Only a few textual witnesses exist for the Aramaic incantation type before the appearance of Mandaic and other Aramaic incantations on earthenware bowls, metal, and leather from Mesopotamia. Due to the non-survival of writing material like leather, vellum, and papyrus, early Aramaic incantation texts have not come down to us. The most important and only exception so far has been an incantation in cuneiform script from Uruk. It shows specific elements (structure, language features) that already point to the later Aramaic incantation type known from the fifth through the eighth centuries A.D., originating from central and southern Babylonia as well as Khuzistan, in addition to a few minor finds from northern Mesopotamia. In textual set-up the Aramaic incantations distinguish themselves from the richly documented Akkadian ones. Therefore we cannot expect to find exact parallels and phrases in the form of direct Akkadian transmission or translations. Most of the Aramaic incantations testify to a parallel development of the Akkadian and Aramaic magical texts over several centuries in the same geographical environment. The Aramaic incantations may reveal Akkadian influences in contents and terms, often evident in demonology, phraseology, and lexicography, but to a very limited extent.
The Aramaic incantation represents an independent and completely different type of textual structure. The most notable difference is the use of the first person singular in reference to the active client instead of a higher demonic figure, whereas the Akkadian incantation employs the second person. This is prominent in the beginning of Aramaic demon stories where with the help of higher beings the active demon fights antagonists of all kinds, such as demons, slander, evil eye, diseases, and so on--see below M145; AMB amulet 15, bowl 12a, b; Muller-Kessler 1996; 1999a; 2002a; 2002b. In contrast stands the curse type that prefers to employ the impersonal address (see below M102, CBS 2971 = AIT 28; BM 91771, YBC 2393, VA 2492 in Muller-Kessler forthcoming). Such Aramaic historiolas are still in the minority in comparison to the stereotyped incantation where simple accounts of evil entities to be warded off with the help of higher beings, gods, and their gnostic representatives are listed. The Mandaic incantations show many different types of such stories, and some were translated into neighboring Aramaic dialects, among them Babylonian Jewish Aramaic and what is now termed koine Babylonian Aramaic. Even when translated there remained isoglosses and phrases which are so specifically Mandaic that they are recognizable in whatever script and dialect they appear. Often such orthographic hints are misunderstood and then emended by modern editors who have not been trained in the eastern dialectology of Babylonia (see below M145, M102, Muller-Kessler 1999a; 2001b; 2002a; Ford 2002b).
Why is this important? The script or the Aramaic dialect does not determine the textual affinity of a magic formula. It is only the doxological prologue and epilogue (frame), and the insertions (e.g., biblical quotations) which point to the obvious background and religious affinity of the scribe of such texts. They can differ in dialect and language from the incantation formula itself. Harviainen (1993) published a convincing study on this issue. The only exception here are the Mandaic magical formulas, which are kept in one dialect except for a few Hebrew textual loans, e.g., "Amen, Amen, Sela" at the end of an incantation.
Far less representative is the client's name, which is in most cases Iranian and rarely (c. 10%) of Semitic (Aramaic and Hebrew) origin. Identical family names with graphic variants can be found in different bowl scripts and dialects with diverse religious affinities which say more about the beliefs of the writer--in most instances probably a priest--than about the religion of the client. This has been proven with some certainty for the Nippur finds. Clients procured magic bowls for their houses in various scripts and with diverse religious contents (Epstein 1921; 1922; Muller-Kessler 2005). The archaeological data speak for this as well (Hunter 1995). However, it is debatable whether the clients' names are their actual ones. They are probably aliases, so-called malwasa (zodiacal) names. Perhaps this explains the use of matronymics instead of the patronyms expected in Semitic societies.
Despite the prevailing controversy among scholars concerning the religious background of magic text formulas in various Aramaic scripts and dialects, certain bowl texts show undoubtable Jewish contents and lore, although not all Aramaic square-script bowl texts contain Jewish themes. This publication of magical bowls from the private collection of Shlomo Moussaieff presents some specimens certainly produced by Jewish scribes probably trained in Rabbinic schools and who might even have been Rabbis themselves. It has long been doubted that the latter group produced magic texts, since according to the Babylonian Talmud sorcery was not acceptable. Priests and professional scribes were long the only group within the population able to read and write, and had studied the religious texts which tend to be quoted often in magical texts. There have even surfaced bowl texts containing only biblical quotations (Kaufman 1973; Muller-Kessler 2001b; 2005: 12-15). Since one is dealing here with apotropaic texts, protective or white magic, the production of incantations by priests, even Rabbis, is not so questionable as it is often regarded in scholarly works on Jewish magic of Late Antiquity.
The edition of selected bowls from the Moussaieff collection under review presents new texts. It comes at a time when several collections from museums and other institutions are being published. In contrast to the long overdue publication of the complete finds of incantation bowls from Nippur and other legally excavated specimens, the private collections are being presented to the scholarly public in a rather swift fashion. Levene explains this in the following way: "The present author was fortunate to have been given access by Dr. Shlomo Moussaieff to his collection of magic bowls." (1) This remark describes more or less how accessible such private collections really are. However, the reviewer does not want to go into the matter of the questionable provenances of those objects, since museums' policies in acquiring objects and allowing scholars access to bowls and metal amulets over nearly 150 years have not been much better. It is, therefore, not surprising that scholars have been prepared to work with private collections instead of their legally excavated counterparts.
This monograph, although this fact is not mentioned in the acknowledgements and introduction, is nearly identical to a Ph.D. thesis which was accepted and approved by the Department of Hebrew and Jewish Studies, University College London under the supervision of Markham J. Geller in 2000. Taking the book by its title, A Corpus of Magic Bowls, the reader would have expected a vast collection of magic texts, but this turns out not to be the case, since only a few of the most legible specimens of the Moussaieff collection were chosen, of which none have secure provenances. M163 was auctioned by Christie's in 1997 and had previously been published in 1999 by Levene, and again in the same issue of the journal in a modified version by Shaked (1999b). Some of the bowl texts (M103, M119, M123, M138, M142, M145, M149, M155, M156) have very close parallels in the Moussaieff and Martin Schoyen collection, which is painstakingly demonstrated in the synoptic tables. This proves in detail that a major find of incantation bowls from an ancient site in Iraq or Iran was split up between these two dominant collectors. Considering the fine scribal skills in evidence, one might consider Sura or Pumbeditha. Both were home to Talmudic academies, but have not been satisfactorily located to this very day. Major sites like Babylon, Borsippa, Nippur, and others where the bowls were found in situ with the exception of a few specimens have yielded rather poorly written texts. Quite a number of bowls were picked up on the surface in Kutha by Hormuzd Rassam and are housed today in the British Museum. These texts show a much superior writing style, but they have only recently come to our notice. (2)
This edition has some deplorable shortcomings. First of all, it has not been updated from the Ph.D. version, which is noticeable insofar as recent literature has not been incorporated, especially in the area of linguistic and lexical studies. Relevant standard works (e.g., Dictionary of Deities and Demons in the Bible, Reallexikon der Assyriologie) on the issues of demons and deities could not have been consulted, or such wild explanations for a goddess like Istar or the old Mesopotamian demoness Lilith would not have been suggested (see below). Other literature was obviously not carefully scrutinized for parallels and different views, nor were older publications or the British Museum incantation bowl catalogue by Segal 2000, although the latter is mentioned a few times. Even if one considers Segal's publication wanting, one must still quote the texts according to their latest publication (photograph), or should at least indicate a preferable edition. To dwell on publications like Ellis 1853 or Cowley 1923 is rather unsatisfactory in the year 2003, when those texts have recently been presented in new editions. In the case of diverse and debatable readings all opinions should always be considered.
Also striking is the incorrect usage of scientific terms. For example, Levene speaks consistently of "transcription," although he presents the texts in Hebrew transliteration. Unusual and unexplained is his transcription of [TEXT NOT REPRODUCIBLE IN ASCII] in the translations as a' or A', which should be 'a or 'A. The term "duplicate" is often...