Ancient History & Civilisation


Wherever possible (i.e. except where they fail to include a text), I have translated the texts found in Diels’s and Kranz’s edition of the fragments and testimonia (number [1] in the Select Bibliography, pp. xxxvi-xlii); any places where I differ from the text they provide are marked in the translation with an obelus, which refers the interested reader to the Textual Notes (pp. 337–44). Since Diels/Kranz is an anthology, I have also concluded each extract in the book with a precise reference to the location of the original text in the standard edition, or at least in an accessible edition.1 However, only in cases where Diels/Kranz fail to include a text should this concluding reference to another edition be taken to imply that I have translated the text of that edition; in all other cases, to repeat, I have translated the text found in Diels/Kranz.

The heading of each translated piece usually also includes a few numbers, which give a conspectus of the numbering of that fragment (F) or testimonium (T) in the most important editions. Thus, for instance, you might find this heading: F20 (DK 31B17; KRS 348, 349; W 8; I 25). This means that the fragment of Empedocles (whose prefix number is 31 in Diels/Kranz) which is numbered 20 in my translation, is number B17 in Diels/Kranz (in whose edition, by and large,2 testimonia are signalled by the prefix A and fragments by the prefix B), numbers 348 and 349 in Kirk/Raven/Schofield [2], number 8 in Wright’s edition of Empedocles, and number 25 in Inwood’s edition.3 These coded conspectuses will be complex, therefore, only where the thinker has received the benefit of a number of standard editions, whose numbering of fragments differs from that of Diels/Kranz. More normally, you will find only DK and KRS entries, for example: T12 (DK 12A1; KRS 94). This means that the testimonium of Anaximander numbered 12 in my translation is numbered 1 in Diels/Kranz and 94 in Kirk/Raven/Schofield. Rarely, an entry reads no more than, say, T23; this means that the passage does not occur in Diels/Kranz or in any of the standard editions (which in any case are generally editions of the fragments rather than testimonia). The amount of text I have translated in any particular instance, especially where testimonia are concerned, may be longer or shorter than what is to be found in Diels/Kranz or in any other edition.

The following abbreviations have been used:


Coxon (Parmenides)


Diels/Kranz [I]


Inwood (Empedocles)


Kahn (Heraclitus)


Kirk/Raven/Schofield [2]


Lee (Zeno)


Marcovich (Heraclitus)


Taylor (Atomists)


Wheelwright (Heraclitus) or Wright (Empedocles)

Note that some books which count as standard editions preserve the numbering of fragments found in Diels/Kranz, and so do not need a separate code. This goes for Lesher’s edition of Xenophanes, Kirk’s edition of Heraclitus’ cosmological fragments, Robinson’s edition of Heraclitus, the editions by Gallop and Tarán of Parmenides, Huffman’s edition of Philolaus, and Sider’s of Anaxagoras. Note also that although there are in existence some fine editions of some of the Presocratics in languages other than English, I have not given them codes because I decided to restrict my bibliography strictly to the English language. But I should like to mention especially A. Laks’s edition of Diogenes of Apollonia (Cahiers de Philologie, 9; Lille: Presses Universitaires de Lille, 1983); S. Luria’s of Democritus (Leningrad: Scientific Publishers, 1970), J. Bollack’s of Empedocles (Paris: Les Éditions de Minuit, 1965–9), and J. Bollack and H. Wismann’s of Heraclitus (Paris: Les Éditions de Minuit, 1972). Finally, note that all the works that have been coded are editions of the Greek texts (which invariably include translations); I have not referred in this way to other translations, however widespread their use may have become.

If you find an error please notify us in the comments. Thank you!