The Conceptual Unity of Aristotle's Rhetoric

The standard view--that the Rhetoric lacks conceptual unity--has strong and prestigious support, stretching over most of the century. To David Ross in 1923 the unity of the Rhetoric was practical, not theoretical; to misunderstand this fact was to see this work, mistakenly, as "a curious jumble of literary criticism with second-rate logic, ethics, and politics, and jurisprudence, mixed with the cunning of one who knows well how the weaknesses of the human heart are to be played upon" (285). So unaltered has been this view over the years that, in 1995, Jonathan Barnes could turn Ross's metaphor into an argument concerning the Rhetoric's lack of intellectual focus: it presents, not one art, but

fragments of three arts, and of three arts that exist quite independently of rhetoric. The sections on logic have the closest connections with dialectic and in part overlap with what Aristotle says in the Topics; the sections on the emotions are linked to the ethical and psychological writings; Book III . . . shares its subject matter with the Poetics. Rhetoric . . . is a magpie, thieving a piece of one art and a piece of another, and then botching a nest of its own. (Companion 263-64)

To Ross and Barnes, while logos, êthos, and pathos are essential to rhetorical invention, and style and arrangement are essential to rhetorical performance, they have nothing conceptually in common. Insofar as disciplinary identity depends on such conceptual unity, rhetorical studies lack such identity.

William Grimaldi offers an alternate view; beneath Ross's jumble, he sees as the work's fundamental unifying principle, the enthymeme, which serves both as a form of argument and as a source for arguments about fact, character, and emotion. On this view, the factual content of a rhetorical argument is structured logically, a structure that is then molded linguistically by means of êthos and pathos to suit the requirements of the particular case: "The content of . . . statements, the manner in which propositions will be expressed, the language, and images in casting the statements will be determined by the topical examination of the ethos and pathos of [a] particular speaker-auditor's situation" (Commentary I, 335n). Grimaldi convinced no one that this was the case; as Jakob Wisse has pointed out, Grimaldi's interpretation takes unacceptable hermeneutic liberties (27-28); nothing in the Rhetoric supports his view of the enthymeme as the central organizing principle of the work.

Although we agree that Grimaldi was mistaken, we would like to argue that he was correct in seeing inferential processes as the key to the Rhetoric's conceptual unity. We would like to claim, however, that, contrary to Grimaldi's view, each of the three pisteis is inference-based, as is the canon of style. Moreover, the canon of arrangement, while not inference-based for Aristotle, plays a vital role in strengthening the patterns of inference that speakers prefer; in addition, it can be inference-based, as Perleman and Olbrechts-Tyteca have shown. So much conceptual unity seems to us consistent with Aristotle's definition of rhetoric; if such unity exists, it would mean that Aristotle was correct in calling rhetoric a tekhnê, which was, like dialectic or medicine, a unified body of knowledge (1.1.2-1354a1-14)

We are not going to argue that the text as we have it is conceptually unified; obviously, no text that contains inconsistent views and serious ambiguities can be unified and there is little doubt that the Rhetoric contains such inconsistencies and ambiguities. With the aid of current scholarship, however, we intend to construct from the text a consistent theory of persuasion with inference as its base. In our construction, we will add nothing to Aristotle. But we will omit some passages--especially, a crucial one in which he seems clearly to spurn emotional and ethical means of persuasion (1.1.3-1354a16-45). We will also prefer some interpretations--especially an...