Empedocles’ Complete Fragments, Volume I

III. The papyri

In 1904 the German archaeologist Otto Rubensohn bought from a Cairo dealer a stiff, collar-shaped strip of papyrus that had served as backing for the gilded copper leaves of a funerary wreath, probably from the necropolis at Panopolis. It reached Strasbourg the following year and lay unidentified for most of a century. Alain Martin recognised the text as Empedocles in the early 1990s, and the edition by Martin and Oliver Primavesi appeared in 1999.

The roll dates from the first century AD, or possibly the second, and was written by a skilled scribe in scriptio continua, without punctuation or word-division. Fifty-two pieces were arranged by the editors into eleven ensembles lettered a to k, of which six are substantial enough to yield connected sense; the frequently repeated phrase "ensembles a–f" refers to those six, not to the whole lettering. The six major ensembles carry something on the order of seventy-four visible lines, of which roughly twenty overlap passages already known from Simplicius. That overlap is the only place in the entire tradition where we can audit a witness against a copy independent of him, and it shows Simplicius' text to be slightly corrupt at points.

The papyrus yielded two results of the first importance. The first is structural. A marginal stichometric mark, the letter gamma standing for three hundred, identifies the three-hundredth line of the book in ensemble a. The columns hold about thirty hexameters each, so the first three hundred lines occupied ten columns and ensemble a is the tenth column of Book I. Because ensemble a overlaps the end of fragment 17, the opening line of that fragment can be fixed at approximately line 232 of Book I — which means that the proem of On Nature ran to well over two hundred lines. Every reconstruction made before 1999 had assumed a short proem.

The second result is doctrinal, and more disruptive. Ensembles d to f contain verses on forbidden foods and on punishment: material that every editor before the papyrus had assigned to the Purifications, purely on the ground of its content. The papyrus places it inside Book I of the physical poem. The criterion by which the fragments had been divided between the two titles for a century is thereby shown to be invalid. Whether this proves that there was only ever one poem is another question, and scholars differ; but the content-based division can no longer be defended as it stood.

A further papyrus has since come to light. P. Fouad inv. 218, identified by Nathan Carlig in 2021 as coming from the same roll and the same scribe as the Strasbourg fragments, was published by Carlig, Martin and Primavesi in 2025 and carries some thirty further verses, bearing specifically on effluences and pores. Its consequences for the theory of perception have not yet been absorbed into the handbooks, and any account of Empedoclean perception written before 2025 should be read with that in mind. Neither papyrus is represented in the Greek text printed here; the reasons are given in section VII.