My observations. 1. What exactly is the problem of the unity of a definition? (1037b8-14)
Ross: "what constitutes the unity of a subject of a definition?"
Bostock: "what is it that holds these parts together, so as to make them parts of some one thing?"
Burnyeat: "Why is two-footed animal one thing, not many?"
Marc Cohen: "Definitions are complex (a definition is always some combination of terms), so what accounts for the definiendum being one thing, rather then many?
My question is: Is Aristotle asking a linguistic question or an ontological question about definition? I mean, is Aristotle asking (a) or (b)?
(a) Why does a definition, although consisting of complex parts, e.g. rational and animal, represent a unitary "formula," e.g. a rational animal?
(b) Why can definition, although consisting of complex parts, e.g. ratioanl and animal, pick out a unitary "thing," e.g. a human being?
It seems to me that Bostock and Marc Cohen go by (a), and Ross and Burnyeat go by (b). I'm inclined to think the latter is right. I don't think what concerns Aristotle is the unity of the parts of the definition, rather, the question why the definition, although consisting of complex parts, can pick out one thing, not many. The method of division (1037b27-38a17) may best support the ontological thesis.
2. Two observations on 1037b14-26
In 1037b14-26 Aristotle rules out the possibility that the unity of the unity of definition could be solved by that the genus participates in its differentiae (e.g. a pale man: the man participates in its attriubute pallor). For (a) the genus does not participate in its differentiae, else the genus would participate contrary attributes at the same time, and (b) if the genus participates in its differentiae, all the differentiae would form a unity. Aristotle goes on and explains "But the things in a definition must in fact be a unity, for a definition is a unitary formula, and must be a formula of some unitary thing since it is a formula of a substance; for a substance is a unitary thing, and signifies a this (37b24-26)."
I have two observations
(1) It seems to me that Aristotle just begs his question here. Here Aristotle seems to hold that the unity of the parts of a definition is due to the corresponding fact that what a definition picks out is one substance, a this. But the question is just why a definition can pick out one thing not many (if the ontological thesis can be appropriately held).
(2) Another observation about this paragraph is that: Bostock proposes that in Zeta 12 what a definition picks out is not a single particular but a universal particular. But clearly Aristotle talk about "a this" here.
3. About "method of division" (1037b27-38a9)
Aristotle proposes a solution called "method of division." According to this method, one begins with the broadest genus that contains the species to be defined, and divides the genus into a subgenus by means of some differentia. Then one divides again by adding a further differentia of the differentia. And so on one continues until the last differentia, only a single species, not further divisible, remains. This is what the unity of definition is meant to be.
Two observations on the paragraph of the method of division:
(1) What is the relation between the genus and its differentiae?
In 1038a5-9 Aristotle mentions "If, then, the genus does not, in any unqualified sense, exist apart from the forms of the genus, or if it exists but only as matter--for voiced sound is the genus and the matter, and the differentiae make from this the forms of sound and the phonetic elements--then it is clear that the definition is just the formula composed of the differentiae."
Bostock points out that here we are offered two versions of the argument for the claim that the definition is "the formula composed of the differentiae." (a) The first is that the genus does not exist apart from the species, but what does it mean? (b) The alternative is that the genus exists only as "matter" to the species. Again, what does it mean?
(2) About the last differentia
Even if the differentia is the last or completing one and there remains only a single species which cannot be further divided, we are still faced with the problem why genus+differentia constitutes a unity not plurality.
4. Observations on the reasons why Zeta is an instertion:
Jaeger: Z12 is an insert.
Burnyeat:
(a) The recapitulation at the end of Z11 would lead smoothly into Z13
(b) Just like Z7-9, Z12 is not recalled in the summary of H1.
(c) But it is not so clear that the insertion was due to Aristotle himself, but probably a later editor.
Bostock:
(a) In PA I,2-3 Aristotle himself attacks the kind of definitions described here, i.e. those that are obtained by a divsion, as being both usellss and impossible. So Z12 was written before PA I,2-3, and even further, Z12 antedates practially all of Aristotle's serious work in comparative biology, perhaps not much after the Analytics to which it refers.
(b) Aristotle's view of the soul as the form of the body (Z10-11) superseded some earlier views. One may reasonably conjecture that Z12 antedates Z10-11.
