At Metaphysics Zeta (Book VII, Chapter 17, Bekker 1041b10-30), there is a clear anticipation of Bradley's Regress and an interesting formulation of what may well count as the fundamental problem of metaphysics, the problem of unity. What follows is the W. D. Ross translation of the passage. It is a mess presumably because the underlying Greek text is a mess. The Montgomery Furth and Richard Hope translations are not much better. But the meaning is to me quite clear, and I will explain it after I cite the passage:
Since that which is compounded out of something so that the whole is one, not like a heap but like a syllable-now the syllable is not its elements, ba is not the same as b and a, nor is flesh fire and earth (for when these are separated the wholes, i.e. the flesh and the syllable, no longer exist, but the elements of the syllable exist, and so do fire and earth); the syllable, then, is something-not only its elements (the vowel and the consonant) but also something else, and the flesh is not only fire and earth or the hot and the cold, but also something else:-if, then, that something must itself be either an element or composed of elements, (1) if it is an element the same argument will again apply; for flesh will consist of this and fire and earth and something still further, so that the process will go on to infinity. But (2) if it is a compound, clearly it will be a compound not of one but of more than one (or else that one will be the thing itself), so that again in this case we can use the same argument as in the case of flesh or of the syllable. But it would seem that this ‘other’ is something, and not an element, and that it is the cause which makes this thing flesh and that a syllable. And similarly in all other cases. And this is the substance of each thing (for this is the primary cause of its being); and since, while some things are not substances, as many as are substances are formed in accordance with a nature of their own and by a process of nature, their substance would seem to be this kind of ‘nature’, which is not an element but a principle. An element, on the other hand, is that into which a thing is divided and which is present in it as matter; e.g. a and b are the elements of the syllable.
My interpretation is as follows.
The syllable ba is a whole of two parts b and a. But it is more that the mere 'heap' or collection of its two parts. If the whole exists, then the parts exist. But if the parts exist, it does not follow that the whole exists. Therefore, the whole is not identical to its parts: it is its parts and something more. The problem is to specify what this 'something more' is.
The same problem arises for many sorts of wholes, though we ought to dismiss Aristotle's example of fire and earth composing flesh. That is primitive physiology, but the problem, being one of general ontology, survives the rejection of the primitive physiology.
Aristotle speaks of "something else" as that which make the difference between the mere sum of vowel and consonant, on the one hand, and the syllable, on the other. Call this "something else" U, for 'unifier.' U is that which unifies b and a. Now U is either an element (a simple) or a compound composed of elements.
Suppose U is an element. Then what we have is b, a, and U. We have a collection of three items. But the existence of these three items does not entail the existence of ba. Therefore, ba cannot be identical to these three items: it must be something more than them. Now if you introduce U* to do the job that U failed to do — the job of unification — then the problem arises once again. You have four disconnected items. Adding a further element U** obviously does no good and issues in an infinite regress.
An infinite regress may or may not be vicious, but in this case it appears to be vicious. For what we want is an account of unity, but if we try to account for unity by postulating a special unifying constituent, we are left with the problem of explaining the unity of this unifier with the constituents it is supposed to unify. At no stage in the regress do we arrive at something more than a mere 'heap' or collection.
So U cannot be an element. But if U is a compound, then the same problem arises as arose regarding ba, namely the problem of explaining how it is more than a mere 'heap' of its constituents.
What then is U if it is neither a simple nor a compound constituent of the thing whose unity it is supposed to explain? Aristotle's answer is that the substance of a thing is that which unifies it. This substance is not an element but a principle. Aristotle's answer in terms of substance or nature is obscure, as obscure as his conception of substance. But that is a huge topic and not my present concern.
What interests me is not Aristotle's (defective) answer, but his clear grip on a central problem of metaphysics -- one that no one has satisfactorily solved -- and his equally clear anticipation of a form of argumentation that in recent times flies under the flag, 'Bradley's Regress.' Bradley of course approaches the problem differently as one about relational facts involving what have come to be called external relations. But it is the same general problem, namely, that of the unity of a complex.
If anyone is interested, I explore this problem in painful detail in Chapter VII of A Paradigm Theory of Existence.