This work argues that the presence of "alchemical" and "magical" material found in papyri dated to the third or fourth century CE must be related to the declining local aristocracy of the Roman empire and to their educated clients.

Jacco Dieleman, Ian Moyer and mostly David Frankfurter, first described the Greek and Demotic Magical Papyri (part of the Anastasi papyri collection) as the work of Egyptian priests responding to the wishes of Greco-Roman aristocrats who traveled to Egypt in search of mageia . The conclusion of my work suggests two alternative modifications to the current interpretive framework: either (1) the Anastasi papyri were not solely the work of Egyptian priests but also of a social group of "dependent scholars," men trained in Greek learning but economically dependent, and thus associated with a patron; or (2) from a socio-economic point of view, there was practically no difference between many late antique "Greco-Egyptian priests" and "dependent scholars" who used handbooks of recipes to maintain or promote the status of their patrons (alternatively the patrons could very well have used the books themselves).

Persian philosophy (mageia) was a popular topic of study among late antique philosophers and we should expect ancient education to have generally accepted research on the magoi as a valid topic of inquiry (chapter one). A misunderstanding of the subject of "magical" study in turn pushed patrons to ask their dependent scholars for curses and divination (chapter two).

Scholars (e.g. Apion "Pleistonikes") who were known as sorcerers were generally attached to patrons, involved in sympotic entertainment (esp. in the form of paignia) and in natural philosophy (chapter three). Comparing dependent scholars with each other shows that they read or wrote recipes that were considered to be magical by more established scholars (chapter four).

The presence of paignia and of other performance-related recipes in the Anastasi papyri strongly suggests the kind of sympotic environment and of intellectual patronage referred to in the preceding chapters (chapter five). The same conclusion must be drawn a fortiori for P. Leid. and P. Holm., which are constituted of recipes for the counterfeiting of luxury items (chapter six).