The Sortes Astrampsychi are one of the less well-known documents from antiquity. It is a second (or third) century CE oracular collection of both questions (92 in total) and answers (1030 in total) in Greek, bearing the name of the legendary magician Astrampsychus (on which see now C. Harrauer, “Astrampsychos,” Der Neue Pauly 2 [Stuttgart 1997] 121-122). It enjoyed a wide-spread popularity in the world of late antiquity. Its system worked as follows (it is explained in detail in the preface of the document but is here presented in a somewhat simplified manner). The enquirer first looks in the list of 92 numbered questions to find his question or the one most like the question he wants to raise. Then he chooses by some kind of sortition or selects in his mind a number between 1 and 10 and adds it to the number of his question. The sum thus reached has now to be looked up in a list of oracular gods with the concordance that follows the list of questions. The concordance indicates by means of a number after the god’s name the ‘decade,’ i.e., the section with ten possible answers. In that decade the answer is found under the number that was chosen by lot or selected. For example, your question is, “Will I get the woman I want to have?” This is question no. 29. You draw by lot or select the number 7, so the total is 36. In the list of oracular gods you find under 36 the name Hephaestus, and after this name the concordance number 27. Decade 27 has under number 7 the following answer to your question: “Yes, you will get the woman you want, but much to your detriment!” Since it is a god, Hephaestus, who directs the whole process, the answer cannot but be correct, for the theory behind this method of consultation was that the god’s action put the proper number in the mind or hand of the consultant. (For the wider cultural context of this phenomenon the reader is referred to P.W. van der Horst, “Sortes: Sacred Books as Instant Oracles in Late Antiquity, in L.V. Rutgers, P.W. van der Horst, H.W. Havelaar, L. Teugels (eds.), The Use of Sacred Books in the Ancient World, Leuven: Peeters, 1998, 143-174.)
One of the most interesting things about this lot-oracle is that we have it in both a pagan and a Christian version. Early papyri from the third and fourth century have the names of pagan gods in the concordance list, but in the many medieval manuscripts of the Sortes the names of the gods have been replaced by those of biblical persons. Although the Christians had also changed a few of the questions (e.g., “Will I be reconciled with my girlfriend?” has now become “Will I become a bishop?”), by and large they were using the same book as the pagans did not long before them.
The original work has not survived, but we now have it in two ‘editions’ which do not coincide with the pre-Christian and the Christianized version. About the ecdosis prior, published in the same series by Gerald M. Browne ( Sortes Astrampsychi, vol. I, Leipzig: Teubner, 1983), Stewart says that as the product of “a botched attempt to restore a damaged copy of the work, [it] preserves for the most part the phraseology of the archetype, whereas the second edition is more faithful to the structure of the archetype while altering its wording considerably” (see his contribution “The Oracles of Astrampsychus” in W. Hansen (ed.), Anthology of Ancient Greek Popular Literature, Bloomington and Indianapolis: Indiana University Press, 1998, 287). The ecdosis prior is contained in only one manuscript from the 13th century, whereas the ecdosis altera is found in no less than 8 papyri from the 3rd to the 5th cent. (5 of them from Oxyrrhynchus) and 11 manuscripts from the 14th to 16th cent. The first but also latest edition of this recension is from 1863 (by Hercher)! So Stewart’s new Teubner edition is most welcome, and that for more than one reason. This carefully excuted edition, based on all the evidence including the papyri Hercher could not yet know, was not only long overdue, but Stewart also makes a brave attempt to restore the original. In a section called ‘Archetypi reconstructio coniecta’ (87-127) he offers, with due caution (‘non est editio critica!’ p. xvii), a reconstruction based on the insight that the ecdosis prior by and large is faithful to the original phraseology (although one finds many traces of christianization there as well) while the ecdosis altera preserves the structure. Of course this is a very hypothetical reconstruction and one can quibble with Stewart about many details, but on the whole the present reviewer was impressed by the well-balanced judgment that underlied his choices.
This edition of the ecdosis altera is very complete. It not only has the standard critical apparatus in which the textual evidence is set out in admirably full detail (gratefully also the readings in the only manuscript of the ecdosis prior are recorded); it also has a second apparatus where only the obvious scribal interpolations are presented. Spot-checking I found hardly any errors, the only one being a funny mistake in the introduction: At p. xxiii the abbreviation mg is explained not as ‘in margine’ but as ‘in margarine’!
Apart from this valuable edition, Stewart has also provided us with a number of studies about the textual transmission and the date of origin of the Sortes and with the first English translation of this document (in the volume edited by Hansen, mentioned above); we are also promised an investigation of the fascinating biblical onomasticon in the medieval manuscripts of the Sortes. All students of late antiquity are in his debt.