The Unity of Aristotle’s Metaphysics

The anthology that the ancient editors of Aristotle’s manuscripts entitled literally After the Physics is the original paradigm for “meta level” inquiry in general. Medieval writers like Avicenna, Aquinas, and Duns Scotus took it as an authoritative treatise on Being.

The German classical scholar Werner Jaeger greatly influenced early and mid-20th century readings, with claims that apparently inconsistent statements in the Metaphysics reflected different stages in the development of Aristotle’s thought. Earlier, the Marburg neo-Kantian Paul Natorp had proposed that parts of the Metaphysics reflected an immature stage of Aristotle’s thought, and should be removed from the text.

The mid-20th century Catholic scholars Joseph Owens and Giovanni Reale, who both produced valuable major studies of the Metaphysics, were prominent opponents of these developmental approaches, emphasizing instead the dialectical character of Aristotle’s thought. My favorite translator, Joe Sachs, strongly agrees with them. Just because Aristotle says apparently conflicting things on the same topic is no reason to assume that he changed his mind.

That Aristotle’s thought is in general highly coherent ought to be clear to serious students. That it is highly dialectical is easy to establish. Aristotle frequently makes preliminary statements that are easy to grasp, and then substantially corrects them later within the same text. If we are serious about interpretive charity, we ought to try this sort of reading first.

But neither is its dialectical coherence proof that the Metaphysics, despite appearances, was originally written by Aristotle as a single work, structured in the order in which it has come down to us. The coherence of his thought is one question; the composition of the manuscript is another. Owens, Reale, and even Sachs tend to write as if the fact that it is dialectically coherent meant that the surviving text must be basically in the form in which Aristotle wrote it (but see quote below). This is also too strong.

Many different linear orderings of presentation may reflect the same underlying dialectical coherence, so the finding of coherence is not sufficient to establish that the order of the text is Aristotle’s.

Sachs notes that “Its first two books are both numbered one (with upper and lower case alphas)…. But in content, the Metaphysics begins over again much more than twice. Of its fourteen books, only books VIII, IX, and XIV are not new beginnings. The eleven sections of the whole inquiry are not set end-to-end like bricks in a row, but are woven together like threads in a complex design” (p. xv). “And there is no question that the composition of the Metaphysics was not a single act; the work is compiled from a number of separately composed pieces” (p. xii).

I like to imagine that my many blog posts are “woven together like threads in a complex design”, even though they were all written separately, and can be read in many different orders. If I were to turn them into a book, it would not consist of all the posts in chronological order, but would be thematically organized in some way. That is part of the art of editing. I do change my mind from time to time, but the great majority of differences in treatment of related topics have to do with differences in context, or different “places” in a more abstract dialectical development.

I see the various “sections” of the Metaphysics in the same way — as originally separate writings that nonetheless cohere, because the thought in them coheres.

What makes this question important is its relative effect on various points of interpretation. For example, I see Aristotle’s two brief discussions of “being qua being” as relatively isolated responses to what would be an important Platonic question that Aristotle himself decisively moves beyond (especially in books VI through IX), and thus as far from defining the subject of the Metaphysics as a whole. Even if the text were a continuous whole, the claim that the whole is adequately characterized as about being qua being ought to be viewed as at best highly contentious.

I very much like Sach’s top-level summary quoted in Long Detour?, which emphasizes not being qua being, but the dependencies of being on forms and the good.