30 April 2006

Splitting the Atom

I wonder if anyone saw the significance in the passage from the Categories which I quoted the other yesterday:

Some are in a subject but are not said of any subject. (By 'in a subject' I mean what is in something, not as a part, and cannot exist separately from what it is in.) For example, the individual knowledge-of-grammar is in a subject, the soul, but is not said of any subject; and the individual white is in a subject, the body (for all color is in a body), but is not said of any subject. (Ackrill)
It's a common view that, in the Categories, Aristotle treats substances as unanalyzable 'atoms'. One finds this claim repeated again and again. S. Marc Cohen's statement, from his SEP article, also quoted already (in part), is representative:
The concepts of matter and form, as we noted, are absent from the Categories. Individual substances — this man or that horse — apart from their accidental characteristics — the qualities, etc., that inhere in them — are viewed in that work as essentially simple, unanalyzable atoms. Although there is metaphysical structure to the fact that, e.g., this horse is white (a certain quality inheres in a certain substance), the fact that this is a horse is a kind of brute fact, devoid of metaphysical structure. This horse is a primary substance, and horse, the species to which it belongs, is a secondary substance. But there is no predicative complex corresponding to the fact that this is a horse in the way that there is such a complex corresponding to the fact that this horse is white.
But this is false, isn't it?

Look again at that passage from the Categories. There we find, at the very beginning of the work--when Aristotle is setting down some fundamental distinctions, between substance and accidents, and between particulars and universals--he gives as examples, not "individual knowledge-of-grammar is in the individual man" or "individual whiteness is in the individual man", but rather "individual knowledge-of-grammar is in the soul", and "individual whiteness is in the body."

That is, from the start, Arisotle is prepared to analyze primary substances, when they are living things at least (and these are agreed to be paradigmatic primary substances, for Aristotle) into body and soul.

Cohen is simply incorrect that "this is a horse is a kind of brute fact, devoid of metaphysical structure." No, 'this is a horse' is analyzed into 'this is a soul and a body'.

Tell a story, if you wish, about how Aristotle's use of the form-matter distinction in analyzing substances is a development of the soul-body distinction. But that Aristotle recognizes the possibility and appropriateness of analysis, from the very start, is undeniable.

29 April 2006

My Accomplishment Consists in This, That I Am Aware That I Lack Accomplishment

From what has been said, then, it is clear that intellectual accomplishment is a combination of systematic knowledge and intelligence, with the things that are highest by nature as its objects.
I don't imagine anyone could guess in advance that this sentence was about sophia. It's a translation of Nic. Eth. 6.8 1141b2-3. e0k dh_ tw~n ei0rhme/nwn dh~lon o3ti h( sofi/a e0sti\ kai\ e0pisth&mh kai\ nou~j tw~n timiwta&twn th|~ fu&sei. (And 'highest' is an undertranslation, to be sure.)

And consider this one, the sentence that follows:
That is why people call Anaxagoras and Thales and people of that sort 'accomplished', but not 'wise', when they see them lacking a grasp of what is to their own advantage; and they say that people like that know things that are exceptional, wonderful, difficult, even superhuman--but useless, because what they inquire into are not the goods that are human. It is wisdom that has to do with things human.
This is almost unintelligible and fails dramatically to convey Aristotle's sense-- "Ah, there goes Thales again, falling into a well. That's what happens when you are accomplished but lack any wisdom."

I take both of these sentences from Rowe and Broadie, who render sophia as 'intellectual accomplishment' and phronesis as 'wisdom.' Their treatment of the intellectual virtues, as a result, becomes so confusing that I can hardly recommend their translation, even if in other respects it may be the best available.

Work in Progress

This is something odd I noticed about S. Marc Cohen's entry on "Aristotle's Metaphysics" in SEP.

The entry amounts to about 12 pages on my screen. Of these, 2 pages are introductory. 9 concern book Z. One page deals with material elsewhere in the Metaphysics. There is no discussion of Lambda.

How is this a suitable encyclopaedia entry on this topic? On his homepage Cohen describes the entry as 'a nearly complete draft'. But since when does SEP publish drafts? (I see nothing in the SEP editorial policy acknowledging that it publishes drafts.) And if some of its entries are 'drafts', shouldn't these be identified as such?

True, SEP is a self-described 'dynamic reference source'. But shouldn't that mean that finished drafts will be revised as called for, not that drafts get posted as if they were finished?

A Schizophrenia in Aristotle Scholarship

When people lecture on Aristotle's De Anima, it is common, I think, that they say something like this:

Aristotle takes a hylomorphic view of the relationship between soul and body. That is, he thinks that soul stands to body as form to matter. But forms depend on the matter which they inform: when the bronze is melted, the statue is destroyed at the same time. Thus, Aristotle's hylomorphism places him generally at odds with Plato. Hylomorphism is inherently incompatible with dualism. It is true that Aristotle leaves open the possibility that some souls, or some functions of some souls, exist independently of a body. And possibly he even believes that himself. But, if he does, this looks like a strange and perhaps even unaccountable divergence from what is the basic outlook of his philosophy of mind.
Yet when they lecture on the Metaphysics, they might say something like the following:
The central books of Aristotle's Metaphysics are a careful examination of substance as form. In these books, Aristotle looks largely at 'sensible' substances, that is, substances that have matter. Yet he thinks that Metaphysics in the proper sense is 'first philosophy', which is concerned with, as he puts it, non-sensible and unchangeable substances. His discussion of sensible substances is meant to prepare the way for his examination in book Lambda of separately existing, non-sensible substances. These are pure forms, which exist without matter.
That is, these lecturers ascribe to Aristotle the incompatible view that (i) forms can exist without matter, and that the most important substances are like that (those which, from the point of view of knowledge and explanation, are most basic); yet (ii) forms depend on matter for their existence, and it would be strange and unaccountable to regard some forms as independent of matter.

Chris Shields, in his SEP article, is not so crude as the lecturer I imagine above:
In a fairly direct way, though, the question of whether soul and body are one loses its force when it is allowed that it contains no implications beyond those we establish for any other hylomorphic compound, including houses and other ordinary artifacts.

One way of appreciating this is to consider a second general moral Aristotle derives from hylomorphism. This concerns the question of the separability of the soul from the body, a possibility embraced by substance dualists from the time of Plato onward. Aristotle's hylomorphism commends the following attitude: if we do not think that the Hermes-shape persists after the bronze is melted and recast, we should not think that the soul survives the demise of the body. So, Aristotle claims, “It is not unclear that the soul -- or certain parts of it, if it naturally has parts -- is not separable from the body” (De Anima ii 1, 413a3-5). So, unless we are prepared to treat forms in general as capable of existing without their material bases, we should not be inclined to treat souls as exceptional cases. Hylomorphism, by itself, gives us no reason to treat souls as separable from bodies, even if we think of them as distinct from their material bases. At the same time, Aristotle does not appear to think that his hylomorphism somehow refutes all possible forms of dualism. For he appends to his denial of the soul's separability the observation that some parts of the soul may in the end be separable after all, since they are not the actualities of any part of the body (De Anima ii 1, 413a6-7). Aristotle here prefigures his complex attitude toward mind (nous), a faculty he repeatedly describes as exceptional among capacities of the soul.

Yet even Shields' discussion leaves something to be desired, because he writes as if the independence of mind from body is a special exception to the doctrine of hylomorphism, pressed upon us, perhaps, by attention to human psychology. There is no suggestion in the above paragraph--as there should be--that Aristotle thought that a careful examination of hylomorphism itself would show that forms could in principle exist independently of matter. (The sentence, "At the same time, Aristotle does not appear to think that his hylomorphism somehow refutes all possible forms of dualism", is misleading through understatement.)

Souls Are Particulars

Chris Shields considers this neat argument, in his SEP article on Aristotle's psychology, that, according to Aristotle, particular forms exist:

  1. All souls are forms.
  2. Some souls are particulars.
  3. Thus, some forms are particulars.
(It could be that all souls are particulars. The weaker claims suffice.)

Since 1. is so entrenched in Aristotelian psychology, the only plausible way to deny the conclusion is to deny that Aristotle accepts 2. Shields admits that in this controversy it is difficult to find a 'proof text'. Yet he does not consider Categories 1a25.

It is conceded on all sides that the Categories distinction between 'said of'/'not said of' is meant by Aristotle to be a distinction between universals and particulars: 'This particular instance of grammar', since it is an individual, is not 'said of'' someone, although 'grammatical', a universal, is. ('Socrates is grammatical' makes sense; 'Socrates is an instance of grammatical knowledge' does not.)

Presumably, too, a particular which exists in something, exists only in a particular.

Thus, when Aristotle gives as an example:
Some are in a subject but are not said of any subject. (By 'in a subject' I mean what is in something, not as a part, and cannot exist separately from what it is in.) For example, the individual knowledge-of-grammar is in a subject, the soul, but is not said of any subject; and the individual white is in a subject, the body (for all color is in a body), but is not said of any subject. (Ackrill)

ta_ de\ e0n u(pokeime/nw| me/n e0sti, kaq' u(pokeime/nou de\ ou)deno_j le/getai, e0n u(pokeime/nw| de\ le/gw o4 e1n tini mh_ w(j me/roj u(pa&rxon a)du&naton xwri\j ei]nai tou~ e0n w|{ e0sti/n, oi[on h( ti\j grammatikh_ e0n u(pokeime/nw| me/n e0sti th|~ yuxh|~, kaq' u(pokeime/nou de\ ou)deno_j le/getai, kai\ to_ ti\ leuko_n e0n u(pokeime/nw| me/n e0sti tw|~ sw&mati, a3pan ga_r xrw~ma e0n sw&mati, kaq' u(pokeime/nou de\ ou)deno_j le/getai:
This shows that he regards some souls as particulars, in just the same way that he regards bodies as particulars. (One may resist the conclusion by invoking developmentalism, but it is not clear that that recourse would escape the charge of circularity.)

28 April 2006

Sorted Details

I'm following three threads, now: (i) the status of primary substances in the Categories; (ii) the nature of the investigation in Met. Z; and (iii) how Met. Z contributes to the 'first philosophy' of Lambda. Here is something on that second thread.

Recall that I took a hint from Aquinas in De Ente et Essentia and began to think of Z as asking not, as is commonly supposed, 'What is substance?', but rather, 'What makes something such that it can be sorted?'--it being presumed that substance does that. I said that, according to Aquinas, Aristotle regarded this investigation as suggested by the doctrine of the categories. I gave Aquinas' argument for this, based on the fact of the categories, and I gave an alternative argument, based on the priority of substance among the categories. I said that I saw no evidence that Aquinas' argument was in Z1, but that I did see evidence that the other argument was in Z1.

I want to take that back. I now believe that both arguments are put forward at the opening of Z1. To see that they are requires that we parse the opening of the chapter appropriately. Here is the relevant passage, parsed as I think it should be:

(A) There are several senses in which a thing may be said to 'be', as we pointed out previously in our book on the various senses of words;' for in one sense the 'being' meant is 'what a thing is' or a 'this', and in another sense it means a quality or quantity or one of the other things that are predicated as these are.
(B) While 'being' has all these senses, obviously that which 'is' primarily is the 'what', which indicates the substance of the thing.

(i) For when we say of what quality a thing is, we say that it is good or bad, not that it is three cubits long or that it is a man; but when we say what it is, we do not say 'white' or 'hot' or 'three cubits long', but 'a man' or 'a 'god'.
(ii) And all other things are said to be because they are, some of them, quantities of that which is in this primary sense, others qualities of it, others affections of it, and others some other determination of it. (And so one might even raise the question whether the words 'to walk', 'to be healthy', 'to sit' imply that each of these things is existent, and similarly in any other case of this sort; for none of them is either self-subsistent or capable of being separated from substance, but rather, if anything, it is that which walks or sits or is healthy that is an existent thing. Now these are seen to be more real because there is something definite which underlies them (i.e. the substance or individual), which is implied in such a predicate; for we never use the word 'good' or 'sitting' without implying this. Clearly then it is in virtue of this category that each of the others also is. Therefore that which is primarily, i.e. not in a qualified sense but without qualification, must be substance.)
I take (i) and (ii) to provide reasons for the inference from (A) and (B).

But--and this is crucial--I understand (B) to be, not the assertion of the priority of substance (which has been said many times earlier in the treatise), but rather the added idea that this category which is prior indicates 'what' a thing is and that this is called 'substance'. That is, substance is that which makes something such that it is identifiable as a 'what' and can be sorted.

Argument (i) in support of this does seem to be Aquinas' argument. Commentators puzzle over why Aristotle starts talking about the category of quantity here. "Aristotle's object being to distinguish quality from substance, not from the other categories," Ross comments, "'three cubits long' is irrelevant and was suspected by Bonitz. It was, however, read by Alexander, and apparently by Asclepius, and Aristotle is not incapable of such irrelevancies..." Better, I should think, to conclude that Ross is not correct about Aristotle's object. Suppose his object is to make a point about sorting. Suppose he wants to make the subtle and complex observation that, to say anything at all (white, three cubits) presupposes a fundamental act of sorting, viz. into one of the categories; and yet it's not as if all of the categories are on a par, since that which we explicitly recognize as telling us 'what' a thing is gets priority. (Odd, because all the categories tell us what a thing is.)--This, then, would be an argument like that which Aquinas puts forward.--And in this connection the mention of a quantity would hardly be an 'irrelevancy'.

Argument (ii) is the argument that I proposed.

Note, by the way, that if by 'substance' we mean that which makes something such that it can be sorted, then 'substance' would have to indicate some part or aspect of a thing (and so we would want to talk about the substance of a thing) because, clearly, we do not count everything about an individual as relevant to sorting it.

27 April 2006

Getting to the Substance of Things

"Thanks for the question, especially because it made me consider something I hadn't thought about distinctly. Wouldn't it have to be that primary substances are 'theoretical substances'? Though it's difficult to conceive of a 'this man' without his accidents, if a first substance were to include the accidents, then wouldn't an accidental change, by that conception, actually be a substantial change? I mean, if a primary substance includes the accidents, then a change in the accidents would make the 'this man' now a different 'this man'. For example, if Socrates as a primary substance includes the color of his hair, then wouldn't dying his hair make him a new instance of man?"
In my informal polling, that's a typical response. I find that people have not thought clearly about this; yet, when the distinction I raised is posed to them, then they are compelled to say that a primary substance in the Categories is 'an instance of a nature' (as I put it), not a commonsense 'thing'.

What's interesting about an 'instance of a nature', is that, as the quotation above notes, it would lack definite determination of its matter. An instance of human nature would surely be an instance of a soul-with-body, and yet the body would have to be conceived--somehow--as not having a definite size. One would be tempted to say that an 'instance of human nature' was somehow simply a form, that it did not involve matter (on the grounds that matter had to be definite).

Far from this being an objection, I find this difficulty promising as regards interpretation, because it shows that a consideration of the relation of form to matter in primary substances would follow naturally, and almost inevitably, from the very conception of a primary substance understood as an 'instance of a nature'. Think of a primary substance as human nature without accidents, and then inevitably one wants to know how this human nature could involve matter (the body) if at all.

I think it's possible to show that various passages in the Categories require that primary substances be 'theoretical substances'. I'll try to post more on this.

Can Atomists Be Happy?

The following occurred to me when teaching Nic. Eth. 10.7-8 today, in connection with my previous post on Aristotelian monotheism.

Prove: An atomist cannot be happy.


  1. Happiness is activity in accordance with the virtue of sophia (1177a12).
  2. Thus, no one who lacks sophia can be happy.
  3. Sophia is knowledge of first causes (1141a19-20).
  4. No one who has false beliefs, and lacks true beliefs, in that respect has knowledge (1139b20-21).
  5. Thus, no one who has false beliefs, and lacks true beliefs, about first causes, has knowledge of first causes.
  6. Atomists have false beliefs, and lack true beliefs, about first causes.
  7. Thus, no atomist has sophia.
  8. Thus, no atomist can be happy.


'Atomist' of course stands proxy for that class of people who, on Aristotle's terms, lack knowledge of first causes.

We might grant that there are some philosophers who lacked sophia, but who nonetheless approached happiness in some way, because their beliefs were insightful approximations to the truth (e.g. Plato on the Good, Anaxagoras on Mind).

What the above argument does, is to raise a question about the importance of truth for happiness. Does happiness require that we attain to truth about the most fundamental things? Aristotle's remarks in the opening of the Metaphysics, and his definition of sophia as knowledge, would suggest that it does. If 'all men by nature desire to know', then those who fail, fundamentally, to know, will have failed to attain a 'natural good'--and 'a happy person has all the natural goods' (NE 1169b19).

The argument also takes seriously the fact that happiness is objective, not subjective, for Aristotle. A person can think he is happy, and not be so (see Kraut, "Two Conceptions of Happiness"). Similarly, a person who thinks he has wisdom, but does not, may think he is happy, but is not. (And note the Socratic resonances in this.)

An Aristotelian Shema

I want to answer my question from the previous post: What argument, if any, in the first philosophy of Lambda, seems intended by Aristotle to draw upon results or material from Z?

There is only one such argument, I believe:

Evidently there is but one heaven. For if there are many heavens as there are many men, the moving principles, of which each heaven will have one, will be one in form but in number many. But all things that are many in number have matter; for one and the same definition, e.g. that of man, applies to many things, while Socrates is one. But the primary essence has not matter; for it is complete reality. So the unmovable first mover is one both in definition and in number; so too, therefore, is that which is moved always and continuously; therefore there is one heaven alone. (1074a31-36)

o3ti de\ ei[j ou)rano&j, fanero&n. ei0 ga_r plei/ouj ou)ranoi\ w3sper a1nqrwpoi, e1stai ei1dei mi/a h( peri\ e3kaston a)rxh&, a)riqmw|~ de/ ge pollai/. a)ll' o3sa a)riqmw|~ polla&, u3lhn e1xei (ei[j ga_r lo&goj kai\ o( au)to_j pollw~nm oi[on a)nqrw&pou, Swkra&thj de\ ei[j): to_ de\ ti/ h}n ei]nai ou)k e1xei u3lhn to_ prw~ton: e0ntele/xeia ga&r. e4n a1ra kai\ lo&gw| kai\ a)riqmw|~ to_ prw~ton kinou~n a)ki/nhton o1n: kai\ to_ kinou&menon a1ra a)ei\ kai\ sunexw~j: ei[j a1ra ou)rano_j mo&noj.
This makes use of the notions familiar from Z of to ti en einai, eidos, and hule; and it makes use, too, of a principle from Z, that matter individuates form. And it is the only such passage.

Now, I can't decide whether this is a significant argument or not:
It is insignificant. Aristotle mentions it in passing. It seems a digression. It looks like an ancillary argument, against those who have claimed that there are 'many worlds'.

It is significant. Although the argument is put forward almost in passing, what is important about it, is the principle on which it depends, and which Aristotle affirms, (using results from Z), namely, that the 'first essence' or 'unmoved mover' --God-- is one.
It seems an odd suggestion, and yet there it is: Could it be that the main purpose and point, in Aristotle's eyes, of all of the build-up to first philosophy, in Z and in the other central books of the Metaphysics, is the simple assertion of monotheism? And then the 'wisdom' of first philosophy becomes: the intellectual recognition of a single, supreme first cause.

On this interpretation, the Metaphysics is, as it were, Aristotle's treatise-length and philosophical equivalent of the shema.

A strange idea, I agree, but I do like it better than the 'philosophical fizzle' interpretation.

26 April 2006

A Philosophical Fizzle

I pointed out in an earlier post that the attention that Aristotle gives to perceptible substances in the central books of the Metaphysics, and in Lambda 1-5, raises a difficulty: How does this investigation differ from physics?--since, for Aristotle, physics is distinguished from purely formal disciplines on account of its considering matter together with form. Burnyeat's answer was that Aristotle was 'doing second philosophy conceived as first philosophy', yet it was unclear what this might be.

Here I wish to draw attention to an additional difficulty that arises even if we grant that it makes sense to say that, in those texts, Aristotle is 'doing second philosophy conceived as first philosophy.'

The difficulty is that there still must be some way in which 'second philosophy conceived as first philosophy' is useful for first philosophy proper. After all, why doesn't Aristotle simply launch into a discussion, from the start, of the existence and number of imperceptible substances? Why exactly do we need any preparation for this? Aristotle's own answer is that it helps for us to begin first with things 'more knowable to us', before we move on to a discussion of things 'more knowable in themselves', and (he says) perceptible substances, which have matter, are more knowable to us. Good--but what precisely is the help that we get from them? What are the big lessons, ideas, or principles that we take away from this study and then use when we are doing first philosophy proper?

Let me put the question more concretely: Can you point to a single argument in Lambda, and some argument in Z, such that, in your view, Aristotle regards that argument in Lambda as receiving confirmation or support, which derives from that argument in Z? Or, again: What in Lambda is such that, as Aristotle sees it, a reasonable person should be more disposed to accept it, precisely because of something previously examined in Z? (Or say, if you wish, "because of something like what is examined in Z", to allow for a sometime independent composition of the treatises.)

Burnyeat's Map, despite its consideration of "The place of Z in Aristotle's Metaphysics" (ch. 6), gives almost no insight to this question. The following seems the only relevant remark:

H reworked the concepts of matter and form in terms of potentiality and actuality, these being the concepts we need to make the transition to the non-sensible realm where neither matter nor form apply. Q then undertakes a careful analysis of potentiality and actuality as such. (121)
That is, according to Burnyeat, the entire 'yield' of the central books of the Metaphysics, for the purposes of first philosophy, is the analysis of potentiality and actuality carried out in Q. (One would think that, on the terms of his interpretation, it would be an important result, somewhere, in the central books, that form need not depend on matter for its existence. But I do not see that Burnyeat emphasizes this.) Presumably, then, the consideration of form and matter in Z is 'second philosophy conceived as first philosophy' only to the extent that it prepares the way for the discussion of potentiality and actuality in Q.

We may waive the difficulty of how a 'reworking of the concepts' of A and B could be of any use in an investigation of 'a realm in which neither A nor B apply'.

A more relevant difficulty is that it is not at all clear how Z prepares the way for or is necessary for a consideration of actuality and potentiality. Moreover, if Burnyeat's view were correct, wouldn't that be a disappointing and unsatisfactory result? All of the dialectic in Z, all of its careful analysis and subtle distinctions--this would be of no more philosophical consequence, for Aristotle's project, than to prepare the way (somehow) for a discussion of actuality and potentiality, which, gets used in some of the arguments of first philosophy proper?

That looks like a lot of build-up for a very disproportionate payback. A philosophical fizzle, really.

An Argument for Theoretical Primary Substances

Is there a way to determine whether, in the Categories, Aristotle by 'primary substance' means what I called a 'commonsensical' or a 'theoretical' substance? It would be desirable to establish the latter, I think, since this then brings the Categories into greater harmony with the Metaphysics. In the former work, too, Aristotle would be concerned with 'the substance of a thing', and he would be conceiving of it as that which places that thing in a system of genus-species classification.

It proves difficult to find arguments based on language alone. One consideration is that proper names do not figure in Cat 5, whereas they are used freely in the postpraedicamenta when they are needed to make a point-- which gives some support to the suggestion that primary substances are not what are indicated by proper names.

Here is an argument based on the reduplicative idiom qua.

"Socrates qua quality" asks us to consider some quality of Socrates, say, that he is pale, and thus we may say "Socrates qua quality is pale." In such a sentence, the word 'pale' predicates paleness of Socrates, and in the context the expression "Socrates qua quality" picks out, not Socrates, but this quality which Socrates has.

Similarly, if we say, "Socrates qua quantity is five feet tall", then "five feet tall" predicates five feet of Socrates and "Socrates qua quantity" picks out, not Socrates, but this quantity that Socrates has.

By parity, then, if we say "Socrates qua substance is a man", then "a man" predicates humanity of Socrates, and "Socrates qua substance" picks out, not Socrates, but this substance that Socrates has.

This substance that Socrates has would not be a secondary substance (note that we do not wish to say that "Socrates is man", but rather "a man", indicating an 'atom' or instance of humanity).

The point is that the reduplicative idiom has a use as regards the category of substance, just as it does for the other categories, and then what it picks out is not the individual, Socrates, nor the species or genus, but rather some instance of a nature. The claim would be what this picks out would go undesignated, and something else would be given two designations, unless 'primary substance' indicated this.

25 April 2006

Two Arguments from the Categories

Yesterday I considered the suggestion that the key question of Metaphysics Z is not, as is commonly held, "What is substance (ousia)?" but rather "What makes something such that it can be sorted?" --it being supposed that ousia does this.

I took this suggestion from Aquinas' De Ente et Essentia, approaching that work as a brief commentary on Z and Lambda. One advantage of this interpretation, I pointed out, is that it seems to make good sense of the topics actually considered in Z.

Yesterday I also noted that Aquinas thinks that Aristotle arrives at this question through a consideration of the categories. Now there would seem to be two ways in which one could take the categories to lead to the investigation of Z so conceived.

The first way, one might say, is from the very fact that there are categories; the second is from the priority of one category over the others.

The first way is what, it seemed to me, Aquinas was proposing in De Ente et Essentia. I gave a formulation yesterday, but here is another way of putting it:

We seem to incapable of considering that something exists, without supposing that it exists in one of the categories. We cannot even entertain the existence of something, without first thus classifying it. For instance, to consider the height of a horse, is already to be considering something distinct from the color of the horse. We couldn't be confused about whether we were investigating a quantity as opposed to a quality of the horse. This suggests that there is a deep connection between asserting the existence of a thing, and classifying it. We might then conclude that to assert the existence of a thing, just is to say that there is something about it such that it can be classified. Call this its ousia.

I asked whether there was warrant in Z1 for attributing an argument like this to Aristotle. After looking again at the chapter with this question in mind, I do not think that there is (although please feel free to disagree with me from your own examination of the chapter).

What I do find warrant for attributing to Aristotle, in Z1, is the following argument, which seems different from what Aquinas proposes. This argument hinges, not on the mere fact of the categories, but on the priority which the category named ousia enjoys relative to the others. This argument would proceed as follows.

To ask what ousia is, is to ask what truly or really (ontws) exists. We may understand this as: what especially (malista) exists. Let us approach this question through a consideration of the categories, since these involve predications of what 'is'. We see in that case that, although there are ten categories, the category of ousia is prior. In fact it is prior in all the ways in which one might wish to assert priority (as Z1 takes pains to observe). Thus ousia especially exists. However, a predicate which identifies the ousia of a thing tells us what, that is, what sort, it is (1028a14-18). Thus, what really exists is what makes a thing such that it can be sorted. To inquire into ousia, then, is to inquire into what it is about a thing which makes it such that it can be sorted. (And thus follows the investigation of Z.)

This latter argument does seem to me to be Aristotle's and to be found in Z1.

And this seems to me the correct way of approaching Z.

24 April 2006

Two Concepts of Substance

I find that in discussions of the Categories there tends to be a lack of clarity about what Aristotle means by a primary substance. I shall therefore draw a distinction and then, by way of illustration, give a few texts in which this distinction seems to be fudged. I shall refer to the distinction as that between 'commonsense' and 'theoretical' primary substance.

A commonsense primary substance is what we would call a 'thing'. It is a particular to which we might give a proper name. E.g. this particular horse ('Secretariat') or this particular human ('Socrates').

A theoretical primary substance is a notion we arrive at by beginning with a commonsense primary substance, but then conceiving of it (somehow) in abstraction from any of the accidents that might belong to it. E.g. this instance of horse nature; this instance of human nature.
It makes sense that, on Aristotelian terms, there should be such entities as theoretical primary substances. After all, the category of substance is distinct from that of quality, quantity, relation, and so on. Therefore, the entity which is a substance would be distinct from the entities which are qualities, quantities, relations, and so on, inhering in that substance. The substance, then, would presumably be something that existed distinct from these other things.

Of course, one might wonder how there can be an instance of human nature which was not, however, determined yet in accordance with the other categories.

Now which of these does Aristotle mean when he talks of 'primary substances' in the Categories? The language he uses is indeterminate as between them. ho tis anthropos can mean either "The particular human" (in the ordinary sense of the term) or "The instance, man" (in the theoretical sense).

Now, as I said, I find unclarity when scholars refer to the primary substances of the Categories. Dan Devereux, for instance, refers to them as 'concrete particulars' and says such things as the following:
Although the species 'horse' has two substance-making features while the individual horse has only one, the individual horse (and not the species) is a primary substance because it is more of an underlying subject (162).
It looks as though by 'individual horse' he means a commonsense thing, such as Secretariat; and yet his language, I suppose, could be taken in the other way (although, admittedly, 'concrete particular' definitely suggests the commonsense notion).

S. Marc Cohen seems to go back and forth between both notions in his SEP article on Aristotle's Metaphysics. When he says:
The concepts of matter and form, as we noted, are absent from the Categories. Individual substances — this man or that horse — apart from their accidental characteristics — the qualities, etc., that inhere in them — are viewed in that work as essentially simple, unanalyzable atoms. Although there is metaphysical structure to the fact that, e.g., this horse is white (a certain quality inheres in a certain substance), the fact that this is a horse is a kind of brute fact, devoid of metaphysical structure.
He seems to be understanding primary substances as theoretical. And yet when he says, in the same article:
At the top (or trunk) of the tree are the most generic items in that category (e.g., in the case of the category of substance, the genus plant and the genus animal); branching below them are universals at the next highest level, and branching below these are found lower levels of universals, and so on, down to the lowest level universals (e.g., such infimae species as man and horse); at the lowest level — the leaves of the tree — are found the individual substances, e.g., this man, that horse, etc.
He seems to be understanding them as commonsensical. Surely 'this man' is something like Socrates, and 'this horse' is something like Secretariat.

Which is Aristotle's notion? This much, at least, is clear: if Aristotle's notion is of a 'theoretical' substance, then it would be natural for him to speak of 'the substance of' a thing (as he does, without explanation, in the Metaphysics)--because the substance of Secretariat would be the horse nature that Secretariat has; the substance of Socrates would be the human nature that Socrates has; etc.

I believe the most common view is that by 'primary substances' in the Categories Aristotle means the commonsensical sort. But that view has a big problem--doesn't it?--namely that a commonsensical substance is not only a substance: it has lots of accidents thrown in with it.

The Significance of Ousia

I'll give here the argument I mentioned in the previous post.

According to Aquinas, Aristotle begins Z by invoking the categories (or so one may speak, if one reads De Ente et Essentia, as I have suggested, as a brief commentary on Met. Z and L), in accordance with the strategy of beginning with what is 'more knowable' by us. Aquinas suggests that the behavior of the predicate o)/n, as shown in the ramification of its uses into the various categories, is more knowable than the behavior of ou)si/a:

Quia vero ex compositis simplicium cognitionem accipere debemus et ex posterioribus in priora devenire, ut, a facilioribus incipientes, convenientior fiat disciplina, ideo ex significatione entis ad significationem essentiae procedendum est.
But then, according to Aquinas, Aristotle uses the fact of the categories, as the basis for an argument about the signficance of ou)si/a. Here is his extremely economical presentation of the argument:
(*) Et quia, ut dictum est, ens hoc modo dictum dividitur per decem genera, oportet quod essentia significet aliquid commune omnibus naturis, per quas diversa entia in diversis generibus et speciebus collocantur, sicut humanitas est essentia hominis, et sic de aliis.
Aquinas then mentions various near equivalents, as he sees it, of essentia, namely, quiditas, quod quid erat esse, forma, natura:
Et quia illud, per quod res constituitur in proprio genere vel specie, est hoc quod significatur per diffinitionem indicantem quid est res, inde est quod nomen essentiae a philosophis in nomen quiditatis mutatur.

Et hoc est quod philosophus frequenter nominat quod quid erat esse, id est hoc per quod aliquid habet esse quid.

Dicitur etiam forma secundum quod per formam significatur certitudo uniuscuiusque rei, ut dicit Avicenna in II metaphysicae suae.

Hoc etiam alio nomine natura dicitur accipiendo naturam secundum primum modum illorum quattuor, quos Boethius in libro de duabus naturis assignat, secundum scilicet quod natura dicitur omne illud quod intellectu quoquo modo capi potest. Non enim res est intelligibilis nisi per diffinitionem et essentiam suam. Et sic etiam philosophus dicit in V metaphysicae quod omnis substantia est natura.
These then gets fairly naturally connected with 'definition':
Quiditatis vero nomen sumitur ex hoc, quod per diffinitionem significatur.
And it can be seen that Aquinas thereby wishes to explain which topics get covered in Z.

But what especially interests me is the argument (*) above. It seems to be something like:
  1. The various predications of 'is' correspond to the highest sorts.
  2. Thus, what is signified by 'is' is that which makes something such that it can be sorted (at all).
  3. Ousia is what is signified by 'is'.
  4. Thus, ousia is that which makes something such that it can be sorted.
Met. Z then becomes an investigation of what it is by which things are sorted; the above account explains why, in light of this, it deals with the topics it does; 'matter' (hule) would get brought in insofar as it is relevant to sorting something; and genos and katholou have to be dealt with, additionally, because they might look like they work in that way (What sorts something if not its genus--the very sort it belongs to?).

It may be seen that Aquinas does not take Z1's appeal to the categories, as does Burnyeat, to be a restriction of our attention to the first category: he thinks that the question, 'What is it which makes this such that it can be sorted?', may be asked of a thing in any category, although, to be sure, it is most properly asked of things in the first category.

Let me underline a point: argument (*) seems to me one of the more startling, interesting, and (if correct) important arguments I have encountered. I wonder if there is warrant in Z1 for ascribing it to Aristotle.

On Existence and Substance

This last weekend I picked up Thomas Aquinas, De Ente et Essentia, on the hunch that somehow it would be relevant to what I had been thinking about. I confess that I hadn't understood before the significance of this book, since I had in a fairly foolish way allowed myself to get misled by the common translation of the title (On Being and Essence).

This time I was astonished to find that Aquinas uses ens for o)/n, and essentia for ou)si/a. A scholar of Greek philosophy, then, would wish to understand the title as, "On Existence and Substance", or, perhaps, "On Being and Substantial Being".

When its theme is understood in that way, the opuscula acquires the appearance of, in effect, a brief and elegant commentary on Z and its relation to L. Furthermore, it is a commentary which is enlightening insofar as it builds on a tradition very different from ours, that is, the tradition of Arabic commentators.

I want to share with you an interesting argument from its beginning.

The opening chapter of Z, as you know, seems simply to repeat doctrine about the priority of substance among the categories. It is a puzzle, I think, how this chapter relates to the rest of the book, and I think it is fair to say that most commentators today treat Z as if it begins with chapter 3. Burnyeat for instance, although he comments on ch. 1, says that the chapter merely directs us to what sort of being should be the principal object of the following investigation, namely, 'substantial being': "The chapter explains why a treatise on being should concentrate on primary being" (Map 12).

Yet Aquinas in De Ente et Essentia thinks that there is an important argument implicit in Aristotle's opening appeal to the categories. On Aquinas' reading, this argument sets the direction for the investigations of Z and, indeed, even helps to explain what I've been wondering about, viz. precisely on what grounds perceptible substances are given so much attention in Z.

I've whet your curiosity, I hope. In a subsequent post, I'll say what that argument is.

23 April 2006

In a Swamp, and Calling for Help

I've wandered out into the morass of Met. Z and find myself wondering how I got here. Looking back, I see the following. (What follows is a 'map into the swamp'. You might just wish to surf in fresher waters.)

[i] I was writing a review of a volume (OSAP Winter 2003) in which Dan Devereux's paper appeared on the relationship between Zeta and Eta. (That review should be coming out shortly in BMCR.) It was not that I was looking to find flaws in that paper. Devereux's paper involved some relatively close and interesting matters of interpretation, which it was not possible to discuss in the review; therefore, I took up some of them here.

[ii] I looked at two sorts of arguments that Devereux put forward, involving (i) redaction criticism and (ii) philosophical interpretation. I found Devereux's arguments interesting, as I said, and worthy of serious consideration--indeed, that is why I spent time on them. Yet it seemed to me in the end that all of these arguments were unpersuasive, and I gave my reasons why. (By the way, dear reader: I rely on you to point out where I am wrong. Do not forget that you would do me a favor by refuting me.)

[iii] Since Devereux had relied frequently upon Burnyeat as an authority, I looked briefly at Burnyeat's Map of Metaphysics Z. (I had read the Map several years ago, when it was a lengthy paper; I had not previously looked at it in book form.) I found that, as regards redaction-criticism, one of the unsound arguments that Devereux had drawn from Burnyeat (about en allois) had also been endorsed by Burnyeat. But I found that another argument (about the significance of the H1 summary) had effectively already been replied to by Burnyeat.

[iv] As regards philosophical interpretation, on the other hand, a difficulty in Burnyeat's general approach appeared to me, which I've tried to clarify in the last few posts. The problem has to do with why Aristotle spends the bulk of Z examining perceptible ('sensible') substances. Here is another way of putting the problem.

[v] As is well known, Aristotle describes what he is doing in the Metaphysics in two ways. He describes it as the study (i) of being qua being, and (ii) of imperceptible, separable substances. Now, as regards (i), it would seem misguided to restrict oneself to perceptible substances only, as this would be to 'cut off only a part of being' and study only that, rather than all of being. And, as regards (ii), clearly, the study of perceptible substances is not the study of imperceptible substances.

[vi] It is not clear to me that Burnyeat's interpretation can respond to this difficulty: Burnyeat says that Aristotle introduces the form-matter distinction in Z to reply to logicizing philosophers (as we may call them); but then we are left wondering how Aristotle's examination there of substances as having matter counts as metaphysics rather than physics. I used Burnyeat remarks on Lambda 1-5 to show that he does not succeed in explaining this.

[vii] The difficulty, in my view, becomes even worse, when one considers that Burnyeat argues in his Map that Zeta-Eta-Theta were written as a unit, that Lambda was tacked on later, and that it is, so to speak, extrinsic to those other books. It is unclear how, on this view, Aristotle in Z could be doing 'second philosophy for the purposes of first philosophy'. If there is no immediate and intended application of the results of Z to imperceptible substances, then how is it first philosophy?

[viii] I suppose Burnyeat could say that, on his interpretation, the conclusion 'form is substance' is exactly what is needed for the study of imperceptible substances--that is, Z establishes that substances can exist as pure forms, in establishing that form (alone) is substance. But then I wonder why Aristotle would take Platonizing or logicizing philosophers as his target in arguing for this sort of a conclusion. (Is that what they would find difficulty with?) Or, if you say that there is some special notion of form, not shared by them, that Aristotle wants to arrive at in Z, derived from the way that that notion gets used in the Physics--then, once again, the problem arises of how Z is different from physics. Also, it's not clear that there is any 'special notion of form' in Z except insofar as form is considered in its relation to matter--and yet then it appears to be a difficulty that Burnyeat wants to treat Z7-9, where form in its relation to matter is given thorough consideration, as an incongruent interpolation.

[ix] We could of course say that Aristotle is confused about his own distinction between first and second philosophy, or that the distinction doesn't amount to much in the end.

[x] And perhaps these are illusory problems, something that is hanging me up momentarily and that isn't very serious. My difficulties could easily be based on a misunderstanding. I present them simply as something troubling me for the moment.

[xi] Finally, when discussing Devereux on Z3, I proposed my own interpretation of the famous reductio there. I said that Aristotle was wishing there to draw a distinction between two senses of 'hypokeimenon'

(i) where the subject of predication is what is predicated of it (hypokeimenon 'as form' or 'as composite'); and
(ii) where the subject of predication is not what is predicated of it (hypokeimenon 'as matter');
and I said that the reductio of Z3 was meant to drive home the point that any intelligible use of (ii) presupposed a prior use of (i).

[xii] This suggestion (which, by the way, corresponds nicely to the back reference about the hypokeimenon, at Z 13.1038b2-6, and receives confirmation from that back reference), leads to an appealing simplification of the structure of Z, which is this.

[xiii] Of the four marks of substance which Aristotle presents at the beginning of Z3 (hypokeimenon, to ti en einai, katholou, genos), it is clear that Aristotle wishes to deal with the last two negatively, in order to reject them. What really needs to be explained, then, is Z3-11, which my distinction does:
Z3: distinction in senses of hypokeimenon
Z4-6: hypokeimenon as form (to ti en einai, called by Aristotle 'form')
Z7-9: hypokeimenon as composite and as proximate matter
(Z10-11: ancillary discussion of definition)
[xiv] My post pointing out that 'form' and 'matter' are not used by Aristotle in Z4-6 was meant to support this interpretation: 'form' is not used in Z4-6, I would suggest, because hypokeimenon 'as form' is precisely what Aristotle understands as the to ti en einai and is discussed through that phrase; and 'matter' is not used, because his discussion of hypokeimenon 'as (proximate) matter' and 'as composite' begins at Z7.

Now that I know where I am in this swamp, I'm radioing to be air-lifted out.

22 April 2006

The Metaphysical Study of Perceptible Substance

Here's the problem. Aristotle's official characterization of 'first philosophy' is the study of form as truly separable from matter. Yet ZHQ and also the first five chapters of L are about forms in matter. So how do these count as 'first philosophy' (metaphysics) rather than 'second philosophy' (physics)? One would like to say that there is some way of studying form-in-matter that counts as metaphysical.

But explaining this proves difficult to do.

One may see this in Burnyeat's discussion of L .1-5 (see Map, pp. 133-4). He rejects the view of Ross and others that these chapters are 'physics'. And yet he seems to give no clear reason why these are not physics, as the following analysis shows.

Here it is relevant to recall the evidence of Z11 that there is a first-philosophical way of studying sensible substantial being as well as a second-philosophical way.
That is exactly what is at issue. What is the difference?
The statement 'Sensible substantial being is liable to change' (L .1.1069b3) can be read as an invitation to study change. That is a topic of Aristotelian physics. But the statement can also be read as an invitation to study the type of substantial being we find in the sensible world-- changeable substantial being as opposed to the unchanging type that first philosophy aspires to understand...
But presumably, as Aristotelian scientists, we appropriately study change by studying the changes of substantial beings. Also, if we are not studying the 'unchanging type' of beings which are the proper object of first philosophy, then how are we studying first philosophy?
The way to read the early chapters of L is as a first-philosophical use of the factors invoked in Physics I to explain change (matter, form and privation).
Indeed. But what is this 'first-philosophical use'? This has not been explained.
These now reappear as the principles that explain the substantial being of sensible, hence changeable, substantial beings.
I don't understand. Is it that we should study these 'changeable substantial beings' as if they don't change? Or study them when they are not changing? What would that be? And in any case, Aristotle devotes much attention in the Metaphysics to how they come into and out of existence.
Just this is what happened when H1 resumed after the summary. That passage continues with a reference to the physical works (1042b8:e)n toi=j fusikoi=j), which confirms that H1 itself is not conceived as a contribution to physics.
We may grant the fact, but what is it for H1 not to be 'conceived as a contribution to physics'? (Also: Isn't it a fallacy to argue that any work that refers to the Physics isn't of a piece physics? Compare: "The Politics refers to the Ethics, which shows that it is not conceived as a contribution to ethics." Also: note too that here Burnyeat does grant that H1's discussion of perceptible substance resumes an earlier one.)
Similarly, when L 1 announces a comprehensive study of all three ranks of substantial being (sensible and perishable, sensible and eternal, unchanging and non-sensible), and says that because the first two involve change, they are the subject-matter of physics, this does not mean that are L 1-5 conceived as physics rather than first philosophy.
Granted, once again. Yet, once again, we still don't know what it means to study the objects of physics but not in the way that physics does.
L defined its task as that of finding the principles and causes of substantial beings (L 1.1069a18-20), where 'substantial beings' (we soon learn) covers all three ranks of substantial being enumerated at 1069a30-b2. If L starts from sensible substantial beings, this is because they are better known to us than the other two ranks of substantial being (cf. Z 3.1029b3-12).
To be sure; but, still, in what way is its 'starting from sensible substantial beings' distinct from physics?
The point of reminding us, here and at the beginning of L 6.1071b3, that the first two ranks of substantial being are studied by physics is that, since they are also studied by physics, the reader can be expected to come to first philosophy already familiar with matter and form and other factors in the Aristotelian analysis of change. Compare A3.983a33-b1, which says that the four causes have been sufficiently explained in the Physics; the knowledge we are assumed to have is now to be put to a new use.
Again, that's not controversial. But what is that new use? Once again, this is unexplained.
Readers versed in Aristotelian physics can be expected also to be familiar with the contrast between the perishable substantial beings of the sublunary world and the imperishable, eternally circling heavens. Crucially, they wil have the understanding of human psychology which is presupposed in the attempt of L 7 and 9 to extrapolate to the mind of God. First philosophy is first in order of understanding, but last in the order of learning.
But now we are on to 'first philosophy' in the proper sense, without its ever having been explained to us how L 1-5 did not count as physics.
To put it another way, the task of finding the principles and causes of all three ranks of substantial being has the universal scope assigned to first philosophy at the end of E4.1028b3-4: to find the principles and causes of being itself qua being.
But if we append, in front of a treatise, "What follows should be understood as having universal scope", have we thereby turned it into metaphysics? That's a bit too easy, I should think. Not that this would be the case anyway:
True, the rubric 'being qua being' is not found in L. But nor is it found in ZHQ. Instead, both Z1 and (more briefly) L 1 use the focal analysis of being to explain why the inquiry will concentrate on substantial being rather than the dependent categories. Because the dependent categories are dependent, it is substantial being that explains the rest. The study (to coin a phrase) of substantial being qua substantial being is the first part (cf. 1069a20: prw=ton me/roj) of the universal study of being qua being. It is this part that is begun, but not completed, in ZHQ, begun and completed in L. From beginning to end L is first philosophy.
I can't say that I'm convinced.

First and Second Philosophy

I raised the question in an earlier post: If Aristotle, in Met. Z, appeals to the form/matter distinction in order to reply to Platonist-type philosophers with a taste for the abstract, wouldn't we expect him to be using that distinction to draw attention to the importance of matter, not form? After all, form is what Platonists truck in.

Perhaps you gathered that, underlying my question, was a concern that a usual way of setting up Z --Aristotle is trying to figure out what substance is, and his answer is that it is form--was off the mark; that Aristotle isn't asking this question or answering it in that way; that we were missing what he was truly interested in.

Someone might say in response: "If his point in introducing the form/matter distinction was to draw attention to the matter, then his discussion in Z would be no more 'second philosophy', that is, physics. But it is supposed to be 'first philosophy'. And it's by drawing attention to the form that his discussion counts as first philosophy."

But I don't see how. If discussions appealing to matter count as second philosophy, why not also discussions appealing to form-in-matter? Or, if you say that it's the fact that a discussion of form-in-matter leads up to a discussion of separate substance, form on its own, which makes it 'first philosophy', then I can say, just as well, that it's the fact that a discussion of matter leads up to a consideration of form on its own which makes that 'first philosophy'.

In fact, this whole question of the distinction between 'first' and 'second' philosophy seems confused to me, as I'll explain in my next post.

21 April 2006

Perceptible Substances All Have Matter

"But now let us resume the discussion of the generally recognized substances. These are the perceptible substances, and perceptible substances all have matter."

nu~n de\ peri\ tw~n o(mologoume/nwn ou)siw~n e0pe/lqwmen. au{tai d' ei0si\n ai9 ai0sqhtai/: ai9 d' ai0sqhtai\ ou)si/ai pa~sai u3lhn e1xousin.:
Thus H 1 1042a24.

might mean 'let us start with' (as Burnyeat wishes to take it: Map 63). But it can alternatively mean 'return to' (cf. NE 1172b8), as Ross seems to understand it. Suppose it does and that we read this in conjuction with the methodological passage at the opening of Z 4 (transposed now typically to the end of Z3):
Some of the perceptible substances are generally admitted to be substances, so that we must look first among these. For it is an advantage to advance to that which is more knowable. For learning proceeds for all in this way-through that which is less knowable by nature to that which is more knowable; and just as in conduct our task is to start from what is good for each and make what is without qualification good good for each, so it is our task to start from what is more knowable to oneself and make what is knowable by nature knowable to oneself. Now what is knowable and primary for particular sets of people is often knowable to a very small extent, and has little or nothing of reality. But yet one must start from that which is barely knowable but knowable to oneself, and try to know what is knowable without qualification, passing, as has been said, by way of those very things which one does know.
Then it would be natural to take the above to be saying: we need to consider each of the marks we have mentioned (hypkeimenon, to ti en einai, katholou, genos), with a view, first, to how they work as regard perceptible substances. But these have matter. Thus we must consider how they work as regards things with matter.

That is, the above methodological passage gives Aristotle's reason for introducing the form-matter distinction. It answers the question: Why does Aristotle introduce the form-matter distinction in Z? And it does so consistently with what we should antecedently expect.

And then, as I remarked in an earlier post, chs. 7-9 do not appear to be out of place. They are the commensensical deployment of the form-matter distinction to perceptible substances insofar as they come about by art, nature, or spontaneously.

An Interesting Pattern

The following is a chart of the frequency of occurrence of two words in Met. Z 3-17. The blue numbers represent Z 4-6, and the red numbers represent Z 7-9. Can you guess what words these are? (I give the answer in the comments box.) Also, what lesson, if any, should we draw from this?

1029a 11 2
1029b 0 0
1030a 0 0
1030b 0 0
1031a 0 0
1031b 0 0
1032a 3 1
1032b 4 6
1033a 5 2
1033b 2 7
1034a 4 6
1034b 2 3
1035a 12 12
1035b 8 4
1036a 6 3
1036b 3 6
1037a 9 0
1037b 3 0
1038a 2 4
1038b 1 0
1039a 0 1
1039b 2 0
1040a 0 0
1040b 1 1
1041a 0 0
1041b 3 1

Substance Over Form

Suppose you are an Aristotelian. I am a philosopher with Platonist tendencies, and I approach you, all excited, that I have developed a new theory of pleasure and pain. "Pleasure and pain," I say, "are actually ratios of numbers. If the ratio is greater than 1, then it is a pleasure; but if it is less than 1, then that is a pain."

As an Aristotelian, in considering this theory, you might very well object that it was deficient in very basic way, such that it could not possibly be a correct account, yet, of pleasure and pain in living animals. "Your theory, if correct," you might say, "would be an account only of the 'form' (eidos) or definition (logos) of pleasure and pain. But, if you wish to give an account of these as they occur in living creatures, you must also explain their 'matter' (hule). That, indeed, is what distinguishes a natural philosopher from a mere dialectician or logician--attention to the matter. After you propose also an account of the matter, then we might begin to evaluate the correctness of what you say."

I tell this simple tale (adapted of course from De Anima 1.1 403b) because, to my mind, it establishes an 'antecedent probability' for the interpretation of Metaphysics Z, with which, it seems to me, Myles Burnyeat's famous account is prima facie at odds.

Recall that Burnyeat holds that Z consists of four 'non-linear' and non-cumulative discussions, each of which examines one of the marks for substance set down at the beginning of Z 3 (hypokeimenon, to ti en einai, katholou, genos). According to Burnyeat, each of Z's four discussions is carried out on 'two levels' , starting with one of these marks considered as proposed in a merely 'logical' way (the first level), and then moving on to a consideration of the mark in terms of the form-matter distinction (the second level).

It is a familiar enough view that, in the Metaphysics, Aristotle goes beyond concepts he had used in the Organon by deploying the (supposedly) later form-matter distinction developed in the Physics and in his other writings on natural science. What is distinctive about Burnyeat's version of this thesis, I take it, is that Burnyeat understands Z as, so to speak, polemical rather than merely developmental: he proposes that its discussions are pointedly directed at Platonists, or philosophers with a similar outlook, who consider that one should do metaphysics relying solely on notions in logic or mathematics. (Burnyeat quotes Philip Merlan to bring home this point: "Whenever we use the word 'metaphysics', we should hear its Platonic or anti-Platonic overtones: metaphysics and not metamathematics", 81n6.)

What is especially appealing, I think, about Burnyeat's understanding Aristotle's target in this way, is that it seems true to life. In our own time, for instance, we are familiar with major philosophers, metaphysicians, who similarly believe that the right way to approach questions of ultimate reality is to ask what the world must be like for logic and mathematics to be possible.

But to my mind there is a basic difficulty in Burnyeat's interpretation, which becomes clear if we pose the question: Precisely why does Aristotle use the form-matter distinction against these Platonists and 'logical' metaphysicians? Why reply to them in that way, rather than in some other way?

I don't see that Burnyeat gives a satisfactory answer to this. He speaks as if the form-matter distinction is simply a kind of philosophical wringer which all theories must pass through (while admitting that Aristotle does not deploy the distinction in major treatises such as the Ethics or Politics).

[The procedural peculiarities of Z] are Aristotle's solution to the problem of how to encourage his readers to start at the right place, with reflection on form and matter as the first principles of physics. (81)
But why is this the 'right place'? Burnyeat doesn't say. Or, if he does give an answer, it has to do with the result which, according to Burnyeat, Aristotle wishes to draw from his deployment of the form-matter distinction, namely, that 'substance is form':
From the perspective of physics, substantial form is nature as the internal principle of change and stability characteristic of natural things. The very last sentence of Z brings us back to substantial form in nature... (7)
That is to say, if Burnyeat gives an answer to the question of why Aristotle introduces the form-matter distinction in Z, it is that Aristotle wishes to force the conclusion that substance is to be identified with his notion of form.

But the point I wish to make--to go back to the tale I told at the beginning of this post--is that the antecedent probabilities work in another direction. It is antecedently likely, I consider, that, in a discussion between Aristotle and a merely 'logical' philosopher, if Aristotle introduced the form-matter distinction, this would be, rather, to make a point about matter, not form. Isn't it precisely what is wrong with a merely 'logical' account, that it considers the form, and only that? Antecedently, we'd expect that Aristotle's lesson, in Z, would be the opposite of what Burnyeat proposes: that matter must be considered, too, in any correct account of substance.

19 April 2006

The Significance of the H 1 Summary

This may be my last post on Devereux's redaction criticism; perhaps enough is enough.

Yesterday I drew attention once again to Devereux's argument that Z 3 is a later addition to Z, because the summary in H 1 does not mention or recount the dialectical discussion of Z 3. Yet this argument does not hold up under scrutiny, as even a brief examination of the relevant passage from H 1 will show.

That passage begins by stating a theme of the Metaphysics which recurs repeatedly from almost the beginning of that treatise. Moral: the summary is not meant to be simply a review of Z:

We have said that the causes, principles, and elements of substances are the object of our search.
It then gives a large amount of space to mentioning the 'commonly agreed upon substances'. These were not part of the argument of Z. Moral: there is a disproportion between the summary, and the content of Z; it would be unsound to argue from the size or significance of a mention in the summary, to the size or significance of the corresponding treatment in Z:
And some substances are recognized by every one, but some have been advocated by particular schools. Those generally recognized are the natural substances, i.e. fire, earth, water, air, &c., the simple bodies; second plants and their parts, and animals and the parts of animals; and finally the physical universe and its parts; while some particular schools say that Forms and the objects of mathematics are substances.
Then it turns to the considerations which occupy the bulk of Z. It mentions 'essence' (to ti en einai). Aristotle's discussion of this in Z occupies chapters 4-6. This summary simply mentions that essence is regarded as a candidate for substance; it does not summarize any argument from those chapters or even say what conclusion was reached by those arguments. Moral: we have no reason to think that the similar mention here of 'substratum' (hypokeimenon) does not stand, likewise, for a lengthy discussion in Z:
But there are arguments which lead to the conclusion that there are other substances, the essence (to ti en einai) and the substratum (hypokeimenon).
Next we have simply a mention of the two other criteria for substance (genus, species), and an explanation of why other things (Ideas, definition, formula) were dealt with as well, as related to these. Moral: no arguments of Z are presented in this summary.
Again, in another way the genus seems more substantial than the various species, and the universal than the particulars. And with the universal and the genus the Ideas are connected; it is in virtue of the same argument that they are thought to be substances. And since the essence is substance, and the definition is a formula of the essence, for this reason we have discussed definition and essential predication. Since the definition is a formula, and a formula has parts, we had to consider also with respect to the notion of 'part', what are parts of the substance and what are not, and whether the parts of the substance are also parts of the definition.
Next we do have a definite conclusion being mentioned, but this is only in order to defer until later (books M and N) a particular discussion:
Further, too, neither the universal nor the genus is a substance; we must inquire later into the Ideas and the objects of mathematics; for some say these are substances as well as the sensible substances.
Thus, there is simply no basis in this summary for concluding that Z 3 was not originally part of Z.

But this point was made already by Burnyeat in his Map of Metaphysics Z. In the chapter on 'Signposts', after reviewing the correspondences between the summary (and drawing lessons from it, although not quite the same lessons as those I draw above), Burnyeat at first observes:
There are, however, two apparent omissions (apart from Z7-9 and Z12) that call for explanation. The discussion of the subject in Z3 is not separately recorded, and the summary does not mention Z17.(63)
But then he later argues, clearly rightly:
If Z3 is cited without express mention of its rather brief discussion of substantial being as subject, whey complain? The summary reminds us that subject [hypokeimenon] was discussed, but the only thing more to say would be what conclusion was reached, and positive conclusions are something the summary avoids.(65)
Strange that Devereux in his article should frequently cite Burnyeat's Map as an authority and yet insist, against Burnyeat's commonsense point, and without giving an argument in reply, that there is a significant omission here.

18 April 2006

Puzzlement for Me

I'm very puzzled now. Recall Devereux's argument in redaction criticism, which I posted here:

Scholars have argued...that chapters 7-9 and chapter 12 are 'later insertions' in Z since (i) they are intrusive in their context and (ii) there is no mention of them in the summary at the beginning of H. Further, in H 3 Aristotle says that it was shown elsewhere--not in Z or H--that forms are neither generated nor produced (1043b16-18). But Z 8 contains an argument for this thesis, so it seems that this chapter was not a part of Z when the reference in H 3 was written. These details suggest that there was an early, 'first' version of Z which is summarized at the beginning of H, and a later expanded version containing the 'inserted' chapters, Z 7-9 and Z 12.
This argument largely rests on the presupposition that in H 3, when Aristotle wrote the phrase, "in other discussions", e)n a)/lloij, then (if he really did write this) he could not have meant some other book in the Metaphysics: according to Devereux, he could not, from within H, have referred back to something already existing in Z by using that phrase. I pointed out that this argument was almost worthless. But I didn't do the obvious thing then, namely, check for other uses of e)n a)/lloij in the corpus. I argued on the basis of antecedent probabilities and analogies.

Well, I checked just now, and it turns out there are at least two other examples where Aristotle uses that phrase in one book of the Metaphysics to refer to something in another book. At 1017b9, in 5.8, that phrase is used to refer to 9.7 (o&te de\ dunato_n kai\ po&te ou1pw, e0n a1lloij dioriste/on))) and at 1046a5, in 9.1, that phrase is used to refer to 5.12 (o3ti me\n ou}n le/getaipollaxw~j h( du&namij kai\ to_ du&nasqai, diw&ristai h(mi=n e0n a1lloij)

So Devereux's argument is entirely worthless, isn't it?

What's bizarre is that Myles Burnyeat endorses the argument also in his Map of Metaphysics Z. In a section entitled, "Various reasons for thinking Z 7-9 a later insertion" he writes:
[iii] H3.1043b16, refers to Z8 as 'in another discussion' (e0n a1lloij). This proves the one-time independence of Z7-9 provided H is continuous with Z, which there is anyway good reason to suppose.
'Proves' is a strong word. I wonder if the appearance of proof comes from Burnyeat's argument having the character of tautology: if H is continuous with Z, then of course a reference to part of Z as if it were not continuous would show that that part was not within Z. That's true 'by definition', given an appropriately strong meaning of 'continuous'. On the other hand, one might just as well say that the H3 reference shows that Aristotle never regarded that discussion (or indeed any discussions in the Metaphysics), as 'continuous' in that very strong sense.