In this paper, I discuss the current thesis on the modern origin of the ad hominem-argument, by analysing the Aristotelian conception of it. In view of the recent accounts which consider it a relative argument, i.e., acceptable only by the particular respondent, I maintain that there are two Aristotelian versions of the ad hominem, that have identifiable characteristics, and both correspond to the standard variants distinguished in the contemporary treatments of the famous informal fallacy: the abusive and the circumstancial or (...) tu quoque types. I propose to reconstruct the two Aristotelian versions (see sections 1 and 2), which have been recognized again in the ninteenth century (sec. 3). Finally, I examine whether or not it was considered as a fallacious dialogue device by Aristotle and by A. Schopenhauer (sec. 4). (shrink)