Ancient Magic (The Prayer of Jacob)

(online lecture by J. Davila on 8 April, 1997)

Since there will be no realtime class session to correspond to this online lecture, I will let myself go on considerably longer than I usually do. Ancient magic and theurgy are quite important for my current research, so I’m seizing this opportunity to organize my thoughts about them and to review the current state of the question. I’ll begin with some remarks on theoretical and methodological problems with defining the term “magic,” then I will discuss in turn Greco-Roman magic, Jewish magic, and magic in the OT Pseudepigrapha.


The word “magic” is probably the most difficult and elusive word we will grapple with in this course. It is open to debate whether it can be defined at all or whether it is a useful term for the critical study of ancient history and religion, but the scholarly discussion about the problem (whether it is insoluble or not) is certainly of great interest for our purposes. In this section I will survey some approaches that have been tried and will make a stab at coming up with a rough working definition. As much as possible I will sidestep the debate about the relationship between magic, science, and religion, since it isn’t terribly relevant for the texts we will be looking at.

One approach, exemplified by Daniel O’Keefe’s book _Stolen Lightning_, could be called historical. O’Keefe draws on anthropology, psychoanalysis, and history to try to explain the origins of magic in human societies. He defines magic (“in the strict sense”) as sacred institutions related to religion but often of an illicit or peripheral nature and mostly based on the relationship between practitioner and client rather than on a community relationship. His thesis is that primitive human beings had (or have) a much weaker sense of self than modern Westerners and magic emerged out of religion, developing to protect this fragile self from potentially fatal social pressures such as “voodo death,” “soul loss,” and even anxiety. By seizing the offensive and giving the weak self a sense of control, the self was preserved more often and the resulting dialectic between magic and religion actually renews religion.

Now this is an enormous and enormously complicated book and any one-paragraph summary is bound to be an oversimplification, but I think the above is a fair distillation of O’Keefe’s main points. The historical approach is fascinating and may well give us insights into the origins of magic and religion, but I have two fundamental criticisms of it. First, to all intents and purposes it is nonfalsifiable. Given our current knowledge of Neolithic and earlier human society we simply don’t have the evidence to test this sort of historical theory (and we probably won’t have it anytime soon, if ever). Second, I don’t find theories about the origins of magic terribly useful for increasing our understanding of the magical texts and traditions from antiquity to the present that we actually have, which is my real interest.

For this reason my preference is for what could be called the functional approach, that is, an attempt to understand how extant magic texts and traditions functioned in their own societies and historical contexts. (I should note that O’Keefe also has a lot to say about functional issues along the way and I find this material in his book very helpful.) The most basic functional definition of magic involves a listing of things usually done by a magician and the contexts in which the are done. An example is John Middleton’s comments in his article “Theories of Magic” in the _Encyclopedia of Religion_:

“Magic is usually defined subjectively rather than by any agreed-upon content. But there is a wide consensus as to what this content is. Most peoples in the world perform acts by which they intend to bring about certain events or conditions, whether in nature or among people, that they hold to be the consequences of those acts. If we use Western terms and assumptions, the cause and effect relationship between the act and the consequence is mystical, not scientifically validated. The acts typically comprise behavior such as manipulation of objects and recitation of verbal formulas or spells. In a given society magic may be performed by a specialist.” (vol. 9, p. 82)

Thus far it looks as though we’re moving toward a possible definition: magic is the manipulation of physical objects or recitation of formulas and incantations by a specialist on behalf of him/herself or a client to bring about action by the divine world. The hitch with this definition is that it can be applied too widely: Is a Catholic priest praying the rosary for a parishoner a magician? How about Lubavitcher Rebbe teaching Kabbalah to his disciple? O’Keefe adds the proviso that magic is illicit or peripheral, and this is an important point. The term magic (in whatever language) has generally meant religious cult whose legitimacy is rejected by the speaker. In other words, it is a theological or confessional term, not an objectively descriptive one. For this reason Jonathan Smith suggests that magic is not a legitimate category for “second order, theoretical, academic discourse” (p. 16) and it should be replaced with more specific and useful terms such as healing, divination, exorcism, etc. Perhaps he is right, but I still think the category magic has some heuristic value, so I’m not quite willing to give it up entirely.

For me the most helpful recent discussion of magic has been the introduction to _Ancient Christian Magic: Coptic Texts of Ritual Power_ by Marvin Meyer and Richard Smith (pp. 1-9). They find the centre of these magical texts to be their manipulation of power through ritual, ritual whose form is in turn is determined in part by its social function in individualistic Hellenistic society. Sifting through these various insights and caveats, I propose the following very provisional definition of magic for our use as we begin to approach the magical literature of the OT Pseudepigrapha.


  • Magical texts are ritual texts that manipulate divine powers for the benefit of the user or clients and that are generally looked on as illegitimate by official or mainstream cults in the society.

(Comments on this definition are, as always, welcome.)


This corpus of magical literature is of great interest to us in this section of the course. The material is preserved in various forms and genres, including amulets, curse tablets, curse figurines, incantations, and handbooks from the Hellenistic period to late antiquity. The last category, “handbooks,” includes the great magical papyri in Greek and Demotic collected in late antiquity. The volume of translated texts edited by Betz, et al., includes these MSS. A look at the contents of this collection reveals a bewildering variety of spells and charms, including, inter alia, rites for acquiring familiar spirits, restraining spells against spirits, spells for divination and obtaining of revelations, love charms, numerous healing spells for various ailments, curses (especially to inflict the victim with insomnia), spells for victory at games and competitions, and even contraceptive spells.

Dr. Gideon Bohak, of the University of Michigan, has posted a web page on magic that includes examples of similar material from the Library collection. I suggest you check the web version of my lecture for links to excellent pictures of amulets and gemsrecipe books (= handbooks), and some aggressive magic texts. (Note that clicking on the images will enlarge them.)

In general the classical philosophers looked down on magic and its practitioners. But the popular Greco-Egyptian philosophy of the Hermetic corpus contains magical elements even after purging by the Byzantine tradents, and the later Neoplatonists such as Plotinus renamed their magical practices “theurgy” and distinguished them from the (according to their view) vulgar magic we find in the papyri.

The question of the relationship of magic to earliest Christianity is quite important, but I want to defer it until next year when we explore the relationship of the traditions about Solomon the magician to Jesus the exorcist in the online course on Divine Mediator Figures. Suffice to say here that although the volume edited by Betz consists entirely or almost entirely of pagan magic, there is a vast corpus of Christian magic from the same period, especially in Coptic. It is not significantly different in form or content from the pagan material, apart from frequent references to specifically Christian doctrines and figures. Much of this material is translated in the book edited by Meyer and Smith cited above.

For more on Greco-Roman magic see the article “Magic in Greco-Roman Antiquity,” by Hans Dieter Betz in the _Encyclopedia of Religion_ vol. 9, pp. 93-97.


Jewish magic is also of interest to us, since both the texts we will be looking at from Charlesworth have been influenced by it to a greater or lesser extent. There is an ever growing corpus of Jewish magical texts, extending now back to the turn of the era with the publication of a number such documents from the Dead Sea Scrolls. These include 4Q186 (a Hebrew physiognomic and astrological text written in a cryptic script and strikingly similar to some medieval documents in Hebrew); 4QMess ar (a fragment, perhaps of the lost “book of Noah,” with themes related to physiognomic and Merkavah traditions–see below); 4Q318 (an Aramaic “brontologion,” a genre of omen that predicts the future by interpreting thunder); 4Q511-12 (“Songs of the Sage,” songs for protection against demons); 4Q560 (a fragmentary Aramaic apotropaic incantation anticipating much found in the Babylonian incantation bowls–see below); and some exorcism incantations included in 11QPsalms. incantation bowls. (Note that some in this display are written in a “pseudo-script,” i.e., realistic looking markings that aren’t actually letters or words. A cut-rate product? Or a consumer fraud? Also have a good look at #39, which has a tacky drawing of demons typical of the bowls.)

Somewhat later, from roughly the fourth to seventh centuries, come the Jewish Aramaic incantation bowls (along with Syriac and Mandaic bowls), found in Iraq and Iran. Many of these were excavated in Nippur and others have been sold on the antiquities market. These bowls contain inscribed spells, usually in a spiral inside the bowl, mostly having to do with protection from demons and other nuisances. Bohak’s web page has a number of

From roughly the same time period come Jewish amulets inscribed on metal (usually lead, copper, or silver) in Aramaic. These have been found mostly in Syria-Palestine and Asia Minor, and Naveh and Shaked have published thirty-two of them in their two volumes of magical texts. The purposes of these amulets are typical and include healing, exorcism, domination of others, protection of pregnant women, etc.

A great many magical documents are also known from the Cairo Geniza and some have been published by Naveh and Shaked, Schiffman and Swartz, and Shaked and Peter Sch<“a>fer. They include amulets, incantations, and short handbooks covering a vast range of spells for every imaginable purpose. (Unfortunately, at present the Cairo Geniza web page of the University of Cambridge contains no pictures of magical texts.)

In addition, a number of longer magical handbooks have been passed down in their own MS tradition. These include Sepher HaRazim (the Book of Mysteries), Harba deMosheh (the Sword of Moses), and Havdala deR. Akiva (The Havdalah of R. Akiva). Only one of these (Sepher HaRazim) is available in a convenient English translation (by Michael A. Morgan), although it should be noted that some scholars feel that the original editor of the Hebrew text, Mordechai Margaliot, exercised an unduly heavy editorial hand in connecting the various fragments of the work into a single document.

Finally, a word should be said about the important and, until recently, little-studied Jewish literature from late antiquity and the Middle Ages on astrology, physiognomy and chiromancy, and mystical theurgy. A Hebrew document called Baraita deMazzalot (the Baraita of the Constellations) is a comprehensive astrological treatise along the lines of hellenistic astrology but with a Jewish theological slant. A number of texts purport to be able to evaluate personal character and destiny on the basis of palm reading and observation of facial appearance and other bodily traits. Some of these also draw on astrology. Connected with these on some level are the early mystical tractates known as the Hekhalot literature (Merkavah mysticism) that claim to give instructions on how to ascend through the seven celestial palaces to view the throne of God (the Merkavah) and sing with the angelic choir or, conversely, to compel angels to descend from heaven to teach the mystic esoteric secrets of Torah. Near the end of this course we will look in detail at 3 Enoch, one of the later Hekhalot texts.

It is worth noting in passing the extraordinarily conservative streak in some of the Jewish magical literature. I mentioned above that 4Q186 shares ideas and terminology with the medieval Jewish physiognomic tradition. One of the Hebrew texts from the Cairo Geniza (T.-S. K. 1.157 fol. 1a/14-23, published by Sch<“a>fer and Shaked) has a passage that is almost a direct translation of a Greek incantation for ascent of the uterus also found in the Greek Magical Papyri (PGM VII.260-71). The Havdalah deR. Akiva contains Aramaic material in chapter 4 that has close parallels with the much earlier Babylonian magic bowls and material in chapter eleven that seems to echo the Watchers myth in the Enoch literature. And the earlier chapters of 3 Enoch clearly have a close relationship to the Similitudes of Enoch (which will be explored later this semester). This evidence supports the contention of Professor Reeves in his earlier online lecture that a good bit of material from the second temple period and late antiquity survived in Jewish circles into the Middle Ages.


So at last we come to the texts to be read for this course. The Prayer of Jacob (PGM XXIIb.1-26) is an odd incantation, even given the bizarre company it keeps among the Greek magical papyri. It purports to be a prayer by the Israelite patriarch Jacob summoning God and demanding wisdom and deification for the reciter. Charlesworth takes the prayer to be Jewish, since it specifies that the reciter should be “from the race of Israel” (line 14) but this seems like a typical bit of magical window dressing on which we probably shouldn’t place much weight. (Contrast the reference to the “serpent gods” in line 8.) Still, the text anticipates themes in the Hekhalot literature, so we shouldn’t be too quick to assume it was pagan either. Hekhalot texts do tend to summon angels (never God, I believe) in a very premptory way, and they ask for various types of esoteric wisdom and theurgic power. It seems clear also that the Merkavah mystics were deified (turned to angelic creatures of fire) before the throne of God, at least for the duration of the heavenly journey. Enoch received this deification permanently when he was transformed into the angel Metatron in 3 Enoch. But perhaps not too much should be made of this either, since deification rites are also known from the pagan magical papyri (e.g., PGM IV.475-829) and human claims to godhood go back to the Egyptian funerary Pyramid Texts. On strictly methodological grounds it is very difficult to assign the Prayer of Jacob with any certainty to a Jewish or pagan (or conceivably even Christian) background. There is a reference to Osiris-Michael later in the papyrus, which hardly settles matters, but otherwise the content is generic.

For what it’s worth, there are a number of other texts among the Greek magical papyri that are also indebted to Judaism for some of their content. There is a “Charm of Solomon that produces a trance” in PGM IV.850-929, but its religious content is otherwise pagan. Various versions of the “Eighth Book of Moses” appear in PGM XIII.1-343; 343-646; 646-734, followed by a “Tenth (?) Hidden [Book of] Moses” in 734-1077, but the content of these too, is almost entirely pagan.

The most interesting passage for our purposes is the “tested charm of Pibechis for those possessed by daimons” (PGM IV.3007-86), which also terms itself a “Logos Hebraikos” near the end. In my opinion this text has as much or more claim to be included in the OT Pseudepigrapha as the Prayer of Jacob (although I disagree with M. Gaster that there is any realistic possibility that it was composed in Hebrew). The exorcism adjures “the god of the Hebrews, Jesus” (cf. Acts 19:13) and makes mention of various events surrounding the exodus of “Osrael” from Egypt, including the ten plagues upon Pharaoh and the crossings of the Red Sea and the Jordan. It also refers to “the seal which Solomon placed on the tongue of Jeremiah,” a category of demon called “Pharisee,” “holy Jerusalem,” “the fiery Gehenna,” and numerous other Jewish themes. Gaster has drawn parallels to the oath of creation in 1 Enoch 69:16-25 and Daniel Sperber has shown that the author was familiar with elements of rabbinic exegesis. It is not likely to be a Jewish work (Gaster had to emend away the reference to Jesus in order to make this claim). More likely it is by a pagan author who drew on exotic (for him or her) Jewish and Christian traditions to make the spell more potent. In any case, the context of the passage in PGM IV shows that, whatever its origin, it was used by pagan magicians.

I will close with these observations. Next week we will be looking at another magical treatise in the OT Pseudepigrapha, the Testament of Solomon. But I will let Richard Wayman tell us about it in due course.


(Works mentioned in this lecture but not listed below can be found in the annotated basic bibliography for DI3216, The Old Testament Pseudepigrapha [1997].)

Quotations from the Greek magical papyri are from the translation edited by Hans Dieter Betz.

James R. Davila, “The Hekhalot Literature and Shamanism,” _Society of Biblical Literature 1994 Seminar Papers_ (Atlanta, Ga.: Scholars Press, 1994) 767-89

Garth Fowden, _The Egyptian Hermes: A Historical Approach to the Late Pagan Mind_ (Princeton: Princeton University Press, 1986)

M. Gaster, “The Logos Ebraikos in the Magical Papyrus of Paris, and the Book of Enoch,” _Journal of the Royal Asiatic Society_ 33 (1901) 109-117

“Magic,” in _Encyclopedia of Religion_, ed. Mircea Elide (New York/London: Macmillan, 1987) vol. 9, 81-115. Eight articles on various subtopics relating to magic.

Peter Sch<“a>fer and Shaul Shaked, _Magische Texte aus der Kairoer Geniza_, vol. 1 (T<“u>bingen: Mohr [Siebeck], 1994)

Jonathan Z. Smith, “Trading Places,” in _Ancient Magic and Ritual Power_, ed. Meyer and Mirecki, pp. 13-27.

Daniel Sperber, “Some Rabbinic Themes in Magical Papyri,” _Journal for the Study of Judaism_ 16 (1985) 93-103

(c) 1997
Reproduction beyond fair use only on permission of the author.

Contact details

St Mary’s College
The School of Divinity
University of St Andrews
South Street
St Andrews
Fife KY16 9JU
Scotland, United Kingdom

Tel: +44 (0)1334 462850 
Fax: +44 (0)1334 462852