than magic, Augustine argued, be-
cause they were vulnerable to manipulation for evil ends. As proof, Au-
gustine referred to Porphyry’s story about a Chaldean who complained
that his efforts to purify souls had been thwarted by someone with greater
supernatural power. “For if this business was concerned with ‘good gods,’
the good man who undertook to purify a soul would undoubtedly have
prevailed over his malevolent opponent” (City of God 10.10).

47. One of the few scholars to have addressed this issue, Festugière, recognizes both Pro-
clus’ commitment to traditional gods and his “genuine” piety (1971:1581–89). Festugière
tries, with only moderate success, to distinguish between theurgic operations (®nerg¸mata)
and traditional divine manifestations (®ntyx¸mata). Late Antique writers do not abide by
these distinctions. Proclus connects these two types of practices, just as he combines Fes-
tugiére’s “reflective piety” and “popular piety.”
48. The Liturgy of Mark (Brightman 1965:142), English translation in Neale 1976.
49. For an English translation of the corpus and a recent bibliography, see Luibheid and
Rorem 1987.
14 Icons of Power

In effect, Augustine argues here that good supernatural powers should

always win out over bad ones—that is, if one is good the supernatural
forces of one’s enemies will be less effective than one’s own forces. This
was a slippery position for Augustine to take, because it depends on the
definitive triumph of “good”; Christian acts could be disqualified by the
same reasoning. The strategy was useful, however, in his attempt to de-
molish Porphyry’s system of classification. By introducing the invincibility
of good powers, Augustine undermined the distinctions between theurgic
and fraudulent rituals made by earlier writers. Every rite, or at least every
pagan rite, is now simply “magic.” The fantastic success of this rhetorical
stance must be noted; it continues to function today in the many discus-
sions of theurgy as a form of magic. For the Neoplatonic practitioners,
this would have been nothing less than malicious slander.
Another set of debates about ritual shades into the magic versus
theurgy debates, although it is cast in slightly different terms. For want of
a better phrase, I call this dimension of ritual the aesthetics of ritual. Aes-
thetics is concerned with notions of beauty, and, I argue, rituals were eval-
uated in this manner. That is, in Late Antiquity different modes of
interacting with supernatural figures were not all considered equally ele-
gant and pleasing to the deity/deities. The criteria used for evaluation
were again shared across what we think of as the boundaries of traditions.
One common Late Antique mapping of ritual aesthetics points toward a
low evaluation of or skepticism about the use of objects in ritual. For ex-
ample, Origen argued that Moses used both a divine rod and a verbal for-
mula (logos) so that he would not be suspected of magic, which would
involve only a rod (Selection on Exodus [PG 12.284B]).
The use of props seems to have been taken as evidence that the ritual
employed lower-level supernatural powers such as daimons. A rabbinic
anecdote, for example, uses the two appearances of the biblical term for
“magic” spelled slightly differently in arguing for the existence of two
kinds of magic, one that uses props and one that does not (bSanhedrin
67b). The person who uses props is using daimons, and hence doing an in-
ferior type of practice.
If props are evidence of baser rituals, language-only rituals were placed
higher on the scale. Justin, for example, argued that the use of the name
 Late Antique Theories of Efficacy 15

of God is better than the “incantations and incense of pagan and Jewish
exorcists” (Dialogue with Trypho 85.3). John Chrysostom argued that
Christians should rely on words, and words alone, in their rites. He im-
plicitly acknowledges a hierarchy and tries to place Christian practice at
the top. The only “praxis” employed by good Christians is making the
sign of the cross (Homilies on Collossians 8.5 [Col 3:5).50 Hilarion brags
that the simple recitation of a name is enough, all by itself, for a healing
(Weltin 1960:87). Similarly, the recitation of words alone was understood
to be sufficient to transform the Eucharist (Weltin 1960: 83, 89).
Thus the very best rituals in the eyes of some Jews, some Christians, and
some Greco-Roman practitioners would be those directed to the highest
powers and employing the fewest objects. Just beyond the top of the ritual
scale would be the divine person characterized by his ability to do miracles
based on his indwelling power alone (M. Smith 1978:74). The claim was
made that Apollonius of Tyana did not need sacrifices, prayers, or even
words to perform miracles (Philostratus Life of Apollonius 7.38).51
The stance of Plotinus represents perhaps the purest form of this aes-
thetic snobbery, rejecting any notion of external rite at all. Some rituals
were blatantly ridiculous to Plotinus, such as those that included
“melodies, shrieks, whisperings and hissing with the voice” (Enneads
2.9.14).52 Beyond that, he disdained prayers altogether, resulting only in
“poi¸seiq/fabrications” (4.4.26).53 Although he considered prayer dé-
classé, Plotinus did not deny its efficacy. The gods do answer prayers, and
he can explain why: sympathy (4.4.26). Putting the cosmic forces of sym-
pathy into play is no better than using daimons or nonsense sounds. All of
this is inferior to the use of mind to commune with the deity.

50. This sign should be employed, for example, when going near a synagogue to ward off
the daimons who live there (Against the Judaizers 8.8; 940).
51. For supernatural power without spells, see Theodoret [Pseudo-Justin] Answer to the
Orthodox 24 and Arnobius Against the Nations 1.43–44.
52. Rist notes that Plotinus equated prayer with magic (1967:27). In a partisan statement,
Dodds claimed that Plotinus’ notion of mystical union “is attained, not by any ritual of evoca-
tion or performance of prescribed acts, but by an inward discipline of the mind which involves
no compulsive element and has nothing whatever to do with magic” (1947:58). This claim re-
peats Plotinus’ own claim to spiritual superiority, as does Majercik’s characterization of Ploti-
nus’ goal as “a genuine contemplative experience free of external manipulation” (1989:39).
53. Iamblichus uses the same term in his criticisms of Egyptian sacrificial traditions (On
the Mysteries 5.8; 208.15).
16 Icons of Power

A similar notion may be behind the position of some early Christians

who, Irenaeus reported, were against the use of water in rituals. “[T]he
mystery of the ineffable and invisible power ought not to be performed by
means of visible and corruptible created things, nor the inconceivable and
incorporeal by means of what is perceptible and corporeal.”54 It is not sur-
prising that their preferred means of redemption was using the mind to ac-
quire special knowledge, an entirely “spiritual” endeavor.55
The reason for the disdain Plotinus had for much standard ritual be-
comes apparent in the infamous temple scene reported by his biographer.
Here Plotinus permits his power to be revealed and it turns out that his in-
dwelling supernatural power is a god and not a lowly demon. He does not
need to implore other divine powers; he himself is one. Plotinus’ rejection
of prayer stems from his notion of the divine soul—to whom could he pray
to except himself, and why pray to himself? The same attitude is displayed
in his reply to a request to go to a religious ceremony: “It is for the gods to
come to me, not for me to go to them” (Porphyry Life of Plotinus 10).
For Plotinus, “contemplation (uevr®Áa) alone stands untouched by
magic” (Enneads 4.4.44).56 This appears persuasive to some modern writ-
ers, who see him as a holdout against irrational magic and ritual.57 His
goal, however, the quest for being “god-like” (1.6.9), is not rational, and
neither are his methods. Plotinus may appear “rational” because he ad-
mits ignorance rather than fill in the gaps in his theoretical system by
telling myths (one of the sins he attributes to the “gnostics”). He also tries
to argue from principles, but that does not make him rational if the prin-
ciples he argues from are irrational. Even his “philosophical” approach
involves techniques. In the case of Plotinus, we find contemplation based
on mental imaging of the sphere that is the world soul.58

54. Irenaeus Against the Heresies 1.21.4. See Epiphanius’ account of the “Archontics,”
Medicine Chest 40.2.
55. Irenaeus denounces these individuals as “gnostics,” one of the great polemical terms
from Late Antiquity translated best in this context perhaps as “know-it-alls.”
56. On the mental process of contemplation, see also Enneads 1.6. Shaw points out that
scholars have long identified Plotinus’ stance with Neoplatonism in general, even though he
appears to be in the minority on this point (1985:3).
57. This is especially true of the classic work of E. R. Dodds.
58. Enneads 1.6.9 suggests “Withdraw inside yourself,” followed by the analogy of mak-
ing a statue. See Rappe 1995 for a discussion of Plotinus’ mental imaging techniques.
 Late Antique Theories of Efficacy 17

When Olympius, a jealous opponent of Plotinus, sends “star spells”

against him, Plotinus is able to reflect these rays back on the opponent.59
He is successful because of his divine status, not because of some pre-
sumed rational stance of his. To see superiority in Plotinus’ ability to ward
off the star spells without having to actually “do” something is to adopt
the aesthetic stance of Plotinus.60
Most Late Antique thinkers had a higher opinion of rituals. Iamblichus
is a good example, because he opposed cutting oneself off from the mate-
rial world. He believed that the lower and upper worlds were intimately
connected and that the material world was not entirely evil. So too for Sal-
lustius: “Prayers divorced from sacrifices are only words, prayers with
sacrifices are animated words (On the Gods 16),” and for Albinus both
contemplation (uevr®Áa) and action (pr˙jiq) lead to divine likeness (In-
troduction 3). In all these cases the positive evaluation of ritual action
may have seemed misguided to the “gnostics,” who favored internalized
rituals, and to Plotinus, who polemicized against any form of prayer.
However, these rituals had the authority of being extremely old. They
were thought to be connected with being connected with ancient sacrifi-
cial traditions and revelations from the gods.
The terms “magic” and “theurgy” are similar in that they are best un-
derstood as ways of evaluating and characterizing rituals, and not just
simple references to a specific set of rituals. Anything and everything
could be “magic” because the world was full of unexplained events that
were threatening to someone. Seen through the lens of “theurgy,” both
old and new rituals spoke to a central religious concern of the Late An-
tique period: overcoming the gulf between humans and gods.61 Why then
coin a new term?62 In the context of self-conscious reflection on the func-
tion of rituals about ritual that appear to have flourished in the second

59. Porphyry Life of Plotinus 10. For the sage’s ability to ward off spells, see Enneads
4.4.43.
60. When Andrew Smith argues that the differences between Plotinus and Iamblichus
were more semantic than substantive (1974:9), he is struggling with the same elusive issues
that I call aesthetics. See the comment by Shaw (1985:7).
61. See the comments by Hirschle 1979:12ff.
62. Bidez believed the term was invented to denigrate theology and theologians, who can
talk only about the gods (1965:369 n. 8), but we have no evidence that those who champi-
oned “theurgy” spurned theology.
18 Icons of Power

century and later, a word was needed to parallel “theology.” Just as words
about the divinity are called “theology,” so too now actions that involve
divinity can explicitly be referred to as “theurgy.” Rituals, including tra-
ditional ones, now have a label that carried with it an implied efficacy.
Whatever the original impetus for the term “theurgy,” its usage varied
widely. Most important, “theurgy” was used in highly strategic ways.
Modern scholars should not flatten these rich usages with oversimplified
characterizations of theurgy as referring to a specific, often extremely nar-
row set of rites. Nor should theurgic rites been seen as any more suspect
than other rituals employed in these centuries.
We have found a spectrum of modes for constructing rituals: the spec-
trum runs from “words only” rituals that consist primarily of the recita-
tion of verbal formulas, to rituals that combine the recitation of formulas
with the use of some types of objects, and finally to rituals that do not in-
clude any verbal formulas at all. We begin analyzing this spectrum by
looking at the most powerful words: at divine names.
 2
The Divine Name as
Effective Language
In every place where I cause my name to be remembered,
I will come to you and bless you.
—Exodus 20:24

What Is a Divine Name?

O
ur investigation into effective language begins with one small
but very special area of biblical interpretation: interpretation of
God’s name, and especially as it intersects with the notion of a
special mode of divine speech. A prime locus of discussion about divine
speech is exegesis of the creation story in Genesis, a story that implied to
ancient readers that speech could create literal reality.1
The dramatic manner in which the deity’s speaking is an effective and
ordering force in the universe was perceived by later interpreters to be a
central message of the Genesis text. In the process of discussing the cre-
ation story, the exegetes tried to capture the creativity of the deity’s words
and employ these pragmatics for their own ends. The particular twist that
concerns us most is the emerging idea that the word(s) God spoke when
creating the world was his name. Unpacking the pragmatic implications
of the divine name lays bare for us the pragmatics of ritual language spo-
ken by humans. The efficacy of words spoken by the deity will lead us to
the efficacy of words spoken by humans, which in turn derive power from
their divine source.
In the Hebrew Scriptures the deity has a distinct name presented as the
four letters “YHWH.” Since the text appears without vocalization, the

1. Numerous writers have discussed this theme. For one recent discussion that intersects
with this chapter, see Tambiah 1985:17–59.
20 Icons of Power

exact ancient pronunciation of the letters is not known.2 This name dis-
tinguished him from other deities in the earliest biblical texts (e.g., Baal).
Some biblical texts imply a secret, perhaps powerful, name; just as not
everyone can see the deity, not everyone can know his name (Ex 3:14–15).
With an increasingly monotheistic vision, a name for God became redun-
dant; because he was the only deity, he could simply be referred to as
“God.”
All these biblical notions about the deity’s name were extravagantly re-
worked and developed by Late Antique Jews and Christians. Recasting
ideas about the divine name was only one small part of the exegetical sys-
tems of nascent rabbinic Judaism and early Christianity, both of which
were founded on intricate systems of scriptural interpretation. These tex-
tual interpretations, as the exegetes themselves were aware, ranged from
defining obscure words to reading an entire text as allegory. The basic
process in all these cases was one of using language to understand, define,
and make explicit the language of the text.3 Scriptural interpretation is full
of statements—for example, clarifying the meanings of archaic words and
giving often fanciful etymologies of names. In addition to statements
about the meanings of words, some statements are about the functions of
words.
We do not find as many statements in the Bible about the special roles
of words as we would like. For example, the efficacy of biblical blessings
and curses remains somewhat a mystery. Blessings and curses represent a
particular subgroup of verbal utterances that have an “automatic” qual-
ity about them; they cannot be taken back once spoken.4 The reasons for
this are never explained, and we have to infer that it may be because such
utterances are understood to come directly from the deity, and therefore
stand distinct from the intention of the speaker.5

2. The name is often transliterated as “Yahweh.”

3. Such words-about-words are, in the terms of linguists, metalinguistic. As Jakobson
states, “An ability to speak a given language implies an ability to talk about this language.
Such a ‘metalinguistic’ operation permits revision and redefinition of the vocabulary used”
(1972:162).
4. Hence the need to use euphemisms; see Mitchell 1987:160–64.
5. See Mitchell’s interesting observation that for the utterance to be automatic either God
or a representative of God must be the speaker (1987:7–8, 174).
 The Divine Name as Effective Language 21

In the case of the divine name, however, we find a steady supply of

statements about the role and function of that name, giving us much to
analyze. The Targums, the Aramaic exegetical translations of the He-
brew Scriptures, offer many rich examples of comments, explications,
and expansions of words.6 These translations combine literal rendering
of the Hebrew with various amounts of interpretative material. It is
generally agreed that the Targums functioned as translations for Jews
who could no longer read Hebrew. Targums were clearly also opportu-
nities for exegetical ideas to be spun out and new meanings added to
the text.
Our specific example is from Targum Neofiti, one of the Palestinian
Targums.7 A translation of the specific Hebrew phrases that interest us is
offered here, followed by a translation of the Aramaic translation, with
the additions highlighted:

Moses said to the Lord,

“Behold, I am going to the Children of Israel and say to them,
‘The God of your fathers has sent me to you,’ and they will say,
‘What is his name?’ Who should I say sent me?”
And the Lord said to Moses,
“I am who I am.” And he said, “And thus you will say to the chil-
dren of Israel, ‘I Am sent me.’”
(Exodus 3:14–15 [Hebrew, emphasis added])

And Moses said to the Lord,

“Behold, I am going to the Children of Israel and say to them,
‘The God of your fathers has sent me to you,’ and they will say,
‘What is his name?’ What will I say to them?”

6. For brief introductions to the Targums, see Alexander (1990, 1992) and Flesher
(1995).
7. Alejandro Diez Macho, who discovered the text in the Vatican Library, posited a first-
century date. He argued (1960) that the presence of “anti-halakhic” (legal) material, inclu-
sion of early geographical names, New Testament use of Targumic material, and
disagreement with the Masoretic text are all evidence for an early (that is, pre-Christian)
date. His method for dating remains controversial; more scholars posit a second- to third-
century dating (York 1974; Wernberg-Moller 1962; Flesher 1995:44) or no later than the
third to fourth century (Alexander 1992).
22 Icons of Power

And the Lord said to Moses,

“I am who I am.” And he said, “And thus you will say to the chil-
dren of Israel, ‘The one who spoke and the world was there at the be-
ginning, and who is to speak to it “exist” and it will exist sent me.’”
(Targum Neofiti Exodus 3:14–15 [Aramaic, emphasis added])8

This translation is much wordier than the original, and full of the types of
circumlocutions that occur throughout the Targums. Common among
them is the translation of “the Lord did x” as “the memra [Aramaic:
word] of the Lord did x.” The phrase “I will be there with you” is trans-
lated, for example, “I will be there, my word, with you” (Targum Neofiti
Exodus 3:12). The standard explanation that memra is added in order to
avoid anthropomorphism fails because it is not consistent with other as-
pects of the translation (Klein 1972). The inclusion of “word” is in part
an extension of the biblical theme whereby God’s name is a substitute for
his presence9—that is, the word “word” was chosen specifically to present
the divine presence as a continuation of the emerging biblical theme that
the deity is represented on earth by a substitute (an angel; a name).10 The
theme is developed well beyond the biblical presentation and, as we shall
see, is used as a key for interpreting the biblical text itself.
The implicit claim of this translation of God’s revelation of his name in
Exodus is that the new Aramaic version has the same meaning as the orig-
inal Hebrew text. In the Targumic translation, however, the deity’s name
is given a new characterization. If we compare the Hebrew original with
the Aramaic translation, we find that the Aramaic has expanded the de-
ity’s name from the simple “I am” to the more complex “The one who
spoke. . . .” The deity’s revelation of his name to Moses is a choice op-
portunity for interpretation. Names present ready-made, hidden texts for
exegetes in that the semantic content of names can often be explained in a
variety of ways. Proper names logically need have no obvious semantic

8. This translation is from the edition of Diez Macho 1979.

9. See Mettinger 1982.
10. The ideology is articulated by the Deuteronomistic writers/editors. See, for example,
1 Kings 8, where “the name of the Lord,” not the deity himself, is said to dwell in the earthly
Temple.
 The Divine Name as Effective Language 23

content—that is, meanings that emerge systematically from linguistic

structure.11
It is no surprise that the exegetes chose the deity’s name to embody cre-
ative speech. The notion that a name is closely linked to the person it
stands for is widespread. Every name seems to be a mini-text waiting to be
explained, and few exegetes can refrain from discussing names. Some bib-
lical names are based on Hebrew roots, presenting obvious guides to ex-
egetes in unpacking their meaning. Exegetes also supply additional
meanings far beyond these roots. If names of humans properly explained
reveal important information, then divine names, when properly ex-
plained, tell us about the very nature of the deity.
As modern scholars have noted, the Hebrew text of Exodus seems al-
ready to contain a short meditation on the meaning of God’s name. The
name usually translated as “I will be what I will be” is playing with the
connection between the name YHWH (hvhy) and the root “to be (hvh).”12
This interpretation, extended in the Aramaic translation as God’s name, is
expanded to reveal something essential about God’s divine nature. God
not only exists himself as the creator of the world; he is also the source of
all existence. Thus, in order to refer to him and to explain who he is, the
simplest way is to describe God in his unique role as the speaker who cre-
ates. The deity’s creative act of speaking has become his proper name.
At one level this form simply tells us what the name is, just as a defini-
tion tells us the meaning of a word (“My name is x”). However, we al-
ready know that the meaning of the name is not limited to semantics. It is
also has pragmatic implications because the deity’s name is context-
related speech. The definitional form therefore has an additional, less fa-
miliar, function based on its connection to the speaking situation. Divine
utterances are context-related in the most extreme sense imaginable,
much more so than standard examples of effective speech.
When we are aware of effective speech, as for example the binding lan-
guage of a contract, this effective language usually operates in a very cir-

11. There are, of course, names such as “Hope,” which do have obvious semantic con-
tent, but that is not necessary.
12. This is not to imply that the root of the name was in fact “to be,” but only to state
that the Exodus text appears to play with this connection. See de Vaux (1970).
24 Icons of Power

cumscribed arena. The pragmatics of divine speech is distinct. The act of

speaking created the world, and thereby the very possibility of speaking to
the world.13 The “creativity” of all other words and usages pales by com-
parison. All other creative speech is only secondary, reflected power that
is dependent on the primal creative speech that established creation itself.
Divine language sets the standards for the creative power of language, and
the most important word in the divine language is the name of the deity.
Peircean semiotic analysis offers us a term that accounts for the specific
“standing for” of the divine names. The name is not an arbitrary word
chosen to stand for the deity, hence it is not a symbol. Instead, it represents
the deity in the less-familiar way in which an icon “stands for” its object.
Just as a line is formally linked with what it represents (it must have the ex-
act form of a line), so too here the divine name is understood to have a
formal, motivated relationship with what it represents (the deity).
God’s report of his name in the Exodus translation is an exegetical key
to other parts of the text. This key sheds light on a general principle of the
Targums; it highlights for us the status of the entire Torah as divine re-
ported speech. We are told that God’s words function differently from hu-
man words because they are icons of divinity in all instances of divine
speech. The exegetical key from the Targum highlights that God is mani-
fest on earth, in the words of the Torah, or, to put it more concretely, in
the icons of divine speech contained in the Torah scroll itself. The subjects
God talks about are secondary to the fact that the Torah exists as the di-
vine presence made manifest on earth for humans.
If we return now to the multiple additions of the word “word” to the
Targums, we can now see that it is the repeated articulation of an exeget-
ical principle about the status of the text itself. The references to “word”
repeatedly draw our attention to the fact that God’s words are present not
just as words spoken in the past but also as reported speech in the text. Di-
vine speech, in the objectified state of the text itself, embodies the divine
presence on earth. Each time the translation focuses on God’s word on
earth, it tries to incorporate back into the text the implicit claim of the
text, that God’s words work (create) on earth. Because the deity’s words
are present in the Torah, the words of the text are not just any collection

13. See Silverstein 1976 and 1981.

 The Divine Name as Effective Language 25

of words, but are the “creative word” manifest in the midst of the com-
munity. The word as divine presence is manifest physically in front of
them in the scrolls, scrolls that contain actual tokens (examples) of divine
speech.14 The Aramaic translation also emphasizes that the creative words
used by the deity are contained in the text, literally. This in turn highlights
the latent power of the text. The words by which the world was created
are found in the document, along with many other examples of divine
speech. Each “thus says the Lord” is an utterance of divine speech, and
the text is a collection of these utterances.
At this point it is worth reflecting on what we have learned about the
status of the text in the eyes of the exegetes. We have seen that holiness ex-
ists in the text because it is reported speech of a particular type, reported
speech of divine language. Here the exegete, the student of God’s word,
learns both about what the deity says and how he talks. If people could
talk like the deity, they would be able to create as well.
The translation represents a major shift in thinking with regard to the
Hebrew original. The scriptural text is based on an ideology of the spo-
ken-ness of divine language (“Thus says the Lord”).15 Divine language is
manifest in the examples of God’s speech and in the fact that his speech al-
ways turns out to be true. In the Targums, the focus of the translation is
by necessity on a collection of written words. These written words are ex-
plicitly construed as divine words, and thus as a collection of divine words
now present in their written form. It is as if the translator was saying that
the words spoken by the deity also exist for the reader in a written form,
and it is in that form that they are accessible currently just as they were ac-
cessible in the past in spoken form. This ideology articulates the power of
the written tokens of divine speech, introducing us to the idea that words
written on a piece of paper can have their own special power if they are
speech of the deity. These ideas can be used in rituals, as we shall see in the
next section.

14. Compare this with the widespread Neoplatonic doctrine of the “synu¸mata/tokens,”
which appears in Oracles 2.3 and 109.3 (Majercik 1989) and is discussed by Lewy
1978:190–96 and Shaw 1995: 48–50, 162. These tokens also are iconic representations.
15. Except for engraving the Ten Commandments on a stone tablet, the deity is not de-
scribed as doing any writing. On the development of the idea that scriptures are the written
work of Moses, see Janowitz 1991.
26 Icons of Power

The Interpreted Name: A Rabbinic Naming of the Name

The link forged between the divine name and the creation story is only
part of a rich complex of Jewish exegetical ideas about creative lan-
guage.16 Given the sheer breadth of rabbinic exegetical literature (third
through seventh centuries), numerous and even contradictory themes on
this topic appear. This section examines a few examples of the rich ex-
egetical traditions about the divine name that place it at the center of cre-
ative language.17
Looking backward for a moment, the creative power of the divine
name is not a biblical notion—that is, it is not explicitly stated in the bib-
lical text that the deity’s name is an automatic source of divine power.18
The creativity of God’s word is mentioned in a few scriptural references,19
and his name is conceived as a representation of his presence.20 There are
also injunctions against the improper use of God’s name, but the Scrip-
tures contain no prohibition against merely stating it.21 The biblical stories
still present the name of the deity primarily as something he reveals in or-
der to make himself known to the Israelites and other nations.
While the creative function of the deity’s name was not biblical, the no-
tion predates its appearance in numerous rabbinic anecdotes.22 One of the
first examples of the automatic power of the divine name is Artapanus’
story that Moses uttered the name of his deity into Pharaoh’s ears with the
result that Pharaoh had to be brought back to life by Moses.23 The name

16. In his survey of rabbinic literature, Idel (1992) delineates five ways in which language
and creation are connected, each represented by at least one anecdote. He differentiates be-
tween (1) the Torah as the paradigm for creation, (2) human letter-combining as the mode
of creating the Tabernacle, (3) God speaking divine names, (4) God speaking creative words,
and (5) letters as the structural elements used in creation. Themes 2 and 5 are discussed in
Chapter 4 and themes 1, 3, and 4 are covered in this section.
17. Yet another mode of creative language, heavenly liturgy, is examined in Chapter 5.
18. Scholem noted that what he called the power of the divine name was not a biblical no-
tion (1972a:63).
19. See, for example, Ps 33:6, 9 and Sir 43:5.
20. As mentioned briefly above.
21. The closest parallels in biblical texts are blessings and curses.
22. For more discussion of this topic, see Fossum 1985:245–53, who cites many of the
sources listed below, and Janowitz 1989:25–26.
23. Artapanus is usually dated to third to second century b.c.e. (Collins 1985:890–903).
This particular story is preserved in both Eusebius Preparation for the Gospel 9.27 and
Clement of Alexandria Miscellanies 1.23; see Collins 1985:901 and Holladay 1983:219.
 The Divine Name as Effective Language 27

is clearly understood to be a powerful word in this story, with the auto-

matic power to kill. The setting in Artapanus points to the wider Hel-
lenistic context. No Greco-Roman texts developed these ideas as much as
the later Jewish texts, and particularly rabbinic texts.
The Book of Jubilees 36:724 and the Prayer of Manasseh 1:325 both re-
fer to an oath that contains the divine name that was used in the creation
of the world. These examples are significant because they connect the di-
vine name with creation and because this theme reappears in numerous
rabbinic anecdotes.26 They are limited, however, in that the power of the
name may derive in part from the status of the utterance as an oath.
The link implicit in the Targums between the divine name and the act
of creation is more explicit in numerous rabbinic elaborations about the
power of the divine name. Disparate stories about God’s name, coming
from various rabbinic texts edited over hundreds of years, are all united in
their semiotic mode of presentation of the name. The iconic status of the
divine name is established and embellished in numerous exegetical texts
that recount its unique function and inherent power. Some stories explore
its power, and hence the need to limit drastically its use, and even its men-
tion. A particularly rich anecdote states that the deity’s name was not
supposed to be uttered in court by a witness to blasphemy (mSanhedrin
7:5)—that is, even in the highly defined setting of a witness report the
name cannot be uttered, because it is impossible to separate out the re-
porting of the deity’s name from its use.
Others ask about the definitional limits of the name (What is the name?
Is there only one?). As for the definitional limits of the name, some would
extend punishment for blasphemy even to those who substituted divine
attributes for the divine name (bSanhedrin 56a). This ideology endows the
attributes with the same status as the divine name itself. So too the letters
of the divine name are as powerful as the name; a midrashic text discusses

24. The Book of Jubilees is usually dated to the first century b.c.e.
25. This text is usually dated to the second to first centuries b.c.e.; see Charlesworth
1985:627.
26. See bSuccah 53a, jSanhedrin 10:2, bMaccot 11a, Targum Pseudo-Jonathan to Exod
28:30, bSanhedrin 29a, and Midrash Samuel 26. Odes of Solomon 4:7–8 mentions a seal but
does not refer directly to the divine name. In midrashic texts the abyss is sealed with the
Torah; see Sperber 1966:173.
28 Icons of Power

the destruction of the world by two letters from the divine name (Genesis
Rabba 12.10).27
The power of the divine name was harnessed by select biblical figures
who used it for protection and even for violent acts of aggression.
Solomon uses a ring with the divine name on it to subdue a demon (bGit-
tin 68b); Moses kills an Egyptian with the divine name (Exodus Rabba on
1.29). The divine name could animate lifeless images and statutes, a no-
tion that was later to be popularized with the story of the golem.28 Entire
histories of the divine name or, rather, a series of conflicting histories were
composed by the rabbinic exegetes. The distant past and the future were
both portrayed as times when people knew (or would know) how to use
the divine name, as the knowledge of the name became a metaphor for the
presence of the deity on earth and for his interaction with his people. For
example, one anecdote states that the divine name was once entrusted to
the entire nation of Israelites, given to them during their journey through
the desert. It was taken away, however, because of the Israelites’ worship
of the Golden Calf.29 Perhaps the most famous “history” is that although
at one time the name was widely known, growing corruption led to in-
creasing restrictions until the name was not used anymore.30
Once this interpretation of the divine name had been developed, the
name was itself “named.” The name was given a name as a shorthand ref-
erence that both included a reference to the fact that it is a name and re-
ferred to its complex content. The divine name is now named the
“Meforash/Explicit Name (wrvpmh ow)” based on the root “wrp”/explain,
make explicit.”31 Even though the root is commonly used to refer to the
process of interpreting and explicating, the Explicit Name is restricted in
its use. Already in 1901, Max Gruenbaum argued that this phrase had
connotations of secrecy (Gruenbaum 1901:244). He noted the similar us-

27. See bMenahot 29b; Genesis Rabba 1:10; 3 Enoch 13.1; Masseket Hekalot, Jellenik,
Bet Ha-Midrash, 2.46. See also Alexander 1983b:265 and Fossum 1985:253–55.
28. In bSota 47a and bSanhedrin 107b, Gechazi animates Jeroboam’s calves by placing
the divine name in their mouths, and they begin to speak.
29. Song of Songs Rabba 1:2, 5:1, 8:1 (on Exod 33:6) recounts two theories about the di-
vine name, one that the name erased itself, and the other that an angel descended and erased
it.
30. See bYoma 39b; jYoma 3.7. This explains why prayers are no longer effective
(Midrash on Psalms 91:8).
 The Divine Name as Effective Language 29

age of the root in the Targumic translation of Judges 13:18. In the Hebrew
text, an angel refuses to tell Manoach his name because it is “wondrous
(yalp)”; in the Aramaic translation, the name is described as “Meforash.”
Unfortunately the appearance of the doctrine about the Explicit Name
is very difficult to date. Scholem stated that “Meforash,” connoting “se-
cret,” appeared “from the 2nd or 3rd century onwards,” but he did not
go into any more detailed discussion of the topic (1972a:68). The Hebrew
texts in which the term appears are all difficult to date, and we lack clear
parallels to help us with a more precise dating.
It is tempting, but wrong, to guess exactly which name this name-for-
the-name refers to. As a name for the divine name, the term cannot be ex-
plained by pinpointing a unique name to which it refers. The term
crystallizes in a single unit the ideas discussed so far—that is, the term sig-
nals that a secret, powerful name exists, a name that itself is an object of
speculation and investigation.
Not only is the divine name a text to be studied, but it is also a power-
ful text. The term “Explicit Name” is a shorthand reference for the
process of divine naming, both as the name refers to the deity and as it is
an instrument of creativity. Exegesis of the Divine Name is exegesis of all
the deity’s power bundled in a single word. The choice of name is simple,
again because “name” most closely “stands for” the deity.
With the “Explicit Name,” we arrive at the heart of the rabbinic ideol-
ogy of ritual language.32 According to rabbinic tradition the Name can be
uttered only once a year by the High Priest when he is in the Holy of
Holies on Yom Kippur—that is, it can be spoken only by the holiest per-
son at the holiest time of year in the holiest place. All rabbinic blessing
formulas (“Blessed are you, Adonai”) are built around the recitation of a
divine name that, while it is not the “Explicit Name,” is a highly restricted
and powerful divine name that derives from the Explicit Name.
The iconicity of the divine name is seen in the focus on form as an in-

31. Classic discussions of the Explicit Name include Bacher (1901), Gruenbaum (1901),
Marmorstein (1927), and Cohon (1951). This name is contrasted with the nickname in
bSota 38a.
32. This is also the heart of their theory of textuality as seen in the additions of the word
“word” in the Targums.
30 Icons of Power

tegral aspect of the name. The structure of the object (deity) is revealed in
the structure of the word (name/text). The shape of the written form must
be preserved exactly, because it is part of the significance of the name,
hence the emphasis on the shape of the letters in copying the name.33 The
scriptural text derives its value to the extent that it is a copy of this pat-
tern. An analogy has been established between name and text; simply put,
because the name can be a text, a text can be a name.34 Just as the Torah
is said to have existed before creation, so too did the divine name, as the
Torah and the name become two mirroring versions of the primordial
text.35
The extremely dramatic relationship of the text to its context is played
out in the anecdote that the addition or subtraction of a single letter will
destroy the world.36 This anecdote is also a clear indication of the name’s
iconic power. Other stories recount the necessity of rearranging the man-
ifest content in order to find the content that must be hidden due to its po-
tential use and misuse. If the correct order had been given, anyone could
use it to wake the dead or perform miracles (Midrash on Psalms 3.2). Syn-
agogues are in turn holy places because they house these divine manifes-
tations.37 The name is literally manifested in the scroll, which becomes a
written incarnation. The meaning of the text is no longer limited to the se-
mantic meaning of the words it contains.38
Because God created the world by speaking his name, this “washes

33. This emphasis on the form led to preserving the Hebrew letters even in Greek trans-
lations, where they would be read backward.
34. In the thirteenth century Nachmanides stated, “We possess an authentic tradition
showing that the entire Torah consists of the names of God and that the words we read can
be divided in a very different way, so as to form names” (Commentary on Bereshit 1:1). He
cites the image of black fire on white fire as proof that the Torah can be read either in the tra-
ditional manner or as a series of names. The black fire on white fire image is found in
Midrash Konen, Jellenik, Bet Ha-Midrash 2.23–24. See Idel 1981.
35. Pirkei de R. Eliezer, chap. 3. cited by Scholem 1972a:70. Idel’s theme 1 (see note 16
above), Torah as paradigm of creation, posits that the world as we know it is a copy of a
written text. This is a bold attempt to give priority to text, and it also fits in with the shift
from spoken divine language to written divine language outlined above.
36. Attributed to Rabbi Ishmael in bErubin 13a. See Scholem 1965a:39.
37. See the brief comment by Lightstone 1984:118 on the scrolls as holy “relics.”
38. In linguistic terms, the name becomes less and less semantic as it is encoded in the text
into a calibrated, hyperstructured type of discourse that is at basis metapragmatic.
 The Divine Name as Effective Language 31

out” the semantic content of the divine utterances in Genesis 1—that is,
the content of the creative speech act in Genesis is not limited to the ut-
tering of “Let there be light” and so on. Divine speech is not primarily
about anything so much as it is the example par excellence of God’s cre-
ative power (in our terms, not primarily semantic, but pragmatic).
This stunningly complex ideology of the divine name will reappear in
still later Jewish texts and will continue for centuries to make both writ-
ing and speaking the divine name subject to extensive taboos. To unsym-
pathetic outsiders, the divine name will suffer the same fate as “voodoo,”
and that which is most holy will be held up by outsiders as the model of
the lowest forms of ritual practice. However, to those who understand the
pragmatic implications of using the divine name, numerous rituals will be-
gin to make sense.
 3
Thinking with the Divine Name:
Theories of Language in
Christian Exegesis
My reverence, Protarchus, for the names of the gods is profound.
—Plato Philebus 12C, cited twice by Origen
(Against Celsus 1.25, 4.48)

C
hristian exegesis paralleled Jewish explication of the divine name
and its role in creation. At the same time, Christian exegesis de-
veloped its own distinct theories.1 First Clement, for example,
claims: “Your name is the primal source of all creation” (61.59.2). In this
chapter we consider two Christian writers—Origen (third century) and
Dionysius the Areopagite (fifth century)—both of whom wrote specifi-
cally about divine names. Origen, who had close ties to Jewish exegetes,
shares many of the same ideas we found in the rabbinic anecdotes.2 For
example, in his commentary on the Song of Songs (Homilies on the Song
of Songs 1.6) and in Against Celsus (4.34) he attributes the interpretation
of divine names to “Hebrews.”3 As de Lange has pointed out (1976:118),
the term “tvmw wrvd/name interpreter” attributed to Rabbi Meir and
Rabbi Joshua ben Qorlah (Ruth Rabba 2.5 and Genesis Rabba 42.5) is
similar to Origen’s term “interpretes nominum.” Origen’s discussion,
however, is not limited to anecdotes, for he outlines some of the philo-
sophical ideas behind the power of names. As such, and because it is pos-

1. Fossum discusses similar Samaritan texts (1985:76ff., 97ff.).

2. For discussion of Origen’s contact with Jews, see G. Bardy 1937:221–23, Gögler 1963,
N. de Lange 1976, H. Bietenhard 1974, David Halperin 1980, and Paul Blowers 1988.
Bardy lists sixty-nine examples of traditions he believes Origen borrowed from Jews. Bi-
etenhard cites more than forty examples, most of them also cited by Bardy, and includes spe-
cific citations from Hebrew exegetical writings for most of his examples. Origen’s particular
amalgam of ideas about names was influenced, as he states, by the ideas of Hebrew name in-
terpreters.
3. See Homilies on Exodus 5.1 and Homilies on Genesis 12.4.
34 Icons of Power

sible to date his writings much more specifically, his presentation is par-
ticularly valuable when it intersects with and expands on the Jewish anec-
dotes discussed in the preceding chapter.
Dionysius the Areopagite, a fifth-century Christian theologian, offers a
slightly modified theory of divine name;4 such names encode in themselves
specific manifestations of the deity. He rejects the notion that divine
names represent the deity iconically, which is familiar from Origen, but
elaborates his own version of a special connection between a divine name
and the deity. For him as well, divine names are far from arbitrary. He also
incorporates some striking Greco-Roman ideas about divine names,
which gives us the opportunity to briefly return us to some of the Neopla-
tonic thinkers discussed in Chapter 1.
For all these writers, regardless of their primary religious tradition, di-
vine names are not completely arbitrary. The names, based on different
theories of their specific modes of “standing” for divinity, can be used in
a variety of ritual transformations.

Divine Names in Origen

Origen mentioned divine names in a number of places,5 but he goes into

greatest detail in his Against Celsus where he defends against the attack by
the pagan Celsus that Christians practice magic. Origen’s answer to this
charge is that Celsus fails to understand the true philosophy of names.
Language is not of human origin, and names are not arbitrary. Using Je-
sus’ names to heal is a practice attested to by Jesus himself, as Origen can
attest from the New Testament.
He broaches the topic of language and of names in particular by men-
tioning and rejecting various philosophical options (Against Celsus 1.24).
He criticizes first Aristotle’s theory that there is no inherent relationship
between a word and what it represents. Aristotle’s influential theory dis-

4. For an introduction to Dionysius the Areopagite and a complete English translation of

his writings, see Luibheid and Rorem 1987.
5. Origen discusses the significance of names in Selections on Genesis 17.5; Homilies on
Numbers 25.3, 27.5, and 27.13; and Homilies on Joshua 13.2 and 13.4.
 Thinking with the Divine Name 35

tinguished between the culturally varying words used to describe com-

mon, even universal experiences. As Parmentier explains, “For Aristotle
(De Interpretatione 16.1), mental experiences are direct images of things
all men experience uniformly, while spoken words are variable symbols of
these universal images” (1985:360).6 Origen rejects this theory because it
negates any possible divine dimension of language and makes it a social
construct. Equally weak, according to Origen, is the Stoic theory of “first
utterances being imitations of the things described and becoming their
names.” Although at first blush the materialistic bent of the Stoic theories
might seem closer to his, their ideas about the way words are “imita-
tions” of objects does not permit a divine origin for language. The Epi-
curean argument of social conventions based on emotional reactions to
objects, which Origen also rejects, must have seemed ludicrous and out-
rageous to him.
Particularly worrisome to Origen was Celsus’ articulation of the com-
mon trope that the diverse divine names used by humans all refer to the
same deity.7 According to Origen, this argument was planted by daimons,
who “attribute their own names to the supreme God so that they may be
worshiped as the supreme God” (Exhortation to Martyrdom 46). The
special power of names makes Origen hesitant even to discuss the issue.
The nature of names, he writes, “is a profound and obscure question,” as
well as “a certain, mysterious divine science.” Indeed, it is a “consistent
system which has principles known to a very few” (Against Celsus 1.25).
Names point to the deepest meanings of objects, signifying their nature
and revealing their essence (On Prayer 24.2).
Origen finds proof of the nonarbitrariness of names in the success indi-
viduals have in summoning supernatural forces—why else do they re-
spond to humans? Angel names, for example, are based on the activities
that God gave them to do and that they may do for humans (Against Cel-
sus 1.25).8 So too, in a rabbinic exegesis of Judges 13:18, an angel replies

6. Note the presupposed transparency of his system, with different words being simply
different names for the same things. Origen’s remark does not go into any detail about Aris-
totle.
7. See Against Celsus 5.41, 8.69 and note 21 below.
8. See On First Principles 1.8.1, Homilies on Joshua 23.4, Homilies on Numbers 14.2.
36 Icons of Power

to a request for his name with “I am not able to tell you my name, for the
Holy One, Blessed be He, gives us a name to accord with the errand on
which he sends us” (Numbers Rabba 10.5).
The category of divine names, as we saw in the Jewish exegesis, in-
cludes many different names. The Late Antique cosmos was populated by
many types of divine characters, enlarging the pool of potential effective
names. Origen casts a wide net that included names of all kinds of divine
beings. “Jesus,” of course, is an efficacious and divine name (Against Cel-
sus 1.67). Names of patriarchs are efficacious because the patriarchs each
have a divine character (4.34).9 The name “Sabaoth” derives its meaning
from the divine beings called “Sabai,” with “Sabaoth” being their ruler,
and Ephesians 1:21 is evidence that there are other divine beings whose
names are not known (Commentary on John 1.34). According to Celsus,
Christians, among others, can achieve special powers by employing all
these names, including the names of daimons (Against Celsus 1.6).10 Ori-
gen counters that Christians use only Jesus’ name.
The power of a divine name is automatic and not based on the inten-
tion of the speaker, a theme that is familiar to us by now.11 Moses knew
enough about these “secret doctrines” to prohibit mentioning the names
of gods, according to Origen’s reading of Exodus 23:13 (Against Celsus
5.46). Thus, for Origen, a Christian must be careful not to speak these
names of other heavenly powers, for the power would still be turned on.
A Christian would rather face death than utter the word “Zeus,” which,
whether he meant it or not, might produce miracles (4.33–34). While
Zeus may not be God, he might be a powerful daimon, and therefore an
active threat.12
Part of the power Scripture has over humans, and its ability to influ-
ence them toward good and away from evil, is due to the automatically ef-
fective names it includes. In his Homilies on Joshua, Origen states that

9. The angelic status of the patriarchs was a common theme in Jewish texts found,
among other places, in Philo and the Prayer of Joseph.
10. See Against Celsus 6.40 and Homilies on Joshua 20.1.
11. This is contrary to predominant features of present-day theories of language, where
intentional models often supplement semantic models.
12. For similar Christian prohibitions against uttering divine names, see Tertullian’s dis-
cussion in On Idolatry 20–21 and Teachings of the Apostles 21.
 Thinking with the Divine Name 37

“more powerful than any incantation is the naming of names” (20.1).13

The phrase from Psalm 102:1, “Praise the Lord, my soul, and let all the
things within me praise his holy name,” means that things (that is, pow-
ers) inside a person benefit from hearing the Scriptures read, regardless of
the intention of the hearer.
Some of the power of names comes from the special qualities of He-
brew, which for Origen is the original language.14 Names, and the prayers
that contain them, lose their efficacy in translation. For example, the
phrase “God of Abraham” translated as “God of the chosen father of the
echo” will have no effect,15 as Origen explains: “In this case the recitation
would have no effect, as it would be no different from the names which
have no power at all” (Against Celsus 5.45). The words must be said in
Hebrew or they will be “weak and ineffective” (1.25).16
Because of the problems translation presents, Origen decides it must be
“the qualities and characteristics of the sounds’ (Against Celsus 1.25) that
give names power.17 While he does not use the term “icon,” he under-
stands that the semantic reference of a name is not what is important. The
iconic aspects of the name would be lost in translation to other languages,
because in another language a word does not have the same form.18 Ori-
gen does not explicitly state, as the rabbis do, that the power of the
sounds comes from the fact that they are the sounds of the word used to
create the world. However, his interpretation of the prologue of the
Gospel of John develops a Christian parallel to the Jewish theories. “In
the Beginning was the Word” (John 1:1) establishes the creative “word”
as neither “Let there be light” nor the divine name, but instead the Greek
word “løgoq /word.” Origen comments on this text: “Because he regards
Him as the Creator of all things, he gives him the name of ‘the word.’”

13. Translations are adapted from Dillon 1985:214–15.

14. Adam spoke Hebrew, which is proof that it the original language (Homilies on Num-
bers 11.4).
15. The reason Origen translated the name “Abraham” in this fashion is not clear. “Av”
means father, but his choice of “echo” is mysterious.
16. See also Against Celsus 8.37.
17. The power of sounds is discussed at length in the next chapter. Neither Chadwick
(1965) nor Dillon (1985) considers possible Jewish influences on this aspect of Origen’s
thought.
18. Compare Dillon’s discussion of sense and reference, 1985:214.
38 Icons of Power

Again we find a package of ideas about creative language and divine

names that is as dense and complex as the Targumic renaming. Here Jesus
as the “logos” stands for the deity in his creative aspect. How then, we
wonder with Origen, does Jesus “stand for” the deity? How does his
name, “Word,” represent deity? Origen’s major divergence from the
stance of the Hebrew name interpreters is that those interpreters elevated
Hebrew as the scriptural language while he dealt with Greek scriptures.19
Origen leaves us with the question of how the Greek word for “word”
stands for the deity.20
Dionysius investigates this very question, and in his answer the special
role of Hebrew is gone. The manner in which names, including the name
“Word,” stand for God was as complicated as the manner in which the
material world reflects the transcendent deity. The question was the sub-
ject of a lengthy treatise by Dionysius entitled Divine Names, which he be-
gan with an ironic comment about his ability to reach his goal through a
discussion of the divine names. The “utterly transcendent” that he was
seeking, “it is gathered up by no discourse, by no intuition, by no name”
(Divine Names 1.1 588B). The idea that God was beyond any name was
by then quite old.21
The same point occurs later in the treatise where Dionysius cites the fa-
mous story from Judges 13:18 where an angel, asked his name, rebukes
the questioner: “Why do you ask my name, seeing it is wonderful?” (Di-
vine Names 1.6 596A). For Dionysius, the rebuke means that God is so
wonderful he does not, in truth, have a name. “Theologians praise it by
every name—and as the Nameless One” (1.6 596A). God’s name is won-

19. A similar problem was faced by Philo, who exegeted a Greek translation. He there-
fore developed the story of the inspired translators in order to establish the divine status of
the translation. See Janowitz 1991.
20. For Augustine’s explication of a theory of how words stand for objects, see Janowitz
1991.
21. On the theme of God being beyond names, see “God has many names” (Pseudo-Aris-
totle On the World 7); “No name is properly used of me” (Philo Life of Moses 1.75); “God
needs no name” (On Abraham 51); “We rely on names for the nameless” (Maximus of Tyre
8:10); “Who is greater than any name” (Corpus Hermeticum 5.1); “Anyone who dares to
say there is a name raves” (Justin First Apology 61.11); “No one can utter the name of inef-
fable God” (First Apology 63.1); “No name given” (Justin Second Apology 6.1); “God’s
name was not sent into the world” (Martyrdom of Isaiah 1:7). See also Josephus Against
Apion 2:167. See Marmorstein 1927:111.
 Thinking with the Divine Name 39

derful—that is, above every name to such an extent that he is nameless.22

For the Targumic translators discussed above, the root used to translate
“wonderful” is the same root that serves as the root of the “Explicit
Name.” Their understanding of the rebuke is: Do not ask my name be-
cause it is too powerful and too secret to utter aloud.
If the deity is beyond names, why write an entire treatise called Divine
Names? The answer is that names are actually not entirely arbitrary. A
major problem in Dionysius’ cosmology is how to conceive of the rela-
tionship between the highest, transcendent level and the lowest level, the
world of matter. In what we could call an “emanationist” theory of lan-
guage, the answer to this problem is supplied by the model of divine
names. The divine names articulate the specific relationship between the
lower levels (physical sounds/world) and the highest deity (what the
sounds represent), calibrating the cosmos.
In this system of cosmic emanations, while at one level of understand-
ing God is thought to be beyond names, divine names still supply crucial
evidence about the deity. In part this crucial evidence is at the level of se-
mantic meaning. Thus the Divine Names of Dionysius explains the mean-
ing of each of the deity’s names, starting with Good and Light. Each
name, as he now demonstrates at great length, is a mini-text that reveals
the deity’s relationship to the world and explains his nature. As we read
the lengthy treatise, which explains one name after another, we must
grasp not the particular content of each name he discusses, but instead the
connection between each divine name and the divinity named. In fact, for
Dionysius, exegesis of the revealed names, combined with the cosmologi-
cal picture developed in the other treatises, is enough to understand the
entire world, all the way from the far-off God to the lowest level of mat-
ter. Dionysius wrote systematically, and in each of his four extant treatises
he attacked a distinct philosophical issue. Each treatise examines a differ-
ent layer of the standard Late Antique multilayered cosmology. As I. P.
Sheldon-Williams notes (1972), in the treatise the Most Mysterious The-
ology Dionysius deals with God as transcendent, and in the Celestial Hi-
erarchy and the Ecclesiastical Hierarchy he deals with the heavenly and

22. This theory is reminiscent of Celsus’ position discussed above.

40 Icons of Power

earthly orders. In Divine Names, he writes about “God as First Cause”

(1972:68).
The linguistic unit “name,” as noted above, offers a particularly rich
exegetical opportunity for Dionysius to use in spinning his emanationist
theory. He exploits the very specific linguistic role of “name.” In the case
of the attributes he discusses, all have semantic content, such as “good.”
Names, as we have noted repeatedly, primarily have reference—that is,
they refer to the deity named. Thus they all point directly to the deity’s be-
ing. For Dionysius, “God” is not his name, and anyone who confuses the
name of God with God himself is obviously misguided (and magical).
However, God in fact is his name, because each name is a manifestation,
or a revelation, of the higher levels of reality now available on earth.
Therefore, divine names are a striking exception to Dionysius’ symbolic
theory of language, according to which most words are arbitrary.
The most idiosyncratic description that concretizes this point is the
statement of Dionysius that divine names are “statues.” At the start of
chapter 9, he writes: “Greatness and smallness, sameness and difference,
similarity and dissimilarity, rest and motion—we must examine also all
that is manifested to us from these statues that are divine names” (Divine
Names 9.1 909B).23 For him divine names “stand for” the deity in the
same complex ways statues do. He is trying here to point to the kind of
“standing for” relationship that Peirce calls iconic. This is another iconic
theory of divine names. A statue is not identical to the deity it represents,
but it is not arbitrarily depicted either. Just as a statue can be a formal rep-
resentation of divinity, so too can a divine name be a representation of di-
vinity too.
This striking linguistic theory, that divine names are statues, is found
scattered throughout the writings of Proclus, whom we encountered
above. Dionysius was greatly indebted to Proclus and borrowed many
philosophical ideas from him.24 Proclus’ theory of divine names was part
of his general theory of language, articulated in his commentaries on

23. Luibhied translates this phrase “divinely named images” (1987:115).

24. Proclus probably did not invent the notion of divine names as statues of God, but it
appears that he popularized it. Its absence in Iamblichus and Julian points to the early fifth-
century Athens, when Syrianus and Plutarch of Athens were teaching there (Saffrey
1982:70).
 Thinking with the Divine Name 41

Plato.25 Language, while it is essentially human, is not based on chance,

but instead manifests the essences of things. Proclus’ additional refine-
ment of these common ideas is the parallel he draws between the act of
naming and intellectual knowledge; beings higher up on the scale have
greater capacity for both. Humans can demonstrate their divine knowl-
edge when they give names to objects, because the process of naming
is participation in divine power (Commentary on the Cratylus 42, p.
13.19–27) and intellectual activity. “There are many grades of names, as
of knowledge. . . . And some are names we too can utter, others are ex-
plicit. In short, as the Cratylus has taught us and before it divinely in-
spired tradition, both knowledge and name-giving among the gods
transcends ours” (Commentary on the Parmenides p. 853.1–3). Just as a
divine presence illuminates a statue in a ritual, so too the human mind,
when it expresses a divine name in ‘articulated sounds’ reveals the hidden
being of the gods” (Platonic Theology 1.29.123–24).26
This theory has implications for rituals; it motivates the use of names
and the sounds of names in rituals. Thus in Proclus’ rituals, referred to as
theurgy, the officiant may speak incomprehensible sounds that are in fact
divine names. These sounds are “natural” and not based on social con-
vention, and are therefore directly connected to the supernatural world.
Dionysius not only shares Proclus’ extremely distinctive “names are
statues” definition, but also uses the word “theurgy” to describe the Eu-
charist.27 In one of his letters, he writes that Jesus passed on not only para-
bles but also “the mysterious rites of theurgy through the symbolism of a
table” (Letter 9.1 1108A).28 Dionysius is therefore able, through this rite,
to manifest heavenly workings on earth, to connect the many layers of the
cosmos, and to demonstrate the uppermost secrets in the earthly “sym-
bols” of the bread and the wine by their transformation.
The bread and the wine are representations of divine power in the same
way that divine names are “statues,” though the manifestation of heav-

25. Proclus’ most extensive discussion of language is in his Commentary on the Cratylus
71, pp. 29–35). See Trouillard 1975.
26. For other references to names as statues, see his Commentary on the Cratylus 51, pp.
18.27–19.17 and Commentary on the Parmenides, p. 851, 8–10.
27. See Chapter 1 for an extended discussion of the term “theurgy.”
28. Luibhied translates this as “mysteries of his divine activity” (1987:283).
42 Icons of Power

enly workings in the Eucharist is more dramatic and more directly effica-
cious than that of divine names. Even the teachings of the saints are called
“theurgical lights,” also revealing divine power on earth for individuals
who can read the teachings correctly (Divine Names 1.4 592B).
Translation from one language to another is not a problem for either
Proclus or Dionysius. Proclus states that the divine names of the Egyp-
tians, the Chaldeans, the Indians, and the Greeks all have their own effi-
cacy.29 The “grounding” of effective language is based on the universal
relationship of divinity to world, which has an international status and
transcends any particular language.
At first glance, the theory of Dionysius is best described, in Peircean
terms, as symbolic and not iconic. Dionysius will state mockingly, “Is it
wrong to call twice times two four?”30 on the basis that they are simply
two different names for the same thing. If Dionysius does not explicitly la-
bel Origen’s strange-sounding Hebrew names with their innate power as
“magical,” someone with his point of view will sooner (Celsus) or later
(Dillon 1985) make that exact assessment.
At the same time, Dionysius will argue that names are specific mani-
festations of divine power on earth and that language has a transforma-
tional role in rituals. He will, for example, never say that a priest can say
whatever he wants over the Eucharist, for the transformation at the Eu-
charist is dependent on having the exact formula spoken over the bread
and the wine.
Moreover, the words and the “symbols” of the Eucharist are transfor-
mational. Dionysius repeatedly states that the Christian rituals “divinize,”
showing that the object of the rite is the transformation of not only the
bread and the wine but also the participants. In a statement that is shock-
ing by modern standards, he claims: “Sacred deification occurs in him di-
rectly from God” (Ecclesiastical Hierarchy 2.1).31 Thus Saffrey translates

29. Commentary on the Cratylus 71, p. 32. Proclus also writes here that the amount of
power unleashed depends on the status of the divine being who is invoked, a point reminis-
cent of Origen.
30. Dionysius employs this argument to support his exegesis (Divine Names 4.11 708C).
The meaning of the text can be represented by either the text or its exegesis.
31. Thus Luibhied translates “theurgy” as “God’s works,” a switch from objective to
subjective genitive (1987).
 Thinking with the Divine Name 43

Dionysius’ use of “theurgy” as “the mysteries by which one is made a

god” (1982:72). By explaining to us his theory of names and his most se-
cret theology, Dionysius provides us with model for manipulating and
transforming earthly elements to reveal divine ones. His use of language
may be more familiar to us, perhaps, than Origen’s, but in the end divine
names are still formal representations of divinity on earth. Language is no
less transformational than for Origen, and can manifest heavenly ele-
ments from earthly.
 4
The Meaning of Letters: From
Divine Name to Cosmic Sounds
We use not just words, but sounds full of efficacy.
—Corpus Hermeticum 16.2

T
he use of letters in rituals is not a simple or obvious topic. The no-
tion that letters, which have no meanings, have cultural functions
is even stranger perhaps than the idea that words do. Analyzing
divine names, however, has prepared us for this idea, because names do
not always have meaning (semantics). It is a small step from divine names
to the letters and sounds of those names.
For this chapter, the first example of the meaning of sounds is a divine
revelation about the creation of the world attributed to the second-
century Christian thinker Marcus. The revelation about the origin of the
cosmos turns out to be a complex narrative about language, with names,
words, letters, and sounds being the vehicles of and models for creation.
Marcus’ narrative explains how the divine name created the world, and in
that process he reworks standard distinctions between words and letters.
That reworking provides us with numerous models for the meanings of
letters, many of which were probably of Jewish origin.
For Marcus, divine speech makes accessible for humans that which is
unspeakable (divine). At the same time, as in the Targum example, divine
language is not primarily language about anything (semantics) and there-
fore may not be easily comprehensible to humans. The nonsemantic di-
mensions of language reveal subtler and deeper information about reality
and divinity if properly understood—that is, if the way in which letters are
signs is understood. Divine language literally embodies all the characteris-
tics of divinity, as confusing as that may be to humans. The sounds of let-
ters are heightened language because they are cosmic sounds linked to the
structure of the divine realm.
46 Icons of Power

Marcus’ discussion offers a point of departure for examining in turn a

Hebrew text about letters, the Book of Creation (Sefer Yetsira), a dense
meditation on the pragmatics of letters. The text is a barely comprehensi-
ble investigation into the linguistic processes by which the world was
made. The distinction between speculative and ritual texts is blurred, be-
cause the speculation is about inherently powerful letters. Finally, we shall
turn to some Neoplatonic texts that offer additional theoretical grounding
for the pragmatics of letter sounds.
Marcus’ ideas are presented at length, and roundly denounced, by his
opponent Irenaeus (c. 130–c. 200 c.e.), who viewed Marcus as a heretic
and threat to Christianity.1 The presentation is somewhat cryptic, proba-
bly due to its inherent complexity and not to any distortions introduced
by Irenaeus. Marcus claimed to have had a unique revelation from a heav-
enly figure who told him the “origin of all things, which it has never re-
vealed to any one either of gods or men.” This claim to secret knowledge
appears to have annoyed Irenaeus tremendously.
The model for creation is speech that takes the ineffable and makes it
effable. The female revelatory figure begins by telling Marcus that God
created the world by manifesting himself in the pronunciation of his
name. “When first the unoriginated, inconceivable Father willed to bring
forth that which is ineffable in him, and to endow with form that which
is invisible, he opened his mouth, and sent forth the Word similar to him-
self who standing near, showed him what he himself was” (Against the
Heresies 1.14.1).
Jewish influence on this text, which is based on a four-letter divine name
mentioned shortly after this introduction, has long been noted.2 The Christ-
ian version secures the dating of this influence. The particular and by now
familiar twist here is that the deity created the world by a “Word” which is
“similar to himself.” He could then see this linguistic unit, which had the
same form as he did. This is a clear description of a Peircean icon. While it is
difficult to think of a word as being visible, this one is specifically said to be.

1. Irenaeus Against the Heresies 1.13–22 (paralleled in Epiphanius Against the Heresies
34.4.1–7 and followed in Hippolytus Refutation of All Heresies 6.37). Translations of Ire-
naeus are taken from Roberts and Donaldson 1979, with slight modifications. For discus-
sions of Marcus, see Dornseiff 1922:126–33 and Morray-Jones 1993.
2. Gaster 1971, 2:130–35, Dornseiff 1922:126–32.
 From Divine Name to Cosmic Sounds 47

The revelation that follows presents a number of specific models of the

primordial utterance, and its exact role in creation. The first model is of a
single word splitting into many pieces, which explains the complexity of
the material world. Just as in the example from the Targums, the divine
name becomes a complex of many “words,” in this case four words of 30
letters (4 + 4 + 10 + 12).3 With so many segments, the extended name
comes to include all the letters of the alphabet. The words forget their uni-
tary origin, each thinking that it is the whole sound.
The old Latin translator of Irenaeus understood the word “beginning”
from the phrase “which was the beginning” not as a temporal marker,
but as the first word spoken and then referred back to in the next phrase
(“and that utterance consisted of four letters”). He therefore preserved
the Greek word for “beginning” untranslated in his Latin text, as if he
sensed that part of the meaning of the word was its exact form.4
Marcus’ second model posits that humans can find the form and con-
tent of the creative name in unexpected places. The “emblem of this ut-
terance”—that is, of the reunited sounds, is the word “amen” (Against the
Heresies 1.14.8). “Amen” has no simple meaning; it marks the assent of
the audience to the praise offered. In this model, the word is yet another
icon of cosmic sound. Thus far in the text a number of linguistic units
function as icons (name, beginning, the four words, amen), but none of
these linguistic units is intended to function primarily for communication;
other nonsemantic linguistic units can function similarly. According to
Marcus, the cry of the infant leaving the womb “is in accordance with the
sound of every one of these elements,” an image based on “Out of the
mouth of babes and infants you have perfected praise” (Ps 8:2). So too
when “The heavens declare the glory of God” (Ps 19:1), also cited by
Marcus, the heavens do not necessarily speak in simple declarative sen-
tences. These emblems are the sounds that “remain below as if cast out-
side” when the letters ascend back up to the heavenly world. The sounds

3. We are never told the specific four words that were uttered first. The four-letter names
point to a Hebrew original, and perhaps the four words are not included because they did
not translate well from a Hebrew original to the Greek version.
4. The Hebrew name for the deity was sometimes preserved untranslated in Greek texts,
leading to great confusion among later readers.
48 Icons of Power

of letters are still available to humans even after the bodies, which make
letters manifest in the lower world, return to their origin.
The third model Marcus offers is familiar from Jewish sources as well,
and adds a fourth dimension to any text: all the hidden words manifest
only by their first letters. The letter “delta,” for example, is itself a word
that can be broken down into the name for each letter: delta, epsilon,
lambda, and so on.5 Once this happens, the difference between words and
letters collapses. Each word (name) is made up of letters, which are them-
selves revealed to be words made up of letters. These hidden words mir-
ror the complexity of the cosmos, which, as we saw, stems from the many
different sounds generated from the primordial word(s). Here the distinc-
tion between word and letter dissolves, and we see how a single word gen-
erates multiple other words.
Marcus’ fourth model is the part of his revelation cited most often, the
description of a female “Body of Truth.” “Behold, then, her head on high,
Alpha and Omega; her neck, Beta and Psi; her shoulders with her hands,
Gamma and Chi; her breast, Delta and Phi, her diaphragm, Epsilon and
Upsilon; her back, Zeta and Tau; her belly, Eta and Sigma; her thighs,
Theta and Rho; her knees, Iota and Pi; her legs, Kappa and Omicron; her
ankles, Lambda and Xi; her feet Mu and Nu” (Against the Heresies
1.14.3). The giant female figure incarnates divinity simultaneously in a
human-like body and in letters. This description is a fascinating nuance on
the trope, familiar from the Hebrew Scriptures, of describing the deity and
other heavenly beings as having human-like bodies. Some Jews view this
practice with skepticism or reject it outright as being an unacceptable
mode of thinking about divinity. In fact, the anti-anthropomorphic stance
has been so influential that it is sometimes difficult for modern scholars to
admit that Jews attributed a body to their deity.6
The substance of the divine figure, the pieces from which it is con-
structed, are letter-pairs,7 which embody, ironically, “the expression of all

5. See Dornsieff 1922:129 n. 2, where this practice is called writing (plhrvmatik©q), which
he compares to the Hebrew terms “yvlymb” or “wrvpm.” This exegetical method is found in later
Jewish texts. See Jerome Letter to Marcellus for the divine name written Yodhayvavhay.
6. See Graetz’s comment cited in Chapter 5 note 45. For a recent discussion of this topic,
see Wolfson 1994:13–51.
7. The twelve letter-pairs are similar to the letter-pairs used as the twelve zodiac signs, a
point made by Franz Boll (1903:471).
 From Divine Name to Cosmic Sounds 49

that is unspeakable.” These giant letters literally bridge the gap between
the heavenly realm and the human realm.8 The body stretches across the
cosmos, linking the upper and lower worlds. Descriptions of giant divine
figures abound in Late Antique texts, with everyone from Moses to Zeus
appearing with bodies that fill up the entire cosmos.9 Sometimes the figure
appears to be the deity, other times an angel.10 The ambiguity may be on
purpose—that is, one way of dealing with queasiness in assigning a body
to a deity is to shift the body language from the deity to an angelic figure
or leave the text so ambiguous that it is difficult to tell exactly which fig-
ure is being discussed.
As in Origen and the rabbinic texts, this text refers to numerous differ-
ent divine names, some more secret than others. The giant letter-body
speaks a name that, unlike the hidden name, “we do know and speak.”
The highest-level name, as we learn later in the text, is beyond human
comprehension and cannot be uttered by humans (Against the Heresies
1.15.1). The name that can be uttered is “Jesus Christ.”
Marcus’ fifth model divides the alphabet into three types of letters
based on formal characteristics of pronunciation: mute, semi-vowels, and
vowels. Each group of letters is the image of one of the three “powers
that contain the entire numbers of the elements above” (1.14.5). The
mute vowels, for instance, represent Pater (father) and Aletheia (truth)
“because they are without voice, that is, of such a nature as cannot be ut-
tered or pronounced” (1.14.5).11 Language can have formal resemblances
with cosmic powers even when both the sign and the object it stands for
are unutterable. Language is iconic even at the unpronounced level of
sounds.
Introducing still another model, the sixth, for cosmic letter-sounds,
Marcus states that the heavens utter the sounds of the vowels.12 Many an-
cient writers aligned the seven planets and the vowels, though the specific

8. Similarly, the letters from which the manifest world was formed are said to be infinite
(Irenaeus Against the Heresies 1.14.2).
9. See Nilsson 1963 and Ezekiel the Tragedian.
10. See, for example, the description of Sandalphon (Scholem 1991:24 and M. Smith
1963:151). Alcibiades mentions a ninety-six-mile-high angel, complete with dimensions,
who reveals a book to Elchasai (Luttikhuizen 1985 and Baumgarten 1986).
11. Subdivisions of letters into groups also appears in the Book of Creation, discussed be-
low.
12. See Dornseiff 1922:82–83 and Blau 1914:130.
50 Icons of Power

schemes were not identical.13 According to Marcus, “The first heaven pro-
nounced Alpha, the next to this Epsilon,” and so on (Against the Heresies
1.14.7). When the cosmos speaks, it utters identifiable sounds but not se-
mantically meaningful words. Humans can then make use of these
sounds, as the soul calls out the letter Omega in order to get help from the
upper world (1.14.8).
The name “Jesus” also has a numerical value based on the system known
in Hebrew as gematria, where each letter has a numerical value; this is Mar-
cus’ seventh model. The derivation of the term gematria is unclear.14 This
numerical system for letters was originally Greek, based on the Greeks’
standard use of letters to signify numbers. The practice was no doubt older
than the coining of the term. According to Dornseiff, the oldest Greek ex-
ample is found in eighth-century b.c.e. Miletus, and among Jews, the coin
of the high priest Simon, 138–135 b.c.e. (1922:91–118).15 According to
Tertullian, Marcus claimed that Jesus said he was the Alpha and the Omega
in order to introduce this system of exegesis (Against the Heretics PL 70A).
Elsewhere Irenaeus reports that the Orphites and Sethites classify divine
names by numerical value, with names that add up to less than 100 being
“material” and hence of lesser value. Gematria is yet another nonsemantic
use of language, and one that distinctly points to letters as the significant
unit. Generating the numerical value of letters and words opens up endless
interpretive possibilities and adds new levels of meaning to the text.
Some Jews, some Christians, and some Greco-Roman writers consid-
ered the method of gematria scientific. Artemidorus used the system for
investigating dreams, for example (Interpretation of Dreams 2.70).16
While gematria flourished all the way from medical to philosophical
uses,17 it was adopted with particular enthusiasm in rabbinic circles.18

13. Godwin (1991) attempts to find a standard schema based on the partial agreement
between Marcus and Porphyry (Saturn/omega, Jupiter/upsilon, Mars/omicron, Sun/iota),
but even these two writers do not entirely agree. For the identity of this Porphyry, who wrote
a commentary on Dionysius of Thrace and is not the same as the famous Neoplatonist, see
Bidez 1964 (1913), appendix, p. 72. See the comparative chart of Plutarch, Porphyry, and
Lydus made by Ruelle 1889:42.
14. For some speculations, see Levias 1905.
15. See also Stambursky 1976.
16. See Gersh 1978:300.
17. For the use of gematria in medicine, see Censorinus Concerning the Day of Birth 7
and 14.
 From Divine Name to Cosmic Sounds 51

Not everyone accepted this mode of analysis, and attacking gematria

was a way of ridiculing enemies. For example, Hippolytus, in his Refuta-
tion of All Heresies, cast scorn on this popular method as a means of cas-
tigating his opponents. He used the examples of computing the value of
the names Agamemnon and Hector, presumably avoiding Christian ex-
amples so as not to confuse his rhetorical stance (4.14). The entire theol-
ogy of Egyptian religion is based on a theory of numbers (4.43–44), and
the Egyptians even calculated the numerical value of the word “deity.”
Hippolytus, galled by the notion that these computations were used to
make predictions, summarized: “I think that there has been clearly ex-
pounded the mind of arithmeticians, who, by means of numbers and of
names, suppose that they interpret life” (4.15). According to Hippolytus,
using numbers and names to make predictions is as silly as looking at peo-
ple’s foreheads, a practice he also denounces but that had its supporters.19
While Theodorus found gematria valuable for investigating, for exam-
ple, the meaning of the word “soul,” his fellow Neoplatonist Iamblichus
denounced it (On the Timaeus, frag. 57.9–22).20 Iamblichus could easily
point to the large number of words with the same numerical value but
lacking any other connection. He did claim that Pythagoras taught a
form of divination through numbers and that it was a great improve-
ment over animal sacrifice as a means of divination (On the Pythgorean
Life 54.21–25). How exactly this divination system worked is not stated.
Theodorus’ system was too contrived for Iamblichus, though theoretically
numbers were divine.
Finally, for Marcus, the beginning of the story points to the end; when
the sounds are reunited, a linguistic reunification that mirrors the cosmic
reunification will be achieved. So too a rabbinic text claims that in the
world to come everyone will pronounce the divine name (bPesachim 50a).
By the time Irenaeus has finished his summary of Marcus’ ideas, entire
sets of exegetical systems have been paraded in at least seven different
models, each one of which can endlessly find cosmic meanings in a word

18. For Talmudic references, see bSuccah 28a, bSuccah 52b, and bShabbat 7a. See Levias
1905, Lieberman 1962:69–74, and Scholem 1972b.
19. Including, it should be noted, Pythagoreans and rabbis.
20. See the discussion by Shaw 1995:206–8.
52 Icons of Power

or letter. These hidden, nonsemantically encoded meanings are the most

important dimension of the holy text because they, and not the simple
meanings of the words, represent divinity. Each image carries its “mes-
sage” in its very structure. This does not mean that semantics does not
play a part; it does, as soon as the system reaches the level of words.
Most significant for us is that all sorts of nonsemantic sounds and
words now have a special connection to divinity and thus instantiate it on
earth. These units can all be used in ritual to put divine power into play,
pointing to a system of ritual language that relies not on first-person
verbs, for example, but on the creative words and sounds that already
were used in establishing the world itself.
The Jewish sources Marcus used are not preserved per se. We do have
one extant text that intersects very closely with his ideas and expands
them in a slightly different direction (toward divine names). The Hebrew
text, The Measurement of the [Divine] Dimension (Shi’ur Komah),21
closely parallels Marcus’ description of the Body of Truth. This time it is
a male body with names for each limb, instead of a female body with let-
ter-pairs.22 In general outline the texts contain a limb-by-limb description
of the deity, with the astronomical size of each limb (“the height of his
neck is 130,000,000 parasangs”) and the limb’s name. The textual tradi-
tion of Shi’ur Komah is immensely complex with extensive editing; Mar-
cus’ discussion demonstrates that Jews developed this material long
before the textual tradition was solidified.23 The phrase shi’ur komah ap-

21. Translations of Shi’ur Komah are cited according to Cohen’s edition (1985), includ-
ing the specific text and line. Some of the translations have been adapted slightly. For a gen-
eral introduction, see Scholem 1991:15–55.
22. Scholem, following Gaster 1971, cited Marcus’ system as a parallel to Shi’ur Komah
(Scholem 1954:65 and 1965b:37–39). Cohen (1983:23–25) rejects this comparison because
of the difference between letter-pairs and name, a difference that the text itself mitigates. See
Morray-Jones, forthcoming.
23. Scholem believed that this text was “the oldest possession of Jewish Gnosticism”
(1954:66) and argued that the original version of the text was the version found in Merk-
abah Rabba (Scholem 1965a:6 and 1965b:27). Cohen has subsequently argued that the
original text was composed in the early Gaonic period in Babylonia; it is best represented by
British Museum manuscript 10675, which he dated to the tenth century (1985:5). Schäfer
examined the manuscript and believes that it is probably from a much later date; he rejects
the idea of a single original text (1988b:75–83). For an extensive critique of Cohen’s recon-
struction of the text history, see Herrmann (1988), who depends closely on Schäfer’s as-
sumptions (1988a). For another view, see the attempt to reconstruct some of the history of
the editing process in Morray-Jones (forthcoming).
 From Divine Name to Cosmic Sounds 53

pears to refer to a category of knowledge (size/names of divine limbs) as

much as to a specific text.24 As if to clarify this confusion, all five textual
versions used in Martin Cohen’s recent edition use the device of question
and answer. One states, for example, “I asked, teach me the measurement
of our Creator . . . he said to me the shi’ur komah.” Most important for
our discussion, the variations among the textual versions show that the
pragmatic implications of these texts are expressible in manifold specific
wordings.25
The text as we have it includes two distinct levels of editing, each with
its own pragmatic implications. At the first level, the text implicitly asks
the question “What is divinity?” and answers it by presenting a divine
language, by now familiar to us, of names and numbers. The names are
individual icons of the limbs and, when they are arranged in order in the
text, an icon of the overall body as well. The litany of limb names instan-
tiates an entire language of creative words; each name embodies the prag-
matic force of that particular “part” of the deity. God is conceived of as a
collection of powerful forces (limbs), each force being captured in the
name of the limb. Thus, the reciting of the list of names is a series of mini-
transformational moments that “construct” the divine body “here” for
viewing.26 At each step they highlight the similarities between the divine
body and the human body.27

24. Idel argues that the term is used generally for a secret dimension of Torah, mention-
ing yet another variant in the Alphabet of R. Akiba, “the measurement of the dimensions of
the names of God” (1981:38). This material appears in texts with similar but distinct titles
such as the Book of the [Divine] Dimension (Sefer Ha-Komah) and the Book of the Mea-
surement (Sefer Hashi’ur), and in the hekhalot (palace) texts, discussed at greater length in
Chapter 5. Three of the five versions used by Cohen are part of hekhalot texts.
25. R. Ishmael said to him, “How much is the measurement of the body of the Holy One
which is hidden from all men?” (Sefer Raziel 98–99); “This is the size of the [divine] body as
stated in the Book of the Dimension” (Sefer Hashi’ur 1); R. Ishmael says, “Whosoever
knows this measurement of his Creator and the glory of the Holy One?” (Sefer Ha-qomah
120–21); “And I said to him: ‘Teach me the dimensions of the Holy One’” (Siddur Rabba
55–56); “I said to the Prince of the Torah, ‘Rabbi, teach me the measurement of our creator,’
and he said to me, ‘the shi’ur komah’” (Merkabah Rabba 4–5).
26. See Morray-Jones 1992 and forthcoming.
27. In 3 Enoch 9–13 and 2 Enoch 2 we find similar “cosmicizing” of a human body. In 3
Enoch, Enoch-Metatron narrates that he was taken up from earth by the deity. The deity be-
stowed wisdom on him and then “laid his hand on me and blessed me with 1,365,000 bless-
ings. I was enlarged and increased in size until I matched the world in length and breadth”
(9:1–2; Alexander ed.). Enoch receives a new name, which contains the divine name, “The
Lesser YH” (“My name is in him”) (9:12), and a crown inscribed with the words “by which
54 Icons of Power

At the second level, frames (added introductory comments) correlate

knowledge of the text with specific goals. This correlation led Joseph Dan
to characterize the text as “redeeming knowledge that assures its possessor
happiness in this world and a share in the next—the world of redemption”
(1979:72).28 These frames seek to capture the perceived power of the names
and to direct that power to a variety of goals. This mode of supplementa-
tion through framing is common; several of the texts examined below in-
clude later frames that redirect the pragmatic implications of a text.29
Sometimes the frames add new goals that appear more mundane than the
original goals. In this case Marcus might feel that the new frames somewhat
degraded his revelation, because the revelation that was given to him alone
is now being adapted for more everyday concerns. Such is the fate of many
texts when the pragmatic implications are “recycled” for new purposes.
The most important extant Jewish text about the pragmatic implica-
tions of letters is the Book of Creation.30 This short and almost com-
pletely cryptic text is so difficult to decipher and offers so many
interpretations that analysis seems pure conjecture. Translations vary
widely because it is hard to construe the Hebrew terminology and phras-
ing. Of significance to us in this discussion is not so much the interpreta-
tion of the details of the text as articulating its overall strategy. In short,
the text investigates the Hebrew letters as the building blocks of creation.
If we understand its strategy better we may understand why the text is so
difficult to understand.
The subject of the Book of Creation is the creation of the world by the de-

heaven and earth were created” (13:1). Fossum (1985:245), following Scholem, notes the
parallel passage in chapter 41 where the same letters are engraved on the throne of glory. In
this text, the deity transforms Enoch by blessing him with a cosmic-size number of blessings.
The transformation is marked by the bestowal of a divine name and the literal wearing of
the effective letters. Reciting those letters permits a human to use the same transformational
language as the deity.
28. See Cohen 1983:68–71, 167–85.
29. See the added frames in the Mithras Liturgy, discussed in Chapter 5.
30. For the dating debate, see, among others, Scholem 1954:75 and Pines 1989. Gruen-
wald (1971) published a critical edition, though others would reject his endeavor. For dis-
cussion of textual problems and attempts to create a stable text, see Gruenwald 1971 and
1973b and Sed 1973. Pines’ thoughtful discussion points to Late Antique parallels for a dat-
ing of the basic ideas, and to a series of subsequent redactions after the sixth/seventh century
(1989).
 From Divine Name to Cosmic Sounds 55

ity using a combination of what appears to be letters and numbers. Creation

is described generating letter pairs from the divine name, which includes in
it all letters and sounds. Some of these are familiar themes, but the text is so
condensed that it demands interpretation in order to make any sense at all.
The building blocks of creation are described using the common He-
brew root “rps, count/number.” The text states: “He created His World
in three books:31 rps (writings?) rps (numbers?), rps (speech?).” A new
term is coined from this root sefirah (plural: sefirot), the implications of
which are obscure. The ten sefirot appear to be a combination of the four
elements (spirit, air, water, fire) with the six dimensions.
Scholars have struggled with exactly how best to translate the obscure
sefirot and the various terms that play with the root “count/number.”
Translations of the former vary from “forms of expression” to “book.”32
The import of these phrases is that these creative elements are all related
but distinguishable. This is itself a model of creation, where the same let-
ters (root) can be used to build the various dimensions (“meanings”) of
the created world. When the ten sefirot are combined with the twenty-two
letters of the alphabet, the result is the “thirty-two secrets paths of wis-
dom” by which the world was created. The text also includes one explicit
reference to the creative power of the name, a theme now familiar to us:
“and that is why everything which is created and everything which is spo-
ken proceeds from one Name” (2.5).
The basic process described in the text is the movement from letter to
natural element. The text explains how each letter instantiates a cosmic
element. “M” represents water, “S” fire, and “A” air. Again, these are not
symbolic modes of representation; the letters include in themselves the
form and the sound of what they represent. These letters are identified as
the “mother” letters, differentiated from the seven double letters and
twelve simple letters.33 The division of letters into subgroups is familiar
from Marcus. Here the emphasis is on the elemental aspects of the letters.

31. Left untranslated, following Pines 1989.

32. Pines opted for the translation “book,” although in this study he draws parallels with
the Greek term ‘kstasiq/extension (1989:81). Dornseiff related it to the Greek word for
“sphere” (1922).
33. Another part of the text divides the letters into five groups based on the part of the
mouth used to produce the sounds. This section may be a later addition to the text.
56 Icons of Power

Creation can be brought about via letters because the letters are cosmic el-
ements, and cosmic elements are letters.
A useful way of approaching the text is to view it as a dictionary of let-
ters—that is, a dictionary in which the entries are letters and not words.
Such a text presents immense challenges of translation and interpretation.
The text must define the entries, the twenty-two letters of the alphabet
and the ten sefirot, and then explain them.
To make the task even more complicated, the dictionary is not about
semantics, which is the usual role of a dictionary. Letters have no seman-
tic meanings. Instead, each entry tries to explain the creative (or prag-
matic) implications of each letter. The entire text is therefore
metapragmatic, delineating for the reader the basis by which letters can be
used as pragmatic forces. It offers details as to how letters become mani-
fest multidimensionally and ultimately “congeal” into the basic cosmic el-
ements such as water and fire.
In general, the text is much denser than the already dense rendering of
Marcus’ system, because it attempts to give us a conception of language that
is quite distinct from our own. It outlines the basis of a process that is not
easily put into words. It is as if Marcus wrote a manual to answer the many
questions his summary raises. For example, how specifically do the creative
letters mediate between the unspoken divine world and the spoken material
world? How does invisible speech transform into the visible world?
The text borders on incoherence because the processes it describes are
beyond coherence. The best the text can do is hint at the creative
processes. It presses beyond standard notions of semantic meaning, set-
ting up analogies between the forms of language and the forms of natural
elements. A dictionary that explicates the nonsemantic meanings of cre-
ative speech can do no more or less.
Before turning to comparative Greco-Roman examples, it is worth not-
ing that the text intersects with two famous rabbinic anecdotes, both of
which are tantalizing evidence of rabbinic speculation about the creative
function of letters. The first states that Bezalel made the Tabernacle by
combining letters (bBerakot 55a).34 The brief comment gives only a bare-

34. The saying is assigned “Rav Judah said in the name of Rav,” a third-century figure.
The saying is followed by a short meditation on “wisdom” and “knowledge” being used in
 From Divine Name to Cosmic Sounds 57

bones description of the technique used by Bezalel. The story reveals little
about the questions we are interested in (how did he use the letters? what
exactly did they do?). We do learn that the making of the Tabernacle par-
allels the creation of the universe, which was also done via letters.
The second rabbinic anecdote refers to a work called “The Book of
Creation”: “On the eve of every Sabbath Rab Hanina and Rab Hoshaiah
used to study the Book of Creation and created a three-year-old calf and
ate it” (bSanhedrin 65a). It is extremely unlikely that their manual was the
same as the extant work by that name, but the story does point to rabbinic
figures who dabble in a written cosmogonic text. This story, along with
several other anecdotes, holds out the possibility that rabbis can do the
same creative work as the deity.35
Returning to the Neoplatonists, Theodorus of Arsine, a disciple of
Iamblichus, found multiple meanings for a word based on analysis of the
letters.36 All his theories establish in different ways the iconic status of let-
ters. Theodorus outlines for us in a systematic manner the meanings of let-
ters, and their formal resemblance to cosmic reality. His ideas are
preserved for us by Proclus, who scrutinized Theodorus’ ideas at some
length because their common mentor, Iamblichus, had criticized them.37
Theodorus argued for reasoning about the soul, from “letters, charac-
ters and numbers,”38 a phrase reminiscent of the Book of Creation. He ar-
ticulates the meaning of the word “soul” using three main approaches: the
phonetic, the graphic, and the arithmetical.39 According to a phonetic
analysis, even the “rough breath” that begins the word “one” can be filled
with meaning. It is an exact image of that which cannot be spoken, the
supreme principle (Commentary on the Timaeus 2.274.16–23).40 Just as in

creation (Exod 35:31 and Prov 3:19). Scholem argued that the letters were understood to
come from the divine name, though that is not explicit in the anecdote (1972a:7).
35. This theme is discussed further in the Concluding Note.
36. For an introduction to Theodorus, see Praechter 1934.
37. Proclus Commentary on the Timaeus 2.274.10–278.25. On this, see Gersh 1978, ex-
cursus, pp. 289–304.
38. I follow here Gersh’s suggestion for translation (1978:289 n. 2) contra Festugière
1967:318 n. 2.
39. Gersh dismisses as less important a fourth mode of analysis, noted by Festugière,
which is based on the position of a letter in a word (1978:290 n. 5).
40. See Hadot 1968:97. Plotinus also mentions the “rough breath” at the beginning of
the word “One,” seeming to imply that it is commonly held to have this special meaning
58 Icons of Power

Marcus’ revelation, the linguistic representation of “a nature as cannot be

uttered or pronounced” must itself be something that cannot be uttered.
The linguistic unit, in this case the rough breathing, mirrors remote reality.
The highest reality cannot be spoken, and neither can the rough breathing.
Theodorus’ second mode of analysis is the graphic mode. Graphic
analysis reveals that the circular motion of the soul is represented by the
shape of the lines of y, which is spherical (Commentary on the Timaeus
2.276.2). Similarly, Theodosius Alexandrinus states that “A” is the “origin
of multiplicity” because of its two diverging lines (Gersh 1978:277). Two
additional examples are (1) the gematria of “E” is half of theta because E
is half of its shape and (2) E’s shape symbolizes the beam Pythagoreans are
told not to step over (Theologic Arithmetic 40.9). The X is described in
Proclus’ On the Republic as the shape that occurs when the Demiurge di-
vides the psychic material and then crosses it to form the “world soul”
(Commentary on the Republic 2.143.20–24).41 In the Book of Creation,
the shape of the letter was significant, and here too the key to the creative
process by which the demiurge made the world is encoded in letters.
The third type of meaning found in letters, according to Theodorus, is
arithmetical, familiar from the examples of gematria. Theodorus’ particu-
lar usage reduces a letter to its bases (subnumbers), generating new letters
for analysis. Manipulating the numerical value of letters is one of the eas-
iest modes of finding meaning in letters, given the common use in both
Greek and Hebrew of letters for numbers. It is no coincidence that this
method of analysis was much more widely accepted than Theodorus’
other modes of analysis.
Iamblichus’ criticism of Theodorus, as presented by Proclus, rejects
some of the connections between letters and reality that Theodorus pre-
sumed to be meaningful. For Iamblichus, the connections are based on so-
cial convention;42 only those that are “natural” (iconic) can tell us
something about reality. For example, Iamblichus argued, as noted above,

(Enneads 5.8.4, 6). He also mentions the special representational status of Egyptian hiero-
glyphs (agalmata). He himself, however, favored the mind’s intuition as a mode of appre-
hending the One (5.5.5.19–27). See Hirschle 1979:39–42 and Ferwerda 1982:57.
41. See Gersh 1978:297.
42. In the Peircean terms of this study, signs that represent based on social convention are
called symbols.
 From Divine Name to Cosmic Sounds 59

words could have the same number of letters without this having any sig-
nificance. The shapes of letters change over time, showing that they are
based on social convention. To label someone else’s icons as mere conven-
tional symbols, as in Iamblichus’ attack, is to sever Theodorus’ claims
about a connection to reality. In Peircean terms, the signs Theodorus con-
siders to be icons are merely conventional symbols for Iamblichus and
therefore cannot represent the highest levels of reality.
As always, linguistic theories will have ritual implications. Iamblichus re-
jects certain ritual uses of language, such as the idea of standing on characters
while reciting words (On the Mysteries 3.13; 129.14–132.2) Proclus cate-
gorically states: “It is unsafe to argue dialectically from characters” (Com-
mentary on the Timaeus 2.278.9). The characters do not have any formal
resemblance to any level of reality, so they cannot have any special function.
Iamblichus preserves his own iconic notions of language and embraces
some connections between language and reality. It is not surprising that
central among them are names. Even when they do not appear to make
sense to humans, they are not without sense to gods (On the Mysteries
7.4.254.17–255.6). Here Iamblichus joins Origen and the many other
thinkers who posit that it is not the semantic content of names that gives
them significance, but instead their power to manifest divinity on earth.43
The key for Iamblichus is that names are able to do this because they are
“natural” representations of the deity and not based on convention.
Nicomachus, whose early use of the word “theurgy” was discussed
above, also justifies the use of vowel sounds in rituals.

And the tones of the seven spheres, each of which by nature produces
a particular sound, are the sources of the nomenclature of the vow-
els. These are described as unpronounceable in themselves and in all
their combinations by wise men since the tone in this context per-
forms a role analogous to that of the monad in number, the point in
geometry and the letter in grammar. However, when they are com-
bined with the materiality of the consonants just as soul is combined
with body and harmony with strings—the one producing a creature,

43. See Shaw 1995:110, 179.

60 Icons of Power

the other notes and melodies—they have potencies which are effica-
cious and perfective of divine things. Thus whenever the theurgists
are conducting such acts of worship they make invocation symboli-
cally with hissing, clucking, and inarticulate and discordant sounds.44

Language is a metaphor for the state of the cosmos as both material and
immaterial at the same time. The effective dimension is that which more
closely resembles the soul and not the body.
The Hebrew name theory and the Greek vowel theory merge when the
Hebrew divine name is thought to consist of only vowels. A specific nexus
between the four-letter Jewish divine name and the seven Greek vowels
appears, for example, in Eusebius’ statement that the Jewish divine name
consists of seven vowels reduced to four (Preparation for the Gospels
11.6). As both vowels and sounds of the divine names, these four letters
are doubly effective. A more general theory of the efficacy of vowels is
found in Philo, who simply stated that vowels are the best and most ef-
fective letters (Allegorical Interpretation 1.14).45
All the theories outlined here are put to work in various Late Antique
ritual texts where vowels are employed. The use of vowel sounds in ritu-
als was widespread enough that we find diverse kinds of evidence for it.
Demetrius of Phalerium believed that the pleasing use of vowels in rituals
came from Egypt, reversing the common trope that bad ritual practices
(magic) comes from Egypt: “In Egypt the priests, when singing hymns in
praise of the gods, employ the seven vowels, which they utter in due suc-
cession; and the sound of these letters is so euphonius that men listen to it
in place of flute and lyre” (On Style 71).
A Jewish-influenced prayer from the Greek papyri (mid-third to mid-
fourth century) connects hidden divine names, vowels, and cosmic forces
(PGM 13.760–932).46 The basic thrust of the prayer is identification with

44. Greek Writers on Music 276.8–18. Jan; translation from Gersh (1978:295). See
Dornseiff 1922:52.
45. Philo may also have written a treatise, now lost, on the meaning of letters (Wutz
1914).
46. See the next chapter for more uses of vowels in rituals. Secret vowel-based divine
names also appear in several of the Nag Hammadi texts, including the Gospel of the Egyp-
tians (3.2.43–44, 66) and Discourse on the Eighth and Ninth (6.6.61; 60.17–61.17).
 From Divine Name to Cosmic Sounds 61

a divinity by the reciter who knows the secret name. The prayer is entitled
“Here is the instruction [for recitation] of the heptagram,” a term that ap-
pears to refer to the seven vowels. The prayer begins with a reference to
the name being “hidden and unspeakable” for humans (764–65). Mid-
way in the text, it invokes the divine name of vowels: “Yours is the eter-
nal processional way in which your seven-lettered name is established for
the harmony of the seven sounds [of the planets that utter] their voices ac-
cording to the 28 forms of the moon sar aphara aphara” (PGM
775–80).47
Further on in the prayer the secret name reappears as “a unique phy-
lactery in my heart,” which guarantees identification of the reciter and the
divinity (you are I, and I am you) (795–96). The name also resides in the
reciter’s soul (800). In the closing, the reciter claims to have “the power of
Abraham, Isaac and Jacob” and “the great god demon Iao.”48 This prayer
asks for quite an array of blessings, including “health no magic can harm,
well-being, prosperity, glory, victory, power, sex appeal” (804).
Repeatedly we have seen that letters and their sounds are significant be-
cause they have more than simple, worldly (semantic) meaning. Instead,
cosmic sounds are important because they are in fact cosmic. They repre-
sent the divine world on earth, being icons of some aspects of that world.
Sounds can “do things” because they are icons of divinity, audible and ac-
cessible bits of the inaudible and inaccessible divine world. Though dis-
tinctions between words and letters are confused, language does not have
to be destroyed in order to find reality. Instead, language is reality, and
particularly at the level of an individual letter and its sounds. The basic
structure of the cosmos is contained in, and not distorted by, the basic el-
ement of language: sound.

47. The translation cited here is by Morton Smith (Betz 1986:190). Gersh also mentions
this phrase in his discussion of phonetic analysis below (1978:294). See Dornseiff 1922:37.
48. On the popularity of the divine name Iao, see M. Smith 1973:233. See Dieterich
1891:68–71 on PGM 5:460–85. According to Pistis Sophia 136, Jesus spoke the name “Iao”
to the four directions.
 5
Using Names, Letters, and Praise:
The Language of Ascent

Tell the mystery, and seal yourself

with this seal, this is his name
—The Second Book of Jeu 52 (GCS 13.315–21)

T
he heavenly ascent by Rabbi Nehunya in Hekhalot Rabbati has
become something of a classic, first championed by Gershom
Scholem and since then the subject of much discussion.1 This text
is part of the loosely defined collection of Hebrew texts referred to as
hekhalot (palace) texts or merkabah (chariot) texts, which describe the
heavenly realm, the liturgy of the heavenly chorus, rites for calling down
the angelic figure the Prince of the Torah, and assorted other esoteric tra-
ditions.2 In Rabbi Nehunya’s ascent,3 he separates his body from his soul
and traverses the fiery and terrifying heavens, describing them as he goes.4
In another ascent text, the Mithras Liturgy, we witness the bodily trans-
formation during ascent of a human to heaven, who then becomes a heav-
enly being. Employing special language functions is the heart of the
efficacy of both these ascents, including such now-familiar techniques as
reciting divine names and letter-sounds. The rituals depend on an addi-
tional mode of divine language: descriptions of the angelic cult in heaven
and its liturgy, found already in the Dead Sea Scrolls. Harnessing divine
language permits the human practitioners to transcend their earthly exis-
tence and become part of the heavenly world.

1. See especially Scholem 1954 and M. Smith 1963. See also the studies cited below.
2. For summaries, see Schäfer 1992, and on the angelology, see Elior 1994.
3. The Hebrew term used is “descent.” For Scholem’s speculations on this term, which
have not been superseded, see 1954:46–47 and 1965a:20 n. 1.
4. James Tabor outlines the basic pattern of Late Antique ascent (1986:87).
64 Icons of Power

Ascent to the heavens was a common postbiblical literary theme re-

flecting the elaborate systems of heavens that emerged after the third cen-
tury b.c.e.5 Revelatory texts attributed to biblical characters were often
cast as ascents.6 Characteristically the enigmatic phrase “Enoch walked
with God” (Gen 5:24) was expanded in later texts to a lengthy and de-
tailed presentation of what Enoch saw and learned on an extended tour of
the cosmos.7
Jewish descriptions of ascent were by no means unique; numerous
Greco-Roman and Christian texts were structured as trips taken by indi-
viduals to the upper realms.8 The formats of these ascents varied widely.
Heroic figures were portrayed as being snatched up into the heavens, usu-
ally in the distant past. Some ascents happened during the character’s life,
some after death;9 the setting could be a dream, a vision, or even part of a
philosophical investigation.10
Although its origins are somewhat murky, ascent emerged as a ritual
practice by the first century c.e., when it was no longer something at-
tributed only to figures from the distant past. Exactly how, or when, the
ritual practice of ascent emerged is not clear. The complex relationship
between the literary themes of ascent and the development of rituals is
perforce shrouded in mystery because so few texts survive.11 Ascent was
no longer conceived of as something that happened only to past heroes.
In the Dead Sea Scrolls we see claims that humans can see the heavenly
world, making ascent a goal that could be actively sought out and af-
fected by human action.12 New types of evidence about ascent emerge by
the first century c.e., including first-person claims of ascents by historical

5. See Nilsson 1948:96–110 for a summary of the Greek version of this cosmology; for
Jewish versions, see the many descriptions of the multiple heavens in the pseudepigraphical
texts and rabbinic texts.
6. See Segal 1980 for a review of texts that include ascents.
7. See 1 Enoch in Charlesworth 1985, as well as the other Enoch texts.
8. For additional Greco-Roman references to ascent, see Johnston 1997:116 n. 4, and
for ascent as a theme in earlier Greek texts, see ibid., 176 n. 35.
9. For ascent after death attributed to the “Orphians,” see Against Celsus 6.24–38.
10. See M. Smith 1981.
11. The Qumran texts give us a hint as to how much of ritual and liturgy is lost to us.
12. Human volition does not negate the need for divine permission and even help in the
highest realm.
 Using Names, Letters, and Praise 65

figures,13 liturgical texts structured as ascents,14 and deification-by-ascent

rituals.15
The possible reasons for and meaning of the ascent rituals were diverse.
In the multiheavened Late Antique cosmology, a trip to the heavenly realm
was a way to gain access to the angelic court and its occupants. Ascent was
sometimes envisioned as a means for a human either to return to his “home”
or to achieve a new, heightened status by the trip.16 Sometimes ascent was
believed to happen only after a person died, and therefore the ritual would
be enacted only after death. The divinization of the Roman emperors was a
highly visible example of ascent as transformation after death.17
Ascent could be a goal in itself, as in the philosophical life,18 or a
process undertaken in the name of other goals, such as learning about the
future.19 Some Jews wanted to ascend because they, like other residents of
the Greco-Roman world, wanted to have closer contact with the deity
who resided in the seventh heaven.20 Ascent could have been undertaken
in order to gain immortality.21 The widespread belief in astral immortality
held out the possibility that someone could ascend into the heavens and
become an immortal star (4 Macc 17:5). Undertaking an ascent while still
alive could function as proof that someone was godlike. For Philo, ascent
was a means of divinization (Questions on Exodus 40).
Literary scholars often have limited specific interest in issues of ritual ef-

13. See the claim of Paul in 2 Corinthians 12:2–4 and the discussion of this text as an as-
cent by Paul in Tabor 1986.
14. See the Songs of the Sabbath Sacrifice, discussed below.
15. On the deification rituals of the Roman Emperors, see Bickerman 1929 and more re-
cently Price 1984 and 1987.
16. See J. Z. Smith 1974:749.
17. See the Roman senator’s report that he saw Augustus ascending to the heavens (Sue-
tonius Augustus 100.4 and Dio Cassius History 56.46.2).
18. For Apollonius of Tyana’s ascent, see Philostratus Life of Apollonius of Tyna 8.30
(M. Smith 1981:409).
19. This is a theme in some of the texts discussed below.
20. Martha Himmelfarb opts for a philosophical, almost existential, meaning of ascent
(1993). She turns to Goodenough’s view of Philo, for whom ascent was a way to solve the
problem of the perception of God’s inaccessibility. Here she follows earlier scholars, such as
J. Z. Smith and Morton Smith. This analysis is more convincing than elsewhere in her book,
where she posits that the daily lives of Jews may have been unsatisfactory and that that may
have led them to imagine themselves “like the glorious ones” (1993:114).
21. Among hekhalot texts, this particular claim is found most clearly in claims about the
world to come.
66 Icons of Power

ficacy. The result is that the explanations of ascent rituals are based on very
narrow concepts of efficacy. The most common explanation of ascent ritu-
als is that they work by means of a trance.22 Certain procedures, usually de-
privational, induce a trance during which an individual thinks he is
traversing the heavens.23 Trance is sometimes presented as the best rational
explanation for seemingly irrational rituals (people become dizzy and think
they are ascending). These explanations, like comparisons with shamanistic
“trips” in oral cultures, tend to obscure the historically specific nature of rit-
uals.24 As appealing as this idea seems, there is little direct evidence of trance
in the texts.25
Morton Smith astutely observed that when Rabbi Nehunya’s compan-
ions want to ask him a question they must carefully interrupt him because
he is not simply talking about an ascent but ascending while he talks
(1963:145). How does his “talk” make him ascend? At first glance, this is
not a promising clue about techniques. We find no complex instructions,
such as having to go to a particular place, wear certain clothing, or bring
some special objects.26 However, the previous chapters have prepared us
for ideas about the efficacy of words, which we now see here in action:
ascent is achieved primarily based on the repetition of vowel sounds,
divine names, and, in this case, heavenly liturgy.27

22. This explanation goes back as far as Bousset 1901.

23. Philip Alexander uses the notion of trance to explain how ascent works in 3 Enoch.
He states: “Given the rigorous preparations for the ascent, the sense of anticipation and the
aura of mystery and awe with which the act was surrounded, this repetition would have
been potent enough to send the adepts into trance. Its efficacy would have been magnified if,
as Hai Gaon asserts, the mystics recited the hymns with their heads between their knees, in
the manner of Elijah on Mount Carmel, thus constricting their breathing” (1983b:233).
24. It seems impossible to me to distinguish between a “trance ascent,” which appears be
a pseudo-ascent, and an ecstatic rapture, where someone actually travels somewhere”
(Schäfer 1992:154–55).
25. Support for the trance theory is found in later Jewish texts, such as Hai Gaon’s tenth-
century commentary (note 23 above). These later explanations of technique may represent
reinterpretations of ascent developed when the earlier basis of efficacy was no longer under-
stood or had simply changed.
26. Ma’aseh Merkabah has a single reference to fasting in an Aramaic section (19).
27. Some of Scholem’s heirs have been eager to find evidence of ascent techniques in any
text that mentions ascent. Martha Himmelfarb has cautioned against these arguments,
pointing to the literary nature of many references to ascent (1993). Her caution is warranted
in the case of those narrative texts where ascent is only a literary topos. A clear distinction
should be kept in mind between a literary reference to a person ascending and a ritual text
where ascent is structured into the text as a practice.
 Using Names, Letters, and Praise 67

The technique of reciting words in order to ascend appears in the first

extant ascent text, Songs of the Sabbath Sacrifice (4Q400–407).28 The
Songs are among the earliest extant Jewish liturgy and display some of the
strategies Jews used in creating liturgy.29 In fact, they are precious evidence
because no public liturgy besides the Qumran scrolls is extant from the
first centuries c.e.30 The Songs is a liturgical collection of thirteen hymns
containing intricate descriptions of the angelic cult in heaven. The hymns
detail the seven priesthoods along with other participants in the heavenly
cult, ending with a description of the heavenly temple (based on Ex
40–48). All the participants in the heavenly court are busily engaged in
singing praises to God.
The image of a heavenly chorus goes back to the Hebrew Scriptures.
The most famous and oft-repeated instance is Isaiah’s report that the an-
gels recite “Holy, holy, holy is the Lord of Hosts, the whole earth is full of
his glory” (Isa 6:3). The image of a heavenly court, complete with liturgy
and other cultic activities, was extensively developed in Jewish and Chris-
tian texts. The heavenly cult superseded the earthly cult, with angels of-
fering their own versions of sacrifice complete with continual praise.31 As
we would expect based on all the descriptions of heavenly sounds exam-
ined in the previous chapters, the heavenly sounds were both much more
extensive and much more cryptic than Isaiah’s “Holy, holy, holy.” For ex-
ample, Apocalypse of Abraham 15:7 describes the sound of heaven as “a
sound of words I knew not.”32

28. This text was found in multiple copies. See the critical edition by Newsom 1985.
29. Newsom believes, despite the extreme ambiguity of the case, that the scales tip in fa-
vor of the document’s not being a product of the sectarian community. The bases for this is
that a copy was found at Masada, and the particular manner in which God’s name is
recorded (Newsom 1990b). It is a mistake to marginalize the text as more “sectarian” than
other contemporary texts.
30. Talmon writes: “The absence of codification, definition and binding formulation may
account for the fact that the text of not even one Jewish public prayer from before the de-
struction of the Second Temple, or for that matter from before the middle of the first mil-
lennium c.e., has been preserved” (1978:271). Some rabbinic circles may have actively
repressed liturgy that they felt was not sufficiently under their control or did not represent
their interests.
31. In addition to the Qumran descriptions, rabbinic descriptions of the angelic cult,
complete with its sacrifice of souls, are collected in Aptowizer (1930–31).
32. For other descriptions of the heavenly chorus and its language, see Origen Against
Celsus 7.9, Lucian Alexander the False Prophet 13, Philo Who Is the Heir 259–66, Testa-
ment of Job 48–50, and Corpus Hermeticum 1.26.
68 Icons of Power

The Songs describe the cultic heavenly world in detail. Newsom con-
cludes: “What characterizes the Sabbath Shirot is description—insistent
and often vivid description of angelic praise and blessing, of the heavenly
temple, its structures and appurtenances, the merkabah, the procession of
the angels, the splendid vestments of the celestial high priests, and so
forth” (1985:64). The seventh Song, for example, describes how the
cherubim and ophanim, the mythical beasts who carry the deity’s throne,
and even the heavenly furniture praise the deity:

And all the crafted furnishing of the debir [inner sanctum] hasten
(to join) with wondrous psalms in the debi[r . . .] of wonder, debir
to debir with the sound of holy multitudes. And all their craft fur-
nishing. [. . .] And the chariots of His debir give praise together, and
their cherubim and thei[r] ophanim bless wondrously [. . .] the
chiefs of the divine structure. And they praise him in His holy debir.
(Songs of the Sabbath Sacrifice 4Q403 1.2.13–15 [Newsom
1985:229])

Based on the preliminary publication of fragments of the text, Morton

Smith claimed that the text was an ascent text (1981:412). The editor of
the critical edition of the fragments, Newsom, came to a similar conclu-
sion with a different nuance.33 That the text is liturgy is widely accepted;
that the liturgy was used as part of a ritual aimed at ascent is more con-
troversial.34 According to Newsom, the description of the celestial high
priest is the climax of a step-by-step ascent. The ascent by the individual
reciting the text culminates in a vision of the innermost heavenly temple
and its cult. The language of the text creates “the experience of being pres-
ent in the heavenly temple and in the presence of the angelic priests who
serve there” (1990b:115).35 Baumgarten (1988) similarly observes that the
worshiper is transmitted progressively through the heavens by the text.
How exactly can we account for what these scholars are claiming—that

33. See Newsom 1985:59–72 for a discussion of its liturgical use.

34. Schiffman (1982) and Himmelfarb (1993) reject the notion that ascent was a ritual
practice in this text.
35. Newsom (1985:63–64) also discusses the shared priestly community of angels and
humans.
 Using Names, Letters, and Praise 69

is, what is the basis of the perceived efficacy of the Songs to transport their
reciter into the heavenly world?
Both the structure and the language of the text contribute to the effect
of seeing the upper realm, which, given the cosmology, necessitates an as-
cent. Surveying the Songs quickly gives one an impression of their struc-
ture: an exact image of the structure of the heavenly world, mapping out
the increasingly holy and central level of the cult.36 Patterns of seven lay-
ers structure the text:

Praise [the God of elim,

O you seven priesthood of his inner sanctum . . .] loftiness,
seven wondrous territories according to the ordinances of His
sanctuaries [the chiefs of the princes of the (wondrous)
priesthoods] [. . .]
seven priest[hoods] in the wondrous sanctuary for the seven
holy councils. . . .
And the tongue of the first [angelic prince] will grow strong
sevenfold [joining] with the tongue of the one who is second
to him.
And the tongue of the one who is second with respect to him
will grow strong sevenfold from [the sound of] the one who is
third with respect to him.

(Songs of the Sabbath Sacrifice 4Q403

1.2.20–28 [Newsom 1985: 230])

Such repetitive patterns are the heart of many ritual texts since they out-
line the “culturally derived aspects of the cosmologies” (Tambiah
1985:142). In this case the culturally derived aspect of the cosmology is
the entire cosmology, because the “here” of the ritual takes place in the
heavenly realm. The entire ritual “takes place” in a place that has been
constructed by reference to the reciter’s systematic mapping of it.37

36. This is also true of later ascent texts, as, for example, the layer after layer of flaming
chariots depicted in Ma’aseh Merkabah.
37. The repetitious nature of the text is sometimes presumed to point toward trance, as
noted above. The point being made here is that ritual texts are repetitious in part because
that repetition provides the essential poetic structure for the ritual context.
70 Icons of Power

The ascent is put into action by the act of reciting the liturgy. As New-
som states, the imperative verb forms at the start of each hymn activate
that particular hymn (that is, “Praise . . .”), coordinating human activity
and divine activity. Finishing the recitation is linked with having seen the
most holy level of cult. The Songs are a special form of socially conceived
action by which the context of the heavenly world, complete with the an-
gelic chorus and cult, is invoked as a place where the reciter has been.
Recitation of the hymns results in construction of the special (heavenly)
context for human liturgy. The projection moves from the language of the
ritual to the cosmos, modeling in the Songs not only every detail of the
heavenly cosmos but also the interactions that take place within it.
At the same time, the Songs have the semblance of being free of any
particular earthly context—that is, they do not seem to be tied to a spe-
cific time or space. Recitation of the Songs seems to take place suspended
in time and abstracted from a clear historical setting. The earthly context
in which the hymns are sung is set aside and not mentioned.38 The very
lack of any earthly dimension to the cosmology emphasizes the heavenly
world, which is the only reality for the ritual.39
Priestly status seems to be a key to this ascent, because the reciters are
the earthly parallels to the heavenly priests.40 The priests who recite the
text, presumably as a group,41 overcome the horizontal divisions between
earthly beings and heavenly beings enabling them, literally, to watch the
heavenly cult unfold. Pointing out the superior status of the highest angels
over the human priests preserves some humility.42 However, the special
human and angelic cohort is higher than both other humans and other
heavenly groups because they are in the presence of the highest cult.
When the ascent practitioner is not a priest, the importance of angelic

38. See Newsom on avoiding explicit references to the human community (1990b:115).
39. Richard Parmentier explains: “In many cases the denial of referential specifically en-
ables rituals to concentrate on reference to eternal or universal truths, in much the same way
that, as Mukarovsky argued, the aesthetic function of a work of art is freed from particular
denotational value” (1994:131).
40. Elior has written most extensively and persuasively on the importance of priestly
themes in this literature (1999). Descriptions of heavenly liturgy do not necessitate an origin
or use only in priestly circles, though in the Songs the priestly connection is clear.
41. Note the frequent use of the plural.
42. See the self-critical comment made by the priests in the second Song.
 Using Names, Letters, and Praise 71

priests is, not surprisingly, diminished. A more general model of heavenly

liturgy as angelic but not priestly praise will be exploited in later ascent
texts.43
The purpose of the ascent in the Songs is never made explicit. To the ex-
tent that the cult was understood to take place in heaven, having intimate
knowledge of that cult was a necessity for any priest on earth. The hymns
also undoubtedly reinforce the authority of those who used them (New-
som 1990b). In another Qumran text (1QSb), blessings transform a person
into a being like the “angel of the presence.” It is difficult to imagine that
being part of the angelic chorus, or being like the angel of presence, is any-
thing less than a mark that the reciter of the hymns has become immortal.
Yet another Qumran text (4Qma War Scroll variant) describes an indi-
vidual who becomes one of the “gods” (oyla) via an ascent (M. Smith
1990). The phrase “I shall be reckoned with gods and established in the
holy congregation” (frag. 11, col. 1, 13–14) is a claim that, as Morton
Smith points out, makes the most sense coming from someone originally
mortal. This text is a report of a transformation of a human who is es-
tablished in the heavenly cult, though we are not told exactly how this
happened.
In all these examples, the Qumran texts present critical evidence about
the use of “talk” to transform an earthly speaker. Sometimes such talk is
explicit descriptions of the heavenly cult, other times it seems to be a more
general use of blessings, which are another subcategory of heavenly
liturgy. All this talk is not human talk, but instead the copying of the lan-
guage of the upper realm.
Against the contours of the Songs, the strategies used in Rabbi Ne-
hunya’s ascent emerge as variations on the same themes of divine “talk.”
The rabbi’s ascent appears in Hekhalot Rabbati, yet another Jewish text
that presents foreboding challenges.44 Like all the hekhalot texts, this one

43. In the Book of Secrets, chap. 6, no priestly motifs remain in the heavenly liturgy. In
parallel rabbinic texts, the priestly themes remain (for these, see Elior 1990); new rabbinic
criteria of association ultimately supplant them, though rabbis are always willing to lay
claim to the priestly authority.
44. The Hebrew text is available in Schäfer 1981:81–280 and the German translation,
Schäfer 1987–91, vol. 2. On its redaction, see Schäfer 1988a:63–74 and Gruenwald
1980:150–73. The text begins with a long series of hymns introduced as being appropriate
72 Icons of Power

is difficult to date, difficult to translate, and exhibits particularly thorny

problems with regard to the manuscript traditions. None of the hekhalot
texts are extant from Late Antiquity; they must be reconstructed from
much later, mainly thirteenth- and fourteenth-century manuscripts
(Schäfer 1981). Given the controversy that surrounds these texts, it is nec-
essary to pause and discuss the nature of the evidence before examining
Rabbi Nehunya’s ascent.
Scholars continue to debate the viability of reconstructing Late An-
tique texts from late medieval manuscripts, a debate loaded with the
added burden that the material is viewed with a general suspicion. Ger-
shom Scholem consistently pressed for not only a Late Antique but even
a pre-Christian date, to implicate a Jewish origin for the material (1954).
To bolster his argument and to make the exotic texts appear more main-
stream, whenever possible he cited parallels to the normative rabbinic
texts, such as the Mishnah and the Talmud (the earlier the rabbinic text,
the better).
Many of Scholem’s points have become clichés, with widespread
agreement that Late Antique Jews engaged in practices and had beliefs
that were, pre-Scholem, denounced as nonsense or late medieval degener-
ation of Judaism. No one argues anymore that these texts were borrowed
from Islamic material, for example, nor would anyone describe them, as
Graetz did, with such terms as “childish,” “degenerate,” or “monstros-
ity.”45 However, not everyone is ready to accept Scholem’s early dating of
the hekhalot texts (first through third centuries) or to place them at more
than the very fringes of Judaism.46
Rejection of Scholem’s early dating of the material in the hekhalot
texts revolves around the status of the textual evidence. Peter Schäfer,

for ascent, and then shifts to a series of apocalypses that give “historical” settings for specific
ascents, such as the story that after the Romans had seized several scholars R. Ishmael was
then sent to descend and find out why God had permitted this to happen (par. 107, sec. 5).
The text returns to hymns again, Shi’ur Komah material, more hymns, descriptions of a
prayer ritual in heaven, and lists of divine names. Only then does the main ascent begin (par.
198, sec. 13). After the main ascent, the text concludes with some final hymns and two ap-
pended stories about enhancing memory.
45. Graetz on Shi’ur Komah, cited in Cohen 1983:vi.
46. See also the discussion of the marginalizing of selected texts in modern scholarship in
the Introduction.
 Using Names, Letters, and Praise 73

for example, questions whether the texts ever existed in “original” edi-
tions, given the high number of variants found in the various manu-
scripts (1981). He abandoned the idea of a single original text and
presents the text in synoptic form instead. This fluidity of tradition is
seen as evidence that it is impossible to reconstruct Late Antique ver-
sions of the texts. Hence, it is unknown what, if anything, goes back to
the first centuries.
Scholars have also attacked the connections Scholem attempted to
draw between the hekhalot material and the more familiar rabbinic
texts.47 For example, David Halperin (1980) argued, based on redaction
criticism, that the Mishnaic, Talmudic, and Midrashic references to
hekhalot material appear in the later strata of the rabbinic texts and thus
are not evidence of early rabbinic practice.48 Thus the classic Talmudic
story about four rabbis who went to Paradise was not originally a story of
a heavenly ascent vision, as Scholem argued (1965a:14–19), but simply
an allegory about the practice of biblical exegesis. Only in later interpre-
tations was the story understood to refer to an ascent practice. Early rab-
binic Judaism remains orthodox. In a similar vein, Schäfer posits a
historical context of a fringe group of “mystics” versus “normative” rab-
binic Jews, and this reconstruction determines the contours of all the
themes he elaborates from the hekhalot material.49
Despite the claims that the methods of textual criticism hold out seem-
ingly scientific proof of the development of traditions, the continuing de-
bates prove that it is possible to come to more than one conclusion about
the history of the texts. Recently, there has been resurgence in support for
the position that the texts have sufficient coherence to be viewed as iden-
tifiable texts from the first centuries. Morray-Jones (1992, 1993) offers an
alternative reconstruction of the editing of the rabbinic texts that is more

47. See Swartz 1996:214 for a brief comparison of purity issues in these texts and in other
rabbinic texts.
48. See the review in Elior 1990 and the critique by Neusner 1985:172–95.
49. Schäfer’s view (1991) that the person ascending rejects traditional means and tries in-
stead to storm heaven depends upon an extremely narrow definition of traditional practices.
He claims that it is accepted by all scholars that ascent moves between two poles: the pure
form and the magical. Yet even Schäfer’s pure form, ascent via the use of hymns, is “magi-
cal” by his own criteria. On his narrow view of traditional practices, see Lesses 1998:35–43.
74 Icons of Power

favorable to Scholem’s thesis.50 Davila (1993) has recently argued for a

critical edition of Hekhalot Rabbati.
Perhaps the textual controversy has been asked to bear too much
weight.51 Whether the reference to a “chariot” vision is late or early in a
specific rabbinic text tells us only about the redaction of that particular
text.52 Jews engaged in extensive speculation about the heavenly chariot
and palaces for centuries before it was recorded in rabbinic texts. Rachel
Elior maintains that “an uninterrupted line can be drawn from the reli-
gious-literary activity of the last centuries b.c.e., associated directly and
indirectly, with the merkabah tradition, to the mystical works of the first
centuries c.e. known as hekhalot and merkavah literature” (1999:105).53
Rabbinic texts begin to polemicize about ascent practices in the third cen-
tury, but that does not mean the practices began then.54 This does not
make the esoteric material a late intrusion into a monolithic rabbinic Ju-
daism. It does point to contemporary power struggles about religious rit-
uals directed toward heavenly and earthly power. The editing and
reshaping of ritual texts probably continued for centuries, driven by
forces far removed from the earlier speculative traditions.
Morton Smith dates one of the apocalyptic sections of Hekhalot Rab-
bati to the mid-fourth century, but he also argued, following Scholem,

50. Morray-Jones has been critiqued in turn by Goshen-Gottstein 1995, who argues that
the “paradise” story about Jewish mysticism appears only in the Talmudic version. See also
Alexander’s discussion of the relationship between 3 Enoch and Talmudic traditions
(1987b).
51. See, for example, the statement by Goshen-Gottstein that “the answer to the question
of how much mysticism existed in rabbinic Judaism depends on the interpretation of this
story” (1995:69). Schäfer does not demonstrate that the hekhalot materials fall outside the
spectrum of variants found in other manuscript traditions. In the case of other texts, con-
siderations of “orthodoxy” spurred the creation of modern normative critical editions that
obscured the vast number of variants. Perhaps the lesson from the hekhalot texts is that
many other Late Antique texts would benefit from synoptic presentation.
52. In addition, rabbinic texts are not simple reflections of first-century practices, esoteric
or otherwise. With the compilation of the Mishnah dated to the mid-third century, and its
origins back in the previous two centuries obscure, rabbinic literature generally is dated
“later” and thus, oddly enough, closer to the common dating of the hekhalot texts.
53. For a recent collection of the evidence, see Elior 1999. Whether all this literature was
produced, as she claims, by secessionist priestly circles is open to debate. Speculation about
the heavenly Temple probably occurred in many circles.
54. See Goshen-Gottstein 1995 and his careful argument that the Tosefta shows debates
over the legitimacy of ascent practices.
 Using Names, Letters, and Praise 75

that the other sections contained material going back to the first century
(1963:148). References to similar techniques and ritual practices in,
among others, Greco-Roman55 and Christian writers56 point to the middle
of the third century as a possible time frame for composition of the ascent
sections. By this time the strategy of the Songs had evolved considerably.
The ascent is part of a composite text containing material that was origi-
nally written in Palestine and later expanded in Babylonia.57 Much denser
poetic constructions than the Songs, the hymns pile up words of praise so
intricately that they are difficult to parse and translate.

King adorned, clothed in royal tribute,

Honored with the embroidery of song,
Wreathed with majesty and honor,
A wreath of sublimity and a crown of awesomeness.
Because his name is pleasing to him,
and his remembrance is sweet to him,
And his throne is adornment to him,
and his [palace] is honor to him,
And his glory is desirous to him,
and his magnificence is becoming to him,
and his strength is pleasing to him,
and his servants sweetly sing to him.

(par. 252)

Scholem’s heirs have attempted to refine his rudimentary structural analy-

sis by either further explicating the theme of heavenly praise58 or detailing

55. Parallels with the Mithras Liturgy are cited below.

56. Use of the names of the gatekeepers to travel around the heavens is mentioned
by Origen. Arnobius of Sicca Against the Nations 2.62 mentions the use of hymns in as-
cent.
57. Note the references to “the mangers of Caesarea” and “the gates of Caesarea” (par.
214). Alexander believes that the story in sections 5–6 is also of Palestinian origin
(1983b:222).
58. Grozinger expands Scholem’s point, outlining the centrality of images of song in the
hymns (1987).
76 Icons of Power

the structures of the hymns.59 The latter task is particularly difficult be-
cause there are many possible ways of arranging the short praise-phrases
and noun-pairs found throughout the hymns.60 The praise-phrases and
noun-pairs can be built up into seemingly endless kinds of larger struc-
tures based on the inclination of the editor. Some editors particularly liked
names and extended the name sections.
If it was possible to ascend with the Songs, what does the later ascent
liturgy show in terms of new strategies and notions of efficacious lan-
guage? The Songs consist of descriptions of the heavenly cult but include
no direct citations of angelic praise.61 In contrast, the hekhalot hymns are
repetitions of the exact words of the heavenly liturgy. Scholars often miss
the sheer cleverness of this strategy;62 the hymns are even more self-refer-
ential than the Qumran texts, talking about what they are doing.63 This
shift in content makes the ascent of the reciter more obvious to the reader
than in the Songs.64 While the Songs link the reciter of the liturgy with the
heavenly cult, the hekhalot hymns make him more directly part of the ac-

59. Maier, for example, delineates a distinct “hekhalot style” that includes the accumula-
tion of verbs or nouns, the extension or elaboration of series of words, and the litanies of a
noun repeated with varying adjectives (1973). The difference between accumulation, repeti-
tion, extension, and elaboration is not clear. For Maier these structural elements are a sub-
set of the general structure of rabbinic prayer, which in turn is defined as combining rhythm,
loose parallelism, and the use of construct pairs to designate nouns.
60. Prose lines that look like complete sentences alternate with and merge into shorter
praise phrases that look more poetic. Hence Alexander’s judicious conclusion that “the large
structures of the Heikhalot liturgies are formally ill-defined” (1987a:53). What coherence of
structure there is exists in the smaller units, because, as Alexander states, “rigorous symme-
try occurs only in the micro-forms” (1987a:53). See his careful review of Carmi’s attempt
(1981) to make Hekhalot Rabbati par. 159–60 into stanzas (Alexander 1987a).
61. Newsom notes the complete lack of cited praise (1985:16).
62. Martha Himmelfarb (1988, 1993) limits technique to imperative verb forms (for ex-
ample, “show the seal,” “recite” names). Imperative verb forms are not however coextensive
with technique; they stand out because they are taught by means of direct order (hence the
modality of speech) and not by example. Similarly, Schäfer finds scant evidence of ritual
techniques in the texts, stating “The hekhalot literature does not provide us any indication
as to how the heavenly journey actually is carried out, or even if it is practiced at all as a
‘truly’ ecstatic experience” (1992:155).
63. Newsom notes that the hymns do not claim that the human priests conduct or par-
ticipate in the heavenly cult (1990b:114). Thus they talk about what other beings are doing.
The closest overlap is the general action of praising.
64. And less controversial among scholars. The role of reciting praise to join the human
chorus to the heavenly chorus was noted as far back as Bloch 1893 and Altmann 1946, al-
though the mechanisms were not explained.
 Using Names, Letters, and Praise 77

tion by expanding the link to another dimension.65 The recitation of an-

gelic praise by the individual indexes him directly as a participant in the
heavenly chorus and not just a spectator of heavenly cult. The literary de-
vice of quoted speech or, in this case, quoted praise shifts the praise of the
heavenly speaker to the earthly one.66 The hymns literally embed the heav-
enly praise in the human’s mouth, making the text iconic at another
level.67
The hymns include a few instances of the standard rabbinic blessing
formula “jvdb/blessed,” as in “Blessed are you . . . ,” in this case com-
pleted by the phrase “wise in secrets.” The act of blessing the deity has
been called “the most conspicuous form of Jewish prayer” (M. Smith
1982:203). Declaring the deity blessed is what the angelic choruses do; the
appropriation of this action by humans brings heavenly efficacy to earthly
requests. The rabbinic blessing formula summarizes the specific role of the
deity addressed in the prayers, in this case the deity is “wise in secrets.”
The blessing formula includes a divine name and its recitation, and so is
restricted in use. The formula is awkward in that it includes both a third-
person description of the deity as blessed, and use of the second person
(“you”). This phrasing permits the formula both to declare that the deity
is blessed and to instantiate the deity’s presence as audience.68 While this
enigmatic formula is worth noting because of its common occurrence in
rabbinic liturgy, in this text the few appearances of the formulas are an-
cillary to the main thrust of the text.

65. In more technical linguistic terms, this move completes a trajectory toward fully re-
flexive calibration across the human and divine realm.
66. In other hekhalot texts, we are overhearing a report of a successful ascent where, due
to the automatically effective language, the didactic role and the experiential role collapse
into one (Janowitz 1989).
67. Himmelfarb stresses the similarity between the hekhalot hymns and the synagogue
liturgy in order to negate any sense that the hekhalot hymns have any ritual efficacy. This
comparison does not negate the role of the hymns in ascent, but it does remind us that we
have not explained the role and purpose of the more common liturgy. With great insight she
remarks, “The hekhalot literature is intended for an audience of potential ascenders who
need to know the right words to say to be like angels” (1988:77). We are trying to explain
how the right words make someone like an angel.
68. As Heinemann noted in his extensive discussion of the formula, “One of the hall-
marks of the liturgical berakah in the statutory prayers is the address of God in the second
person, especially in the ‘introductory clause’ . . . although the ‘main content clause’ usually
continues in the third person. . . . The style is awkward, for the liturgical berakah uses the
78 Icons of Power

In addition to fusing the reciter with the heavenly chorus, the hymns in-
clude the special power of divine names and letters with no clear distinc-
tion between the two.69 Divine names occur less often in this text than in
some of the other hekhalot texts.70 The power of the divine name is also
used in the ascent in the guise of angel names that derive from it. The
recitation of divine names has stringent taboos, as seen in the fight that
Rabbi Azariah has with Rabbi Nehunya when the latter fails to give the
names of the angels in the seventh heaven. Nehunya explains that these
names “have the name of the King of the world in them”—that is, they
end with “la/god”—and therefore Nehunya does not want to say them
aloud.71 However, the ascent is incomplete without them, so he finishes the
list. There is no successful completion of the ascent without the recitation
of the names, and no recitation of the seventh-heaven names without
completion of the ascent.
The power of names is illustrated in a slightly different fashion in para-
graph 204, where Ishmael is told that if he calls on Suria, the prince of the
countenance, and vows a complex name of the deity 112 times, “immedi-
ately he descends and has power over the chariot.” Here the names appear
to be a functional equivalent to the recitation of hymns. The section is a

word baruch (blessed) in an adjectival sense, and not as a verb. Grammatically, then, it can-
not be followed by a prepositional phrase.” In order to explain these anomalies, Heinemann
argued that the formula did not originally include the “you.” The formula is a fusing of the
form “Bless God who . . .” with “Blessed is God,” as the latter became the hallmark of bless-
ing (1977:100). He writes: “The same attributes of praise in the active participial form
which were gleaned from the Bible came to serve a number of different liturgical functions.
They were combined, as stated above, with the introductory formula ‘Baruch Atah Adonay,’
and in this manner the new pattern of the berakah was created. . . . A pattern is needed in
which the praise will be voiced in the present tense, and for this purpose the new pattern, in
which the ‘specific praise’ is phrased in the form of an active participle, is most appropriate”
(1977:91–93) (see Schlüter 1985).
69. Sometimes the text reads either “letter” or “letters” instead of a name. See Grozinger
1987:54.
70. See Schäfer 1992:140 n. 2. In the ascent hymns in Ma’aseh Merkabah the power of
the divine name more directly supplies the “fuel” for the ascent. See Janowitz 1989. Some
manuscripts (NY 8128) include more names, as if the copyist had a special preference for
this technique.
71. This is very useful and clear articulation of the basis of the pragmatic implications of
signs—that is, a metapragmatic statement. The revelation functions as an aside, and perhaps
was added at some point by an editor who believed it was necessary to teach explicitly the
principles that other parts of the text teach by example.
 Using Names, Letters, and Praise 79

conscious reflection on ascent techniques, incorporating ascent as one

more task that can be accomplished by calling on angels.72
Once in the heavenly realm, Nehunya shows seals to the gatekeepers
on the left and the right in order to pass from one level to the next.73 The
practice of using seals is referred to in several hekhalot texts.74 Sealing is
not so much an ascent technique as a protective device used to negotiate
the heavens once in the heavenly realm. The Orphians, according to Ori-
gen, show “symbols” as they pass through the levels of heaven (Against
Celsus 6.27). In the Ascension of Isaiah 10.24 the angels at the heavenly
gates demand to see a “character.” Showing a seal proves to the heavenly
beings that the individual is protected by the highest powers and permits
him to pass to the next level.75 In many cases, the seals appear to include
the divine name, incorporating yet another icon of the name into the rit-
ual.76 It is also possible that a seal consists of letters of the divine names,
as we saw with the spoken permutations of divine names.
In the ascent rituals, the term “seal” may refer to amulets worn by the
practitioner, as noted by Morton Smith (1963) and argued most recently
by Rebecca Lesses (1998:319).77 Another intriguing possibility is that seal-
ing refers to writing directly on the body.78 It is often impossible to tell

72. See the discussion in Chapter 6 about employing angels.

73. Scholem notes that the Hebrew term is parallel to the Greek term sphragis
(1965a:133). For a general review of the imagery of seals, see Fitzer 1971, and for a discus-
sion of seals in the hekhalot texts, see Lesses 1998:317–23.
74. Ma’aseh Merkabah 23, 3 Enoch 48d:5.
75. For sealing as means of gaining protection, see Merkabah Rabba, where the practi-
tioner seals himself and asks for mercy (709).
76. The Book of Jeu states, “Tell the mystery, and seal yourself with this seal, this is his
name” (GCS 13.315–21). According to rabbinic legend, God’s amulet is the “greatest” ver-
sion of the divine name—that is, the Torah.
77. In the Naasene hymn Jesus prays, “For his salvation send me, that I may go down
with the seals in my hands, traverse all the aeons, open all the mysteries, disclose all divine
beings to him, and proclaim to him the secret of the holy way—I call it knowledge” (Hip-
polytus Refutation of All Heresies 5.10). For other examples, see Lesses 1998:318.
78. In addition to the rabbinic texts cited below, a dream revelation from the Genizah in-
structs the practitioner to write divine names on his left hand (Lesses 1998:318). Bar-Ilan ar-
gues that by writing on his skin the practitioner makes himself into a slave of the deity
(1989:44). This practice is also mentioned in a Coptic spell (Lesses 1998:318 n. 212). An-
other practice of “sealing” directly on the flesh is anointing the body with oil. According to
Rudolph the protective property of the sealing is related to the paradisical olive tree, with
which the oil is equated (1983:228, 361).
80 Icons of Power

whether the seals are amulets or letters written directly on the body; very
possibly both methods were combined.79 The practice of writing divine
names directly on the body is attested in several places in rabbinic litera-
ture (Bar-Ilan 1989). Seen through the rabbinic lens, all sorts of potential
problems emerge from the practice. For example, mYoma 8:3 debates
whether a man who has written on his limbs may bathe or stand in an im-
pure place. mMaccot 3:6 states that a man is not liable for “marking his
flesh” unless he writes the name—that is, the name of God. Some rabbis
are presented as opposing the practice (tMaccot 4:15), which other texts
appear to take for granted.
Writing names on one’s body is a rich semiotic practice; the names in
themselves are powerful, and that power is transferred directly to the per-
son’s marked body.80 The human body becomes an icon of the divine body
with its named limbs. Once marked with the name, the human body looks
even more like the divine body, which as we saw in the last chapter was
understood to have names on each limb. Any special mark on the flesh
could be interpreted as a name, and thus a seal. Even the mark of circum-
cision, interpreted as the adding of a letter from the divine name to the
flesh, could seal and thus protect a man.81
The goal of the ascent in Hekhalot Rabbati is a slippery question. Nu-
merous possibilities have been suggested, including a foretaste of the heav-
enly world (Scholem 1965a:17–18), the pursuit of Torah-knowledge
(Halperin 1988:366–87), gaining the protection of the heavenly court,82
and attempting to “storm heaven” or participate in the heavenly praise and
thus achieve “unio liturgica” (Schäfer 1992:152–53). Many of the more
general explanations, such as the compensatory role of ascent in the face of
the loss of the Temple or of life under the Romans, fail to take into account
the widespread occurrence of the phenomenon in the Late Antique world.
It is easy to shift the goal of ascent by adding a new introduction to a

79. Note the ambiguity in the phrases in the Aramaic additions to Ma’aseh Merkabah
where the practitioner fasts and then seals himself, reciting the phrase “seal on my limbs”
(par. 566). Later in the text he again states, “There will be a seal on all my limbs” (par.
569).
80. The Golem was vivified when the divine name was inscribed on his forehead.
81. See tBer 6:24, jBer 9:3, bShabbat 137b, and the discussion by Wolfson 1987.
82. Alexander 1987a:77, based on Hekhalot Rabbati secs. 1.1–2.21.
 Using Names, Letters, and Praise 81

text. Whatever its original goal, the ascent in Hekhalot Rabbati has been
reedited into a new frame.83 This frame, which includes the first several
paragraphs of the text, promises that the ascent hymns will enable the re-
citer to “see the deeds of men” and to know what the future holds. With
this frame the ascent is itself redirected to new goals. The ability to work
the cosmology, going up and down at will, becomes a tool that permits the
ascender to achieve a myriad of decidedly this-worldly goals. The multi-
use idea is built into Hekhalot Rabbati when Nehunya says that knowing
the ritual is like having a ladder in one’s home.
The description of Rabbi Nehunya’s ascent shares many elements with
other Greco-Roman ascent texts. In particular, Morton Smith compared
Hekhalot Rabbati with the Mithras Liturgy, found in an early fourth-
century Greek papyrus.84 The Liturgy follows a soul on a journey through
the heavens, and the ascent in the Liturgy combines recitation of prayers
and cosmic sounds with breathing techniques and kissing amulets; the
efficacy of the rite is dependent on the combination of all these actions.
While the general strategies overlap with the hekhalot ascent, the struc-
turing of the ritual illuminates the contrasts.85
The Mithras Liturgy has a distinct tone, described by Tabor as “a jour-
ney into the self beholding of the immortal nature beyond the bounds of
mortality and Fate” (1986:93). The ascent begins with a prayer combin-
ing letter sounds with creation imagery: “First origin of my origin,
aeeioyo,86 first beginning of my beginning, ppp sss phr[e]” (4.487–88).
The prayer includes references to the building blocks of the created
world—spirit, fire, and water—elements that are located “within me” as
the reciter is identified with the cosmos. The reciter has now been “born
again in thought” (4.510), a new birth better than his first one. Positing a

83. The ascent is introduced directly by the story of an impending persecution by Rome.
Similarly, in one of the three apocalypses near the start of the text, Israel is protected from
Roman persecution by the replacement of the Caesar by Rabbi Teradyon. The story is then
placed in yet another frame, this one directed toward divination.
84. Papyrus 574 of the Bibliothèque Nationale, PGM 4.475–829.
85. For a comparison of the Mithras Liturgy with an ascent ritual in the Chaldean Ora-
cles, see Johnston 1997.
86. Because Hebrew is written without vowels, divine names and cosmic sounds are go-
ing to be strings of consonants and not vowels. When Jewish texts are written in Greek, the
same emphasis on vowels appears in them as well.
82 Icons of Power

new origin and a new beginning is a clever strategy for a text that seeks to
alter bodily existence and create an immortal person (Janowitz 1991).
Recitation of strings of letters equated with cosmic elements incorpo-
rates their transformative power into the prayers (“and the sacred spirit
may breath in me, nechthen apotou nechthin arpi eth” [4.510]). The
individual is then instructed to “draw in breath from the rays, drawing up
three times as much as you can, and you will see yourself being lifted up
and ascending to the height, so that you seem to be in mid-air” (4.539–40).
These breathing techniques, including the making of hissing and popping
sounds, enable the ascender to take off and begin his travel through the
cosmos. He is able to see “himself” as he separates his soul from his body.87
Once traveling toward the upper regions, the ascender tries to pass
himself off as a star—that is, as a natural inhabitant of the heavens
(4.570–75). The process depends in part on identification, with the indi-
vidual literally redefining himself as a cosmic being by reciting “I am a
star” (4.574). In the heavenly realm, he receives a guide to help him
(4.630–60). In this case, the gatekeeper is a young god from whom the as-
cender asks protection. More breathing techniques dot the continuing
passage upward, along with references to kissing amulets. “And at once
produce a long bellowing sound, straining your belly, that you may excite
the five senses: bellow long until the conclusions, and again kiss the
amulets and say . . .” (4.704–8). The practitioner begins to lose his sense
of self, saying, “I am passing away” (4.721), until the reciter is told,
“Now you will grow weak in soul and will not be in yourself” (4.725).
A striking parallel to this ritual strategy is found in the Nag Hammadi
tractate Marsanes, where an ascent is achieved by a combination of recita-
tion of hymns, silences, the invocation of names, and the recitation of
strings of vowels (Pearson 1984). The text combines for us the use of vow-
els with some of the philosophy of language articulated in the Neopla-
tonic texts, with vowels said to be the “shapes” of the souls (25*1–26*12,
27*23–30*2). In particular, spherical shapes are associated with the seven
vowels (25*1–26*12).88

87. The parallel bi-locations in Hekhalot Rabbati are noted above.

88. Pearson relates this to the speculation in Timaeus 35A–36A (1981). He also notes the
work of Dionysius Thrax as discussed by Böhlig and Wisse 1975:16–17.
 Using Names, Letters, and Praise 83

The numerous instructions in the Mithras Liturgy result in the more

tentative mood of the text than in Hekhalot Rabbati. They introduce ac-
tions that are referred to (draw in, produce), and mention props that must
be correctly used (amulets). Teaching is done by talking about action, be-
cause the actions are not only verbal. The ritual therefore has less of the
“automatic” quality of Hekhalot Rabbati. The phrasing of the text is not
the tight didactic structure of Hekhalot Rabbati in which Rabbi Nehunya
simultaneously teaches and ascends. The combinations of types of actions
employed in the Mithras Liturgy preclude such a didactic structure. In ad-
dition to the need to include instructions, the constant shift between first
and second person sets up a distinction between the instructions to say a
prayer (You say X) and the content of the prayers (in the first person). For
example, specific heavenly greetings for each constellation are spoken in
the first person and interlaced with strings of vowels.89 The numerous im-
perative forms, “Draw up breath” and so on, also distinguish the person
teaching from the one attempting the ascent. Again this points to a differ-
ent didactic structure, in which the teaching of the ritual can be separated
from its use. The hekhalot texts, at least in terms of the ascent materials,90
are characterized by the collapsing of teaching into experiencing, a struc-
ture that is inseparable from the internal theory of language with its auto-
matic efficacy.
The Mithras Liturgy is more explicit about its goal than Hekhalot Rab-
bati, stating directly that the process confers immortality. The text begins
with a direct request: “For an only child, I request immortality” (4.476).
The invocation includes a call to “give me over to immortal birth”
(4.501), and the hailing formula declares that the reciter has become im-
mortal (4.647).91 Quite an astonishing array of rituals were thought to
confer divinity, and ascent was one of them.92 Nock considered the

89. Other letter strings are made up entirely of vowels, a technique that is absent from the
hekhalot texts because Hebrew is written without vowels.
90. The hekhalot texts include a variety of materials beyond ascents. For example, the
Prince of the Torah (Sar Ha-Torah) rituals attempt to manifest this particular divine figure in
order to gain his assistance. The structure of these rituals is distinct, with explicit instruc-
tions about fasting and other actions.
91. See the remark by Proclus that ascent brings immortality (Commentary on the Re-
public 1.152.10).
92. See Nock 1966:54 nn. 70–71.
84 Icons of Power

Liturgy a ritual means for achieving divinity while still alive, as opposed
to the notion that divinity is regained only after death.93
As a ritual geared toward making someone immortal, the Mithras
Liturgy was probably originally meant to effect a one-time transforma-
tion. The ritual has been reedited and directed toward divination, much as
the Shi’ur Komah materials were reedited. Like the ascent in Hekhalot
Rabbati, this ascent also has additions that appear before the ascent, and
appended instructions for supplemental rituals.94
Whatever the purpose, ascent was believed to be achieved by means of
identification with the heavenly world, and in particular by means of the
repetition of heavenly sounds. In the tremendously plastic rituals of ascent
we see the process by which humans, either for a moment or for eternity,
become something other than their regular selves. Learning to “talk” like
the heavens produces a set of audible signs that point back to the speaker;
he is now someone who has changed his status and can operate both on
earth and in heaven. We should not be fooled by the stereotypical content
of the words he utters, for they are more powerful than any and all
earthly words.

93. For recovering divinity after death, Nock cites Orphic inscriptions and the gold leaf
found in tombs.
94. An introductory reference to herbs and spices is usually considered an addition. The
end of the text includes instructions for a scarab ceremony and for making amulets, with
two additional compositions appended.
 6
Combining Words and Deeds:
Angelic Imprecations in
The Book of Secrets

T
he Book of Secrets is the classic example for modern scholars of
Late Antique Jewish “magic.”1 The word “magic” does not occur
in the text, though English translations sometimes insert the
word. The text is a collection of approximately twenty-nine recipes3 for a
2

miscellany of tasks ranging from how to make a wall collapse to how to

question a ghost (see Table 1). The collection has the form of a handbook,
similar to the ritual handbooks found in the Greek papyri.4 In this collec-
tion serious requests for healing are combined with the seemingly comic,
such as one recipe promising to fill a room up with smoke in order to im-
press friends. All the tasks are accomplished with the help of angels ob-
tained by means of imprecations and complex recipes.5

1. Modern scholars who label the text “magic” include most scholars who write about
this text: Schäfer (1990), Morgan (1983), Margalioth (1966), Niggemeyer (1975), Dan
(1967–68), Gruenwald (1980:225–34), Maier (1968a), and Swartz (1990). The only excep-
tions I have found are Merchavya (1967), who stresses its “mystical” aspect, and Kasher
(1967), who argues, using rabbinic criteria, that the text is theoretical and not strictly mag-
ical. The Book of Secrets is cited by formula number followed by the number of the heaven
and line number in the Morgan translation (1983). Thus, #28 (5.15) refers to formula #28
from Table 1, which appears in the fifth heaven beginning on line 15.
2. See, for example, Morgan’s translation of “qsi/practice” as “practice magic,” and
“hwim/occurrence” as “magical rite” (1983:21).
3. Several of the recipes are composites that can be counted singly or as a cluster of for-
mulas. Similarly, in the Greek Magical Papyri (PGM) a series of partial recipes is combined
so that it is difficult to tell how many separate rituals are involved.
4. See, for example, PGM, nos. 1, 3, 4, 5, 7, 12, 13, and 36. Morton Smith differentiates
these collections from individual formulas either written for specific persons or with no ref-
erences to specific people (1979:129 n. 4).
5. Morton Smith estimates that 70 percent of the PGM papyri concern the use of such
helpers and projects they can complete (1986:68).
86 Icons of Power

In order to understand these recipes and their perceived efficacy, we

need to look beyond the techniques examined thus far, though some ele-
ments will be familiar. The recipes combine the recitation of verbal for-
mulas, including divine names, with a series of actions, some familiar and
some rather exotic. Most important for this analysis, we shall see how the
use of words and objects establish exact models of the goals toward which
the rituals are directed, encoding in them the purpose of the ritual. Just as
we found that words can “do things,” the combined use of objects is re-
lated to culturally specific notions of effective actions.
The Book of Secrets6 is written in a very literate Hebrew with a few in-
stances of transliterated Greek. The modern editor Margalioth reconstructed
the text from five main manuscripts, supplemented by readings from genizah
fragments, Latin and Arabic manuscripts.7 This fact has been used to chal-
lenge the integrity of the text and to argue that it is simply the construction
of its modern editor. Just as in the case of the hekhalot texts, the controver-
sial nature of the material has led to unusually close scrutiny of the textual
traditions. This question must be viewed within the standards for other
texts. The textual history is no more daunting than that of other Late An-
tique manuals, such as, for example, Apicius’ cookbook (Brandt 1927).
In the case of the Book of Secrets we lack the paleographical evidence
helpful in dating similar Greek texts (Morgan 1983:8). Dating the text is
based in part on a single reference to Roman indiction in section 1.27–28
(“the fifteen year cycle of the reckoning of the Greek kings”). This system
was instituted in the year 312 (Margalioth 1966:24–25) but was not used in
nonfiscal contexts until the second half of the fourth century.8 The single ref-
erence can be supplemented by general comparisons with Greek papyri that
point to a date from the early third century to the late fourth century.

6. This may not have been its original title. See Margalioth 1966:56–62 and Merchavya
1971:1594, who posits that the book may have been named after the angel Raziel or Noah.
7. For the manuscript evidence, see Margalioth (1966:47–55) and Morgan (1983:2–6).
For a depiction of the relationship between the various sources, see the chart in Niggemeyer
(1975) between pages 18 and 19.
8. Alexander (1986:348 n. 15) supplements Margalioth’s discussion, adding that the in-
diction method of counting years began in 312, with a five-year indiction in Egypt perhaps
as early as 287. However, Alexander notes that the use of this method in “non-fiscal con-
texts” did not begin until the second half of the fourth century (following Bickerman
1980:78). If the text is unitary, this date helps place the entire composition.
 Combining Words and Deeds 87

The text is bound with merkabah/hekhalot (chariot/palace) treatises in

five of the six main manuscript copies (Ma’aseh Hekhalot, Ma’aseh
Bereshit, and Shi’ur Komah) and shares many of their cosmological con-
ceptions, such as descriptions of the heavenly realm as the world of fiery
angels who recite endless praise of the deity.9 The text is structured in
seven distinct sections, each one corresponding with a heavenly level and
each containing a number of imprecations.10 The heavenly framework
enumerates the names of the angels in that heaven, their mode of organi-
zation, the tasks over which they are appointed, and how they can be em-
ployed in the various tasks. The seventh and final section of the
framework, the seventh heaven, consists entirely of extensive hymns sim-
ilar to the hekhalot hymns.
The formulas appear in the heavens in no clear progression, other than
a steady decrease in the number of recipes per heaven.11 This is true of the
parallel Greek ritual handbooks (PGM) as well, where the organization of
recipes is not clear. The range of the ritual goals is striking, combining
very serious rituals with seemingly comic ones.12 Both these aspects of the
text are also familiar from parallel Greek texts.13

9. The section on the seventh heaven in the Book of Secrets appears at the end of Sefer
HaQommah in one manuscript. See Morgan 1983:2 and Margalioth 1966:xv.
10. Morgan and other commentators view the recipes as a composition distinct from the
heavenly framework, with disagreement about which is earlier. Morgan (1983:9, 27) notes
that the recipes lack the “flowery descriptive wording and the biblical quotations of the cos-
mological framework” and posits that the recipes were composed first and the cosmological
framework added later. Merchavya argues the reverse (1971:1594–95). However, distinct
wordings for the adjurations and the hymns does not mean they were written by different
authors. Alexander points out that the angel names occur both in the recipes and in the de-
scriptions of the heavens, forming an important link between the sections (1986:347–48).
The recipes also refer directly to the framework, with this interweaving making it difficult to
draw a distinct line between frame and recipe.
11. Healing appears in different sections; #1 (1.29) is a general healing formula, #17
(2.95) is for stroke, and #23 (2.182) is for curing a headache or blindness. Both #3 (1.94)
and #28 (5.15) are for knowing about the future.
12. Compare the recipe for making a room fill up with smoke to impress friends (#26)
with the recipe for making men appear to have donkey snouts (PGM 11b:1–5).
13. As a few examples, compare the ritual for binding lovers (5.13) with PGM
4:296–466, knowing the future (#28) with some of the many divination rituals such as PGM
4:3210 and 7:540–78, the prayers to Helios (#27B with PGM 4:247), and the making of
amulets (#18), with an amulet for a pregnant woman PGM 4:80.
Table 1. Outline of the Book of Secrets
88

No. Placement/Purpose Action / (Ma’aseh) Utterances

First Firmament
1st Encampment:
1 Heal (29) Burn incense (S) Angel names, request formula
2nd Encampment:
2 Against enemy, creditor etc. Gather water in vessels; Request formulas
(composite) (29) smash vessels (A)
3rd Encampment:
3 Know future (94) Put written slips in oil Adjuration (sun)
4th Encampment:
4A Influence king (117) Take heart of lion cub; write Angel names, adjuration to Aphrodite
angel names with blood/incense (S)
4B Enter presence of king (132) Anoint self with 4A mixture; Request formula
hide lion’s heart (S)
5 Bind yourself to great woman (143) Take your sweat; bury flask under Adjuration
Icons of Power

her doorstep (A)

5th Encampment:
6A To speak to moon/stars (161) Kill cock, make cakes, write Angel names,
angel names (S) adjuration or formula
6B Kindness to you (170) Put cakes in wine (S) Angel names, adjuration
7 Question ghost (175) Face tomb, hold oil and honey (S) Angel names, adjuration to
Hermes, release formula
8 Speak with spirits (187) Go to place of killed Adjuration (singsong)
6th Encampment:
9 (Catch a fugitive) (202) Write his name (A) Request formula
7th Encampment:
10 Answer king (217) Go to riverbank, incense (S) Angel names, adjuration,
release formulas

Second Firmament

11 Ask anything of angels in Abstain ———

second firmament (6)
12 Silence powerful people (18) Ashes from idol bread offering (S/A) Angel names backward
Request formulas
13 Make woman love man (30) Lamellae with angel names Request formula put in furnace
or her bath
14 Nullify great man’s intention to you (46) Stand at midnight Angel name, request formula
15 Give enemy trouble sleeping (62) Head of black dog, amulet with Request formula
angel names (S) hide head-with-amulet near him
(release: remove and burn)
16 Light an oven in the cold (81) Write angels names on sulfur lump, Adjuration
put in oven (A)
17 Heal a man with stroke (95) Apply oil, honey to him, burn incense, Patient’s name Angel names
amulet (S/A)
18 Expel lion, etc. (111) Make image, amulet, bury/place it (A)
19 For childbirth (124) Amulets (on house, woman) ———
20 Protect man in war (134) Write angel names on leaves, put in oil, ———
Combining Words and Deeds

anoint him, amulet (S/A)

21 Rescue friend from bad judgment (145) Purify self, amulet Request formula
22 Restore to office (161) Oil, honey, flour in vial, make cake, Angel names Adjuration (to moon)
he eats it (S/A)
23 Cure headache/blindness (182) Brain of black ox, write, angel names, ———
amulet, abstain (S)
89
 90
Third Firmament

24 Extinguish bathhouse fire (17) Put salamander in jar with oil (A) Adjuration (Salamander)
25 Race horses (36) Hide amulet in track Adjuration
26 Give proof of your powers, Burn plant (S) Adjuration (reverse: adjuration back-
to fill room with smoke (47) ward with addition)

Fourth Firmament

27A View sun during day, be pure (25) Abstain, incense Angel names, adjuration
27B See sun during night (43) Purify, abstain, wear white, bow Name of sun, angel names, adjuration,
prayer to Helios (release: adjuration)

Fifth Firmament

28 Know future (15) Put lamellae on oil Adjuration 7 times

Sixth Firmament

29 For journey or to war, Make iron ring, lamellae with angel Angel names, adjuration
Icons of Power

to flee, etc. (composite) (25) names in ring, put in mouth (A)

S = modified sacrifice
A = analogical action
 Combining Words and Deeds 91

Verbal Formulas

The majority of the verbal formulas in the first six heavens are adjura-
tions14 with first-person singular forms of “iybwm/adjure, swear” (Table 1,
column 3).15 Examples include “I adjure you angels of wrath and destruc-
tion (1.70) and “I adjure you O sun that shines on the earth (1.99). This
verb parallels closely the Greek verb “∏rkºjv/swear,” found throughout
Greek ritual texts.16 The range of adjurations—Greek, Hebrew, Coptic,
and Demotic—reflects the widespread trope that humans can gain angelic
assistants.17 A major complex of these rituals occurs in the hekhalot texts,
many of the rituals directed at gaining help from the Prince of the Torah
(hrVth r„).18
Angelic adjurations are a subcategory of the much more widespread
social practice of swearing and adjuring. Oaths were most familiar in the
Greco-Roman world, from the area of law. An oath would be taken when
signing a treaty, setting up a commercial contract or a personal contract,
for marriages, and for swearing in officials of the government and sol-

14. No. 2, 1.68, 1.70; #3, 1.98; #4, 1.126; #5, 1.146; #6, 1.172; #7, 1.179; #8, 1.189;
#10, 1.226; #16, 2.83; #22, 2.166; #24, 3.20, 3.28; #25, 3.38; #26, 3.52, #27A, 4.30, 4.41;
#27B, 4.68; #28, 5.23; #29, 6.35.
15. The only extended stylistic analysis of the formulas is Niggemeyer’s work (1975),
which is limited by its formulaic distinction between “prayers” and “magical language.” His
analysis defines a prayer as a petition plus a request, while a “swearing” is a petition plus a
magical binding formula. To no surprise, he then finds that the addition of a “swearing for-
mula” makes a “prayer” into “magical language.” If the distinctions (prayers, swearing for-
mula) are simply definitional, then the analysis is completely tautological. Otherwise the
import of this terminology is not made clear. Niggemeyer finds a few cases where names ap-
pear in what otherwise should, according to his nomenclature, be a prayer. These usages are
explained as being prayer on its way to magical language, another conclusion completely de-
termined by the initial organization of the terms. In his final section, Niggemeyer lists twelve
elements of magical speech, ranging from the use of specific types of words such as names,
to vague categories, such as “exactitude of language” and “modalities.” He does not explain
the functions of the elements. Perhaps Niggemeyer is hindered in general by a hesitancy to
see the use of hostile language as part of religious practice. The imprecations try to convince
angels to harm people, and thus do not fit easily into everyone’s picture of proper religious
language. See, for example, Alan Segal’s comment about the use of curses tablets: “No one
would have practiced it with the impression he was practicing a legal and wholesome reli-
gious rite, however richly deserved was the damage to the intended victim” (1981:88–97).
16. See, for example, the numerous instances in the PGM, such as 1.134.
17. See the extended discussion in Lesses 1998:279–365.
18. As Michael Swartz has argued, these rituals are distinct from the ascent rites and may
be later additions (1996).
92 Icons of Power

diers, and even by doctors (Hippocratic Oath). A wide variety of deities

could be invoked in these oaths, with Greeks often referring to the triad
of sky, earth, and sea gods, and Romans swearing by Jupiter or simply by
“all the gods” (Dillon 1996:1057).
This practice was also common among Late Antique Jews (Lieberman
1974: esp. 24). For example the Hebrew phrase “I adjure you and may it
come upon you” combines an oath and a curse (tSota 2:1; bSota 18a). Ad-
jurations were made invoking the deity, as in the phrase “by He who es-
tablished the world on three pillars” (Song of Songs Rabba 7.8)19 or
directly by the divine name.20 Such adjurations were often employed in a
time of trouble or stress.
In addition to the adjurations, ritual formulas in the Book of Secrets
contain a variety of other verbs—for example, “I ask you angels . . .”
(#12, 2.21–23) and “I charge you angels . . .” (#9, 1.203) (labeled “re-
quest formulas” in Table 1, column 3).21 Almost all these formulas pre-
serve the first-person verb form22 and are addressed to a specific group of
angels, such as “angels of fury,” or to individual angels, such as the sun
(#27B, 4.61). The ancient editor made no clear distinction between the
adjurations and the request formulas. For example, the goal of speaking
to the moon or stars can be reached using either type (#6A, 1.165 and
1.167).23 Given the preponderance of adjurations, the other formulas ap-
pear to assimilate as a subclass of adjurations.
All these verbal formulas reflect a distinct mode of language function.
Working with angels involves a search for the best authority to bring to
bear on the subject, and the approach used here, which presumably res-

19. See also Song of Songs Rabba 7.9 and Lieberman 1965a:107–8; 1974:24.
20. For an example of an adjuration made “by the Name” transliterated into Greek, see
PGM 2.110, and see tNedarim 1:1, which states that “in the name” is an oath. Alon
1950:33 and Albeck 1952:470.
21. See “I ask you angels” (#13, 2.33), “I request you, great angel” (#21, 2.147), and “I
present my supplication before you” (#27B, 4.64). In a few cases, the speaker states that he
is turning a certain person over to the angels so that they will punish him. “I deliver to you
. . . so that you will strangle . . .” (#2, 1.64).
22. Slight variations are found in “You angels . . . let the terror of me be over . . .” (#4B,
1.136), “Moon moon, moon . . . bring my words before the angels” (#6A, 1.162), and one
case of an angel name followed by “accept from my hand what I throw to you” (#2, 1.57).
23. Similarly, in the PGM adjuration formulas are mixed in with formulas that use other
nonadjuring verbs.
 Combining Words and Deeds 93

onates well with our culture, is to “threaten to sue.” The thrust of the for-
mulas is to put the angels in a situation in which they cannot refuse to help
the person making the request. They can no more pretend that they did
not hear the request than someone can ignore a summons.
Angels and daimons were thought to exert far-reaching influence on
daily life, and life was lived keeping one eye on their activities.24 Complex
social networks of responsibilities and obligations extended from the hu-
man world to the supernatural so that angels could be addressed using the
same linguistic forms as humans. In the Book of Secrets, the speaker ad-
dresses an angel with an oath cast as forcefully as possible in the first per-
son, the use of the second-person addressee indexing the presence of the
angel as intended audience/receiver. The practitioner must know the
names of angels and their taxonomy (which angel does what and works
for whom), all outlined extensively in the text.
Adjuring angels as a legal mode of discourse suggests the degree to
which it was a socially recognized activity. We are in the realm of very spe-
cialized ways of talking that are calibrated with social roles. The person
using the adjuration is presumably at the same social level as the angels he
adjures. The adjurations also incorporate the power of divine names, be-
cause angel names include the letters “el/god.”
An illustrative use of the legal adjuration appears in a rabbinic story
about Rabbi Joshua ben Levi adjuring the angel of death (bKetubot
77b).25 When the angel gives him a tour and permits him a premature
glimpse of paradise, the rabbi jumps down into paradise and refuses to go
back and die. He adjures the angel so that angel will not be able to take
him back. As always in rabbinic stories, the point of the story is about
something else (in this case the importance of fulfilling one’s vows); the
reference to the adjuring of the angel is off-hand and without controversy.
The context here is of a struggle between near equals, a battle of wits and
words in which the rabbi can overcome the natural order (die first, then
paradise) if his adjuration works.26

24. For the widespread belief in daimons, see, among others, M. Smith 1978:126–29,
202–5, and Neusner 1966–70, 4:334–441, 440; 5:183–86, 217–43.
25. In some versions it is R. Johanan. See Lieberman 1974:24.
26. Typically rabbinic, another discussion of oaths states that an adjuration is binding
only if the one adjured accepts it (bShevuot 29b).
94 Icons of Power

The use of adjurations contrasts with other ritual strategies for influ-
encing lower-level supernatural beings, such as use of imperatives or even
having a completely distinct “daimon” language. The latter is found, for
example, in the “language of the demons” of mantras, which Tambiah ar-
gues “is muttered by the exorcist and . . . is not meant to be heard, for it
constitutes secret knowledge” (1985:19). Instead of conveying informa-
tion, the words “connote power” (1985:20).
In the Book of Secrets, the only formula with a special intonation is the
one addressed to spirits who dwell in a place where people are killed. Such
places were believed to be especially replete with spirits of people who ei-
ther died before their time or were killed violently. These spirits were
thought to linger near their graves, and thus be accessible to the living
(Waszink 1950). The officiant is told to speak in a “singsong” voice, a
tone that may be related to the common notion that supernatural spirits
whisper (tShabbat 7.23). Other than that, neither special languages nor
special intonation looms large in the Book of Secrets; instead, the primary
mode of talking with an angel employs legal language.
The relentless focus of the recipes is not on the person being healed or
subject to attack; the formulas are all addressed to angelic helpers. The
formulas have to fulfill two purposes at the same time; they must be per-
suasive to the angel and they must convey sufficient information about the
goal to direct the angelic power. As such, they are complex mediating for-
mulas, mediating both between the human speaker and the angel and be-
tween the human speaker and his goal. As such they also differ from
Greek binding spells, formulas inscribed on small sheets of metal and
placed in tombs, chthonic sanctuaries, and, later, near bodies of water or
even in hippodromes (Faraone 1991). These spells have goals (personal
and financial success) that are similar to the recipes in the Book of Secrets
but differ in their specific formulation: binding spells use direct formula-
tions, such as “I bind NN,” which do not occur in this text due to its re-
lentless focus on getting angels to do the work.
Once in the seventh heaven, the officiant recites hymns. The recipes in
the Greek Magical Papyri (PGM) also include numerous hymns, some of
which are cast in archaic language (PGM 3:215). The hymns in the Book
of Secrets praise the God who dwells on his exalted throne, concluding
with some typical praise:
 Combining Words and Deeds 95

Blessed be his Name alone on his throne

And blessed in dwelling places of His majesty
Blessed be His name in the mouth of all living
And blessed in the song of every creature
Blessed be the Lord forever.
(7.39)

The distinction between the modes of discourse in the lower heavens and
in the highest heaven is based on a clear hierarchy of language—that is, the
mode of speaking relates to the status of the addressee (a standard soci-
olinguistic observation). All cultures have a range of ways of making re-
quests, from the direct command “scalpel” to the maximally polite request
to close a window by stating “My, it’s cold in here” (Ervin-Tripp 1976).
Similarly, for practitioners of all stripes the plethora of supernatural
powers inhabiting the cosmos demanded an etiquette that took into ac-
count their hierarchy; it was necessary to act appropriately with each type
of supernatural figure. Adjurations were never to be used in relation to the
main deity. Attacks could be made against any ritual that appeared to vi-
olate the expected etiquette. Raising one’s voice, for example, or talking
too directly to a deity was not considered proper etiquette. Prayer is bet-
ter than other forms of discourse because it is understood to be a more po-
lite form of address.27 A raised voice should be reserved for lower-level
spirits. Iamblichus, for example, explains to Porphyry that the Egyptians
do not threaten their gods, only daimons (On the Mysteries 7.6).28 On the
other hand, adverse comments were made about being too sycophantic in
relationship to the deity, as in the rabbinic texts that distinguish between
those who flatter their deity too much (pagans) and the superior language
of Jewish prayer (Lieberman 1974:25).
In the Book of Secrets, angelic adjurations are limited to the lower six
heavens. At the top level there is no ordering of angels and no need to pro-
cure various artifacts. Striving to see the deity demands a higher-brow
mode of action than interacting with the lower powers. At the highest

27. See Irenaeus’ statement that Christians use only pure, sincere prayer, not incantations
(Against the Heresies 2.32.5).
28. Some modern scholars have reified Iamblichus’ distinctions, which are based on eti-
quette, into substantive categories (religious versus requests versus magical threats).
96 Icons of Power

level the only language used is hymns, complete with their distinct prac-
tice of declaring the deity blessed. As noted in Chapter 5, the meaning and
purpose of blessing the deity is not immediately clear.29 As we saw in the
ascent texts, the hymns in the seventh heaven are reported speech—that is,
the speaker reports the speech of the biblical angelic chorus that declares
the blessed status of the deity. Humans thus recycle the cosmic praise and
mark their status as participants in the seventh heaven.
The fact that this level has no adjuration does not mean that no de-
mands are made or that nothing is wanted from the deity.30 In other
words, the right to address someone respectfully can sometimes indicate
certain expectations of the addressee. A recent study of politeness in
Samoan respect language points out that use of respect language is not
just a way of being nice, but also places demands on the person to whom
the polite language is addressed. Duranti explains: “Respect is not only
given in exchange for something (for example, request, imposition of var-
ious kinds), it is also a pragmatic force that coerces certain behaviors or
actions upon people and thus indexes speakers’ control over addresses
rather than addressees’ ‘freedom’ of action” (1992:96).31 To declare one’s
deity blessed using the most polite forms may simply be the nicest way
possible to make a request.32 Being able to use the most polite form in the
highest level also reflects the elevated status of the speaker.

What Is a Ma’aseh? Modified Sacrifices and Analogies

29. Reif comments that blessing God “counts as another form of Jewish worship that
continued from earlier times into the axial age” and posits that it is basically a mode of giv-
ing thanks (1993:60). See the comments on the berakah formulas above, in Chapter 6.
30. Recently Jon Levenson hints in a footnote that biblical prayers may be indirect re-
quests. “There is room to wonder, however, whether those affirmations of God’s continual
solicitude, and protection of his loyal worshiper too, are not often theurgic rather than static
and simply descriptive. My suspicion is that they often constitute a hopeful pledge of alle-
giance to YHWH’s ideal reputation, rather than the Pollyannaish reporting of empirical fact
that they seem to be” (1994:xxviii n. 12).
31. See the Concluding Note, at the end of this book, for further discussion of modes of
requests.
32. In linguistic terms, this is a typology of demands with markedness proportional to
register.
 Combining Words and Deeds 97

The Book of Secrets does not contain any rituals that consist entirely of
the recitation of words.33 All the recipes combine verbal formulas with the
use of objects; the composite recipes are called “hwim (act/deed).” The
use of the Hebrew term “act” is very close to the Greek “pr˙xiq (ac-
tion/doing),” translated as “rite” in the recent English edition of the
Greek ritual handbooks (Betz 1986).
Rabbinic texts mention a category of “men of deed,” the most famous
of whom is Honi the Circle-Drawer.34 Honi was able to bring rain by
standing inside a circle he drew and using an adjuration. His behavior is
presented as being very audacious in the rabbinic texts. The combination
of adjuration and standing inside a circle is similar to the rites in The
Book of Secrets.
Many of the “rites” in this book involve more complicated actions
than attributed to Honi. The officiant is told to procure and employ a va-
riety of objects for use with the verbal recitations (Table 1, column 2). For
example, the first recipe in the first heaven combines reciting a formula
while burning incense. Objects needed for the various recipes range from
the common (water, incense, and oil) to the more exotic (a lion-cub heart,
the head of a black dog, the brain of a black ox). These objects are no
more obscure than those employed in the parallel Greek rituals, one of
which demands “fat of dappled goat, embryo of a dog and bloody dis-
charge of a virgin dead untimely” (4.2645–46).35
Just as the verbal formulas include a number of persuasive devices, the
ritual use of these objects is multilayered. The actions appear to fall into
two categories: some are modified sacrifices that include, for example, the
use of incense and blood, while others appear to be constructed as analo-
gies to the desired outcomes of the rituals.36
Looking at the former category first, some of the recipes include burn-
ing incense, making cakes, applying blood, and anointing with oil. These
33. There are a few examples of the reverse, where letters or words are written down and
objects are manipulated with no oral recitation. These will be discussed below.
34. See mTa’anit 3:8, bTa’anit 19a, 23 ab, mSotah 9:5, Genesis Rabba 5.5, 13.7, and
Bokser 1985:42.
35. The presence of these exotic items is part of the reason Merchavya (1966–67) con-
siders the text a purely theoretical exercise.
36. Modified sacrifices are marked (S) and analogical actions are marked (A) in the sec-
ond column of Table 1.
98 Icons of Power

actions are reminiscent of Temple and pre-Temple sacrifices but quite dis-
tinct in their aesthetics. It has long been noted that even though animal
sacrifice was an ancient practice in the Mediterranean basin, attitudes to-
ward the practice shifted and changed over time. Explanations for this
shift remain elusive. A general revulsion against animal sacrifice was in
full swing by the first century b.c.e. and caused numerous traditional rit-
uals to be reworked and reinterpreted.37 This critique does not belong to
any particular author or even a religious tradition, but shaped attitudes
on a more widespread basis.
Describing this change in attitude demands some delicacy, since the
emerging critique of sacrifices was often couched in ethicizing language.
The rejection of animal sacrifice was presented as an improvement be-
cause killing animals, at least as part of cultic practices, was considered to
be of questionable moral status. Ethicizing language was so pervasively
employed for this development that it is difficult to see it as anything other
than the moral evolution of religion (from magical animal sacrifices to
purer internal sacrifices).
If we set aside the moralizing discourse, we can catch glimpses of
changing ideas about what is appropriate or pleasing to divinities. Blood
sacrifices seemed to offend against an emerging set of sensibilities accord-
ing to which deity/deities were no longer thought to “eat.” These shifting
aesthetics were part of the impetus for Jews to turn to a heavenly cult and
to the modified sacrifices discussed above.38
The standard view that prayer replaced sacrifice after the destruction of
the Temple is an oversimplification that even rabbinic texts reject.39 Mod-
ified sacrifices were part of the widespread Late Antique practice of rein-
terpreting ancient animal sacrifice traditions.40 The modified sacrifices
appear to fulfill the gamut of purposes that animal sacrifices did, includ-

37. See Wenschkewitz 1932 and Dahl 1941.

38. Against these shifting contours of thought, some individuals, such as the emperor Ju-
lian, supported animal sacrifices, including the plan to rebuild the Temple in Jerusalem.
Given the tremendous shift away from animal sacrifice, these stances must have appeared
quaint, if not downright backward, in other people’s eyes. Because Julian frequently em-
ployed the term “theurgy” for his favorite rituals, his opponents may have seen this as evi-
dence that “theurgy” referred to backward religious practices.
39. See the survey of rabbinic “histories” of sacrifices in Guttman 1967.
 Combining Words and Deeds 99

ing purification and appeasing or inducing supernatural figures to fulfill

specific requests. For example, Armand Delatte compiled a rich collection
of modified sacrifices performed when collecting harvests and simple herb
gatherings (1961:148–63). Prayers were said, and then offerings of honey,
salt, grains, and even coins were made.
Central to these practices is the idea that sacrifice was directed not at
the main deity but at lower-level figures, such as daimons and angels. This
was the culmination of generations of thought about the meaning and
function of ancient sacrifices. The notion that a deity might need or enjoy
a sacrifice seemed outdated and aesthetically displeasing already with the
writing of the final books of the Hebrew Scriptures.
One part of these reworkings of ancient sacrificial traditions was to ar-
gue that angels and daimons were nourished by the offerings. The Late
Antique trope that daimons feed on sacrifices was “universal” (Chadwick
1965:146 n.1). Porphyry saw this as one of the reasons to reject animal
sacrifice and adopt a vegetarian lifestyle. He claimed that daimons “re-
joice in libations, and the savour of sacrifices, through which their pneu-
matic vehicle is fattened; for this vehicle lives through vapours and
exhalations” (On Abstinence 2.42).41
Origen repeats many of the Late Antique clichés that daimons are “riv-
eted” to “burnt-offering” and blood (Against Celsus 8.62) and “delight in
frankincense and blood and the odors rising from burnt sacrifices”
(4.32).42 In fact, daimons desire sacrifices so much that they will try to
steal the sacrifices designated for higher powers.43 Diverse thinkers agree
on this point. Origen shares it with his opponent Celsus as well as with
many other groups, such as the Pythagoreans (3.28).44 At one point, Ori-
gen discusses a figure called Aristeas who seemed uncomfortably similar

40. J. P. Brown (1979, 1980) comments on the rejection of sacrifice but does not say much
about its reinterpretation. He does point to the important critique of sacrifice preserved in
Porphyry On Abstinence 2.20. The alchemical practices, discussed in the next chapter, are
another form of modified sacrifice.
41. See Philo On the Decalogue 74 and Oenomaus ap. Eusebius Preparation for the
Gospels 5:21.
42. See Exhortation to Martyrdom 45. The specific reference to offerings here is from
Homer Iliad 4.49, 9.50.
43. Origen also argues that evil powers were opposed to Jesus because he robbed them of
their sacrifices (Commentary on Matthew 13.23).
100 Icons of Power

to Jesus. According to Origen, Aristeas is in truth not a god but merely a

daimon who obtained the reward of “drink-offering and burnt offering”
(3.28).
Daimons are not the only supernatural beings who like sacrifices. An-
gels also feed off sacrifices, according to Origen. “Just as the demons, sit-
ting by the altars of the Gentiles, used to feed on the steam of sacrifices, so
also the angels, allured by the blood of the victims that Israel offered as
symbols of spiritual things, and by the smoke of the incense, used to dwell
near the altars and to be nourished on food of this sort” (On First Princi-
ples 1.8.1). This stunning statement of Origen’s was almost certainly bor-
rowed from Jewish reinterpretations of sacrifice traditions. The idea that
sacrifices are “symbols of spiritual things” no doubt circulated in Jewish
circles, which, as Yitzhak Baer pointed out, were among the early devel-
opers of such notions (1975). These reinterpretations were then projected
back onto the founders of the sacrificial system. For example, in Against
Celsus Origen explains that Moses knew about the connections between
daimons and specific animals (4.92–93).
The Book of Secrets explicitly uses these ideas by claiming that it
teaches the proper libations for the angels along with their names (1.9).
The well-documented shift away from temple settings meant that these li-
bations and other modified sacrifices could take place in a wide variety of
locales.45 In the Book of Secrets the locale for a specific rite is usually un-
specified. In one instance the officiant is told to go to the spot where peo-
ple have been killed (#8, 1.187), a standard locale for finding supernatural
beings.46 In the vast majority of cases the use of the sacrificial objects and
the recitation of the formulas establish the site as a place where the offi-
ciants and supernatural helpers interact.
Some of the recipes employ objects in ways that do not seem to have
any connection with sacrificial traditions but that instead operate along
the lines of analogies—that is, words are reinforced by actions that spell

44. Origen cites a Pythagorean on this topic (Against Celsus 7.6) as well as Celsus
(Against Celsus 8.60).
45. For a general discussion of the shift away from temples, see J. Z. Smith 1978:
172–89.
46. Supernatural spirits were thought to live near or in sources of water.
 Combining Words and Deeds 101

out the charge to the angels. The officiant, for example, smashes vessels to
show exactly how the angel should smash an enemy (#2, 1.48).
Scholars have spilled more ink about the role of analogical thought and
action than about any other topic related to ritual. James Frazer’s famous
laws of sympathy and contagion are both based on the notion that mis-
guided analogies underlie “magic.” His laws of contagion and like-effect-
like imply that people mistakenly believe that sticking a pin in a doll will
injure an enemy. As noted in the Introduction, however, this theory has
been thoroughly critiqued in the past decades but is still implicated in
most discussions of “magic.”
Within the limits of this study, it seems most useful to point out that an
important aspect of analogies is that they provide models of the desired
results. The models are incorporated in the rites as formal representation
of the goals—that is, actions can function iconically, just as words can.
The formula against enemies and creditors, for example, asks the angels
to “break his bones, to crush all his limbs, and to shatter his conceited
power, as these pottery vessels are broken” (#2, 1.60). Both the words and
the actions give an exact mapping of the desired end of the rite; they tell
the angel exactly what the officiant wants to have happen to the enemy
and then present a model of the completed action. Even without a verbal
formula, the act of smashing vessels itself incorporates a specific instance
of the action of “smashing” into the structure of the ritual.
In addition to iconic signification, in the context of the ritual, it is then
possible to bring the desired ends into contact with a representation of the
victim. The doll can be given the name of the victim or in some way asso-
ciated with him or her. The verbal formula then directs the action specifi-
cally toward the intended victim, functioning as both an index and an
icon. In term of its semiotics, the infamous “voodoo doll” represents icon-
ically because it has a formal resemblance to the person to whom the rite
is directed. Because the doll is associated with person x, it is an indexical
sign, bringing the sign of a specific person into spatial-temporal contact
with the harming supernatural forces.
These types of iconic actions appear throughout the recipes, both in
this text and in most ritual texts from the Late Antique period. In all these
cases, the use of icons is likely to arouse suspicions in the eyes of outsiders.
102 Icons of Power

Dionysius mockingly remarks that two plus two is the same as four—that
is, he claims that the world is symbolic and not iconic. Without the spe-
cific understanding of the iconic status of elements of the rite, it looks
from the outside as if something (doll/name) is simply mistaken as some-
thing else (victim/deity). In no case, however, is the act of smashing vessels
understood in and of itself to effect the smashing of an enemy; that de-
pends on the successful intervention of the forces that can do so (the an-
gels, deity, and so on).
Four recipes lack oral recitations entirely but include written versions of
angel names.47 These recipes are for amulets to protect a soldier, a woman
in childbirth, a person suffering from a headache or blindness, and a city.
In the first three cases the person wears the amulet, in the fourth case a
lamella is rolled up and placed under the heels of a statue in a special part
of the city.
Jews commonly used amulets for protective purposes, and the practice
was suspect only when the amulet was inscribed with the name of the
wrong deity.48 The amulets in the Book of Secrets include angel names,
lacking the familiar biblical verses mentioned in rabbinic sources.49 As
noted above, the angel names are powerful not as tokens of divine speech,
as biblical texts would be, but as iconic representations of the angels
whose names contain the divine name “El.”
Wearing a divine name on one’s body both instantiates the divine pres-
ence and brings it in direct spatio-temporal contact with the wearer. This
is true whether the name is written on an amulet that is then attached to
the body, or whether the name is written directly on a limb.50 Thus, in the
terms of this study, amulets are icons that also function indexically, bring-
ing the iconic power of the names directly into spatio-temporal contiguity
with the wearer. In fact, this may be a good definition of an amulet, be-
cause what matters is that it directly represents the divine presence in a
specific location (on a person, in a house).
The final point to make about the use of objects in “rites” is the notion
of sympathy, common in Late Antiquity. Biblical notions of the selection

47. No. 18, 2.110–12; #19, 2.123–25; #20, 2.134–38; #23, 2.182–85.
48. 2 Macc 12:40 and 1 Macc 5:67.
49. mShabbat 8:2, bShabbat 61b.
50. See Chapter 6.
 Combining Words and Deeds 103

of a certain animal for sacrifice do not play a great role in the modified
sacrifices, which do not even always use the same animals. Instead, the
rites assume notions of sympathy—that is, of a complex web of intercon-
nections between all the elements employed in the rites and other dimen-
sions of the cosmos. Once again it is the Neoplatonists who give us the
most systematic explanations of these ideas. One of the clearest state-
ments is the brief treatise On the Hieratic Arts composed by Proclus. This
treatise both articulates the theoretical principles and gives a series of con-
crete examples of how these theoretical principles are manifested in the
world.51 The basic principle can be summed up as “Things on earth are
full of heavenly gods; things in heaven are full of supercelestials; and each
chain continues aboundingly up to its final members” (46–47).
This principle motivates diverse religious practices because each entity
in the world relates to divinity in its proper manner—for example, the he-
liotrope flower follows the sun, and humans sing hymns. These specific
techniques come from the gods themselves. The notion of sympathy also
motivates the use of certain objects in rituals. The heavenly world can be
found in earthly objects, so that they can represent specific cosmic forces
on earth in ritual settings. Proclus mentions, for example, the divine con-
nections of the cock, one of the animals used both in the Book of Secrets
and in Greek ritual texts as well. Sympathy establishes a set of relation-
ships between the seen and the unseen, opening up innumerable ways of
using the material world to influence the immaterial. These unseen forces
can both attract and repel, thus influencing the world in many ways that
are not obvious to the human eye.
In Chapter 2 we discussed the ideology that divine speech is repre-
sented in the Torah in written form; the Book of Secrets intertwines the
power of both written and spoken language. Amulets are placed on peo-
ple’s bodies, and formulas are spoken over them so that the goals are en-
coded in both spoken and written words. Each individual rite is a
complex use of a number of iconic representations of supernatural pow-
ers, all of which are used indexically within each specific rite.

51. The Greek text is found in Bidez 1928, 6:139–51. An English translation is available
in Copenhaver 1988:103–5.
104 Icons of Power

The Social Context of the Book of Secrets

If we turn briefly to the issue of social context, the relationship between

the rituals in the Book of Secrets and rabbinic circles is a contested issue.
The modern editor Margalioth characterized the ancient editor as distant
from standard rabbinic circles due to his enthusiasm for talking to the
dead and sacrificing to angels. Others have pointed to specific aspects of
rabbinic thought that might allow someone to study a book such as the
Book of Secrets for the purpose of understanding the practices—for ex-
ample, Merchavya finds permission to study magic in various rabbinic ci-
tations.52 Because Merchavya views the text as primarily theoretical, with
recipes no human could hope to follow, he finds a close fit between the
text and the particular rabbinic permission to study magic.
The text itself makes no claim in relation to rabbinic authority; no
names of rabbis appear in it. The opening of the composition describes the
text as “a book from the books of secrets” given by the angel Raziel53 to
Noah54 before he went into the ark.55 The claim to status as a heavenly
book, a pre-Flood esoteric tradition revealed only to the few, was a com-
mon trope in Late Antiquity found throughout the apocrypha and pseude-
pigrapha.56 It locates the text as a strand of tradition separate from other
revelations, including, for example, the sacrificial laws in the Torah. The
introduction specifically locates the text as a particularly old revelation
with its own history and legitimacy.
While the tendency is to draw a sharp line between rabbinic practices
and those in the lower six levels of the Book of Secrets, its rituals intersect
with numerous anecdotes found scattered throughout the variegated cor-
pus of rabbinic texts. The trope of rabbi as “holy man” has been explored
for several decades;57 Jacob Neusner sums it up neatly: “The rabbi knew
the secret names of God and the secrets of the divine ‘chariot’—the heav-

52. Merchavya points to such rulings as bSanhedrin 17a that the study of magic is per-
mitted in order to understand it. He also points out that the later Hasidic Ashkenazim used
these rulings to permit the study of magic for theoretical purposes (1966–67).
53. The angel name “Raziel” also appears on the seventh step in the second heaven.
54. In a variant, the text is given to Adam.
55. The text recounts a chain of transmission similar to mAvot 1:1.
56. See, for example, the reference to a Noahic book of healing in Jubilees 10:13.
57. See Green 1979; Neusner 1966–70, esp. 2:147–50, 3:102–26, and 4:347–70; and
Levine 1975:86–106.
 Combining Words and Deeds 105

ens—and of creation. If extraordinarily pious, he might even see the face

of the Shekinah . . . he could bring rain or cause drought. His blessing
brought fertility, and his curses death. He was apt to be visited by angels
and . . . he could see demons and talk with them, and could also commu-
nicate with the dead” (1969:10–11). It is easy to expand this list, pointing
to stories where rabbis who were famed as healers58 know the future and
are capable of outwitting the angel of death.59 Some claimed to have su-
pernatural births, being not born of women,60 and could control heat and
fire.61
Even daimon sacrifices are mentioned in rabbinic texts, including
casting bread into the sea for the daimon known as the Prince of the Sea
(oyh rw).62 Lauterbach (1936) is probably correct that this practice was
the basis for the popular Tashlich service, where individuals cast bread
into a body of water for forgiveness of sins. The sea could intercede for
people before God, as in the story of rewards given to a man who regu-
larly sacrificed bread to the demon of the sea. Similarly, the obscure bibli-
cal ritual in which a goat is sent off into the wilderness to Azazel bearing
the sins of Israel was understood by the rabbis to be a sacrifice to Satan.63
Other texts mention sacrifices to demons such as pouring out oil and wine
and scattering both parched grain and coins.64
At the same time, if some of the practices in the Book of Secrets con-
tradict modern pictures of rabbinic practice, perhaps our picture of
rabbinic Judaism remains too narrow. Saul Lieberman showed us the
importance of using, for example, the attacks of Christians and
Karaites as sources for evidence of lost rabbinic doctrines and practices
(1939). Karaite denunciations of the rabbis emphasized their interest in
angels, angel names, and the divine name, exactly as seen in the Book

58. R. Kahana cured R. Aha b. Jacob of jaundice (bShabbat 110b); R. Huna knows the
remedy for tertian fever (bShabbat 67a).
59. bMoed Katan 28a and bMegilla 29a.
60. Samuel says of Rav Judah (bNidda 13a).
61. bTa’anit 21b.
62. See, for example, the reference to worshiping the Ruler of the Sea in bHullin 41b. The
practice of throwing something into a lake is also denounced as a “Way of the Amorites.”
63. Pirkei de Rabbi Eliezer, chap. 46, Lauterbach 1936:268–69.
64. tShabbat 7.16 (compare bBerakot 50b, Semahot 8). In addition to sacrifices, there
were, of course, many other modes of relating to demons, including trying to drive them off
or simply fooling them.
106 Icons of Power

of Secrets.65 Rabbinic practices included, for example, the use of in-

cense66—the exact practice denounced in rabbinic sources when it is
done by women. Also denounced by the Karaites is the practice of put-
ting blood on door posts on the 10th of Tevet and calling it “Korban,”
which is yet another modified sacrifice.67 According to detractors, doing
so is acting too much in the image of the destroyed Temple. In all these
cases the gap between rabbinic practices and the rites in the Book of
Secrets narrows as we notice aspects of rabbinic practice that did not
attract much attention from previous generations of scholarship.
However, many of these practices, such as the daimon sacrifices, clearly
predated the rabbis and therefore may represent traditional practices over
which the rabbis had a difficult time gaining control.68 The struggle to gain
control of supernatural forces seen in the Book of Secrets masks another
struggle, one that takes place entirely on the human plane. The rabbinic
corpus evinces conflicts between rabbis, both about practices they could
not control and with all sorts of competing figures who claimed access to
supernatural forces. Rabbis competed with necromancers, as in the story
of the necromancer who could predict an earthquake.69 Dream inter-
preters beat rabbis at their own game, causing great distress to individual
rabbis by giving them bad dream interpretations that come true. Michael
Swartz raises the possibility that the scribes were another recognizable so-
cial group who had their own sources of legitimacy and power (1996).70
Women were another prime source of competition. Gentile women
have their own powers,71 like the woman who cursed Rabbi Judah’s ship
so that he had to put his clothes in water, thereby partially fulfilling and
thus averting the curse (bBaba Batra 153a). Even their own daughters can

65. See Nemoy 1930 and 1986.

66. See Lieberman 1939:76. This is to say nothing about the even stranger practice of ex-
changing gifts on Jesus’ birthday mentioned in a sixth-century text.
67. See Lieberman 1939:76 and Mann 1931–35: 2:74.
68. Scholars tend to call these practices “popular,” though the connotations of that term
in modern scholarship is in flux. It should not imply uneducated (as the rabbinic polemic
states) and so might best be referred to as extra-rabbinic.
69. bBerakot 59a, bTa’anit 21b.
70. The possibility that they had a role in making the ritual antidemon bowls so popular
in the Babylonian community is especially intriguing.
71. See bShabbat 81b, where a woman binds a boat so it will not move.
 Combining Words and Deeds 107

have powers that could threaten rabbis if out of their control, as in the
case of the daughters of Rabbi Nahman who could stir a cauldron with
their bare arms.72
Clearly the rabbis had no monopoly over the types of rituals the Book
of Secrets presents as “rites.” Honi the Circle-Drawer, the rabbinic com-
petitor mentioned above, is ultimately “rabbinized” with the repeated
retelling of his tale by rabbinic editors.73 But Honi, and all the other fig-
ures mentioned, competed with rabbis in their roles as daimon specialists.
The Book of Secrets may give us a window on one type of Jewish elite try-
ing to keep successful methods from other individuals; these could include
both rabbis opposed to specific ritual practices and nonrabbinic ritual
specialists. These recipes are also being “rabbinized” in their presentation
in the seven-heaven cosmology with the hymns on top, and by the associ-
ation of this text with other hekhalot materials.
In the Book of Secrets we may see the shifting of the social setting for
a series of rituals. The text makes secrets out of rituals that may at some
earlier point have been under the control of other segments of society (not
rabbis). The recipes are literally inserted into the upper realm, and the
way up to the secrets is carefully guarded. Necromancers, rabbis’ daugh-
ters, gentile women—what hope do they have of gaining access to these
recipes now, even if these figures once engaged in the kinds of practices
which the “secrets” represent?
This view gives us a new angle on the perennial question of the “rise of
magic” in Late Antique society. The presumed culprit is usually “oriental”
or even “Germanic” ideas that infect the Greco-Roman culture and doom
its basic rationality. Contra this common view, Tamsyn Barton (1994) re-
cently argued that there was less a “failure of nerve” than a shifting of the
social setting for various types of practices. Specifically she argues that
horoscopes gain a new prominence because the practice of astrology had
gained influence in elite circles. The growing desire by those seeking to
rule to prove their destiny as rulers through horoscopes brings the prac-
tice to the attention of modern scholars in a more dramatic manner.
Similarly, what we see in the Book of Secrets is in part the appropria-
72. bGittin 45a.
73. See Green 1979.
108 Icons of Power

tion of rituals that may at other times have been practiced in diverse seg-
ments of society (women, men of deed, dream interpreters, and so on).
These rituals in turn attract the attention of modern scholars and are ex-
plained as the “rise of magic” or the contamination of Judaism by foreign
practices. One intriguing example of this is the imprecation of the sun.
This practice appears, on broader consideration, to have been much more
widespread in Jewish circles during the first centuries than once thought
(M. Smith, 1982). The types of evidence range from the staircase tower
described in the Temple Scroll (11Q19 col. 30), which has been connected
to sun worship among the Nabateans,74 to references to similar practices
in Josephus (Jewish Wars 2.128)75 and in various rabbinic texts.76 This tex-
tual evidence correlates with the extensive archeological depictions of He-
lios, so prominent in synagogue mosaics. In the Book of Secrets,
imprecating the sun is neatly placed in the fourth heaven. Any individual
who does not have sufficient knowledge to gain access to the text (and the
fourth heaven) is not able to make use of this rite.
If the Book of Secrets had not been copied and thereby preserved, the
recipes would have been lost. This was the fate of many other rituals that
we know of only from hints found in the rich pseudepigraphic and apoc-
ryphal literature, as well as in the Qumran texts. The rabbis, to establish
themselves as the religious elite, had to be daimon experts par excellence
and as such displace others who claimed supernatural powers. The rise of
these figures is the shift away from a more widespread ability of individu-
als to intercede with supernatural powers to a focusing of that power in
the hands of only those with the special books, such as the Book of Se-
crets. This shift is a political shift as well and depends on the very this-
worldly ability of the elite to assert their control, not so much over
daimons as over their all-too-human rivals.

74. See Negev 1973.

75. Note the subsequent attempts to clean up Josephus’ comments, M. Smith
1982:202*–203*.
76. See Lieberman’s comments on bBerakot 9b (1974).
 7
Transformation by Deed Alone:
The Case of Alchemy

Whatever casts off the grossness

of the body becomes spirit.
—Zosimos, On Virture (Ber 3.6.1)

A
mong the marginalized rituals of Late Antiquity, the most neglected
are the alchemical rites.1 In contrast to the rituals discussed in pre-
vious chapters, these do not include any transformational lan-
guage. There is no recitation of hymns or divine names and no descriptions
of the heavenly world. Alchemical rites appear to have been silent rites in
which attention was focused on elaborate devices used in heating metals.
With these rituals we have come to the “deeds-only” end of the spectrum,
far from the rituals that consisted primarily of verbal formulas. The alchem-
ical rites elaborate another model of reinterpreted sacrifices, distinct from
the modified sacrifices found in the Book of Secrets. Once again the key to
the rituals is the iconic status of the metals and, even more important, the
iconic status of the changes made to them. These changes have a formal re-
semblance to the changes the world must undergo on its way to perfection.
There are, no doubt, many reasons for the neglect of these rituals. The
technical aspects of the rituals have frightened away scholars. The texts
are extremely difficult to comprehend, replete with unclear terminology
and descriptions of obscure procedures. The only edition of the manu-
scripts, compiled by Berthelot over a century ago, is an idiosyncratic com-
pilation;2 all the manuscripts must ultimately be re-edited and new critical
editions produced.3

1. For introductions to alchemy, see Riess 1893, 1:1338–55; Berthelot 1885; Festugière
1939 and 1950:217–82; and Forbes 1964:125.
2. Most of the manuscripts come from the tenth century on. The standard edition is
110 Icons of Power

Ignored by scholars of religion, the texts have largely been claimed by

historians of science and technology,4 who have been responsible for what
progress has been made in explaining the texts. The goal of these expla-
nations is limited. The texts are of interest only to the extent that they pre-
figure certain aspects of modern science. In the eyes of historians of
science, these texts are relegated to a few odd footnotes or held up as ex-
amples of good ideas gone wrong.5 Yet the activities described in alchem-
ical texts have scant relationship to the modern social practices and goals
of science.6
The term “alchemy” is itself medieval; in late antiquity these rites were
simply referred to as the “Sacred” or “Divine” Art or the “Great Work.”7
What then was the “Sacred Art”? The cliché of alchemy is that it constitutes
the marriage of Greek philosophy and Egyptian technique.8 In this equation,
Greek philosophy means two or three citations from Aristotle or Plato
about first causes and the primacy of matter; the Egyptian techniques are
metalworking techniques. While it is possible to find precursors of alchemi-
cal ideas in classical Greek philosophical texts, this strategy for understand-
ing alchemy generates a deceptively simple genealogy. This model also does
little to explain how the Sacred Art articulated with Late Antique concerns.

Berthelot and Ruelle 1963 (Ber) (reprint of 1888). Berthelot prepared the Greek edition,
which conflates several manuscripts. Ruelle edited some of the Greek and produced the
French translation, which often varies from Berthelot’s construal of such things as punctua-
tion. The manuscripts are cataloged in Bidez et al. 192428.
3. New editions are currently under way, according to Halleux 1981, but only his one
volume has appeared thus far.
4. The most useful discussions are in Taylor 1930, 1937, and 1949 and in Hopkins 1927
and 1934.
5. For example, Jensen writes that early chemists were “disappointed in their hopes” and
so delivered the art into the hands of the masses, where it “deteriorated into the search for
gold” (cited by Forbes 1948:19). Forbes himself writes: “The young chemistry had the typi-
cal rationalistic traits of the older Greek science. . . . Only relatively late in Antiquity can we
speak of a degeneration when alchemical elements (in our sense) are introduced” (1948:19).
6. See the judicious comments by Tamsyn Barton (1994) about the distortion of ancient
texts in modern histories of science.
7. For discussion of the term “alchemy,” see Forbes 1964:126, and for numerous possi-
ble derivations, many of them fanciful, see Lindsay 1970:68–89.
8. See, for example, Festugière 1950. This view demands an earlier dating of the first al-
chemical texts because it is unlikely that Greek philosophy and Egyptian techniques came
into contact only at the late dating (first century c.e.).
 Transformation by Deed Alone: The Case of Alchemy 111

Instead of viewing early alchemy as the wedding of Greek thought and

Oriental technique, alchemy is better understood as the harnessing of
Late Antique t™xnai (arts/practical instructions) for religious goals. In
the Late Antique period, “arts” were found in many areas of social prac-
tice, including rhetoric and medicine.9 In the “Sacred Arts,” religious
thinkers attempted to co-opt, for their own quite distinct purposes, the
prestige and success of the most up-to-date metal arts available.10 The
arts were put to work speeding up the inherent changes that the natural
world (including humans) was believed to be undergoing. Two elements
were required for the “sacred arts” to arise: first, a widespread percep-
tion that metalworking techniques could in fact succeed, and second, a
well-developed theory of the transformable nature of the cosmos and hu-
man existence.
In approximately 290 c.e., an event occurred that Edward Gibbon
(1946:285) and then Ernst Riess considered the first real event in the his-
tory of alchemy (1893:1341): Diocletian ordered the burning “of the an-
cient books of Chemia dealing with gold and silver” (perÁ xhmeºaq
ßrg†roy kaÁ xryso†).11 It is impossible to tell from the brief citation if
these books were simply craft books or included the types of pursuits that
interest us. Such distinctions may not have been of concern to Diocletian,
who may not have cared whether gold was produced for monetary or
salvific purposes. What this incident demonstrates is that some people, in-
cluding people in high places, saw metalworking techniques as potentially
successful and thus a threat. Techniques for producing gold had attracted
attention in the highest circles and were considered effective—that is the
first element necessary for alchemy.
The imagery we find in the writings of Zosimos (late third / early fourth

9. The arts offered practical instruction in medicine, rhetoric, architecture, and dream-
interpretation, to name a few (Barton 1994:7).
10. “Stealing” metal-working techniques was similar to modern attempts to prove that
penicillin, for example, is already prefigured in the Bible, thereby rubbing off on religion
some of the enormous prestige science currently enjoys.
11. According to John of Antioch, Diocletian burned them so that men might not enrich
themselves by this art and draw sources of wealth from it, which would enable them to re-
volt against the Romans (Suda, s.v. Diocletian). Diocletian also proscribed the “mathe-
matici” (Codex Justinian 9.18.2). See Lippmann 1919:103.
112 Icons of Power

century),12 an active practitioner and codifier of alchemical treatises, res-

onates with Late Antique tropes. These tropes center on stories of the fall
of the first man into the body and his resulting subjection to Fate, and of
striving to transform nature and ascend beyond the body and Fate. In
Zosimos’ version these tropes do not become standard “dualistic” think-
ing; instead, they retain a more positive attitude toward the natural world
because that world is in fact changing and changeable—the second ele-
ment necessary for alchemy.
It is not clear when these elements, which form the basis of alchemy,
had emerged. Like the hekhalot texts, the alchemical texts are all based
on much later manuscripts and have numerous variants and problems.
We are heavily dependent on Zosimos’ compilations for the early period
(pre–fourth century). Much of the writings of Zosimos are citations
from and discussions of Maria the Jewess and Pseudo-Democritus, both
of whom are difficult to date.13 Pseudo-Democritus, is thought to have
flourished anywhere from the third century b.c.e. to the first century
c.e.14 His only extant text is “Physical Things and Mysteries” (fysikÅ
kaÁ mystikå),15 although ancient writers associate other texts with his
name.16

12. Zosimos wrote before the Serapeum was destroyed in the 390s but cited Africanus,
who died in 232. For a concise introduction, see Jackson 1978:7. See Taylor 1937:88 on his
general importance. In general on Zosimos, see Plessner 1976, Riess 1893:1348, Forbes
1964:141, Lindsay 1970:323–57 (which must be used with care); and Hopkins 1934:69–77.
The writings of Zosimos are cited by the name of the treatise numbering in Berthelot’s cor-
pus, which is the same for the Greek and the French translation. This is a different mode of
citation than that followed by Patai, who cites the pagination from the French translation,
which is easy to confuse with the pagination of the Greek original and which he also occa-
sionally cites.
13. For citations of Maria, see Patai (1994:60–91). Unfortunately Patai worked with the
French translations of the texts and constructed his picture primarily from secondary dis-
cussions. This may be why there are numerous citation problems in his chapters on the
Greek alchemists. Patai supplies translations of many quotes from Maria but does not ex-
plain them in the context of Late Antique practices. While doing this would have led him off
the track from his emphasis on Jewish alchemists, it is necessary in order to make any sense
of Maria’s turgid quotations.
14. Forbes conflates Pseudo-Democritus with an equally unknown author, “Bolos,” who
also wrote a text called “Chircometa,” and dates him to 200 b.c.e. (1964:138). Lindsay
takes a similar approach (1970: chaps. 5–6). See Halleux 1981:73.
15. Ber 2.1. This treatise is mentioned briefly in Chapter 3.
16. Pliny mentions a “Chircometa” (Artificial Substances) by Democritus, a term that ap-
pears in Zosimos, where it seems to mean metals produced by art (NH 24.160).
 Transformation by Deed Alone: The Case of Alchemy 113

This particular text combines a mysterious revelation about the nature

of nature with essentially practical methods of making gold-like and sil-
ver-like alloys familiar from contemporary metalworking techniques. It
begins with formulas for dyeing cloth purple and then switches abruptly
into a narrative about Pseudo-Democritus’ search to learn how to “har-
monize” nature. The search ends dramatically with the revelation of a
phrase engraved on a pillar in a temple, “Nature rejoices in nature and na-
ture conquers nature and nature masters nature,” a cryptic saying that
implies that natural processes can be used to change nature, to overcome
nature. This is a very optimistic view, that the means for getting beyond
the natural world are available in the natural world. Based on this prem-
ise, it is necessary to learn everything possible about the natural world be-
cause all answers are contained in it. If nature conquers nature, natural
processes will ultimately lead to the cosmic changes sought by religious
practitioners. This positive attitude toward the natural world is striking
and is crucial to the development of the sacred arts.
Similar techniques also appear in the earliest extant papyri associated
with alchemy, two papyri from a cache of four papyri dating from the
mid-third century found in a tomb at Thebes.17 The other two papyri, in-
cluded as part of the Greek Magical Papyri (PGM), contain a now famil-
iar hodgepodge of rites for healing, for sending dreams, and for causing
love and hatred between people. One treatise, entitled the “Eighth Book
of Moses,” includes several versions of an initiation ritual (Leiden 12,
PGM 13).18 The two alchemical papyri appear on the surface very similar
to handbooks for metalworkers (Leiden 10 and Stockholm).19 Whoever
compiled these recipes was not concerned about making jewelry, the usual
task in metalworking, because none of these recipes appear to yield any
practical results. While jewelry makers regularly succeed at their tasks,
modern scholars speculate that these recipes were doomed to fail.20

17. French translation Berthelot 1938:28–77. The Greek with a French translation and
introduction is available in Halleux 1981. On the Leiden papyrus, see Hopkins 1934:61–63.
Other discussions include Ber 3–8; and Taylor 1930:111 and 1937:34–36.
18. Both papyri (Leiden 12 and 13) have numerous Jewish references, as noted by Patai
1994:56–57.
19. Halleux 1981:12.
20. See Pfister 1935.
114 Icons of Power

In this rich find of papyri we see the adoption of metalworking tech-

niques in new social settings and the severing of the techniques from sim-
ple (and obtainable) goals related to jewelry making. Creating the “Sacred
Art” was in part establishing a new context for the techniques. The new
context might be a collection of esoteric rituals; Greek Magical Papyri in-
cludes a recipe for tincture (¬osiq) of gold right after a short formula for
inducing a dream oracle (PGM 12.193–201). The fact that some alchem-
ical rituals appear in the Greek ritual handbooks is no surprise, because,
as argued below, these rituals are also forms of modified sacrifices. Within
the spectrum of Late Antique rituals, the Sacred Art had close parallels to
the modified sacrifices of Greek ritual handbooks.

The Ritual Techniques of Alchemy

The basic goal of the alchemical procedures was to impart a series of color
changes to the metals; blackening, whitening, yellowing, and, the most
obscure, violet-making. Color was understood to be the spirit, the
pneuma, of the metal,21 indicating the inner nature of the metal to the out-
side world.22 The change in the color of the metal is a sign that the metal
is working its way up the cosmic ladder to a higher level.
Working with metals sets up a rich series of “standing for” relation-
ships in the rites. The metals iconically represent the various components
of the natural world because they are pieces of it, and thus share its form,
and so forth. Each metal is iconic of a specific aspect of the natural world,
as its color is a formal link with that level of the natural world. The
change in coloration is also an iconic link with natural processes. Each
change maps for the practitioner the process of transformation that the
natural world undergoes.
In this thought system, metals stand at the center of a web of associa-
tions and analogies that permit their manipulation to have far-reaching re-
ligious significance. The metals represent the natural world, and changes
in the metal are evidence of natural processes in action. The variety of

21. “All sublimed vapour is a spirit and such are the dyeing qualities” (Zosimos, “The
Four Bodies,” Ber 3.12.4).
22. See the short but important discussion by Forbes 1964:140.
 Transformation by Deed Alone: The Case of Alchemy 115

metals permits the investigation of nature in its various aspects and con-
crete manifestations. Metals are “bodies,” analogous to everything that
has a body, including humans.
Metals are not static because the cosmic system is not static. As part of the
unfolding natural world, they too change, or, as Arthur Hopkins describes
it, they are “striving for perfection.”23 Regardless of its status in the natural
world, its place in the hierarchy, every metal is on its way to being gold
(Hopkins 1934:75). The structure of the cosmos includes perfection as a ba-
sic part of metals, all of which are gradually transforming to gold (Forbes
1964:132). The alchemist speeds up the process by adding a catalyst.
The basic change to which all these procedures is directed is trans-
forming the metal from the state of having a body to not having one.24
Maria the Jewess says, “If the corporeal is not rendered incorporeal, and
the incorporeal corporeal, nothing that one awaits will take place” (“On
the Body of Magnesium,” Ber 3.28.8).The first stage is “blackening,” a
process that results in a primary substance considered to represent matter
in its most basic state (an icon of pure substance). What this substance ac-
tually was is not clear. It could refer to an alloy of lead called “tetrasomy”
(“four bodies”) or an amalgam called “metal of magnesium,” which
could be broken down by being fused with sulfur.25 This substance repre-
sented matter in its most embodied state and thus at the very bottom of
the scale of nature; it was “unidentifiable by particular qualities.”26 Here
individuality of matter was submerged, making a “body” that had no
characteristics except that it was “fusible” (combinable with other mat-
ter) (Hopkins 1934:93). Fusibility was the only quality necessary to move
the matter upward on the bodily scale toward greater liquidity.

23. Hopkins considers this an Aristotelian notion: “Accepting Aristotle’s tendency to-
ward perfection, the alchemists visualized the metals as striving to become as perfect as pos-
sible, to become white as silver or even yellow as gold so perfect that goodness and light
should abound more and more and the ‘sun metal’ should illustrate the triumph of perfec-
tion over the primitive evil of the common ‘earth metals”’ (1934:37).
24. For Zosimos, becoming incorporeal is a means of overcoming fate (“On the Letter
Omega,” sec. 7).
25. See Forbes 1964:140–41. See also references to an alloy of copper and lead called
“molybdocale” (Ber 3.28.2).
26. Hopkins (1934:92) notes Plato’s concept of “first material,” giving no specific quota-
tion.
116 Icons of Power

The alloy was then subjected to the second stage, “whitening.” Ac-
cording to R. J. Forbes, this stage derives from general attempts to fake
silver and its alloys (1964:14). The metal could be whitened by fusing to
it tin or mercury or by adding a small bit of silver called a “ferment.”
Thus, it was possible to “whiten” a mass of metal by adding a small
amount of “white.” Hopkins notes that the addition of a small portion of
silver “was supposed to gather to itself any silver already developed, so
that when the black alloy was flooded with mercury or tin, the surface be-
came glistening white and gave forth the appearance of silver.”27 While
this alloy would be white only on the outside, the very fact that the inside
was yellow would not have troubled an alchemist.28
The third step was yellowing. Fusing could be done with gold ferment
this time, and the metal was yellowed with a substance referred to as “sul-
fur/divine water.” The sulfur water had to be produced before it could be
employed; it seems to have been made somehow from lead. Divine water
was used because “it produces the transformation; by its application you
will bring out what is hidden inside; it is called ‘the dissolution of bodies’”
(Pseudo-Democritus, “Synesius the Philosopher to Dioscorus,” Ber
2.3.6). The resulting yellow substance could turn other substances yellow,
again something that was not considered to be true of common gold.
The final and most obscure stage is ¬osiq, which appears to refer to
giving the substance a purple or violet color, from the root ios (violet or
rust).29 This final step produced an “ios of gold” that was believed to be
the essence with which pure gold could be produced. It represented icon-
ically the highest form of physical existence. “It is the tincture forming in
the interior [of the gold] which is the true tincture in violet, which has also
been called the ios of gold” (Zosimos, “The Four Bodies,” Ber 3.19.3).
If we have a vague understanding of these processes, we have less of a
clear sense of the mixtures on which these procedures were performed.
Basic to the esoteric nature of these texts is a secret language of metals.

27. Hopkins adds that arsenic and antimony were considered “mercury” (1934:95).
28. Because is it an alloy of copper, the alloy might be silver on the inside too, and thus
look superior to regular silver (Hopkins 1934:96).
29. See Hopkins 1934:97–98. Taylor 1949:49–50 says that this stage could be either fur-
ther tinting or simply cleaning of the metal or taking away the rust.
 Transformation by Deed Alone: The Case of Alchemy 117

Metals as we know them cannot affect other metals; only when they are
treated do they become agents of change. The ideology of “ferments”
points to the special properties of the metals employed; none of them was
the same as its mundane counterpart. “Each of them tints according to its
own nature. Gold tints to gold; silver to silver” (Pelagus, “On the Divine
and Sacred Art,” Ber 4.1.8). This gold and silver is not what is commonly
called gold and silver. Everything has its own nomenclature.30 “‘Our lead
becomes black,’ Maria explains, ‘while common lead is black from the be-
ginning’” (Zosimos, “About the Philosopher’s Stone,” Ber 3.29.1).

The Theology of Alchemical Transformation

Most of the techniques are preserved without extensive theological intro-

ductions or commentaries. An important exception is Zosimos’ treatise
“On Virtue,” where Zosimos articulates the analogy between working
with metals and redeeming humans. In this treatise he presents a dream vi-
sion31 in which he sees, as Jung described, “a kind of sacrificial act under-
taken for the purpose of alchemical transformation” (Jung 1967:70). The
Sacred Art is presented here with language familiar from sacrifice tradi-
tions. The alchemical rites do not include killing and eating animals, but
the fire remains and the act of cooking is still there.
Alchemical authors relate their practices to ancient sacrificial traditions.
In the Hebrew Temple, according to Zosimos’ letter to Theosebeia, the sac-
rifices were prefigurations of the Sacred Art. The Hebrews tried to effect
their own tinctures. He advises Theosebeia, “Without being asked, offer up
sacrifices: . . . sacrifices such as those that were recommended by Membres
when he addressed himself to Solomon, king of Jerusalem, and principally
those which Solomon himself described according to his own wisdom.”
The famous King Solomon, known internationally for his wisdom, is
presented here as an alchemist. Because he built the Temple, he is most

30. One result of this is that Ruelle substitutes his notion of the “real” name of the sub-
stance in the French translation for the name in the Greek, such as “sulfur” for “lead.”
31. The dream vision appears in the Berthelot edition in some disarray. For other transla-
tions and commentaries, see Taylor 1949:60–66 and Jung 1967, which includes an English
translation by A. S. Glover of Jung’s German translation.
118 Icons of Power

closely associated with sacrifice traditions. A figure by the name of Mem-

bres does not appear in the biblical stories, and his identity is obscure.
Zosimos continues: “Operating in this manner you will obtain the proper,
authentic, and natural tincture. Make these things until you become per-
fect in your soul. But, when you recognize that you have arrived at per-
fection, then beware of [the intervention of] the natural elements of the
material: descending toward the Shepherd, and plunging into the krater
[bowl], you will thus re-ascend to your origin” (“The Final Reckoning,”
Ber 51.8).32
Familiar Late Antique religious goals—escaping the world below, re-
turning to one’s origin—are envisioned as the potential result of sacrifices
similar to those done by Solomon. The Hebrews in their Temple were
“perfecting” themselves by means of sacrifice, causing matter to ascend so
that it can escape the embodied state and thereby escape Fate as well.
Reascending to the highest heavens, a goal familiar from the extensive dis-
cussion of ascent in Chapter 5, is completed here by descending into the
krater, the bowl used in alchemical rites.
The theology of Zosimos includes complex notions of descent and as-
cent: alchemical secrets come from heaven, either via fallen angels, as
noted above, or via the ascent of a special individual who brings the se-
crets down.33 So too the dream vision begins when Zosimos falls asleep
and sees seven steps leading up to an altar shaped like a krater.34 The
krater is similar to the krater mentioned above in the letter from Zosimos
to his “sister” Theosebeia. Equipped with a krater, the altar is a flexible
implement in which a variety of types of “cooking” can take place, in-
cluding, for example, both dipping and boiling.
In a vivid and stunning reinterpretation of sacrifice traditions, the en-
tity sacrificed in the altar bowl is not an animal but the officiant.35 The

32. This dense imagery is explicated somewhat in Zosimos’ vision, which that is de-
scribed here. Some of it remains unclear.
33. For Zosimos, Nikotheos appears to have been such an individual. See Jackson
1978:3.
34. Jung pointed to the use of a krater in the Hermetic tractate “Poimandres” as a paral-
lel (1967:73).
35. Here the scale of sacrificial substitutes is worked backward and human replaces an
animal.
 Transformation by Deed Alone: The Case of Alchemy 119

priest is both the sacrificer and what is sacrificed.36 The meaning of the
sacrifice is explicated in the dialogue between Zosimos and this officiant.
Zosimos learns that by the “casting away of the grossness of the body”
the priest who stands at the altar has become “in perfection as a spirit.”
The priest serves as the model of transformation; his body undergoes a se-
ries of changes until he reaches the highest goal. The priest’s immersion,
as Jung pointed out, leads to his transformation into a spiritual being
(Jung 1967:73). At this point he is no longer has a body, but is instead
spirit and thus perfect.
On a daily basis, Zosimos works with metals and not human bodies.
By means of fire, the metal makes a dramatic progression upward to an-
other type of existence, exactly as human bodies can. Each sacrifice is in
effect a self-sacrifice in which “metal” bodies are harshly destroyed so
that they can reach a new existence. Metallic bodies are cooked in a series
of stages that led ultimately to gold, the most spiritual being. While Zosi-
mos is working with these metals, a story about human bodies supplies
the explanation for his practices. The Sacred Art, in this case, looks not
toward only the special gold that is produced but toward the perfection of
spirit and, most important, the spirit that is found in human bodies.
The process of change Zosimos sees in the vision is violent, as the priest
“spewed forth all his own flesh.” In the first stage he is burned, a process
that is analogous to the “blackening” stage of working with metals. The
priest is overpowered by a mysterious figure and dismembered: “And he
drew off the skin of my head with the sword, which he wielded with
strength, and mingled the bones with the pieces of flesh, and caused them
to be burned upon the fire of the art, till I perceived by the transformation
of the body that I had become spirit.”
When Zosimos sees the altar again, the single figure has been replaced
by a multitude of people who are undergoing the same treatment. The
priest who sacrifices and is sacrificed has power over the other people who
are being punished. The revelatory figure is now gray, and in the next sec-
tion he becomes white, paralleling another step in the transformation of
metal, the “whitening.” The “white” man is engulfed in flames himself.

36. Noted by Jung 1967:70.

120 Icons of Power

The procedure is completed when a god appears to him, marking the

point of divine revelation. The god tells him that the head must be cut off
and the flesh boiled. At the conclusion, Zosimos states that he under-
stands “the art of the metals.”
In Zosimos’ vision, the processes are sped up before his very eyes so that
he can see the entire sequence of cosmic processes in one dream. “Cook-
ing” makes the stages appear in rapid order. What nature effects slowly by
itself, the officiant in Zosimos’ vision makes happen artificially fast. The
intensified heat of the fire is one of the catalysts for the speed—that is, fire
did not simply destroy, but instead transformed the natural world.
The entire procedure is then reviewed by Zosimos in a summary that
stresses that the models for the transformation come from nature. “How
does nature teach giving and receiving? . . . All things are woven together
and all things are undone together” (“On Virtue” Ber 3.1.4). The rites
were thought to mirror nature closely; the officiant is working with natu-
ral goals and natural processes. Someday all lead will naturally turn into
gold; Zosimos’ special talent is that he knows how to take control of the
process and speed it up.
If the “utopian” vision sees a great gulf between matter and divinity,
this vision evinces the opposite.37 We learned from Pseudo-Democritus
that all the answers can be found in nature. This is a very optimistic view;
the means for getting beyond the natural world are obtainable from the
natural world. Based on this premise, it is necessary to learn everything
possible about the natural world because the answers are contained in it.
If nature conquers nature, even without the interference of humans natu-
ral processes will ultimately lead to cosmic changes. The positive attitude
toward the natural world is striking and is crucial to the theology of the
Sacred Art.
Zosimos sees a continuum between the highest and the lowest elements
in the cosmos. No radical break, no insurmountable gulf, separates the
“grossness” of the body from the highest spirit, and nature itself in the
form of fire can itself overcome that gap. The unseen forces of nature can
be mapped in the changes of the metals, and all of this without any help

37. See Introduction, note 1.

 Transformation by Deed Alone: The Case of Alchemy 121

from language at all. Like the ancient biblical sacrifices, which are pre-
sented at least in some texts as being carried out in silence, the alchemical
rites included no names, prayers, or sounds. At the same time, they could
bring about even more dramatic transformations than the adjurations.
The forces that brought about the transformations were mysterious, be-
cause they were largely unseen by the average human being. Zosimos and
his fellow practitioners considered them divine.
 Concluding Note
This study points to the tremendous diversity of religious ideas at the turn
of the millennium. The Late Antique world presented a plethora of ways
for individuals to interact with, direct, beseech, and thwart gods, angels,
and daimons. Centuries-old rituals (sacrifice) were reinterpreted, and new
ones (ascent) were created. Sometimes these rituals were based on tradi-
tional notions of the alignment between heaven and earth with sacrifices
as the basic model for intercession, purification, and thanksgiving. An-
cient notions of sacrifice were reworked based on contemporary ideas of
the role of the deity and of demons. These reworkings depended on no-
tions of sympathy and antipathy. The practitioners of these rituals re-
mained fundamentally optimistic in that they thought the gap between
earth and heaven could be overcome using tradition ideas of communica-
tion and interaction. Other theological models competed with this view,
including many that pointed to an unbridgeable gap between the corrupt
earthly sphere and the divine immortal world of the highest heavens. In
these cases, rituals had to be as “otherworldly” as possible, and thus cut
off from matter in all its forms.
Speculation and action across the theological spectrum were based on
Late Antique notions of efficacy. These may appear obscure to both an-
cients and moderns, who have their own, often unexamined notions of
how the world works. Lurking behind the ritual texts are issues of power,
and in this case human power. Not everyone had access to the texts, nor
was everyone permitted to engage in the practices.
Religious practitioners in the first centuries c.e. could show their
power by their knowledge of cosmic secrets, divine names, and other eso-
teric traditions. A friend of the highest-level supernatural powers, he (and
in rare cases she) could put into play divine power based on socially con-
ceived notions of efficacy. To have the religious experience of ascent
through the heavens is to be permanently changed; to produce gold is to
124 Concluding Note

thoroughly know and control the natural forces. In Late Antiquity, a hu-
man form could conceal a vast array of powers, and the esoteric rituals
may reveal a human to be a supernatural expert walking around in a hu-
man body.
Particular attention was given in Late Antiquity to exegesis of the name
of the deity. We witness a tremendous exchange of ideas about divine
names across religious traditions. Current scholarly understanding of this
exegesis tends to isolate these ideas as distinct from, and inferior to, other
modes of religious expression. This is not without irony, because the di-
vine name was considered especially holy. God’s name was equated with
his creative power, and, as in the Aramaic Targums, the “content” of that
name was the very act of creative speaking itself. For these Jewish writers,
the divine name was not only the metaphysical origin of all language but
also the source of efficacy in ritual language even at the level of the sounds
of the name.
All these ideas are predicated on the function of the divine name as a
particular type of sign, specifically its function as an icon. It is impossible
to understand this function of the name—and to compare it with other
possible ways in which words and signs can function—without the very
specific semiotic vocabulary of icon, index, and symbol. The larger iconic
structure of ritual texts also turned out to be crucial, not only in the as-
cent hymns but also in the mirroring of the cosmic levels in the Book of
Secrets.
For the Hebrew exegetes, the deity spoke Hebrew, permitting a basic
analogy between divine speech and human speech. This is important for
Origen because it permits a direct link between the levels of meaning in
Hebrew and the divine world, between the process by which the world
was created and the potentials of human language. For other Christians
this analogy was less powerful. Almost metaphorically, Jews become “lit-
eralists” no matter how they read scriptural texts because they read He-
brew (the letter) while the Christians interpret the Greek (the spirit).
For Dionysius the Aeropagite, names did not carry any inherent power
based on their sounds, but they were still crucial guides to the divine
world and capable of transforming the symbols of bread and wine, and
humans as well. For everyone, language merited careful thought and
 Concluding Note 125

analysis because the deity chose it to create the world and convey his rev-
elation. Both Name and Word are mysterious in their ability to manifest
divine reality and power to humans on earth. Many of these ideas were no
doubt influenced by the general culture (Greco-Roman concepts of names
and numbers), because they are not simple biblical concepts. These ideas
were developed by the Jews and then in turn borrowed by other groups
based on stereotypes of who had the oldest wisdom.
All these effective units—names, letters, and sounds—can be harnessed
for diverse ritual ends. Human practitioners can transcend their earthly
existence and enter the divine realm by means of special use of language.
Talking like angels assimilates a human individual into the group of heav-
enly worshipers. Ascent is the model on which so many Late Antique rit-
ual texts are built because it articulates with the contemporary vision of a
multilayered and hierarchical cosmos. This cosmology belonged to no
specific religious tradition, and neither did notions of ascent. The ritual
model of ascent was applied to numerous goals, from the mundane pur-
pose of thwarting enemies to the seemingly heretical goal of transforming
a human into a cosmic being.
The sum-total of techniques in the ascent are familiar: the appropria-
tion of heavenly cultic activities by humans, such as reciting praises, com-
bines with wearing amulets and with the special power of the divine
names. No techniques are unique to Hekhalot Rabbati, or to ascent for
that matter. Instead, the key to ascent is using these techniques within the
structure of ascent.
The proper words were also key to summoning angels in the Book of
Secrets. With angelic aid, practitioners are able to thwart or accomplish a
wide range of goals; this is striking because the rites represent a micro-
cosm of human wishes, from the petty and even vain to the dramatic and
transformative. Employing angels demands knowledge not only of cosmic
secrets but also of the workings of this-worldly legal language. The elab-
oration of the ritual comes in the constant interplay between getting the
angel to go to work, supplying exotic ingredients, and remembering the
formulas. The focus of the text is on the ability of the officiant to insinu-
ate himself higher and higher into the cosmic order and finally arrive at a
level where he does not have to manipulate lower-level supernatural
126 Concluding Note

forces but instead needs only declare the deity blessed. The perceived effi-
cacy of the verbal formulas is clearer to the user than, for example, the
manipulation of objects, because an adjuration transparently appears to
the reciter to “do something.” Because the efficacy of swearing is more
self-evident to users than some of the manipulations of objects, the adju-
rations are more adaptable as effective ritual devices. They may continue
to be copied while many of the actions will fall out of a text. For example,
in the collection of recipes known as “The Sword of Moses,” some sec-
tions consist entirely of verbal adjurations that have lost their other com-
ponents (Gaster 1971:288–337).
In the Late Antique context, the alchemical investigations into nature
sought to uncover the most secret layers of existence and thereby permit
the practitioner to step in and control nature. Instead of reciting names or
using heavenly liturgy, the participant strives to uncover the processes that
shape the entire cosmos. Once these processes are understood at the level
of the natural world, and in particular in metals, anything, even humans,
can be transformed into a higher existence. Again, the key to the trans-
formative aspects of the rites is the complex iconic “standing for’s” that
the rites entail. The metals are icons of the basic structure of the cosmos
in a very literal way, being actual bits of cosmic elements. The base metal
produced in the first part of the alchemical procedures has the exact form
(and even more) of the lowest end of the cosmos, and even more so be-
cause it is more base than any naturally occurring metal. The transforma-
tions are also iconic because they map for us, in a fast and visible way, the
transformations of the natural world.
As for the structure of these rituals, the mode of operation is highly
technical and based mainly on the construction of elaborate ritual devices
for manipulating metals. Ritual technique focuses on the sheer complex-
ity of these devices, and not at all on the correct formulas or blessings. No
ritual formulas are recited at all, which is in striking contrast to the heav-
ily word-based ascent rituals.
The alchemical circles, gathered around their flames, worked with
complex concoctions that are now lost to us. These rites were predicated
on the view that no unbridgeable gap exists between the highest and low-
est parts of the cosmos. However much it might look like “lead,” the
 Concluding Note 127

world was on its way to becoming “gold,” and it was within one’s grasp
to uncover and effect this process. The endeavors must have seemed
wildly optimistic to many of the alchemists’ peers, who believed that the
gap between the world of nature and the world of the mind was too great
to overcome.
Contrast these actions with the rabbinic anecdote about the two rabbis
who studied a Book of Creation and were able to create a cow. Their in-
vestigation involves no search for a lion cub, no collecting of water from
seven streams in pure vessels, no collecting of metals and odd utensils.
This esoteric “praxis” was thoroughly “spiritualized” and included only
manipulation of language. Ritual elaboration in this system comes in the
endless spinning of new permutations of the central ritual word—the
name of the deity—and in endless playing with the letters.
For all these rituals we depend on textual traditions that are fraught
with problems. Fortunately, the rituals were conceptualized in dialogue
with discussions of efficacy found in texts that are somewhat easier to
date. The philosophical grounding of rituals point to two stages. The first
is a nascent stage in the first centuries b.c.e. through the first centuries
c.e., when ideas that would be crucial to later ritual were just developing
(ideologies of the name, the transformational potential of nature, the pos-
sibility of a human ascending). By the third and fourth centuries the ritual
techniques analyzed here had undergone substantial development; we find
movements to edit and collect sets of ritual techniques, both of the al-
chemical type and of the types found in the Book of Secrets. By the third
and fourth centuries, we also see evidence of hostile debate and attack, as
those who want to be thought of as normative and “traditional” establish
their positions by rejecting some of these religious traditions (for example,
rabbinic controversies over ascent).
Ancient thinkers were aware of the contrasting modes of approaching
divinity. Damasicius grouped Plotinus and Porphyry as “philosophers”
against Iamblichus, Syrianus, and Proclus as “hieratics” (Commentary on
the Phaedo). Among those who employed rituals, there were divergent
ideas about how to construct the rituals—whether “words only” rites
were superior to other types of ritual actions.
Late Antique debates about the aesthetics of ritual may in turn mask
128 Concluding Note

the radical claims of many of the practitioners. Some Jewish, Christian,

and Greco-Roman ancient writers were unified on one point: the lofty
goals of their rituals. Socrates is reported to have said that his aim was to
“become godlike” (Theaetetus 176B); whatever he might have meant, it
was his follower the Chaldaean Platonist Julian who prayed that his son
would have the soul of an archangel and who saw himself as one in a
chain of divine beings placed temporarily in a human body.
With the proper intention, the very best rituals can have, for
Iamblichus, the same results as Plotinus’ investigations: ‘nvsiq (union).
Divination, for Iamblichus, “unites us with the divine” (On the Mysteries
10.4; 2.289.3–6). The best rituals, theurgy, played their part by purifying
and saving the soul (1.12; 41.13 and 10.7; 293.5–8). His ultimate ritual
goal was “to put on the form of the god” (4.2; 184.8) and exchange his
life for a new divine existence (3.4; 109, 14–15). For Dionysius, rituals
can both transform the bread and wine and also deify the officiant.
If these notions of ritual efficacy are not relegated to the margins of re-
ligious expression as either “magic” or “mysticism,” some stereotypes of
Late Antique religious expressions fade. For practitioners all across the
spectrum, rituals were a path of transformation from earthly to heavenly
existence. For those who believed in the existence of supernatural powers,
all the rituals we have examined offered both the deadliest and the most
holy potential.

1. See Protrepticus chap. 8, p. 48, 9–21.

 Select Bibliography
Primary Sources
(Texts cited according to the Loeb editions are not included in this list.)

Albinus. Eisagogue [Introduction to Platonic Thought]. In Plato, vol. VI, pp. 147–51. Ed.
Hermann. Teubner: Leipzig, 1936.
Artemidorus. Oneirocritica [The Interpretation of Dreams]. Translation and commentary
by Robert J. White. Park Ridge, N.J.: Noyes Press, 1975.
Censorinus. De die natali [Concerning the Day of Birth]. Ed. N. Sallman. Teubner: Leipzig,
1983.
Book of Secrets. Sepher ha-Razim: The Book of the Mysteries. Trans. Michael A. Morgan.
Chico, Calif.: Scholars Press, 1983.
The Chaldean Oracles. Text, translation and commentary by Ruth Majercik. Leiden: Brill,
1989
Corpus Hermeticum. Trans. A. J. Festugière. Ed. A. D. Nock Paris: Les Belles Lettres,
1954–60.
———. English translation: Hermetica. Trans. Brian Copenhaver. Cambridge: Cambridge
University Press, 1992.
Damascius, Commentary on the Phaedo in Greek Commentaries on Plato’s Phaedo vol. 2.
Ed. L. G. Westerink. Amsterdam and New York: North-Holland, 1976
Dionysius the Areopagite. Pseudo-Dionysius: The Complete Works. Trans. Colm Luibheid.
New York: Paulist Press, 1987.
Greek Magical Papyri in Translation. Ed. Hans Dieter Betz. Chicago: University of Chicago
Press, 1986.
Iamblichus. Iamblichus: On the Pythagorean Life. Translated with notes and introduction
by Gillian Clark. Liverpool: Liverpool University Press, 1989.
———. Jamblique: Les Mystères d’Egypte [On The Mysteries]. Trans. and ed. E. des Places.
Paris: Les Belles Lettres, 1966.
———. Theologumena Arithmeticae [Theology of Mathematics]. Ed. V. deFalco, 1922.
Edited with additions and corrections by U. Klein. Stuttgart: Teubner, 1975
———. In Platonis Dialogos Commentariorum Fragmenta [Commentary on the Timaeus].
Trans. and ed. John Dillon. Leiden: E. J. Brill, 1973.
Masseket Hekhalot. In Bet Ha-Midrasch vol. 4. Ed. Adolph Jellenik. Jerusalem: Wahrmann
Books, 1967.
Merkaba Rabba. Synopse zur Hekhalot-Literatur. Ed. Peter Schäfer, Margarete Schlüter,
Hans-Georg von Mutius. Tübingen: Mohr, 1981.
130 Select Bibliography

———. The Shi’ur qomah: texts and recensions. Ed. Martin Samuel Cohen. Tübingen:
Mohr 1985.
Midrash Konen, In Jellenik 1967, see Masseket Hekhalot.
Musici scriptores graeci [Greek Writers on Music]. Ed. Karl von Jan. Leipzig: Teubner,
1895.
Proclus. Elements of Theology. Ed. E. R. Dodds. Oxford: Oxford University Press, 1963.
———. Commentary on the First Alcibiades of Plato. Ed. L. G. Westerink. Amsterdam:
North-Holland, 1954. English translation by W. O’Neill. The Hague: Martinus Nijhoff,
1965.
———. In Platonis Rem Publicam Commentarii [Commentary on the Republic]. Ed. W.
Kroll. Leipzig: Teubner 1899–1901.
———. In Platonis Timaeum Commentaria [Commentary on the Timeaus]. Ed. E. Diehl.
Leipzig: Teubner, 1903–6.
———. In Platonis Cratylum Commentaria [Commentary on the Cratylus]. Ed G. Pasquali.
Leipzig: Teubner, 1908.
———. Théologie platonicienne [Platonic Theology]. Trans. and ed. H. D. Saffrey and L. G.
Westerink. Paris: Les Belles Lettres, 1968–87.
Pseudo-Dionysius. See Dionysius the Areopagite.
Sepher ha-Razim. See Book of Secrets.
Targum Neofiti. Neophyti 1. Targum Palestinense Ms. de la Biblioteca Vaticana. Ed. A. Díez
Macho. Madrid: Consejo Superior de Investigaciones Científicas, 1968–79.
Targum Pseudo-Jonathan. Targum Pseudo-Jonathan of the Pentateuch. Ed. E. G. Clarke,
W. E. Aufrecht, et al. Hoboken, N.J.: Ktav, 1984.
Theodosius Alexandrinus. See Iamblichus.
Zosimos. “On the Letter Omega.” See Jackson 1978.

Secondary Sources

Albeck, Hanokh. 1952. Shishah sidrei mishnah. Tel Aviv: Mosad Bialik.
Alexander, P. S. 1983a. “Review of Peter Schäfer’s Synopse zur Hekhalot-Literatur.” Jour-
nal of Jewish Studies 34:102–6.
———. 1983b. “Third Enoch.” Pages 223–315 in Old Testament Apocrypha and Pseude-
pigrapha, vol. 34, ed. James Charlesworth. Garden City, N.Y.: Doubleday.
———. 1986. “Incantations and Books of Magic.” Pages 342–79 in The History of the Jew-
ish People in the Time of Jesus Christ, ed. G. Vermes, F. Millar, and M. Goodman. Third
edition. Edinburgh: Clark.
———. 1987a. “Prayer in the Heikhalot Literature.” Pages 43–64 in Priere, mystique et ju-
daisme: Colloque de Strasbourg, 10–12 septembre 1984, ed. Roland Goetschel. Travaux
du Centre d’histoire des religions de Strasbourg 2. Paris: Presses Universitaires de France.
———. 1987b. “Third Enoch and the Talmud.” Journal for the Study of Judaism 18:40–68.
———. 1990. “Jewish Aramaic Translations of Hebrew Scriptures.” Pages 217–54 in
Mikra: Text, Translation, Reading, and Interpretation of the Hebrew Bible in Ancient
 Select Bibliography 131

Judaism and Early Christianity, ed. M. J. Mulder and Harry Sysling. Minneapolis,
Minn.: Fortress Press.
———. 1992. “Targum, Targumim.” Pages 320–31 in The Anchor Bible Dictionary, vol. 6,
ed. David Noel Freedman. New York: Doubleday.
Alon, G. 1950. “In the Name.” Tarbits 21:30–39. Hebrew.
Altmann, Alexander. 1946. “Sacred Hymns in Hekhaloth Literature.” Melilah 2:1–24.
Aptowitzer, Victor. 1930–31. “The Heavenly Temple in the Aggadah.” Tarbits 2:257–87.
Hebrew.
Asad, Talal. 1993. Genealogies of Religion: Discipline and Reasons of Power in Christian-
ity and Islam. Baltimore: Johns Hopkins University Press.
Athanassiadi, P. 1993. “Dreams, Theurgy, and Freelance Divination: The Testimony of
Iamblichus.” Journal of Roman Studies 83:115–30.
Austin, J. L. 1962. How to Do Things with Words. Oxford: Clarendon Press.
Bacher, W. 1901. “Shem Ha-Meforash.” Pages 262–64 in Jewish Encyclopedia, vol. 11, ed.
I. Singer. New York: Funk & Wagnalls.
Baer, Yitzhak. 1975. “The Service of Sacrifice in Second Temple Times.” Zion 40:95–153.
Bardy, Gustave. 1937. “Les traditions juive dans l’oeuvre d’Origene.” Revue Biblique
34:220–40.
Bar-Ilan, Meir. 1989. “Magical Seals Written on Bodies.” Tarbits 40:37–50. Hebrew.
Barton, Tamsyn. 1994. Ancient Astrology. New York: Routledge.
Baumgarten, J. 1986. “The Book of Elkesai and Merkabah Mysticism.” Journal for the
Study of Judaism 17:212–23.
———. 1988. “The Qumran Shirot and Rabbinic Merkabah Traditions.” Revue de Qum-
ran 13:199–213.
Berthelot, Marcellin. 1885. Les origines de l’alchimie. Paris: G. Steinheil.
———. 1938. Introduction à l’étude de la chimie des anciens et du moyen-âge. Paris: Li-
brarie des Sciences et des Arts.
Berthelot, Marcellin, and C. Ruelle. 1963. Collection des anciens alchimistes grecs. London:
Holland Press. Reprint of 1888.
Betz, Hans Dieter. 1986. The Greek Magical Papyri in Translation. Chicago: University of
Chicago Press.
Bickerman, E. J. 1929. “Die Römische Kaiserapotheose.” Archiv für Religionswissenschaft
27:1–31.
———. 1980. Chronology of the Ancient World. London: Thames & Hudson.
Bidez, Joseph. 1964. Vie de Porphyre, le philosophe neo-platonicien: Avec les fragments des
traites Peri agalmaton et De regressu animae. Hildesheim: G. Olm. Reprint of 1913.
———. 1965. La vie de l’empereur Julien. Paris: Les Belles Lettres. Reprint of 1930.
Bidez, Joseph, et al. 1924–28. Catalogue des manuscrits alchimiques grecs. Bruxelles: M.
Lamertin.
Bietenhard, Hans. 1974. Caesarea, Origenes und die Juden. Stuttgart: W. Kohlhammer.
Blau, Ludwig. 1914. Das altjüdische Zauberwesen. Berlin: L. Lamm.
Bloch, P. 1893. “Die ywrdy mrkbh, die Mystiker der Gaonenzeit und ihr Einfluss auf der
132 Select Bibliography

Liturgie.” Monatschrift für Geschicht und Wissenschaft des Judentums 37:18–25, 69–74,
57–266, 305–311.
Blowers, Paul. 1988. “Origen, the Rabbis, and the Bible: Toward a Picture of Judaism and
Christianity in Third-Century Caesarea.” Pages 98–116 in Origen of Alexandria: His
World and His Legacy, ed. C. Kannengiesser and W. L. Petersen. South Bend, Ind.: Uni-
versity of Notre Dame Press.
Böhlig, A., and F. Wisse. 1975. Nag Hammadi Codices III,2 and IV,2: The Gospel of the
Egyptians. Leiden: Brill.
Bokser, Baruch. 1985. “Wonder-Working and the Rabbinic Tradition: The Case of Hanina
ben Dosa.” Journal for the Study of Judaism 16:42–92.
Boll, Franz. 1903. Sphaera: Neue griechische Texte und untersuchungen zur geschichte der
Sternbilder. Leipzig: Teubner.
Bousset, Wilhelm. 1901. “Die Himmelreise der Seele.” Archiv für Religionwissenschaft
4:136–69, 229–73.
———. 1923. Apophthegmata. Aalem: Scientia.
Boyer, Louis. 1980. “Mysticism: An Essay on the History of the Word.” Pages 42–55 in Un-
derstanding Mysticism, ed. Richard Woods. New York: Doubleday.
Brandt, Edward. 1927. Untersuchungen zum Romischen Kochbuche: Versuch einer Losung
der Apicius-Frage. Philologus Supplementband 19, Heft 3. Leipzig: Dieterich.
Brightman, F. E. 1965 (1896). Liturgies, Eastern and Western, Being the Texts, Original or
Translated, of the Principal Liturgies of the Church. Oxford: Clarendon Press.
Brown, J. P. 1979. “The Sacrificial Cult and Its Critique in Greek and Hebrew, Part 1.” Jour-
nal of Semitic Studies 24:159–73.
———. 1980. “The Sacrificial Cult and Its Critique in Greek and Hebrew, Part 2.” Journal
of Semitic Studies 25:1–21.
Carmi, Ted. 1981. The Penguin Book of Hebrew Verse. New York: Viking Press.
Chadwick, Henry, ed. and trans. 1965. Origen: Contra Celsum. Cambridge: Cambridge
University Press.
Charlesworth, James H. 1983. The Old Testament Pseudepigrapha, vol. 1. Garden City,
N.Y.: Doubleday.
Cohen, Martin Samuel. 1983. The Shi’ur qomah: Liturgy and Theurgy in Pre-Kabbalistic
Jewish Mysticism. Lanham, Md.: University Press of America.
———. 1985. The Shi’ur qomah: Texts and Recensions. Texte und Studien zum antiken Ju-
dentum 9. Tübingen: Mohr.
Cohon, S. 1951. “The Name of God: A Study in Rabbinic Theology.” Hebrew Union Col-
lege Annual 26:579–604.
Collins, John. 1985. “Artapanus.” Pages 889–903 in The Old Testament Pseudepigrapha,
ed. James Charlesworth. Garden City, N.Y.: Doubleday.
Copenhaver, Brian. 1988. “Hermes Trismegistus, Proclus, and the Question of a Philosophy
of Magic in the Renaissance.” Pages 79–110 in Hermeticism and the Renaissance, ed. In-
grid Merkah and Allen Debus. Washington, D.C.: Folger.
Dahl, Nils. 1941. Das Volk Gottes: Eine Untersuchung zum Kirchenbewusstsein des
Urchristentums. Darmstadt: Wissenschaftliche Buchgesellschaft.
 Select Bibliography 133

Dan, Joseph. 1967–68. “Review of The Book of Secrets.” Tarbits 37:208–214. Hebrew.
———. 1979. “The Concept of Knowledge in the Shiur Komah.” Pages 67–73 in Studies in
Jewish Religious and Intellectual History: Presented to Alexander Altmann on the Oc-
casion of His Seventieth Birthday, ed. Alexander Altmann, Siegfried Stein, and Raphael
Loewe. Tuscaloosa: University of Alabama Press.
Davila, James. 1993. “Prolegomena to a Critical Edition of the Hekhalot Rabbati.” Journal
of Jewish Studies 44:208–25.
Deissmann, G. Adolf. 1895. Bibelstudien. Marburg: N. G. Elwert.
Delatte, Armand. 1961. “Herbarius: Recherches sur le cérémonial usité chez les anciens
pour la cueilette des simples et des plantes magiques.” Memoires Academie Royale de
Belgique, vol. 54.
Dieterich, A. 1891. Abraxas: Studien zur Religionsgeschichte. Leipzig: Teubner.
Diez Macho, Alejandro. 1960. “The Recently Discovered Palestinian Targum.” Vetus Testa-
mentum Supplement 7:222–45.
———. 1979. Neophyti 1, Targum Palestinense ms. de la Biblioteca Vaticana. Madrid: Con-
sejo Superior de Investigaciones Cientificas.
Dillon, John. 1985. “The Magical Power of Names.” Origeniana Tertia: The Third Interna-
tional Colloquium for Origen Studies, University of Manchester September 7th–11th,
1981, ed. R. P. C. Hanson and Henri Crouzel. Rome: Edizioni dell’Ateneo.
———. 1996. “Oaths.” Pages 1056–57 in The Oxford Classical Dictionary, ed. Simon
Hornblower and Anthony Spawforth. New York: Oxford University Press.
Dodds, E. R. 1947. “Theurgy and Its Relationship to Neoplatonism.” Journal of Roman
Studies 37:55–69.
———. 1961. “New Light on the ‘Chaldaean Oracles.’” Harvard Theological Review
54:263–73.
Dodds, E. R., ed. and trans. 1963. Proclus’ The Elements of Theology. Second edition. Ox-
ford: Clarendon Press.
Dornseiff, Franz. 1922. Das Alphabet in Mystik und Magie. Leipzig. Berlin: B. G. Teubner.
Duranti, Alessandro. 1992. “Language in Context and Language as Context: The Samoan
Respect Vocabulary.” Pages 77–100 in Rethinking Context: Language as an Interactive
Phenomenon, ed. Alessandro Duranti and Charles Goodwin. Cambridge: Cambridge
University Press.
Elior, Rachel. 1990. “Merkabah Mysticism: A Critical Review.” Numen 37:233–47.
———. 1993–94. “Mysticism, Magic, and Angelology: The Perception of Angels in
Hekhalot Literature.” Jewish Studies Quarterly 1:3–53.
———. 1999. “The Merkavah Tradition and the Emergence of Jewish Mysticism.” Pages
101–58 in Sino-Judaica: Jews and Chinese in Historical Dialogue, ed. Aharon Oppen-
heimer. Tel Aviv: Tel Aviv University.
Ervin-Tripp, Susan. 1976. “Is Sybil There? The Structure of American English Directives.”
Language in Society 5:25–66.
Faraone, Christopher. 1991. “The Agonistic Context of Early Greek Binding Spells.” Pages
3–32 in Magika Hiera: Ancient Greek Magic and Religion, ed. C. Faraone and D. Ob-
bink. New York: Oxford University Press.
134 Select Bibliography

Ferwerda, R. 1982. “Plotinus on Sounds: An Interpretation of Plotinus’ Enneads

V.5.519–27.” Dionysius 6:43–57.
Festugière, A. J. 1939. “Alchymica.” L’Antiquité Classique 8:71–95.
———. 1950. La Révélation d’Hermes Trismégiste, vol. 1. Paris: Gabalda Press.
———. 1967. Proclus. Commentaire sur le Timée. Paris: J. Vrin.
———. 1971. Etudes de philosophie grecque. Paris: J. Vrin.
Fitzer, G. 1971. “Sphragis.” Pages 7:939–53 in Theological Dictionary of the New Testa-
ment, ed. Gerhard Kittel, Geoffrey William Bromiley, and Gerhard Friedrich. Grand
Rapids, Mich.: Eerdmans.
Flesher, Paul. 1995. “The Targumim.” Pages 40–63 in Judaism in Late Antiquity, Part 1:
The Literary and Archaeological Sources. Leiden: Brill.
Forbes, R. J. 1948. Short History of the Art of Distillation. Leiden: Brill.
———. 1964. “The Origin of Alchemy.” Pages 125–48 in Studies in Ancient Technology,
vol. 1, ed. R. J. Forbes. Leiden: Brill.
Fossum, Jarl E. 1985. The Name of God and the Angel of the Lord: Samaritan and Jewish
Concepts of Intermediation and the Origin of Gnosticism. Tübingen: Mohr.
Fowden, Garth. 1987. “Pagan Versions of the Rain Miracle of a.d. 172.” Historia 36:83–95.
Garrett, Susan R. 1989. The Demise of the Devil: Magic and the Demonic in Luke’s Writ-
ings. Minneapolis, Minn.: Fortress Press.
Gaster, Moses. 1971. Studies and Texts in Folklore, Magic, Mediaeval Romance, Hebrew
Apocrypha, and Samaritan Archaeology. New York: Ktav.
Gersh, Stephen. 1978. From Iamblichus to Eriugena: An Investigation of the Prehistory and
Evolution of the Pseudo-Dionysian Tradition. Leiden: Brill.
Gibbon, Edward. 1946. The Decline and Fall of the Roman Empire. New York: Heritage.
First publication, 1776.
Ginzburg, Louis. 1905. “Yetsira.” Pages 12:602–6 in Jewish Encyclopedia, ed. I. Singer.
New York: Funk & Wagnalls.
Godwin, J. 1991. The Mystery of the Seven Vowels in Theory and Practice. Grand Rapids,
Mich.: Phanes.
Gögler, Rolf. 1963. Zur Theologie des biblischen Wortes bei Origenes. Düsseldorf: Patmos.
Goshen-Gottstein, Alon. 1995. “Four Entered Paradise Revisited.” Harvard Theological Re-
view 88:69–133.
Graf, Fritz. 1997. Magic in the Ancient World. Cambridge, Mass.: Harvard University
Press.
Green, William Scott. 1979. “Palestinian Holy Men: Charismatic Leadership and Rabbinic
Tradition.” Pages 619–47 in Aufsteig und Niedergang der romischen Welt, ed. W. Haase,
2.19.2. Berlin: De Gruyter.
Grozinger, Karl. 1987. “The Names of God and Their Celestial Powers.” Jerusalem Studies
in Jewish Thought 6:53–69.
Gruenbaum, Max. 1901. Gesamelte Aufsatze zur Sprach- und Sagenkunde. Berlin: Calvary.
Gruenwald, Ithamar. 1971. “A Preliminary Critical Edition of Sefer Yezira.” Israel Oriental
Studies 1:132–77.
———. 1973a. “Knowledge and Vision.” Israel Oriental Studies 3:63–107.
 Select Bibliography 135

———. 1973b. “Some Critical Notes on the First Part of Sefer Yezira.” Revue des Etudes
Juives 103:475–512.
———. 1980. Apocalyptic and Merkavah Mysticism. Leiden: Brill.
Guttman, Alexander. 1967. “The End of the Jewish Sacrificial Cult.” Hebrew Union College
Annual 38:137–48.
Hadot, Pierre. 1968. Porphyre et Victorinus. Paris: Etudes Augustiniennes.
Halleux, Robert. 1981. Les alchimistes grecs. Paris: Les Belles Lettres.
Halperin, David J. 1980. The Merkabah in Rabbinic Literature. New Haven, Conn.: Amer-
ican Oriental Society.
———. 1988. The Faces of the Chariot: Early Jewish Responses to Ezekiel’s Vision. Tübin-
gen: Mohr.
Hayward, Robert. 1974. “The Memra of YHWH and the Development of Its Use in Targum
Neofiti I.” Journal of Jewish Studies 25:412–18.
Heinemann, Joseph. 1977. Prayer in the Talmud: Forms and Patterns. Berlin. New York: De
Gruyter.
Herrmann, Klaus. 1988. “Text und Fiktion: Zur Textüberlieferung des Shi’ur Qoma.” Frank-
furter Jüdaischtische Beitrag 16:89–142.
Himmelfarb, Martha. 1988. “Heavenly Ascent and the Relationship of the Apocalypses and
the Hekhalot Literature.” Hebrew Union College Annual 59:73–100.
———. 1993. Ascent to Heaven in Jewish and Christian Apocalypses. New York: Oxford
University Press.
Hirschle, Maurus. 1979. Sprachphilosophie und Namenmagie im Neuplatonismus: Mit e.
Exkurs zu “Demokrit” B 142. Meisenheim am Glan: Hain.
Holladay, Carl. 1983. Fragments from Hellenistic Jewish Authors, vol. 1: The Historians.
Chico, Calif.: Scholars Press.
Hopkins, Arthur John. 1927. “Transmutations by Color: A Study of Earliest Alchemy.” Stu-
dien zur Geschichte der Chemie: Festgabe Edmund P. v. Lippman, ed. J. Ruska. Vol.
9–14. Berlin: Springer.
———. 1934. Alchemy, Child of Greek Philosophy. New York: Columbia University Press.
Idel, Moshe. 1981. “The Concept of Torah in Hekhalot Literature and Its Metamorphoses
in Kabbalah.” Jerusalem Studies in Jewish Thought 1:323–84. Hebrew.
———. 1992. “Reification of Language in Jewish Mysticism.” Pages 42–79 in Mysticism
and Language, ed. S. Katz. New York: Oxford University Press.
Jackson, Howard M., ed. and trans. 1978. Zosimos of Panoplis: On the Letter Omega. Mis-
soula, Mont.: Scholars Press.
Jakobson, Roman. 1972. “On Linguistic Aspects of Translation.” Pages 155–75 in Selected
Writings 5. The Hague: Mouton.
Janowitz, Naomi. 1989. The Poetics of Ascent: Theories of Language in a Rabbinic Ascent
Text. Albany: State University of New York Press.
———. 1991. “Theories of Divine Names in Origen and Pseudo-Dionysius.” History of Re-
ligions 31:359–72.
———. 2001. Magic in the Roman World: Pagans, Jews, and Christians. London: Rout-
ledge.
136 Select Bibliography

Jevons, F. B. 1908. “Graeco-Italian Magic.” Pages 93–120 in Anthropology and the Classics,
ed. R. R. Marett. New York: Oxford University Press.
Johnston, Sarah. 1990. Hekate Soteira: A Study of Hekate’s Role in the Chaldean Oracles
and Related Literature. Atlanta, Ga.: Scholars Press.
———. 1992. “Riders in the Sky: Cavalier Gods and Theurgic Salvation in the Second Cen-
tury a.d.” Classical Philology 87:303–21.
———. 1997. “Rising to the Occasion: Theurgic Ascent in Its Cultural Milieu.” Pages
165–93 in Envisioning Magic, ed. Peter Schaefer and Hans Kippenberg. Leiden: Brill.
Jung, Carl Gustav. 1967. Alchemical Studies. Princeton: Princeton University Press.
Kasher, M. 1967. “Notes on sefer ha-razim.” Torah shelemah 22:187–92.
Katz, Steven. 1978. “Language, Epistemology and Mysticism.” Pages 22–74 in Mysticism
and Philosophical Analysis, Studies in Philosophy and Religion 5. London: Sheldon Press.
———. 1988 Winter. “On Mysticism.” Journal of the American Academy of Religion
56:751–57.
Kern-Ulmer, Brigitte. 1996. “The Depiction of Magic in Rabbinic Texts: The Rabbinic and
the Greek Concept of Magic.” Journal for the Study of Judaism 27(3):289–303.
Klein, Michael. 1972. The Translation of Anthropomorphisms and Anthropopathisms in
the Targumin. N.p. (Israel).
Lange, Nicolas de. 1976. Origen and the Jews: Studies in Jewish-Christian Relations in
Third-Century Palestine. Cambridge: Cambridge University Press.
Lauterbach, Jacob. 1936. “Tashlich.” Hebrew Union College Annual 11:207–340.
———. 1939. “The Belief in the Power of the Word.” Hebrew Union College Annual
14:287–302.
Lesses, Rebecca. 1998. Ritual Practices to Gain Power. Harrisburg, Pa.: Trinity Press.
Levenson, Jon. 1994. Creation and the Persistence of Evil: The Jewish Drama of Divine
Omnipotence. Princeton: Princeton University Press.
Leveque, Pierre. 1959. Aurea catena Homeri: Une étude sur l’allegorie grecque. Annales lit-
teraires de l’Université de Besançon. 2nd series, Civilisations de l’antiquité, 27. Paris: Les
Belles Lettres.
Levias, C. 1905. “Gematria.” Pages 589–92 in Jewish Encyclopedia, vol. 5, ed. Cyrus Adler.
New York: Funk & Wagnalls.
Levine, Lee. 1975. Caesarea Under Roman Rule. Leiden: Brill.
Lewy, Hans. 1969. “Sentence Fragments and Nouns in Greek in Hekhalot Rabbati.” Pages
59–65 in Studies in Jewish Hellenism. Jerusalem: Bialik Institute. Hebrew.
———. 1978. Chaldaean Oracles and Theurgy: Mysticism, Magic, and Platonism in the
Later Roman Empire. Second edition. Paris: Etudes Augustiniennes.
Lieberman, Saul. 1939. Sheki’in. Jerusalem: Bamberger et Vahrman. Hebrew.
———. 1962. “Hellenism in Jewish Palestine: Studies in the Literary Transmission, Beliefs,
and Manners of Palestine in the I Century b.c.e.–IV Century c.e.” Second edition. New
York: Jewish Theological Seminary of America.
———. 1965a. Greek in Jewish Palestine: Studies in the Life and Manners of Jewish Pales-
tine in the II–IV Centuries C.E. Second edition. New York: P. Feldheim.
 Select Bibliography 137

———. 1965b. “Mishnat Shir Hashirim.” Pages 118–26 in Jewish Gnosticism Merkabah
Mysticism and Talmudic Tradition, ed. G. Scholem. Second edition. New York: Jewish
Theological Seminary of America.
———. 1974. “Some Notes on Adjurations in Israel.” Pages 69–74 in Texts and Studies.
New York: Ktav.
Lightstone, Jack N. 1984. The Commerce of the Sacred: Mediation of the Divine Among
Jews in the Graeco-Roman Diaspora. Chico, Calif.: Scholars Press.
Lindsay, Jack. 1970. The Origins of Alchemy in Graeco-Roman Egypt. New York: Barnes
& Noble.
Lippmann, E. o. v. 1919. Entstehung und Ausbreitung der Alchemie. Berlin: Springer.
Luck, Georg. 1989. “Theurgy and Forms of Worship in Neoplatonism.” Pages 185–225 in
Religion, Science, and Magic: In Concert and in Conflict, ed. Jacob Neusner, Ernest S.
Frerichs, and Paul V. M. Flesher. New York: Oxford University Press.
Luibheid, Colm, and Paul Rorem. 1987. Pseudo-Dionysius: The Complete Works. New
York: Paulist Press.
Luttikhuizen, Gerard P. 1985. The Revelation of Elchasai: Investigations into the Evidence
for a Mesopotamian Jewish Apocalypse of the Second Century and Its Reception by
Judeo-Christian Propagandists. Tübingen: Mohr.
Maier, Johann. 1968a. “Das Buch der Geheimnisse.” Judaica 24:98–111.
———. 1968b. “Poetisch-liturgische Stucke aus dem ‘Buch der Geheimnisse.”’ Judaica
24:172–81.
———. 1973. “Serienbildung und ‘numinoser’ Eindruckseffeckt in den poetischen Stücken
der Hekhalot Literatur.” Semitics 3:36–66.
Majercik, Ruth. 1989. The Chaldean Oracles: Text, Translation, and Commentary. Leiden: Brill.
Mann, Jacob. 1931–35. Text and Studies in Jewish History and Literature. Cincinnati: He-
brew Union College Press.
Margalioth, Mordecai. 1966. Sefer ha-razim. Jerusalem: Keren Yehudah leb u-Mini
Epshtein she-‘al yad ha-Akademyah le-Mada’e ha-Yahadut be-Artsot ha-Berit.
Hebrew.
Marmorstein, A. 1927. The Old Rabbinic Doctrine of God. New York: Oxford University
Press.
Merchavya, Ch. 1966–67. “Review of Margalioth’s Sefer Ha-Razim.” Kiryat Sefer
42:188–92, 297–303. Hebrew.
———. 1971. “Sefer Ha-Razim.” Pages 1594–95 in Encyclopedia Judaica, vol. 13.
Jerusalem: Ktav.
Mertz, Elizabeth. 1996. “Recontextualization as Socialization: Text and Pragmatics in the
Law School Classroom.” Pages 229–49 in Natural Histories of Discourse, ed. Michael
Silverstein and Greg Urban. Chicago: University of Chicago Press.
Mettinger, Tryggve N. D. 1982. The Dethronement of Sabaoth: Studies in the Shem and Ka-
bod Theologies. Coniectanea Biblica, Old Testament Series 18. Lund: Gleerup.
Mitchell, Christopher Wright. 1987. The Meaning of BRK “to bless” in the Old Testament.
Atlanta, Ga.: Scholars Press.
138 Select Bibliography

Mopsik, Charles. 1994. “La datation du Chi’our Qomah d’après néotestamentaire.” Revue
des Sciences Religieuses 68(2):131–44.
Morgan, Michael A. 1983. Sepher ha-razim: The Book of the Mysteries. Texts and transla-
tions, Pseudepigrapha series, 25.11. Chico, Calif.: Scholars Press.
Morray-Jones, C. R. A. 1992. “Transformational Mysticism in the Apocalyptic-Merkabah
Tradition.” Journal of Jewish Studies 43:1–31.
———. 1993. “Paradise Revisited (2 Corinthians 12:1–12): The Jewish Mystical Back-
ground of Paul’s Apostolate, 1 and 2.” Harvard Theological Review 86:177–217,
265–92.
———. Forthcoming. “The Body of the Glory: Si’ur Qomah and Transformational Mysti-
cism in the Epistle of the Ephesians.” Unpublished article.
Neale, J. M. 1976. A History of the Holy Eastern Church. New York: AMS Press.
Negev, A. 1973. “The Staircase-Tower in Nabatean Architechture.” Revue Biblique
80:364–82.
Nemoy, Leon. 1930. “Al-Qirqisani’s Account of the Jewish Sects and Christianity.” Hebrew
Union College Annual 7:317–97.
———. 1986. “Al-Qirqisani on the Occult Sciences.” Jewish Quarterly Review 76:329–67.
Neusner, Jacob. 1966–70. A History of the Jews in Babylonia. Leiden: Brill.
———. 1969. “The Phenomenon of the Rabbi in Late Antiquity.” Numen 19:1–20.
———. 1989. “Science and Magic, Miracle, and Magic in Formative Judaism: The System
and the Difference.” Pages 61–81 in Religion, Science, and Magic: In Concert and in
Conflict, ed. Jacob Neusner, Ernest S. Frerichs, and Paul Virgil McCracken Flesher. New
York: Oxford University Press.
Newsom, Carol. 1985. Songs of the Sabbath Sacrifice: A Critical Edition. Atlanta, Ga.:
Scholars Press.
———. 1990a. “Sectually-Explicit” Literature from Qumran.” Pages 167–87 in The He-
brew Bible and Its Interpreters, ed. William Henry Propp, Baruch Halpern, and David
Noel Freedman. Winona Lake, Ind.: Eisenbrauns.
———. 1990b. “He Has Established Himself Priests”: Human and Angelic Preisthood in the
Qumran Shabbath Shirot.” Pages 101–20 in Archaeology and History in the Dead Sea
Scrolls, ed. Lawrence Schiffman. Sheffield: Sheffield Academic Press.
Niggemeyer, J. H. 1975. Beschworungsformeln aus dem “Buch der Gehemnisse.” Hilder-
sheim: Georg Olms Verlag.
Nilsson, Martin. 1948. Greek Piety. Oxford: Clarendon.
———. 1963. “The High God and the Mediator.” Harvard Theological Review 56:101–20.
Nock, Arthur Darby, ed. and trans. 1966. Sallustius: Concerning the Gods and the Universe.
Hildesheim: Georg Olms.
Parmentier, Richard. 1985. “Semiotic Mediation: Ancestral Genealogy and Final Interpre-
tant.” Pages 359–85 in Richard Parmentier and Elizabeth Mertz eds., Semiotic Media-
tion. Orlando, Fla.: Academic Press.
———. 1994. Signs in Society: Studies in Semiotic Anthropology. Bloomington: Indiana
University Press.
 Select Bibliography 139

Patai, Raphael. 1994. The Jewish Alchemists: A History and Source Book. Princeton:
Princeton University Press.
Pearson, Birger. 1981. Nag Hammadi Codices IX and X. Leiden: Brill.
———. 1984. “Gnosticism as Platonism, with Special Reference to Marsanes (NCH 10.1).”
Harvard Theological Review 77:55–72.
Peirce, Charles. 1940. The Philosophical Writings of Charles Peirce. New York: Dover.
Penner, Hans. 1983. “The Mystical Illusion.” Pages 89–116 in Mysticism and Religious Tra-
ditions, ed. S. Katz. New York: Oxford University Press.
Pfister, R. 1935. Teinture et Alchimie dans l’orient hellenistique. Seminarium Kondako-
vianum. Praha: Institut Kondakov.
Phillips, C. R. 1986. “The Sociology of Religious Knowledge in the Roman Empire to a.d.
284.” Pages 2677–773 in Aufsteig und Niedergang der römischen Welt, vol. 2: Principat,
ed. W. Hasse. Berlin: De Gruyter.
Pines, Shlomo. 1989. “Points of Similarity Between the Exposition of the Doctrine of the Se-
firot in the Sefer Yezira and a Text of the Pseudo-Clementine Homilies: The Implications
of This Resemblance.” Pages 63–142 in Israel Academy of Sciences and Humanities Pro-
ceedings, vol. 7.
Places, Edouard des, ed. and trans. 1971. Oracles Chaldaïques. Paris: Les Belles Lettres.
———. 1983. “Notes sur quelques oracles chaldaïques.” Pages 319–30 in Mélanges
Edouard Delebecque. Aix-en-Provence: Université de Provence.
———. 1984. “Les Oracles chaldaïques.” Pages 2299–335 in Aufsteig und Niedergang der
römischen Welt, vol. 2.17.4, ed. W. Haase. Berlin: De Gruyter.
Plessner, M. 1976. “Zosimus.” Pages 631–32 in Dictionary of Scientific Biography, vol. 14,
ed. C. Gillispie. New York: Charles Scribner’s Sons.
Praechter, Karl. 1934. “Theodoros.” Pages 2–10 in Paulys Realencyclopädie der Classischen
Altertumswissenschaft, 2nd series, vol. 10, 1833–38. Stuttgart: Alfred Druckenmüller
Verlag.
Price, Simon. 1984. “Gods and Emperors: The Greek Language of the Roman Imperial
Cult.” Journal of Hellenic Studies 104:79–95.
———. 1987. “From Noble Funeral to Divine Cult: The Consecration of Roman Emper-
ors.” Pages 56–105 in Ritual of Royalty: Power and Ceremonial in Traditional Societies,
ed. David Cannadine and Simon Price. Cambridge: Cambridge University Press.
Proudfoot, Wayne. 1985. Religious Experience. Berkeley and Los Angeles: University of
California Press.
Rappaport, Roy. 1980. “Concluding Comments on Ritual and Reflexivity.” Semiotica
30:181–93.
Rappe, Sara. 1995. “Metaphor in Plotinus’ Enneads 5.8.9.” Ancient Philosophy 15:155–72.
Reif, Stefan C. 1993. Judaism and Hebrew Prayer: New Perspectives on Jewish Liturgical
History. Cambridge: Cambridge University Press.
Riess, Ernst. 1893. “Alchemie.” Pages 1338–55 in Paulys Realencyclopädie der Classischen
Altertumswissenschaft, vol. 1, ed. Georg. Wissowa. Stuttgart: Alfred Druckenmüller Ver-
lag.
140 Select Bibliography

Rist, John M. 1967. Plotinus: The Road to Reality. Cambridge: Cambridge University Press.
Roberts, A., and J. Donaldson, eds. 1979. The Apostolic Fathers with Justin Martyr and Ire-
naeus, vol. 1 of The Ante-Nicene Christian Fathers, Translations of the Writings of the
Fathers Down to a.d. 325. Grand Rapids, Mich.: Eerdmans.
Rudolph, Kurt. 1983. Gnosis: The Nature and History of an Ancient Religion. Edinburgh:
T. & T. Clark.
Ruelle, C. E. 1889. “Le chant des sept voyelles grecques d’aprés Démétrius et les papyrus de
Leyde.” Revue des Etudes Grecques 2:38–44.
Saffrey, H. D. 1981. “Les Néoplatoniciens et les Oracles Chaldaïques.” Revue des Etudes
Augustiniennes 26:209–25.
———. 1982. “New Objective Links Between the Pseudo-Dionysius and Proclus.” Pages
64–74 in Neoplatonism and Christian Thought, ed. D. O’Meara. Albany: State Univer-
sity of New York Press.
Sathas, C. 1875. “Sur les commentaires byzantine.” Annuaire de l’Association des Etudes
Grecques 9:187–222.
Schäfer, Peter. 1987. übersetzungen der Hekhalot-Literatur. Tübingen: Mohr.
———. 1988a. “The Problem of the Redactional Identity of Hekhalot Rabbati.” Pages
63–74 in Hekhalot-Studien. Tübingen: Mohr. German.
———. 1988b. “Shiur Qoma: Rezensionen und Urtext.” Pages 75–83 in Hekhalot- Studien.
Tübingen: Mohr.
———. 1990. “Jewish Magic Literature in Late Antiquity and Early Modern Ages.” Journal
of Jewish Studies 41(1):75–91.
———. 1992. The Hidden and Manifest God: Some Major Themes in Early Jewish Mysti-
cism. Albany: State University of New York Press.
———. 1996. “Jewish Liturgy and Magic.” Pages 1:531–55 in Geschichte—Tradition—
Reflexion. Tübingen: Mohr.
Schäfer, Peter, Margarete Schlüter, and Hans-Georg von Mutius. 1981. Synopse zur
Hekhalot-Literatur. Tübingen: Mohr.
Schiffman, Lawrence. 1982. “Merkavah Speculation at Qumran: The 4Q Sereckh Shirot
‘Olat ha-Shabbat.” Pages 15–47 of Mystics, Philosophers, and Politicians, ed. J. Rein-
hartz. Durham, N.C.: Duke University Press.
Schlüter, Margarete. 1985. “Untersuchungen zur Form und Funktion der Berakha in der
Hekhalot-Literatur.” Frankfurter Judaistische Beitrage 13:83–146.
Scholem, Gershom. 1954. Major Trends in Jewish Mysticism. Third edition. New York:
Schocken Books.
———. 1965a. Jewish Gnosticism, Merkabah Mysticism, and Talmudic Tradition. New
York: Jewish Theological Seminary of America.
———. 1965b. On the Kabbalah and Its Symbolism. New York: Schocken Books.
———. 1972a. “The Name of God and the Linguistic Theory of the Kabbalah.” Diogenes
79/80:59–80, 164–94.
———. 1972b. “Gematria.” Pages 369–74 in Encyclopedia Judaica, vol. 7. Jerusalem: Ktav.
———. 1991. On the Mystical Shape of the Godhead: Basic Concepts in the Kabbalah. New
York: Schocken.
 Select Bibliography 141

Segal, Alan. 1980. “Heavenly Ascent in Hellenistic Judaism, Early Christianity, and Their
Environment.” Pages 1333–94 in Aufsteig und Niedergang der römischen Welt, II, Prin-
cipat, ed. W. Hasse, vol. 23. Berlin: De Gruyter.
———. 1981. “Hellenistic Magic: Some Questions of Definition.” Pages 349–75 in Studies
in Gnosticism and Hellenistic Religion Presented to Gilles Quispel, ed. R. van den Broek
and M. J. Vermasseren. Leiden: Brill.
Shaw, Gregory. 1985. “Theurgy: Rituals of Unification in the Neoplatonism of Iamblichus.”
Traditio 41:1–28.
———. 1995. Theurgy and the Soul: The Neoplatonism of Iamblichus. University Park, Pa:
The Pennsylvania State University Press.
Sheldon-Williams, I. P. 1972. “Henads and Angels: Proclus and the Ps. Dionysius.” Studia
Patristica 11:65–71.
Sheriff, John K. 1989. The Fate of Meaning: Charles Peirce, Structuralism, and Literature.
Princeton: Princeton University Press.
Silverstein, Michael. 1976. “Shifters: Linguistic Categories and Cultural Descriptions.”
Pages 11–55 in Meaning in Anthropology, ed. K. Basso and H. Selby. Albuquerque: Uni-
versity of New Mexico.
———. 1978. “Language Structure and Linguistic Ideology.” Pages 193–247 in The Ele-
ments, ed. P. Clyne, W. Hanks, and C. Hofhaur. Chicago: Chicago Linguistic Society.
———. 1981. “Metaforces of Power in Traditional Oratory.” Unpublished lecture.
———. 1993. “Metapragmatic Discourse and Metapragmatic Function.” Pages 33–58 in
Reflexive Language: Reported Speech and Metapragmatics, ed. J. Lucy. Cambridge:
Cambridge University Press.
Simon, Marcel. 1986. Verus Israel: A Study of the Relations Between Christians and Jews in
the Roman Empire (135–425). New York: Oxford University Press.
Smith, Andrew. 1974. Porphyry’s Place in the Neoplatonic Tradition: A Study in Post-Plo-
tinian Neoplatonism. The Hague: M. Nijhoff.
Smith, Huston. 1990. “Is There Perennial Philosophy?” Journal of the American Academy
of Religion 55:553–66.
Smith, Jonathan Z. 1974. “Hellenistic Religions.” Pages 749–51 in Encyclopedia Britan-
nica, vol. 8. Chicago.
———. 1978. Map Is Not Territory: Studies in the History of Religions. Studies in Judaism
in Late Antiquity 23. Leiden: Brill.
———. 1982. Imagining Religion: From Babylon to Jonestown. Chicago Studies in the His-
tory of Judaism. Chicago: University of Chicago Press.
Smith, Morton. 1963. “Observations on Hekhalot Rabbati.” Pages 142–60 in Biblical and
Other Studies, ed. A. Altmann. Cambridge, Mass.: Harvard University Press.
———. 1965. “The Ascent of Simon Magus in Acts 8.” Pages 735–49 in Harry Austryn
Wolfson Jubilee Volume on the Occasion of His Seventy-Fifth Birthday, English Section,
ed. Harry Austryn Wolfson. Jerusalem: American Academy for Jewish Research.
———. 1973. Clement of Alexandria and a Secret Gospel of Mark. Cambridge, Mass.: Har-
vard University Press.
———. 1978. Jesus the Magician. New York: Harper & Row.
142 Select Bibliography

———. 1979. “Relations Between Magical Papyri and Magical Gems.” Papyrologica Brux-
ellensia 18:129–36.
———. 1981. “Ascent to the Heavens and the Beginning of Christianity.” Eranos Jahrbuch
50:403–29.
———. 1982. “Helios in Palestine.” Erets-Yisra’el 16:199–214.
———. 1986. “Salvation in the Gospels, Paul, and the Magical Papyri.” Helios 13:63–74.
———. 1990. “Ascent to the Heavens and Deification in 4MQa.” Pages 181–88 in Archae-
ology and History in the Dead Sea Scrolls, ed. Lawrence Schiffman. Sheffield: Sheffield
Academic Press.
Sperber, Daniel. 1966. “On Sealing the Abyss.” Journal of Semitic Studies 11:168–74.
Stambursky, Samuel. 1976. “The Source and Reality of the Term ‘Gematria,’” Tarbits
45:268–71.
Stern, Menachem. 1974–84. Greek and Latin Authors on Jews and Judaism. Jerusalem: Is-
rael Academy of Sciences and Humanities.
Swartz, Michael D. 1990. “Scribal Magic and Its Rhetoric: Formal Patterns in Medieval He-
brew and Aramaic Incantation Texts from the Cairo-Genizah.” Harvard Theological Re-
view 83:163–80.
———. 1996. Scholastic Magic: Ritual and Revelation in Early Jewish Mysticism. Prince-
ton: Princeton University Press.
Tabor, James. 1986. Things Unutterable: Paul’s Ascent to Paradise in Its Greco-Roman, Ju-
daic, and Early Christian Contexts. Lanham, Md.: University Press of America.
Talmon, S. 1978. “The Emergence of Institutionalized Prayer in Judaism in the Light of the
Qumran Literature.” Pages 265–84 in Qumran: Sa piété, sa théologie, et son milieu, ed.
M. Delcor. Paris: Leuven.
Tambiah, Stanley J. 1968. “The Magical Power of Words.” Man 3:175–208.
———. 1979. “A Performative Approach to Ritual.” Proceedings of the British Academy
113–69.
———. 1985. Culture, Thought, and Social Action: An Anthropological Perspective. Cam-
bridge, Mass.: Harvard University Press.
———. 1990. Magic, Science, Religion, and the Scope of Rationality. Cambridge: Cam-
bridge University Press.
Taylor, F. Sherwood. 1930. “A Survey of Greek Alchemy.” Journal of Hellenistic Studies
50:109–39.
———. 1937. “The Origins of Greek Alchemy.” Ambix 1:1–37.
———. 1949. The Alchemists, Founders of Modern Chemistry. New York: H. Schuman.
Thee, Francis C. R. 1984. Julius Africanus and the Early Christian View of Magic. Tübin-
gen: Mohr.
Trouillard, Jean. 1975. “L’activité onomatique selon Proclos.” Pages 239–51 in De Jam-
blique à Proclus. Geneva: Vandoeuvres.
Vaux, R. de 1970. “The Revelation of the Divine Name YHWH.” Pages 40–75 in Procla-
mation and Presence: Old Testament Essays in Honor of G. H. Davies, ed. J. I. Durham
and J. R. Porter. Richmond, Va.: John Knox.
 Select Bibliography 143

Versnel, H. S. 1991. “Magic and Religion—Some Reflections on the Relationship.” Numen

38:177–97.
Waszink, J. H. 1950. “Biothanati.” Pages 391–94 in Reallexikon fur Antike und Christen-
tum, vol. 2, ed. Theodor Klauser. Stuttgart: Hiersemann.
Weltin, E. G. 1960. “The Concept of Ex-Opere-Operato Efficacy in the Fathers as Evidence
of Magic in Early Christianity.” Greek, Roman, and Byzantine Studies 3:74–100.
Wenschkewitz, H. 1932. “Die Spritiualisierung der Kultusbegriffe: Tempel, Priester und
Opfer im NeuenTestament.” Angelos 4:71–230.
Wernberg-Moller, P. 1962. “An Inquiry into the Validity of the Text-Critical Argument for
an Early Dating of the Recently Discovered Palestinian Targum.” Vetus Testamentum
12:312–30.
Wolfson, Elliot. 1987. “Circumcision and the Divine Name: A Study in the Transmission of
Esoteric Doctrine.” Jewish Quarterly Review 78:77–112.
———. 1994. Through a Speculum That Shines: Vision and Imagination in Medieval Jew-
ish Mysticism. Princeton: Princeton University Press.
Wutz, Franz. 1914. Onomastica Sacra: Untersuchungen zum Liber Interpretationis
Nominum Hebraicorum des Hieronymus. Texte und Untersuchungen zur Geschichte der
altchristlichen literatur, 41.1. Leipzig: J. C. Hinrich.
Yadin, Yigael. 1962. The Scroll of the War of the Sons of Light Against the Sons of Dark-
ness. London: Oxford University Press.
York, Anthony. 1974. “The Dating of Targumic Literature.” Journal for the Study of Ju-
daism 5:49–62.
Zaehner, R. C. 1960. Hindu and Muslim Mysticism. London: University of London, Athlone
Press.
 Primary Source Index
* See Bibliography, Primary Sources for edition cited

Old Testament Jubilees

10:13, 103
Genesis 36:7, 27
5:24, 64 1 Maccabees
Exodus 5:67, 101 n. 48
3:14–15, 20–21 2 Maccabees
20:24, 2 12:40, 101 n. 48
35:31, 57 n. 34 4 Maccabees
40–48, 67 17:5, 65
Judges Martyrdom of Isaiah
13:18, 29, 35, 38 1:7, 38 n. 21
1 Kings Odes of Solomon
8, 22 4:7–8, 27 n. 26
Isaiah Prayer of Manasseh
6:3, 67 1:3, 27
Proverbs Sirach
3:19, 57 n. 34 43:5, 26 n. 19
Psalms Testament of Job
8:2, 47 48–50, 67
19:1, 47
33:6,9, 26
102:1, 37 Dead Sea Scrolls (Qumran)

Blessings
Apocrypha and Pseudepigrapha 1QSb, 71
Songs of the Sabbath Sacrifice
Apocalypse of Abraham 4Q400–407, 67
15:7, 67 4Q403 1.2.13–15, 68
Ascension of Isaiah 4Q403 1.2.20–28, 69
10.24, 79 Temple Scroll
1 Enoch 11Q19 col 30, 109
64 n. 7 War Scroll variant
2 Enoch 4QMa frag 11, col 1, 13–14, 71
22, 53
3 Enoch
9:1–2, 53 n. 27 New Testament
9:12, 53
9–13, 53 n. 27 John
13:1, 28 n. 27, 54 1:1, 37
48d:5, 79 2 Corinthians 12, 65
146 Primary Source Index

Rabbinic Sources bHullin

41b, 104 n. 62
PREFIXES: m=Mishnah, t=Tosefta, bKetubot
j=Jerusalem Talmud, b=Babylonian 77b, 92
Talmud bMaccot
11a, 27 n. 26
Semahot bMegillah
8, 104 29a, 104 n. 59
mAvot bMenahot
1:1, 103 n. 55 29b, 28 n. 27
mMaccot bMoed Katan
3:6, 80 28a, 104 n. 59
mSanhedrin bNazir
7:5, 27 5a, 51
mShabbat bNiddah
8:2, 101 n. 49 13a, 104 n. 60
mSotah bPesachim
9:15, 96 n. 34 50a, 51
mTa’anit bSanhedrin
3:8, 96 n. 34 17a, 103 n. 52
mYoma 29a, 27 n. 26
8:3, 80 56a, 27
tBerakot
65a, 57
6:24, 80 n. 81
67b, 14
tMaccot
107b, 28 n. 28
4:15, 80
bShabbat
tNedarim
61b, 101 n. 49
1:1, 91 n. 20
67a, 104 n. 58
tShabbat
81b, 105 n. 71
7:16, 104
110b, 104 n. 58
7:23, 93
137b, 80 n. 81
tSota
2:1, 91 bShevuot
jBerakot 29b, 92 n. 26
9:3, 80 n. 81 bSota
jSanhedrin 18a, 91
7:13, 104 38a, 54, 29 n. 31
10:2, 27 n. 26 47a, 28 n. 28
jYoma bSuccah
3:7, 28 n. 30 28a, 18, 51
bBabaBatra bTa’anit
153a, 106 19a, 23ab, 96 n. 34
bBerakot 21b, 104 n. 61, 105 n. 69
50b, 104 n. 64 bYoma
55a, 56 20a, 51
59a, 105 n. 69 39b, 28 n. 30
bErubin *Targum Pseudo-Jonathan
13a, 30 n. 36 Ex 28:30, 27 n. 26
bGittin *Targum Neofiti
45a, 106 n. 72 Ex 3:12, 21–22
68b, 28 Ex 3:14–15, 22
 Primary Source Index 147

Genesis Rabba Book of Creation (Sefer Yetsira)

1.10, 28 2.5, 55
5.5, 13.7, 96 *Book of Secrets
12.10, 27–8 1.9, 1.187, 99
42.5, 33 1.27–28, 86
Exodus Rabba 1.29, 1.94, 2.95, 2.182, 3.47–60, 5.15,
1.29, 28 87 n. 11
Numbers Rabba Naso 1.48, 1.60, 100
10.5, 36 1.57, 1.64, 91 n. 22
Song of Songs Rabba 1.68, 1.70, 1.98, 1.126, 1.146, 1.172,
1.2, 5.1, 8.1, 28 n. 29 1.179, 1.189, 1.226, 2.83, 2.166,
7.8–9, 91 3.20, 3.28, 3.38, 3.52, 4.30, 4.41,
Ruth Rabba 4.68, 5.23, 6.35, 90 n. 14
2.5, 33 1.70, 90
Midrash Samuel 1.136, 1.162, 91 n. 22
26, 27 n. 26 1.165, 1.203, 2.21–23, 4.61, 91
Midrash on Psalms (Midrash Tehillem) 2.33, 2.147, 4.64, 91 n. 21
3.2, 30 2.110–12, 2.123–25, 2.134–38,
91.8, 28 n. 30 2.182–85, 101 n. 47
Pirkei R. Eliezer 6, 71 n. 43
ch. 3, 30 n. 35 7.39, 94
ch. 46, 104 n. 63 Nachmanides, Commentary on Genesis
1:1, 30

Other Hebrew Sources

Greek and Latin Sources
Hekhalot Rabbati
par 1–94, 80 n. 82
*Albinus, Introduction
par 106–151, 214, 252, 75
3, 17, 151
par 107, 198, 72
Anastasius of Sinai, Quaestiones
par 159–60, 76 PG 89, col. 252ab, 7 n. 21
par 204, 78 Aristotle, On Interpretation
Ma’aseh Merkabah 16.1, 35
section 19, 66 n. 27 Pseudo–Aristotle, On the World
par 566, 79 n. 74, 80 n. 79 7, 38 n. 21
par 569, 80 n. 79 Arnobius of Sicca, Against the Nations
*Masseket Hekhalot 1.43–44, 15 n. 51
2.46, 28 2.62, 75 n. 56
*Merkabah Rabba *Artemidorus, The Interpretation of
4–5 (Cohen), 53 n. 25 Dreams
709 (Schäfer), 79 n. 75 2.70, 50
*Midrash Konen Augustine, City of God
2.23–24, 30 n. 34 10.9, 9
Sefer Ha–qomah 10.10, 13
120–21, 53 *Censorinus, Concerning the Day of
Sefer Hashi’ur Birth
1, 53 chaps. 7, 14, 50 n. 17
Sefer Raziel *Chaldean Oracles
98–99, 53 Fr 2.3, 109, 25 n. 14
Siddur Rabba Fr. 7, 123–24, 132–33, 135, 149–50,
55–56, 53 206, 208, 222, 224, 8
148 Primary Source Index

Chaldean Oracles (cont.) Letter

Fr. 28, 12 9.1 1108A, 41
Fr. 76–77, 87, 8 n. 25 Epiphanius
Fr. 107, 9 Against the Heresies
Fr. 110, 153, 7 34.4.1–7, 46 n. 1
John Chrysostom Medicine Chest
Against the Judaizers 40.2, 16 n. 54
1.6, 1 Eunapius Lives of the Philosophers
8.8, 940, 15 475, 11 n. 38
Homilies on Colossians Eusebius, Preparation for the
8.5 (Col 3:5), 15 Gospels
Clement of Alexandria 11.6, 60
Exhortation 5.10, 10
1, 2 n. 4 5.21, 98
Miscellanies 9.27, 26
1.23, 26 First Clement 61.59.2, 33
The Teacher *Greek Magical Papyri (PGM)
3.4, 2 n. 4 1, 3–5, 7, 12–13, 36, 85 n. 4
Pseudo–Clementine, Homilies 1.134, 90 n 16
5.4–7, 2 n. 4 2.110, 91 n. 20
Codex Justinian 3.215, 93
9.18.2, 111 n. 11 4.80, 4.296–466, 4.3210, 7.540–78, 87
Corpus Hermeticum n 13
1.26, 45, 67 n. 32 4.247, 87
5.1, 38 n. 21 4.2645–46, 96
16.2, 45 5.460–85, 61 n. 48
Council of Laodocia, Canon 11b.1–5, 87 n. 11
35–37, 1 12.193–201, 114
*Damascius, Commentary on the Phaedo 13.760–932, 60–61
1.1721, 127 13.764–65, 13.775–80, 13.795–96,
Demetrius of Phalerium 13.800, 13.804, 61
On Style 71, 60 *Greek Writers on Music (Musici
Pseudo–Democritus Scriptores Graeci)
Physical Things and Mysteries 276.8–18, 60
Ber 2.1, 112 Herodotus, Histories
Synesius the Philosopher to Dioscorus 7.114., 2 n. 7
Ber 2.3.0, 116 2.171, 7 n. 32
Dio Cassius, History Hippolytus, Refutation of all Heresies
56.46.2, 65 n. 17 4.14, 51
*Dionysius the Areopagite 4.15, 51
Celestial Hierarchies 4.43–4, 51
1.121CD, 13 5.10, 79 n. 77
Divine Names 6.37, 46 n. 1
1.1 588B, 38 Homer, Iliad
1.4 592B, 42 4.49, 9, 50, 48 n. 42
1.6 596A, 38 *Iamblichus
4.11 708C, 42 n. 30 Commentary on the Timaeus frag.
9.1 909B, 40 57.9–22, 51
Eccesiastical Hierarchy On the Mysteries
2.1, 42 1.12;41.13, 3.4;109,14–15, 10.7;
3.5 432B, 13 293.5–8, 128
 Primary Source Index 149

2.11;96.13–97.9, 10 n. 35 4.476, 4.647, 4.501, 83

3.13;129.14–132.2 , 59 4.487–88, 81
3.25;160.15, 5.7–9;207.7–210.14, 4.539–40, 4.560, 4.570–75, 4.574,
6.4;244.13–245.10, 11 4.704–708, 4.721, 4.725, 82
4.2;184.8, 10.4;289.3–6, 128 Nag Hammadi Library
5.8;208.15, 15 n. 53 Discourse on the Eighth and Ninth
5.11;215.3–7, 10 (6.6.61;60.17–61.17), 60 n. 46
7.4;254.17–255.6, 59 The Gospel of the Egyptians
7.6;246.17247.2, 94 (3.2.43–44,66), 60 n. 46
10.1–7;285.10–293.13, 5 Marsanes
On the Pythagorean Life 25*1–26*12, 27*23–30*2, 82
54.21–25, 51 Origen
Irenaeus, Against the Heresies Against Celsus
1.13–22, 46 n. 1 1.6, 1.67, 4.33–34, 5.46, 36
1.13.1, 2 n. 6 1.24, 34
1.14.1, 46 1.25, 33, 37
1.14.2, 49 n. 8 1.25–26, 2n.5
1.14.3, 48 3.28, 4.92–93, 99
1.14.5, 1.15.1, 49 4.32, 8.62, 98
1.14.7, 50 4.34, 38
1.14.8, 47, 50 4.48, 33
1.21.4, 16 5.41, 35
2.32.5, 94 n. 27 5.45, 37
Jerome, Letter to Marcella 6.24–38, 64 n. 9
25, 48 n. 5 6.27, 79
Josephus 6.40, 36 n. 10
Against Apion, 2:167, 38 n. 21 7.6, 8.60, 99 n. 44
Jewish Wars 2:128, 107 7.9, 67 n. 32
Justin 8.37, 37 n. 16
First Apology 8.69, 35 n. 7
14.2, 2 n. 4 Commentary on John
61.11, 63.1, 38 n. 21 1.34, 36
Second Apology Commentary on Matthew
6.1, 38 n. 21 13.23, 98 n. 43
Dialogue With Trypho Exhortation to Martyrdom
85.3, 1 n. 2, 15 45, 98 n. 42
Juvenal, Satire 46, 35
6.546, 2 n. 5 Homilies on Exodus
Lucian 5.1 PG 12.327, 33 n. 3
Alexander the False Prophet Homilies on Genesis
13, 67 n. 32 12.4 PG 12.227c, 33 n. 3
The Gout Tragedy Homilies on Joshua
171–173, 2 n. 5 13.2, 34
Marinus, Life of Proclus 20.1, 36–37
22, 28, 12 23.4, 35 n.8
Maximus of Tyre, Philosophical Homilies on Numbers
Orations 11.4, 35 n. 14
8:10, 38 13.4, 25.3, 27.5, 27.13, 34 n. 5
Mithras Liturgy (Papyrus 574 of the 14.2, 35 n. 8
Bibliotheque Nationale, PGM) Homilies on the Songs of Songs
4.475–829, 81 1.6, 33
150 Primary Source Index

On First Principles On Abstinence

1.8.1, 35 n. 8, 99 2.20, 98 n. 40
On Prayer 2.42, 98
24.2, 35 *Proclus
Selections on Exodus Commentary On the Alcibiades
PG 12.284B, 14 357.12, 17
Selections on Genesis Commentary On the Cratylus
17:5 PG 12.116AB, 34 42, pp. 13.19–27, 41
Iulius Paulus, Excerpts 51, pp. 18.27–19.17, 41 n. 26
5.23.14–18, 18 71, pp. 29–35, 12, 41 n. 25, 42
Pelagus, “On the Divine and Sacred Art” 72, p. 36, 12 n. 45
Ber 4.1.8, 117 Commentary On the
Philostratus, Life of Apollonious of Tyana Parmenides
8.30, 15, 65 n. 18 851.8–10, 41 n. 26
7.38, 15 853.1–3, 41
Philo Commentary On the Republic
Allegorical Interpretation 1.152.10, 83 n. 91
1.14, 60 1.255.19, 1.290.14, 12 n. 46
Life of Moses 2.123.9–10, 7
1.75, 38 n. 21 2.143.20–24, 58
Commentary On the Timeaus
On Abraham
1.274.16, 3.155.18–25, 12
5.1, 38 n. 21
2.274.10–278.25, 57 n. 37
On the Decalogue
2.274.16–23, 57
74, 98 n. 41
2.276.2, 58
Questions on Exodus
2.278.9, 59
40, 65
3.20.22, 7
Who is the Heir
3.155.18–25, 12
259–66, 67 n. 32
Elements of Theology
Pistis Sophia
Propositions
136, 61 n. 48
57, 145, p12
Plato On the Hieratic Arts
Philebus 46–47, 102
12C, 33 Platonic Theology
Timaeus 1.25.113, 11 n. 39
35A–36A, 82 n. 88 1.26.63, 11
Theaetetus 1.29.123–24, 41
176B, 129 1.29.124, 12 n. 45
Pliny, Natural History 4.26.220, 11
30.11, 30.13, 2 Psellus
24.160, 112 n. 16 The Golden Chain, 7
Plotinus, Enneads Minor Writings
1.6, 16 n. 56 1.446.8–447.14, 7 n. 22
1.6.9, 16 Minor Writings
2.9.14, 4.4.26, 15 1.446.23, 8
4.4.43, 17 n. 59 Sallustius, On the Gods
5.5.5, 57 n. 40 16, 17
5.8.6, 58 The Second Book of Jeu
Porphyry 52 GCS 13.315–21, 63, 79 n. 76
Life of Plotinus Strabo, Geography
10, 16, 17 n. 59 16.2.43, 2 n. 5
 Primary Source Index 151

Suda 2.276.2, 134

s.v. Diocletian, 111 Theodosius Alexandrinus, The Theology
s.v. Julianus 1.433–34, 6 of Arithmetic
Suetonius, Augustus 40.9, 58
100.4, 65 n. 17 Zosimos
Nero “About the Philosopher’s Stone”
16.2, 2 n. 5 Ber 3.29.1, 117
Teachings of the Apostles “The Four Bodies”
21, 36 n. 12 Ber 3.19.3, 239, 116
Tertullian Ber 3.12.4, 234, 114 n. 21
On Idolatry “The Final Reckoning”
20–21, 36 n. 12 Ber 51.8, 118
Prescription Against the Heretics “On the Body of Magnesium”
PL 70A, 50 Ber 3.28.2, 8, 115
Theodoret [Pseudo–Justin] Answer to the *“On the Letter Omega”
Orthodox section 7, 236, 115
24, 15 n. 51 “On Virtue”
Theodorus, Commentary On the Timeaus Ber 3.1.4, 120
2.274.16–23, 57 Ber 3.6.1, 109
 General Index
adjurations, 87–96 by Proclus, 12
Albinus, 17 by Rabbi Nehunya, 63, 71–72, 81
alchemy, xi–xii, 109–20, 127 rabbinic controversies over, 127
harnessing arts for religious goals, 111 rituals, 8, 12, 64–66, 79, 126
as reinterpreted sacrifice, 109, 114 of the soul, 8
theology of, 117–20 texts, 64–65, 68–84
amen, 47 astrology, 11, 106
amulets, xiv, 79, 83, 88–90, 101–2, 125 Augustine, 9, 13–14
definition of, 101 Austin, J. L., xx
as indexical icon, 101 Azariah, Rabbi, 78
kissing, 81–82
names inscribed in, 101 Baer, Y., 99
recipes for, 101 Barton, T., 106
analogies, 6, 56, 96, 100, 114, 124 Baumgarten, J., 68
misguided, 100–101 Betz, H. D., 96
angels, xi, 1, 9, 68, 70, 79, 85, 88–94, Bezalel, 56, 57
99–101, 104, 123, 125 binding spells, 93
as assistants, xi, 90 blackening (alchemical term), 114–15,
fallen as source of alchemy, 118 119
giant figure as, 49 blasphemy, 27
names of, 78, 88–90, 92, 99, 101 Blau, L., xiv, xv
patriarchs are, 36 blessing, 68, 71, 95
rabbinic interest in, 104 blessings, 71, 95, 126
recitations of, 67, 87 automatic quality of, 20
and sacrifice, 98–99, 103 efficacy of, 20
summoning, 125 rabbinic, 29, 77, 104
animal sacrifice. See sacrifice, animal blood, 88, 96–99, 105
Apollonius of Tyana, 15 body, 49, 59, 101, 119–20
Aramaic translations. See Targums analogous to consonants, 6
Aristeas, 99 of deities, 48–49, 128
Aristotle, 34, 110 divine, 52–53
Artapanus, 26 first man and, 112
Artemidorus, 50 and Jewish deity, 48
ascent, xii, 4, 64, 73–84, 118, 123–25 Julian as angel in human, 128
and cosmology, 125 life in the, xi
effected by speech, 66, 70 losing in alchemy, 115, 119
effected by trance, 66 metals are, 115
first-person claims, 64 and soul, 7, 60, 63
formats of, 64 transformation of, 119
goals of, 65–66, 71, 80–81 writing on, 79–80
language and efficacy of, 63 body and soul, 82
in Nag Hammadi, 82 Body of Truth, 48, 52
154 General Index

Book of Creation, 46, 54, 57–58, 127 building blocks of, 26 n. 16, 53–56, 81
translations of, 54 of a cow. See rabbis create a cow
Book of Secrets, 85–107, 124–25 by divine speech, 19, 24–26, 30, 37,
dating of, 86 45–46, 55, 125
social context of, 102–7 language and, 26 n. 16
table of contents, 87 making of Tabernacle parallel to, 57
breathing techniques, 8, 57, 81–82 models of, 45–46, 55
process of, 124
cannibalism, 2 role of divine name in, 33
Celsus, 34–35, 39 n. 22, 42, 98 secrets of, 55, 104
chariots, xi, 68. See also merkabah texts story in Genesis, 19, 26
chorus. See heavenly chorus Torah and Name existed before, 30
Christian exegesis, 20, 32–43 Torah is paradigm of, 30 n. 37
Christian texts, xxv, 64, 67
Christians daimons, 1, 3, 14, 92–94, 123
against use of water, 16 language of, 93
alleged to practice magic, 34 names of, 35–36
attacking theurgy, 13 rabbis as experts on, 107
denounced Jews, 1 and sacrifice, 98, 104
interpret the spirit, 124 theurgy invention of, 13
should rely on words, 15 Dan, J., 54
circumcision, 80 Davila, J., 74
Cohen, M., 53 Dead Sea Scrolls, 63–64, 67, 71, 76,
color in alchemy, 114–18 107
consecration, 8–10 deification, 42
contagion, law of, 100 by ascent, 65
cosmogony, 45, 57 Delatte, A., 98
cosmology, xi–xii, 39, 69, 125 Democritus. See Pseudo-Democritus
alchemical, 111, 115, 126 des Places, 7
of Book of Secrets, 87 deus abscondus, xv
Classical, xv Dillon, J., 42, 91
Dionysius the Areopagite’s, 39, 41 Diocletian, 111
of Hekhalot Rabbati, 81 Dionysius the Areopagite, 13, 33–34,
Late Antique, 8, 36, 64–65, 125 38–43, 101, 124, 128
locative, xi n. 1 distinction, 48
seven-heaven, 106 divination, 11
of Songs of the Sabbath Sacrifice, divine arts. See alchemy
69–70 divine contact, 9
utopian, xi n. 1 divine figures, giant, 49
Zosimos’, 120 divine forces, 60
cosmos, xi, 10, 48–50, 94, 102, 126 divine limbs, 52–53
Enoch’s tour of, 64 divine name, xiv, 15, 19–31, 34, 41, 77,
reciter identified with, 82 101. See also YHWH
and sound, 61 divine names, 18, 33, 36, 45, 49, 52,
state of the, 60 123–24
traversing, xi, 82 all refer to same deity, 35
creation, 24, 27, 47, 55, 56 n. 35, 74, 81, ascent using, 66
124 automatic power of, 26, 36
abyss of, 27 Christian exegesis and, 32–43
act of, 27 as complex, 47
 General Index 155

consists of vowels, 60 divine speech, xi, xix, 23–25, 45, 101

creative power of, 26–28, 31, 38, 45–46, creative power of, 19, 26, 31
124 embodying divine presence, 24
destruction of world with, 28 not easily comprehensible, 45
Dionysius the Areopagite’s theory of, 34, pragmatics of, 24
39–43 in written form, 102
efficacy of, 42 divine writing, 25 n. 15
evidence about deity, 39 divinity
as formal representation, 40, 43 contact with, 5
function of, xxiv incarnate, 48
Greco-Roman ideas about, 34 divinization, 83–84
hidden, 60 by ascent, 65, 83
as icon, 27, 34, 37, 40, 124 and Christian rituals, 42
incomprehensible sounds as, 41 of Roman emperors, 65
Jesus is a, 36 Dodds, E. R., 3, 12
Karaites denounce rabbinic interest in, dream interpreters, 105, 107
104 drugs, 4
knowledge of, xi
Late Antique ideas about, xx Egypt, 60, 94, 110
letters and sounds of, 45 Elior, R., 74
letters of, 79–80 Epicureans, 35
not arbitrary, 34 Eucharist, xiii, xix, 13, 15, 42
not symbolic, 40 as theurgy, 41
numerical value of, 50 exorcism, xiv, 15, 93
Origen’s theory of, 34–38
as origin of language, 124 fire, 55, 81, 89, 104, 117
permutations of, 79 alchemical use of, 119–20
power of, xiii, 54, 78, 92, 125 letters become, 56
Proclus’ theory of, 40 warming, 4
rabbinic interpretation of, 26–31 Forbes, R. J., 115–16
recitation of, 29, 78, 86, 109 formulas, 87, 90–96, 100, 125–26
reciting, 8, 12, 63 blessing, 90–91
seals include, 79 in the Book of Secrets, 93
secret, 103 dyeing, 113
sounds of, 60 to induce dream oracle, 114
as statues, 40–41 mediating, 93
taken away after Golden Calf, 28 release, 88
unspeakable, 61 request, 88–91
use of as magic, xv ritual, xxi, 91, 126
of vowels, 61 swearing, 90 n. 15
wearing, 101 verbal, xviii, 14, 18, 86, 90–96, 99–102,
is wonderful, 38 109
in world to come, 51 Frazer, J., 100
writing on body, 80, 101
divine power, xi, 1, 89, 107 gematria, 50–51, 58
Christians can achieve, 36 gnostic, 16–17
inappropriate use of, 3 Gnosticism, 120
of rabbis’ daughters, 106 Jewish, 52 n. 23
divine powers. See supernatural powers; Graetz, H., 72
angels; daimons grain, 104
156 General Index

graves, 93, 99 metals as, 114, 126

Greco-Roman texts, xxv, 64 names as, 53
Greek language, spirit of the text, 124 sounds as, 61
Greek Scriptures, 38 use of, 101
Greek texts, xii word as, 47
Grünbaum, M., 28 immortality, 9, 12, 65, 83, 84
imprecations, xi, 85–107
Halperin, D., 73 incense, xxii, 15, 88–90, 96–99, 105
heaven, sound of, 67 index, xxii, xxiv, 100, 124
heavenly chorus, 63, 67, 70–71, 77–78, 95 definition of, xxiii
heavenly temple, 67–68 Irenaeus, 16, 46–47, 50–51
Hebrew language, 21 Ishmael, 78
deity spoke, 124
letter of the text, 124 Jesus, 41, 99
names in, 23, 42, 60 as logos, 38
as the Scriptural language, 38 as magician, 2
special qualities of, 37 names of, 34, 36, 49
special role of, 38 numerical value of, 50
Hebrew Scriptures, xv, 19, 21, 48, 67, 98 Jewish texts, xiii, xxv
Hebrew texts, xii Jews, 21, 48, 74, 91, 125
Hecate, 8, 12 denounced Jesus, 2
hekhalot texts, xv, 63, 71–87, 90, 106, Late Antique practices denounced, 72
112, 125 mystics and rabbinic, 73
dating of, 72 turn from animal sacrifice, 97
Hilarion, 15 used amulets, 101
Hinduism, xvi John Chrysostom, 15
Hippolytus, 51 Johnston, S., 7
holy man, 103 Josephus, 107
Honi the Circle-Drawer, 96, 106 Joshua b. Levi, Rabbi, 92
Hopkins, A., 115–16 Judah, Rabbi, 105
horoscopes. See astrology Julian (of Chaldean Oracles), 6–7, 128
hymns, 60, 63–84, 87, 93–95, 102, 106, Julian (Emperor), 97 n. 38
109, 124 Jung, C., 117, 119
Justin, 14
Iamblichus, 5–6, 10–11, 17, 51, 57–59,
94, 128 Karaites, 104
defense of ritual of, 10 Katz, S., xvi
icon, xxiii, xxiv, 100, 120 Kroll, W., 7
actions as, 100
amulets as, 101 language. See also Greek language,
attacked as mere symbol, 59 Hebrew language
definition of, xxii, xxiii culturally specific, xxi
divine names as. See divine names distinct modes of, 91
divine speech as, 24 efficacy of, 19, 23, 76, 83
example of, xxiii, 46 and elements, 56
of highest form, 116 emanationist theory of, 39
human body as, 80 hierarchy of, 94
letters as, 58 as icon, 49, 59
limbs as, 53 legal, xxi–xxiv
of matter, 115 manipulation of, 127
 General Index 157

metaphor for cosmos, 60 as polemical term, 1

not of human origin, 34 Porphyry’s definition of, 10
pragmatic, xx prayer as, 15 n. 52
rabbinic ideology of, 29 Proclus’ definition of, 12
as reality, 61 rabbinic permission to study, 103
referential model of, xx “rise of,” 106–7
request, 94–95 magicians, 2, 11
ritual uses of, 59 magos, 1
secret, of metals, 116 Marcus, 45–52, 56–58
symbolic theory of, 40 as heretic, 46
Lauterbach, J., 104 Jewish sources of, 52
Leiden Papyri, 113 Marcus Aurelius, 4, 6
Lesses, R., 79 Margalioth, M., 86, 103
letters, 45–61, 78 n. 69 Maria the Jewess, 112, 115, 117
creative power of, 56–57 Membres, 117–18
distinction between words and l., 48 memra, 22
divided by pronunciation, 49 men of deed, 96, 107
Greek, 50, 58 Merchavya, Ch., 103
has no semantics, 56 merkabah texts, 63, 68, 74, 78, 87, 103
Hebrew, 54, 58 Mertz, E., xxi
as icon, 57 metalworking, 110–14
numerical value of. See gematria metalinguistics, 20 n. 3, 78 n. 71
as powerful as name, 27 metapragmatics, xxi, 30 n. 28, 56
pragmatics of, 46 miracle, xvii, 13, 15, 36
shape of, 30 rain, 7
Lieberman, S., 104 Morray-Jones, C., 73
logos. See verbal formulas, Word Moses, 14, 21–22, 26, 36, 49, 99, 113
love-rites, xii, 113 mystical experience, xvii
Luck, G., 4 mysticism, xiv–xvii, 128
definitions of, xv–xvi
magic, 1–4, 6, 14–17, 61, 128 Hindu, xvi
Augustine’s definition of, 13 Jewish, xvii
Christian and Jewish rituals as, 2
Christians alleged to practice, 2 n. 5, Nabateans, 107
34 Nahman, daughters of Rabbi, 106
as criminal offense, 2 name interpreters, 33, 38
dangerous and barbaric, 2 name, four-lettered. See YHWH
definitions of, xiv–xv names, 126
efficacy of, 3 Biblical, 23
Egyptian, 60 divine. See divine names
as fraud, 13 not arbitrary, 35
Greco-Roman rituals as, 2 of patriarchs, 36
Iamblichus’ definition of, 11 semantic content of, 22, 59
Jewish, 85 necromancers, 105–106
Jewish ritual as, 1 n. 2 Nehunya, Rabbi, 63, 66, 71–72, 78–79,
Late Antique debates about, xiii, 3 81, 83
misguided analogies underlie, 100 Neoplatonism, 3, 5–6, 12–14, 34, 41, 46,
modern scholarship of, xiii, xxv, 3 57, 82, 102. See also individuals’
as pejorative, xiii–xiv, 2, 11 entries
Plotinus, modern scholars and, 16 Neusner, J., 3, 103
158 General Index

Newsom, C., 68–70 rabbinic texts, 49

Nicomachus of Gerasa, 5–6, 59 rabbis, 73, 104–106
Noah, 103 competing with other religious figures,
105
oaths, 27, 90–92 create a cow, 57, 127
objects creative work of, 57
manipulation of, 4 as daimon experts, 107
unutterable, 49 daughters of, 106
use of as effective, 86 lost doctrines of, 104
oil, xviii, 88–90, 96–97, 104 names of, 103
Origen, 14, 33–38, 43, 49, 59, 79, 98–99, overcome natural order, 92
124 supernatural capabilities of, 104
theories as magical, 42 Rappaport, R., xix
Orphians, 79 Raziel, 103
revelation, 40, 47, 103, 113
Parmentier, R., xviii–xxii divine, 45, 120, 125
Patai, R., 112 n. 13 of God’s name, 22
Peirce, C. S., xxii–xxiii, 6, 24, 40, 42, 46, of Marcus, 46, 48, 54, 57–58
59 statues give, 12
Peircean linguistics. See Peirce, C. S., and ritual, xxi, 3, 10, 15, 52, 81, 83, 102,
individual entries, e.g., icon 125
Penner, H., xvi–xvii aesthetics of, 127
performance, xxi analogy in, 100
Philo, 60, 65 ascent as. See ascent
Plato, 3, 7, 41, 110 attacks on, 94
Pliny the Elder, 2, 91 Christian, xii–xiii
Plotinus, 3, 5, 15–17, 128 context of, 100
Porphyry, 5, 9–10, 13–16, 94, 98 debates about, 14
concept of ritual of, 9 divine name is word of, 127
pragmatics, xx, xxv, 19, 23, 31, 53, 54 divine presence in, xxiii, 41
of divine language, 19 efficacy of, xvii–xxv, 1–18, 65–66, 124,
of divine speech, 24 128
of letters, 46, 54, 56 Eucharist as, 13
prayer, 4, 8, 12, 17, 60, 83, 90 forces in, xxiv
efficacy of, 15 fraudulent, 9
Jewish, 77, 94 Greco-Roman, xii, xiii
as magic, 15 n. 52, 90 n. 15 Greco-Roman explanation of, 3
Plotinus’ rejection of, 16–17 Greek, 96
replacing sacrifice, 97 how understood, xviii
priests, 60, 68, 70 as icon, 124
angelic, 71 iconic structure of, xxiii–xxiv
Prince of the Torah, 63, 90 icons in, 79
Proclus, 1, 7, 11–12, 40–42, 57–58, 102 ideas crucial to, 127
Proudfoot, W., xvii immortality, 84
Pseudo-Democritus, 27, 112–113, 116, incomprehensibility of, 5
120 initiation, 113
Pythagoras, 51, 99 Jewish, xii–xiii
language in, xxii
Qumran. See Dead Sea Scrolls Late Antique, xviii, 17
 General Index 159

modern scholarship of, xii, xiv, xviii, sefirot, 55–56

xxv, 16 semantics, xx–xxi, 30, 45, 52
objects in, 14, 18, 96 basis of, xxi
participation in, 12 meaning not limited to, 23, 31, 45, 52,
philosophical basis of, 5, 9–10 56
Plotinus’ disdain for, 16 rules of, xx
Proclus’ promotion of, 11 semiotics, xx, xxiv, 27, 80
props in, 14 definition of, xviii
purpose of, 86, 128 modern theory of, xxiv
reality in, 70 Peircean, 24, 124
repetitive patterning of, xix of voodoo, 100
scapegoat, 104 Sheldon-Williams, I. P., 39
social setting of, 106 Shem ha-Meforash, 28–29
structure of, xx, 81, 100 Shi’ur Komah, 52, 84, 87
Tambiah’s aspects of, xix sign
theurgic, 4, 12 conventional, xxiii
words-only, 18 definition of, xxii–xxiii
ritual actions, xix, 5, 8, 17 unutterable, 49
ritual handbooks, 85, 87, 96, 114 silence, 4, 8, 89, 121
ritual language, xx, 29, 70 Silverstein, M., xviii
ritual practices, 31, 75, 106 Smith, H., xvi
ritual setting, xxiii Smith, J. Z., xii n. 1
ritual strategies, 82, 93 Smith, M., 15, 66, 68, 71, 74, 77, 79, 81,
ritual technique, 126 107
ritual texts, xii–xiii, xxv, 46, 60, 68, 69, Socrates, 128
74, 101, 123 Solomon, 117–18
Greek, 90 soul, 9–13, 50, 57–63, 81–82, 118, 128.
ritual wheel, 8 See also body and soul
analogous to vowels, 6
Sabaoth, 36 ascent of, 8
sacred arts. See alchemy divine, 16
sacrifice, 17, 120, 123 liberation of, 10
alchemy as interpretation of, 109, meaning of, 51, 57
117 motion of, 58
animal, 4, 8, 51, 96–98, 102 Plato’s, 7
human, 2 world, 16, 58
internal, 97 sounds, xxii, 39, 45–61, 63, 66, 81–82,
laws of, 103 124–125
metalworking as, 119 connected to supernatural world, 41
modified, 96–98, 102, 105 cosmic, 45, 50, 61, 81
to Satan, 104 cosmin, 47
self, 118–19 divine name in articulated, 41
Temple, 97 heavenly, 67–68, 84
Saffrey, E., 42 as icon, 61
Sallustius, 17 incomprehensible as divine names, 41
Satan, 104 language iconic at level of, 49
Schäfer, P., xiv n. 17, 73 non-semantic, 52, 61
Scholem, G., xvii, 29, 63, 72–75 of divine names, 124
seals, 79–80 power of, 37
160 General Index

sounds (cont.) as magic, 7

seven, of planets, 61 origins of, 6
vowel imitate planet, 6 philosophical basis of, 12
speech-act theory, xx theories of, 3
statues, 3–4, 11, 40–41 translation, xxii, 42
Stockholm Papyrus, 113 Aramaic. See Targums
Stoics, 35 challenges of, 56
Suda, 6 loss of efficacy in, 37, 47
sun, 90, 107–8
supernatural powers, xi, xix, xxiv, 11, 36, vegetarianism, 98
49, 94, 128. See also angels, daimons vessels, 88, 100–101, 127
evil opposed to Jesus, 98 n. 57 violet-making (alchemical term), 114, 116
high-level, 79, 98, 123 voodoo, 31, 100
iconic representations of, 102 vowels, 6, 49, 59–60
individuals intercede with, 107 analogous to soul, 6
language has formal resemblance to, efficacy of, 60
49 Greek, 60
low-level, 95 heavens utter, 49
teach names of gods, 12 in Hebrew divine name, 60
within a human being, 37, 124 as icon, 6
Suria, 78 and planets, 49
symbol, xxii, xxiv, 11, 24, 59, 124 in rituals, 5, 59–60, 66, 82–83
in cross-cultural context, xxiv as sign, 6
definition of, xxiii
unutterable, 5 water, 55, 81, 96, 99 n. 46
sympathy, 11, 15, 102, 123 body of, 93, 104
law of, 100 collecting, 127
divine, 116
Tabernacle, 56–57 immerse clothes in, 105
Tambiah, S. J., xviii–xxi, 69, 93 letters become, 56
Targums, 21–27, 39, 45, 47, 124 ritual use of, 8, 16
new characterizations in, 22, 24, 29, whitening (alchemical term), 114, 116,
38 119
shift in thinking, 25 wine, xxiv, 41–42, 88, 104, 124, 128
Tashlich, 104 women, xxiii, 104–105, 107
Temple (Jerusalem), 99, 105, 117–18 Gentile, 105–106
destruction of, 97 as magicians, 3
Tertullian, 50 Word, 37–38, 46. See also memra
Theodorus, 51, 57–58 similar to deity, 46
theology, 3, 9, 13, 18, 43, 51, 118 words, xiv, xix–xxi, xxiv, 11, 17, 20,
theurgy, 3–4, 5–18, 41–43, 59, 128 23–25, 34. See also divine speech,
as best rituals, 10 language
and Chaldean Oracles, 7 as arbitrary collections of sounds,
Christian attack on, 13 xxii
definitions of, 3, 9, 11, 13 comments on, 21
as effective ritual, 11, 13 efficacy of, xx–xxi, 19, 66
Eucharist as, 41 God’s. See divine speech
as a form of magic, 14 meanings of, 23
Julian and, 97 n. 38 power of, xxi, 27, 93
 General Index 161

in rituals, 15, 18 YHWH, 19, 23, 46–47, 60

sacred, 7 Yom Kippur, 29
transformational capability of, xiii
Zaehner, R. C., xvi
yellowing (alchemical term), 114, Zeus, 36, 49
116 Zosimos, 111–20