Abstract

One of the singularities of Latin exegesis of Aristotle’s Sophistici elenchi, is that it arbitrarily brought together two families of fallacies, the «figure of speech» and the «accident», despite the fact that they are on either side of the divide between sophisms related to expression and sophisms independent of expression, a divide that lays at the heart of Aristotle’s taxonomy of sophistic arguments. What is behind this surprising identification? The talk is meant to show that it actually originates from a curious mistake in Boethius’ translation of Aristotle’s Sophistici elenchi, 22, 178b 36-37 which radically transformed the nature of the argument at stake. While it was originally an example of the fallacies related to the «figure of speech», Boethius’ translation wrongly brings about two arguments instead of one, both related to the «accident». This explains why authors from the Latin tradition came to think that fallacies of «figure of speech» were linked to fallacies of «accident» closely enough to ask whether they actually fell outside expression, even though it does not at first glance appear that such a possibility was allowed or even suggested by Aristotle’s text. This odd question illustrates some of the remarkable features of the medieval archive and how some of its most peculiar problems came to be. It specifically allows us to reconstruct the mechanisms through which a minor disturbance in the letter of the text leads to a whole new way of organising its exegetical material.