Evidence of Plato and Aristotle relating to the Ekpyrosis in Heraclitus 1

Abstract

T HE double problem whether in passages of Plato the theory of a periodical conflagration of the world is denied to Heraclitus, and whether in passages of Aristotle it is attributed to him, is of unquestionable importance as part of the vexata quaestio whether Heraclitus really maintained that view or not. And from one point of view the weightier aspect of the double problem is that concerning Aristotle's testimony. For although in his interpretations he perhaps often distorts other people's theories, as H. Cherniss especially has shown (Aristotle's Criticism of PreSocratic Philosophy, Baltimore, 193 5), yet, as is acknowledged by Cherniss himself, 'he had the books of these men [the preSocratics] presumably in their complete form, while we only have fragments.' And let us add that where Heraclitus is concerned, Aristotle himself declares his full and direct knowledge of the text, when he states in the Rhetoric, 1047 b I I foll., that the sentence he quotes was found at the beginning of the book, and gives it as an example of faults of style by which the whole work was marked. Although then there is today a tendency among critics to deny to ancient witnesses previously considered as unimpeachable authorities any direct knowledge of the texts (here cf. J. B. McDiarmid, Theophrastus on the pre-Socratic Causes, Harvard Stud. in Class. Phil. LXI (X953)), we need not contemplate in Aristotle's case a doubt of the same kind as has been expressed by G.S. Kirk in regard to Plato, when he says that it may be suspected that Plato did not know as many authentic aphorisms of Heraclitus as we do (Natural Change in Heraclitus, Mind N.S. (i 9 S i)). In general, it would be erroneous to think that our ancient witnesses have no more real knowledge of their philosophical predecessors than is shown in their explicit citations. Besides the fact which Kirk himself acknowledges (Heraclitus, The Cosmic Fragments, Cambridge (i 9 54) p. I5), that for obvious reasons they cite only as much as serves the purpose of their discussions, we must take account of numerous allusions not accompanied by the name of the author alluded to, so that they may pass unnoticed, and an attentive study is needed in order to detect them. This applies to Plato as a witness for Heraclitus; and it is interesting