Is There Room for Plato in an Aristotelian Theory of Essence?

Abstract

Aristotle’s discussion of essence in Metaphysics, Z.6 continues his interest in earlier chapters in distinguishing different levels of things that have, an essence. The "Basic Argument" at the beginning of the chapter is based on “received” views and makes no distinction in levels. The Pale Man arguments immediately following deal with entities that are not primary and so have an essence in only a reduced sense; Aristotle argues (as he admits) unsuccessfully that such entities are not identical with their essence. In the "Elaboration of the Basic Argument" that immediately follows, he fixes on a class of items, Platonic Forms, that do count as primary, that enjoy an essence in the primary sense, and that are duly identical with their essences. On this last point, arguably, Aristotle is motivated by the thought that for any set of entities that are fundamental in a given ontology, to the question “what is the essence of this item?” the only possible answer must be: itself. It is argued that the resort to Platonism in the Elaboration of the Basic Argument does not undercut Aristotle’s arguments. In a fresh round of arguments, Aristotle connects a principle from the theory of "Izzing and Having" to show that if the identity of a Platonic Form with its essence fails, then impossibly, no forms will be known, and no essences will exist. The second application of uniformity on which this latter argument depends has been criticized, but Aristotle and Plato are allies against the charge of fallacy.

This chapter was largely finished before the invitation came to contribute to the volume in honor of David Keyt, and it was only after selecting this as the appropriate chapter for the occasion that I realized that it already contained, as so often before, a use of Keyt’s published work (see n. 51 below). It is a pleasure to have found such ready evidence of my debt to him.

Appendix

I append here more detailed comments on two recent contributions to the secondary literature on Z.6.

I am thoroughly in agreement with one main theme of Dahl 2003—that the various inferences in the target passage 1030a28-b3, quoted in Sect. 1.2, should be “appropriately generalizable”—insofar as this means validating them in Aristotle’s own theory as well as in Plato’s. This puts us both on the side of purity in our approach to Z.6.

Beyond this lies disagreement. I am not sympathetic to extending the project of “appropriate generalization” so that the sameness result of Z.6, suitably understood, applies to the composite material substance and what counts as its essence (strictly, suitably asterisked to show that this is a different relation, with different logical properties, its essence*, Lewis 1984, pp. 96–118, Code 1985, pp. 118–19). According to Dahl, if the sameness result includes this further case, then the sameness relation in question must be sameness in substance or formula, but not, also numerical sameness. That is, the relation does not imply identity, as I suppose. But if the sameness result is restricted, as I think, and does not, apply to the compound material substance and its essence, then the sameness relation can be correspondingly more stringent and can imply identity.

Is it right that the conclusion of Aristotle’s Elaborated Argument applies also to the compound material substance and its essence, so that use of the weaker sameness relation is triggered, as Dahl supposes? Although the compound material substance is a substance, it is not a primary substance, and it does have a substance prior to it, namely, its constituent substantial form, which also is its essence*. So the compound material substance apparently is disqualified out of hand for the Elaborated Argument, which Aristotle expressly restricts to entities that are primary,. Dahl is well aware of the difficulty (p. 164) and attempts to counter it by pointing out that (a) the substantial form is itself dependent on its parent compound material substance. He adds that (b) the various passages in Zeta, that apparently make form primary substance can be interpreted differently, and that anyway (c) what goes elsewhere in Zeta, need not hold for this chapter.

I do not think that (c) can bear much weight: for all the fluidity in Aristotle’s thought in Zeta,, we should not want to encourage a Protagorean interpretive stance, where a new Aristotle presents himself in each chapter. And Dahl is not above bringing in the results of Z.17, for example, and the notion of a thing’s essence as the cause of its being, in order to supply assumptions to make his version of the Elaborated Argument go. I am also not in sympathy with (b): think of Z.16, for example, on why an individual substance cannot be a primary definable. At the same time, with respect to (a), Aristotle does not discuss how the existence conditions for a given form introduce a dependency on compound material substances. It bears noting, however, that a given substantial form is not (existentially) dependent on a given compound material substance but requires only that there be some compound material substance or other, in which the form is enmattered. On the other hand, a given compound material substance is (essentially) dependent on a given substantial form—without that very form, the substance could not exist. So the dependency in the two directions is not of equal strength.

At the same time, I also do not agree that the weaker sameness result, with its attendant hospitality to the compound material substance, is required, as Dahl claims it is, if we are to have an acceptable account of Aristotle’s discussion of severance at 1031 b3-11 (Sects. 2.1 and 2.2). There is no such requirement, if the argument using (15) in Sect. 2.2 succeeds.

Finally, a fresh attack on the view that the variety of sameness at work in the Elaborated Argument entails identity has been launched in Charles (2011). On the identity reading, Aristotle means to establish that Plato’s primary entities—which here go proxy for whatever entities are primary in a given ontology—are identical, with their essences. Success here, however, may bring a heavy price. Like its (linguistic) definition, the essence of a given primary entity is composed of parts, and these parts will be in some sense prior to the essence, to which they belong; by the Identity Result, then, these parts will be prior also to the entity that was supposedly primary,. Among the different remedies that might be set in place to meet this objection, Charles suggests (with Dahl 2003) that essential sameness between a primary entity and its essence does not after all require that they be identical. But is it clear that essential sameness is sufficiently weak so that the unwanted inference does not go through? In particular, the relation must not be that of “being one and indistinguishable in being,” SE, 24 179a37; cf. Phys, III.3 202b14-16, which would appear to license this and other troublesome inferences. Be this as it may, Aristotle suggests elsewhere that the thing defined—equally, then, the entity of which the essence is being given—will have parts that correspond exactly to the parts in the definition (Met Z.,10 1034b20-22): so the puzzle remains, independently of the exigencies of Z.,6. At the same time, the proposed solution itself appears to be open to the fresh difficulty that there may now exist a multiplicity of essences associated with a single given Platonic Form—multiple essences of the Good (say)—so that the Form of the Good can have an essence, and that, essence have an essence, and so on, for all we know, ad infinitum (cf. 1031b28-1032a2, leading to the regress at a2-4). These essences for the one Form will all be essentially the same and so, presumably, indistinguishable in content—but they will remain numerically distinct. My own sense is that no definitive answer to these problems is yet in prospect.