Substance and life in Aristotle.

Shields, Christopher

I The Dignity of Substance

W. D. Ross speaks easily of Aristotle as excluding artefacts from attaining 'the dignity of substance'. (1) His doing so may seem fair enough, since no artefact is a living being and Aristotle twice declares, or seems to declare, that only living entities qualify as substances: (Metaph 1041b28-31, 1043b19-23; cf. Metaph 1042a7-8, 1043a5-6, 1070a5-20; Ph 192b32-4). Yet the doctrine Ross ascribes to Aristotle is not an easy one to defend; nor is it even, as expressed, a doctrine readily understood. For in speaking this way, Ross evidently treats the category of substance as somehow normative, or at least as scalar, and one would have thought that the ontological categories of being discriminated by Aristotle were fixed and invariant, determined by exhaustive and exclusive binary metaphysical features: something either is or is not a substance, and whatever is not a substance is something else, say a quantity or a quality.

We may bring the problem of Ross' characterization into sharper relief by reflecting on Aristotle's entrenched practice of introducing both artefacts and organisms as examples of substantiality: he regularly speaks of statues and houses as substances in the Metaphysics (e.g., 1043a29-33, 1070a5). It offers no immediate assistance to suggest that he uses these as mere illustrations, since we would only then want to know why Aristotle should introduce non-substances as illustrations of substantiality. After all, he would hardly offer two cubits as an illustration of the category of location. Moreover, if we think that artifacts are not after all substances, then what are they? Being two cubits, while not a location, clearly has a home in Aristotle's set of categories; it belongs in the category of quantity. By contrast, surely no table is a quantity, and no chair is a quality or a place. Unfortunately, artefacts seem to have no home in the theory of categories if not in the category of substance.

More importantly, artefacts certainly seem to have precisely the claim on substantiality that organisms have. At least according to the standard tests of the Categories, for example, artefacts ought to qualify as primary substances: they, no less than individual organisms, are neither said-of nor in (Cat 1a20-2b6) and they can receive contraries while remaining numerically one and the same (Cat 4b10-21). Hence, as far as these tests go, it would seem utterly arbitrary to deny substantiality to artefacts while awarding it to organisms. (2) If we consider them from the standpoint of the Categories, then, living beings are substances if and only if artefacts are. (3)

Our discomfort should extend further. We may believe that by the time of his writing the comparatively mature Metaphysics VII and VIII, Aristotle had abandoned both the tests for substantiality of the Categories and the examples of substance they yield. We may, indeed, regard the hylomorphic account of substance as positively incompatible with the theory of the Categories. (4) In that case, we would not be concerned about the fact that the tests of the Categories gave us the wrong results; this is, rather, precisely the sort of thing we should expect. Still, much the same worry will develop regarding the theory internal to the Metaphysics. Indeed, there is a perfectly general worry about the philosophical arbitrariness of distinguishing artefacts from substances which will inevitably emerge even within the context of hylomorphism: surely, one may suppose, any defensible criteria for substantiality satisfied by organisms will again and equally be satisfied by artefacts. For, intuitively, if we expect substances to be the sorts of entities which, unlike events, do not occur, and which, unlike properties, cannot be instantiated by other things, then organisms will be substances if and only if artefacts are.

Against these initially reasonable reservations, I argue that Ross' observation about the substantiality of organisms is exactly right: Aristotle does regard living beings as especially suited to be substances. I argue further that at least within the hylomorphic framework of Metaphysics VII-IX, Aristotle has good reason for introducing this restriction. For in that framework living beings do satisfy the criteria of substantiality in a way in which artefacts do not. To the degree that these criteria are independently defensible, then, the right inference to draw is not that these criteria must be altered to admit artefacts into the fold, but rather, along with Aristotle, that living beings alone are offered the dignity of substance.

Commentators have been reluctant to adopt this exclusive attitude, thinking it too radical or otherwise inconsistent with others among Aristotle's metaphysical commitments. These commentators have accordingly sought a deflationary sort of interpretation, primarily by taking the category of substance as scalar, with the result that artefacts qualify as substances, but to a lesser degree than organisms. (5) The more radical interpretation, which positively excludes artefacts from the category of substance, endorsed as we have seen by Ross and extending back at least as far as Aquinas, (6) treats the criteria for substance as binary rather than scalar, with the result that artefacts simply do not qualify. (7) One may distinguish these approaches most fertilely by treating them as variations on a single argumentative pattern. The first yields the exclusive claim (EC):

(1) Necessarily, x is a substance if and only if x is a unified entity capable of existing diachronically as a separate and determinate F.

(2) Only living things are unified entities capable of existing diachronically as separate and determinate Fs.

(3) Hence, only living things are substances.

Those who find (EC-3) understandable but too extreme suggest a way of softening (EC-1) and (EC-2) in a partly conciliatory way. While insisting that although there are non-living substances, these commentators allow that living beings alone are paradigmatically substances. This weakened premise set supports only the paradigmatic claim (PC):

(1) Necessarily, x is paradigmatically a substance if and only if x is a unified entity capable of existing diachronically as a separate and determinate F to the fullest degree.

(2) Only living things are unified entities capable of existing diachronically as separate and determinate Fs to the fullest degree.

(3) Hence, only living things are paradigmatically substances.

(PC-2) is relatively moderate in comparison with (EC-2); for clearly the exclusive claim, and not the paradigmatic claim, results in the striking and disquieting conclusion that only living things are substances. The paradigmatic claim, by contrast, admits non-living things as substances, mainly by allowing degrees of substantiality, or at least degrees of satisfying Aristotle's preferred criteria for substantiality. (8) If this were Aristotle's final position, his view would be much less problematic, though it would still call for explanation and justification. In particular, if he accepts only the paradigmatic claim, it will be necessary for Aristotle to show why non-living things, as a class, are worse placed to satisfy any reasonable criteria for substantiality he might advance. As we have seen, certainly in terms of the criteria adumbrated in the Categories, substances and artefacts stand on par with one another.

The available evidence tells strongly in favor of Aristotle's accepting the more extreme (EC-1) and (EC-2), and so of his endorsing the more radical exclusive claim (EC). The evidence for this conclusion is sufficiently strong that it counsels reflecting on the problems which seem to follow in its wake. I will argue that these problems are not so formidable that they require reconsideration of Aristotle's general theory of substance, as advanced in Metaphysics VII-VIII. On the contrary, coming to appreciate why living things alone qualify as substances augments our understanding of Aristotle's theory in general, and of the demand for the determinate and separate diachronic unity of substance in particular. (9)

II Two Routes to the Exclusive Claim

That Aristotle in fact maintains (EC) can be demonstrated in two ways, one brief and direct, the other comparatively long and complex. I present both, because the direct route, though perfectly clear in its purport, leaves the most critical claims of Aristotle's view unexplicated and undefended. The more complex route, by contrast, places those claims in a context within which appraisal is possible.

First, then, to the direct route. Aristotle begins his richest and most nuanced discussion of substantiality, Metaphysics VII, by posing a series of questions for investigation. Included among them is an extensionally oriented question canvassing what his endoxic method reveals as the primary candidates for substantiality:

 Are only animals and plants and their parts substances? Are natural
 bodies (specified as fire and water and earth and anything of this
 kind) and their parts substances? Are the things constituted by
 natural bodies (the cosmos and its parts, the sun, the moon and
 stars) substances? Should we say that (i) some of these and not
 others, or (ii) some of these and some things not mentioned, or
 (iii) none of these are substances? (Metaph 1028b8-15) (10)

Aristotle answers this question near the end of his investigation in Metaphysics VII 16, which opens with a straightforward denial of the substantiality of the elements:

 It is also clear, then, that among things which seem to be
 substances most are potentialities, including (i) the parts of
 animals (for none of them is something separated, and whenever they
 are separated, they all exist as matter), and also (ii) earth,
 fire, and air. For none of these is one; they are, rather, like a
 heap, (11) until they are fused and something one comes to be from
 them. (Metaph 1040b5-10)

Together, these passages constitute a simple disjunctive syllogism, provided that we are prepared to accept the candidates listed as mentioning all of the alternatives Aristotle himself regards as viable:

(1) If x is a substance, then x is either: (i) an animal, (ii) a plant, (iii) a part of animal or plant; (iv) a simple element; or (v) something constituted by the natural bodies (here specified as the celestial bodies).

(2) No part of an animal or plant is a substance; and no simple element is a substance.

(3) Hence, if x is a substance, x is either (i) an animal, (ii) a plant, (iii) or a celestial body.

This conclusion restricts substances to living things, since for Aristotle celestial bodies qualify as living. (12)

Bracketing that peculiar feature of his cosmology, we can investigate why Aristotle supposes that animals and plants qualify as substances, while the elements do not. For surely his reason for denying substantiality to the elements seems peculiar: he precludes them on the ground that they are not sufficiently unified, treating them instead as mere heaps. This is odd, since one might reasonably have taken the elements to be not only unified, but as distinctly and markedly unified: a heap is a pile of stuff of indeterminate shape, size and composition, whereas water is everywhere and always water. Certainly from a compositional point of view, water ought to qualify as unified if anything ever does?

Seeing why the elements fail to qualify as substantial requires us to reflect more deeply on the features of living things which make them uniquely suited to be substances, and so also on the second, more complex route to (EC). The principal reservation Aristotle has about elements turns out to apply to all non-living entities:

 The substance of each thing, then, is this <the form>, for this is
 the primary cause of <each thing's> being. Since among things, not
 all are substances, while as many as are substances <are> in
 accordance with nature, i.e., are constituted by nature, (13) it
 would be this nature -which is not an element, but a
 principle--that is substance. (Metaph 1041b27-31) (14)

This confirms our earlier finding because of its promoting as successful candidates for substance living things only, thereby setting aside not only all of the other candidates considered in Metaphysics VII 2, but also some others which were not considered, except incidentally and in passing as perceptibles (Metaph 1028b29, 31), namely artefacts and all structured but non-living entities.

This may not seem directly obvious. In particular, it may be doubted whether the phrase 'in accordance with nature, i.e., are constituted by nature' is intended to exclude the four elements, earth, air, fire, and water. When restricting substances in this way, Aristotle evidently means to appeal to a distinction developed most fully elsewhere, most notably in Physics II and VIII. There Aristotle begins by marking a broad distinction relevant to his attitudes toward substance at the end of Metaphysics VII 17:

 Among things which exist, some are by nature, and others because of
 other causes. [Existing] by nature are animals and plants and the
 simple things among bodies, for example earth, fire, air and water;
 for we say that these things and these sorts of things are by
 nature; and all these things seem to differ from those things which
 are not constituted by nature. For each of these has within itself
 a principle of motion and rest, in some cases in respect of place,
 in others of growth and decay, and still others of alteration. (Ph
 192b8-15)

Aristotle immediately contrasts things which exist by nature with artefacts of various sorts, again in terms of their having or lacking internal principles of change :

 But a couch or a cloak, or any thing of this sort, insofar as it
 receives each such predicate and in as much as it is a product of
 craft, has no innate impulse to change; (15) whereas insofar as it
 is co-incidentally <made of> what is stone or earthen or of
 mixtures of these, it has <an impulse to change>, but only to that
 extent. This is due to nature's being a kind of principle and cause
 of being moved or being stable in that to which it belongs
 primarily, in itself, and not coincidentally. (Ph 193a16-23)

It is noteworthy that Aristotle moves here almost directly to much the same conclusion drawn at the end of Metaphysics VII 17, namely that only things with natures qualify as substances: 'whatever has this sort of principle has a nature; and all of these are substance' (Ph 192b32-3).

This discussion from Physics II 1 is immediately relevant to the restriction made at the end of Metaphysics VII 17 in two ways. In the first instance, it expands upon Aristotle's brief contrast between things which are and are not constituted by nature by grounding the distinction in the presence or absence of an innate principle of change and rest. (16) Things formed by nature have such principles while things which lack them do not. At the same time, and this is the second form of relevance, Aristotle makes clear in Physics II 1 that the distinction between having and lacking internal principles of change is partly co-ordinate with the distinction between living and non-living entities. It is only partly co-ordinate, since so far the elements are regarded as having an internal principle of change, (17) even though no one so far as I am aware thinks that this commits Aristotle to their being alive.

Later he clarifies the threshold of what it takes for something to have the sort of internal arche sufficient not merely for existing by nature, but additionally for being alive. In Physics VIII, Aristotle insists that things with natures have internal principles not only of motion, but also of rest or stability which evince some manner of controlled directionality (Ph 192b13-15; cf. Metaph 1015a13-15, 1070a6-9; DA 415b28-16a18).

His doing so goes some way towards explaining the connection he sees between life and substantiality. The additional point about controlled directionality may appear at first inconsequential, but this appearance is misleading. For it is precisely by this addition that he connects life with teleology--and thereby substance with life.

This point emerges in an interesting passage in Physics VIII 4, in which Aristotle contrasts living and lifeless things in terms of their agency both in initiating and stopping their movements. He considers a puzzle in Physics VIII 4 concerning what one ought to say about lifeless things which are moved, but not moved contrary to their natures. He maintains that a lifeless thing moved contrary to its nature is normally moved by force (bia(i)). When they are moved in accordance with their own natures (phusei, contrasted with para phusin, Ph 254b21-2), the agency by which they are moved is no longer clear. What is clear is this:

 It is impossible to say that these things [light and heavy things
 moving by or with nature, phusei, 255a4, as opposed to by force,
 bia(i), 255a3] are moved by themselves. For this is a feature of
 life (zotikon), and it is peculiar to what is ensouled. <If they
 could be moved by themselves,> then they would also be able to stop
 themselves. I mean, for example, that if something is the cause of
 its own walking, then it is <also the cause> of its not walking.
 Hence, if fire's being borne upward were up to it, it is clear that
 it would also be up to it <to be borne> downward. It is
 unreasonable that being moved in one way alone should be up to
 these things, if indeed they move themselves. (Ph 255a5-11)

Living things, those with natures, can initiate motion in themselves, and they can bring that motion to a halt. Although some non-living things can move along with nature, so that they are not always moved by force, it would be impossible for them to initiate this movement in the way that living systems do. For they could initiate their motion only if they could stop it and reverse it. But they cannot.

Aristotle's puzzle about the agency of lifeless things which are not always moved by force reflects a conception of the sort of internal arche of motion required for having a nature of a sort sufficient for being alive and so for being substantial. Such an arche must be, so to speak, cybernetic: the sort of arche distinctive of life (zotikon) is the sort which equips that entity of which it is a source to control the directionality of its movement and to halt the movement altogether when it is unwelcome. Consequently, the sort of capacity for motion and rest requisite for having a nature must be broadly cybernetic, in just this sense. It follows, then, that mere directionality, the sort we find in the elements, does not suffice for life. Also required is an ability to control that directionality along some dimension, presumably in the pursuit the end of which is the good of the agent in question. In these ways, Aristotle sees additional connections between life and substantiality: having a nature is sufficient for being a substance, but this is at least in part because something has a nature only if it has a capacity for self-directing motion. Self-directing, or self-regulating, systems are, however, already close to being living systems. (18)

III Life and Directionality

We have seen that Aristotle distinguishes between substances and non-substances in part by excluding non-living entities from the class of substances on the grounds that they are insufficiently cybernetic. This point can be strengthened and extended if we reflect on the way in which the soul, as the principle (arche) of life in a living entity is a cause of its systemic directionality.

Aristotle is explicit about the fact that the soul is both the cause and principle of a body which is alive, though he does caution that these notions are spoken of in many ways:

 The soul is the cause (aition) and source (arche) of the living
 body. But these [cause and source] are spoken of in many ways.
 Similarly, soul is a cause according to the ways delineated, which
 are three: it is a cause <as> the principle of motion, <as> that
 for the sake of which, and as the substance of ensouled bodies.
 That it is a cause as substance is clear, for substance is the
 cause of being for all things, and for living things being is life,
 and the soul is also the cause and principle of life. (DA 415b8-14;
 cf. PA 647b12-25; Ph 255a6-10)

In this passage Aristotle explains his contention that a soul is a source of life in different ways for different kinds of entities by means of a simple argument.

The claims are direct: (i) substance is the cause and principle of being for all things; (ii) for living things being is life; (iii) the soul is the cause and principle of life. The arrestingly simple identification advanced in (ii), the claim that 'for living things being is life' (to de zen tois zosi to einai estin; DA 415b13), should allow one to conclude that the soul is substance, so long as the analytic-sounding claim in (iii) is also granted. This identification evidently implies that every living entity is essentially (but not therefore necessarily) alive. Minimally, then, all living things have natures whose activities result in patterns of behavior which are deeply characteristic of them as substantial beings.

Strikingly, we cannot point to some one set of activities characteristic of all living things as such; and this will add an instructive complication into Aristotle's account. For Aristotle maintains, as against Dionysius and perhaps Plato, that living too is spoken of in many ways:

 This occurred also with Dionysius' definition of life, if indeed
 this is 'movement of a creature which is nourished, when it [viz.
 this movement] is naturally present within it.' This belongs to
 animals no more than to plants. But life seems not to be spoken of
 according to one form, belonging instead in one way to animals and
 in another way to plants. (Top 148a26-31)

This observation encourages him to introduce his own positive account of life in the de Anima by reasserting the claim that there is no one single property, being alive, which all and only living systems have:

 Let us say, then, in taking up a new starting point for our
 inquiry, that what is ensouled is distinguished from what is
 unensouled by living. But living is spoken of in many ways, and if
 even one of these belongs to something, we say that it is alive,
 that is: thought; perception; motion and rest with respect to
 place; and further motion with respect to nourishment, decay and
 growth. (DA 413a20-5)

So, different kinds of living things will have different natures, and so different causes of being. The substance of various different kinds of living things will therefore also differ.

Now, it is entirely likely that Aristotle believes that life is a kind of core-dependent homonym. (19) This in turn divides into two claims. First, life is non-univocal, and so no synonymous account of it can be given; second, even so, these various accounts will be related around some single principle, which organizes them in an explanatorily co-ordinated way. So, we can infer two correspondingly different conclusions from his contention that life is homonymous. The first, which follows from life's non-univocity, is that as a cause of being, the soul as substance is sortally-determinate. By this I mean that substance is not merely the cause of being in the sense of being responsible for the existence of a thing, but also the cause of its being the F kind of thing it is. The second, which follows from life's core-dependence, is that as a cause of being, the soul will also be, in a more general way, responsible for the existence of that thing as a unified being of any kind at all. The core of life will be a determinable form of end-directedness, where this end-directedness will simultaneously account for a substance's existing as living and a living being's existing as a substance. I consider each of these results in turn.

First, then, the soul as substance will be sortally-determinate. If the soul is the substance of a living being because of its being the cause of that thing's life, where life will consist in different activities for different strata of living beings, it will follow that the soul is the cause of different kinds of beings. For instance, the soul as substance in the case of humans will essentially consist in its being a cause of rational end-directed activities; in the case of animals of sensitive end-directed activities; and in the case of plants of bare nutritive end-directed activities. This suggests that since the substance of each living being will differ from one kind to the next, the soul will be the cause of more than mere being. Rather, the substance of a human being, as opposed to a non-rational animal, will be a cause of more than existence: it will be a cause of the existence of humans as humans, with all the end-directed rational activity that human beings exhibit. It follows further then, that substance cannot be anything other than sortal-determinate: every substance will be the cause of a determinate kind of being, with the result that different kinds of substances will themselves necessarily be qualitatively distinct. Otherwise, we would lack explanatory causes of the differences manifested by different forms of living systems. This implies, finally, that Aristotelian substance is already deeply enmeshed in the program of explaining the unified characteristics of organized entities as particular sorts, and not merely the program of explaining what it is for anything at all, indifferently, to exist.

Still, and this is the second point we infer from life's core-dependent homonymy, all living things will bear an explanatorily rich connection to some core features of living beings as such. Fairly clearly, Aristotle expects every living thing to be an end-directed system of some form or other, and also for all natural living things to engage in behavior which typically eventuates in reproduction and the perpetuation of the species (DA 415a25; cf. GA 730b34, Pol 1252a26). So, in some sense, every living thing will be a kind of intentional system, a system whose activities are orchestrated around and subservient to the attainment of some end-state which is its good. Hence, although sortally-determinate, the soul as substance of living things will also provide internal forms of organization explanatory of the characteristic activities of the living being in question.

This implication suggests a second correlative deep connection between substance and life in Aristotle. Every substance is a unified being with a nature of its own, one not brought about by the external workings of some craftsman. It is precisely because some entities, living ones, are unified in the demanding sense that they have identifiable intrinsic ends which structure their activities as being the living entities they are, that they qualify as substances in the first instance.

This now helps us to explain something I had noted as perplexing in Metaphysics VII 16. There Aristotle had ruled against the substantiality of the elements on the grounds that 'none of these is one; they are, rather, like a heap, until they are fused and something one comes to be from them' (Metaph 1040b5-10). What seemed odd was that the elements were barred for want of being unified. Surely, one wanted to object, they are not only unified, but especially, even paradigmatically unified, especially when compared with diachronic substances like Socrates, who after all sustains all manner of material replenishment and who is in any case rarely if ever qualitatively identical across an earlier and later time. Water, by contrast, is everywhere and always water.

It now appears that the sort of unity denied to the elements is enriched, because of the connection to life. Lives are among other things internally directed series of activities unified by a single identifiable end. Water is not like that. It flows wherever it may flow unimpeded. Nor is fire like that. (20) Fire spreads, but not in any recognizable pattern. It too, so to speak, flows towards the combustible. This is why Aristotle denies that fire is alive in the de Anima (413a14-16). Although it can, as Aristotle says, grab hold of the combustible (GC 322a10-13), it does not make use of what it grasps for any determinate end which is a good for it. In this way too, then, substances are irreducibly teleological systems with specifiable intrinsic goods. (21) All teleological systems with specifiable intrinsic ends are, however, alive.

Finally, then, this is what the elements have in common with heaps. After denying that elements have the unity he expects of substances, Aristotle compares them with heaps, noting they will remain heaps until fused in the right sort of way. (22) Now it should be clear that the contrast with substances is grounded in their failing to be cybernetic: the elements are like heaps in that they lack within themselves their own principles of robust or cybernetic directionality (Metaph 1052a24-5). A heap does not activate the attainment of its own form; it thus lacks an intrinsic principle of unity, with the result that it can be added to and subtracted from haphazardly and indefinitely. Living systems, by contrast, have limited and patterned growth in two dimensions. This, at any rate, seems to be the immediate purport of Aristotle's rejection of Empedocles, whom he represents as maintaining that growth results from the natural tendencies of the elements taken by themselves (DA 415b3016a8). According to Aristotle, with only these resources patterned directionality would be inexplicable: 'Fire's growth carries on without limit, so long as there is something combustible. By contrast, for all things naturally constituted, there is a limit and a formula of both size and growth. These things belong to the soul, and not to fire, i.e., to the formula rather than to the matter' (DA 416a15-18).

Taken together, these remarks suggest that in order to qualify as a substance some being must have a principle of unity intrinsic to it in the sense of its being a teleolonomic system. Its being such will, however, involve its being an individual life. Consequently, Aristotle's demand for unity in substance helps further explain why he excludes non-living beings from the category of substance: they lack the forms of cybernetic directionality which living systems, as such, have essentially.

IV Artefacts and Organisms: an Asymmetry

If these points so far are correct, we may think of Aristotle as in an awkward position. He looks to things formed by nature, with rich and determinate internal sources of change and directionality to be substances, and he finds only living things. Now the result would seem to be that artefacts are not substances at all. If so, he embraces the Exclusive Claim (EC), rather that the Paradigmatic Claim (PC). He maintains, that is, that only living things are substances. We started with the thought that it would be awkward for Aristotle to embrace any such claim. After all, a chair is not a quantity and a table is not a quality. So what are they?

Aristotle will argue in response that this question has a false presupposition, namely that artefacts are anything determinate. Instead, he implies, rightly in my view, that they do not exist as determinate Fs at all. So he will not worry about whether artefacts fit into some category or other, since only things which exist determinately require consideration for categorial membership.

If so, the consequence of excluding artefacts from the dignity of substance may appear all the more dire. For if we deny them this honor, we deny them the dignity of existing determinately at all. And surely this seems counterintuitive. For I am sitting on a chair, something at least reasonably determinate and clearly not nothing at all, and I am typing on a computer, something which not only exists but exists as a highly complex hylomorphic structure. How can Aristotle deny its determinate existence?

The answer lies in his response to the paradoxes of generation, distinctness and growth alluded to in Metaphysics III 5 and developed further in the Physics and especially the Generation and Corruption. Artefacts have unstable identity conditions, whereas living things do not. If so, we have as much reason to deny the existence of artefacts as we do the existence of anything whose identity conditions cannot be stated in paradox-free ways--and that is plenty. It turns out that Aristotle in the Metaphysics is not undertaking to precisify some sort of vague folk ontology; instead, he is conducting first philosophy, a branch of theoretical science whose deliverances may amaze us no less than the claims of the physicists.

V Getting Bigger and Growing: an Academic Conundrum

If Aristotle in not constrained by some common-sensical notions of folk substantiality (if there are such), he nevertheless owes an argument capable of justifying his advancing the exclusive claim (EC). So far, we have been given reason to suppose that Aristotle thinks of substances as having an internal arche, in the strong, cybernetic sense, and that only living beings make that grade.

Still, one might well agree that only genuine unities qualify as substances, and might further allow that some features of Aristotle's teleology provide good reason to include some sorts of entities as unified in the way required of substances and while excluding others without thereby consenting to draw the line as Aristotle does, between the living and the non-living. So, for example, one might easily agree that Socrates is a substance whereas the entity cobbled together out the sole of my left shoe and the swatch of carpet upon which it currently rests is not, without thereby allowing that artefacts as a class fail to be substances. One might agree that the jerrymandered thing--let us call it a soleswatch--has no intrinsic unity, lacks an internal end, and in short, in Aristotle's terms, that it is without a nature, even while insisting that artefacts are not like that at all. The soleswatch, one may well contend, evidently contrasts rather sharply with a full range of artefacts, including, for example, my wristwatch. My wristwatch, unlike the soleswatch, exhibits all manner of intrinsic unity, and it clearly has an end, namely the one we have given it, telling time. Surely, then, Aristotle would be perverse to deny that it has a nature? Surely too he would be perverse to deny that it is a substance, to treat it as more akin to the soleswatch than to Socrates?

The only ground Aristotle could offer would seem to be this: there are certain paradoxes to which both the wristwatch and the soleswatch fall prey, but to which Socrates is immune. That is, he will need to show that despite appearances, a wristwatch cannot be a substance, since its identity conditions are unclear in ways which undercut its ability to be fully separate, or autonomous, as substances, as such, must be. There are, it seems, just these sorts of paradoxes available to him, in two senses: first they seem to have been at play in the Academy where he studied; and they seem to be as troubling as they are surprising.

As for the Academy, we find Socrates, in an autobiographical passage of the Phaedo, recounting his exasperation at his early efforts to make progress in natural science:

 I finally judged myself to have absolutely no gift for this kind of
 inquiry. I'll tell you a good enough sign of this: there had been
 things that I previously did know for sure, at least as I myself
 and others thought; yet I was then so utterly blinded by this
 inquiry that I unlearned even those things I formerly supposed I
 knew, including among many other things, why it is that a human
 being grows. That, I used earlier to suppose, was obvious to
 everyone: it was because of eating and drinking; whenever, from
 food, flesh came to accrue to flesh, and bone to bone, and
 similarly on the same principle the appropriate matter came to
 accrue to each of the other parts, it was then that the little bulk
 later came to be big; and in this way a small human being comes to
 be large. That was what I supposed then. That's reasonable, don't
 you think? (96c-d)

This account does seem reasonable--until one asks, as Plato evidently did, why we should say that the food accrues to the flesh and not the flesh to it. We customarily think of the sort of exchange which occurs in eating as the growth of a body by the accretion of something small, which goes out of existence once it is ingested. Why not, instead, treat the exchange as the growth of the food by the accretion of something large, which goes out of existence once it has been attached to the nourishment?

Fairly clearly there will be a ready answer to this sort of question, one which adverts to the intrinsic end of the organism which ingests the food, making use of it for its own purposes, its own end, its own flourishing. That may be well enough, on the supposition that we can distinguish between those things which have intrinsic ends, and which flourish by realizing those ends, and those things which have ends only derivatively, and which are not easily thought of as flourishing in their own right when those ends are reached.

This is presumably, however, the kind of distinction over which Plato, understandably, felt himself stumbling. For unless we are to make it a matter of convention that we grow by eating food and not it by being eaten, we will need to appeal to the existence of intrinsic, non-conventional ends. This, at any rate, is how Aristotle seems to regard the matter, after spelling out the sort of paradoxes which loom for our conceptions of growth and getting bigger. For he too has Plato's worry, but spells it out in a bit more detail:

 Someone might raise a difficulty concerning which is the growing
 thing (to auxanomenon), (a) the thing to which something is added
 (as for example, if <someone adds something> to a calf, it gets
 bigger), while the thing by which it grows, the nourishment, does
 not. <Or (b), both what is added and that to which it is added
 grow.> So why is it not the case that both have grown? For that
 which is added and the thing to which it is added are both larger,
 just as when you mix wine with water; each increases in the same
 way. Or is it that the substance of the one persists, but not that
 of the other, namely the nourishment? (GC 321a29-35: cf. 322a4-16)

The answer, according to Aristotle, involves an ineliminable appeal to the fact that one thing is a substance and the other is not. The human, as a teleonomic system whose activities are co-ordinated around some end-state which is its identifiable non-conventional good, has a function. This function remains identifiable as such in every process of material transaction through which the substance persists.

That it be its non-conventional good is key here. For the problem would not be resolved if the ends were ascribed to entities conventionally, or as Aristotle says, 'insofar as they happen to receive each <such> predicate' (Ph 192b17). For then we could ascribe a function to the food, treat it as a substance, and regard it as the thing augmented by having a big body acceded to it. If so, one will need to appeal not merely to the presence of an end, but rather to the presence of a non-conventional end which makes the relevant difference. In this sense, my wristwatch is more like the soleswatch than it is like Socrates. For the first two have extrinsically specified ends alone; and these ends can be ascribed and withdrawn in a purely pragmatic way, relative only to our current interests and desires. This is evidently not true of Socrates.

The point then can be summed up in a brief argument:

(1) There is such a thing as growth; and it is possible to say, in a principled way, which thing has grown when two material stuffs mingle.

(2) But in a mingling of two material stuffs, it is possible to specify which has grown only if at least one of them is a substance having a non-conventionally specified intrinsic end.

(3) Hence, substances have non-conventionally specified ends.

(4) Further, it is possible for a substance to have a non-conventionally specified end only if it is a living being.

(5) Hence, in a mingling of two material substances, it is possible to specify which has grown only if at least one of them is a living being.

The point, then, is that it is possible to make discriminations we regard ourselves as entitled to make only if we antecedently accept a commitment to non-conventional substances, which are living systems.

Further, these non-conventional substances can be the only substances, and indeed they must be alive. If our earlier results obtain, then all and only living beings have intrinsic ends, as sources of cybernetic directionality, not contingent upon the attitudes and preferences of entities outside of themselves. Treating instrumentally unified entities as substances in their own right will land us in an untenable and indefensible position as regards the metaphysics of growth. To this extent, then, Aristotle will be justified in inferring something which seems initially disconcerting, that living things alone are substances.

This claim can be made more precise in one final way, a way hinted at in Aristotle's worries about growth, but not developed in any detail. But it is really a way of seeing why one should accept the claim that in a mingling of two material substances, it is possible to specify which has grown only if substances have non-conventionally specified intrinsic ends. For what, after all, is it about growth and getting bigger that appears so paradoxical to both Plato and Aristotle? Plato does not say. Aristotle registers the worry we have already seen about the possibility of making a principled decision about which thing grows in cases of nourishment. What exactly is this worry?

Suppose we are in the military, and our sergeant takes us into the yard and assigns us each a pile of mud to remove from the yard. I might move mine onto your pile and then insist that I have finished my work, on the grounds that it was your pile, not mine, which got bigger. My pile is gone, and so trivially gone out of the yard. You might then disagree, insisting that your pile was not made larger, but is now merely such as to be very much closer to my pile. You might even be a philosopher, and add that your pile could not be made larger by aggregation, since nothing can be made larger by aggregation: your pile remains what it ever was, my pile is still my pile, and any new pile which may have come into existence is just that, a new pile. So it could not have gotten bigger either.

If you respond this way, I may find myself at a loss, especially if I thought that my pile was just a mereological compound. Evidently, the only way of treating my pile as more than that would require my giving it some unity. But once that happens, we are in a situation where the unity of our piles is pragmatically determined, so that we can always wrangle about whether the particular unity is the relevant one for the situation at hand. After all, my purposes may not be your purposes; and this, in effect, is what I was trying to ignore at first.

This sort of worry is precisely what Aristotle has in mind when in his discussion of growth he maintains that in a mingling of two material substances, which remains, if either, is determined by which function (ergon), if either, remains (GC 329a35-b2). But whenever we have a Theseus-ship case, it is not possible to determine uniquely and determinately which function remains. This is because in all such cases, there are a host of possible functions ascribable to the entities whose persistence conditions we are charting. If x and y have only conventionally specified functions, any determination about which function remains will be a decision made relative to the ascription of those functions to x and y. Yet, if a determination about x's function is made relative to the ascription of that function to x, it is always possible to ascribe some other function to x such that in terms of that new function x remains even if it would not have remained, relative to the ascription of its earlier function. Hence, it follows in a mingling of two material substances, it is possible to determine uniquely and determinantly which, if either, remains only if decisions about their functions are not made relative to the ascriptions of those functions to them. Further, decisions about the functions of entities are not made relative to the ascriptions of their functions to them only if substances have non-conventionally specified intrinsic ends. So far this seems unproblematic. But we get just the result Aristotle is after if we add one further claim: in a mingling of two substances, it is possible that at least one has a non-conventionally specified intrinsic end only if it is a living being. With that claim it will follow that in a mingling of two material substances, it is possible to specify definitively and uniquely which remains, if either, only if at least one of them is living.

If so, we will have reason to distinguish between living things and artefacts in a surprising and forceful way: living things, unlike artefacts, are substances. If we suppose otherwise, we open ourselves up to the charge that artefacts are stuck, uniquely, with the ends we ascribe to them at their creations. This assumption, however natural, is unsustainable. By contrast, if we are willing to go along with Aristotle in the judgment that some entities persist in virtue of their non-conventionally given natures, we arrive at a secure justification for discriminating, as we do, between the eaters and the eaten. If we are uncomfortable in going along with him in this judgment, we evidently remain where Plato had been before he made his blinding discovery: incapable of saying why what we know to be true is true.

We are, by contrast, perfectly able to say what we know to be true and also explain why it is true if we accept the following simple argument, whose premises we have now in any case explicated and defended:

(1) If x is capable of existing diachronically as a separate and determinate F, then x has an intrinsic end.

(2) x has an intrinsic end E iff x: (i) has an end E; and (ii) x has that end in a non-conventional way.

(3) Only living things have ends in a non-conventional way.

(4) Hence, only living things are unified entities capable of existing diachronically as separate and determinate Fs.

If we accept this argument's conclusion, we also accept our original (EC-2), and so accept Aristotle's own reason for advancing EC, the claim that only living things qualify as substances. This, finally, is Aristotle's motivation for extending the dignity of substance to living things alone. (23)

Bibliography

Albritton, Rogers. 1957. 'Forms of Particular Substances in Aristotle's Metaphysics'. Journal of Philosophy 54: 699-708.

Frede, M., and G. Patzig. 1988. Aristoteles Metaphysica: Text, Ubersetzung und Kommentar. Munich: Beck.

Gerson, Lloyd. 'Artifacts, Substances, and Essences'. Apeiron 18: 50-57.

Irwin, Terence. 1988. Aristotle's First Principles. Oxford: Clarendon Press.

Katayama, Errol. 1999. Aristotle on Artifacts. Albany: SUNY Press.

Lloyd, G. E. R. 1996. Aristotelian Explanations. New York: Cambridge University Press.

Ross, W.D. 1936. Aristotle's Physics. Oxford: Clarendon Press.

--. 1924. Aristotle's Metaphysics, 2 vols. Oxford: Clarendon Press.

Shields, Christopher. 2007. Aristotle. New York: Routledge.

Wedin, Michael. 2000. Aristotle's Theory of Substance. Oxford: Clarendon Press.

Wiggins, David. 1980. Sameness and Substance. Oxford: Clarendon Press.

(1) Ross 1924, Vol II, 229

(2) This is a point appreciated by Katayama 1999, 13-14.

(3) Similarly, Aristotle treats it as obvious on the basis of a simple survey of cases that everything is either said-of or in primary substances as subjects (Cat 2a34-6), a claim used in an argument to establish the primacy of primary substances relative to secondary substances and non-substances generally (Cat 2b-6b-c). If he were to restrict the category of primary substance to living things, he would have the immediately false result that everything would need to be said-of or in them alone.

(4) For a brief overview of why we might think this way, see Shields 2007, 235-7. For a detailed attempt to show that the Categories and Metaphysics yield complementary rather than competitive accounts of substance, see Wedin 2000.

(5) A very clear example is Irwin 1988, 571 n8. Less determinate, but evidently in agreement with Irwin are Frede and Patzig 1988, Vol II, 322. They note some ambivalence in Aristotle on this score, but then observe that in Metaphysics VII 17 he regards the artefacts which he had earlier introduced as substances no longer to be substance in the 'full sense'.

(6) Aquinas, Comm in Aris. Meta [section][section] 1679-1680; cf. [section][section] 1718-1719, where he unpacks Aristotle's suggestion at Metaph 1043b19-21 by contending that 'in artifacts the matter alone is held to be substance, whereas the forms of artefacts are accidents.' See also Albritton 1957, 704 n10, Wiggins 1980, 90-9.

(7) One could in principle accept a non-scalar conception of substance while accepting artefacts as substances--if they were to satisfy the criteria--just as one could accept a scalar approach and rule them out--if they were to either fall too low on the scale or (in principle) if they were not even to fall onto the scale at all. Although these possibilities remain open as a point of logic, in fact they have (for good reason) failed to attract the attention of commentators.

(8) It is important to distinguish two ways that one might regard the category of substance as scalar. One might think that something could be more or less a substance or that everything which is a substance is fully a substance, even though some substances have satisfied the criteria more fully than others. If every student who receives a cumulative mark of 70 graduates with first-class honours, then while no-one graduates with higher or lower firsts, we might yet discriminate between those who have earned marks of 79 and 70.9. As applied to substance, the first, but not the second, approach allows that something can be more or less a substance; the second approach, by contrast, insists that nothing is more or less a substance, while conceding that some instances are especially distinguished instances of the type.

(9) If this is correct, then my arguments will equally apply to the weaker paradigmatic claim. That is, if someone regards Aristotle as accepting (PC-1) and (PC-2) instead of (EC-1) and (EC-2), then the demand for unity which privileges living things in the case of the exclusive claim will also explain why Aristotle should think of living things as paradigmatically substantial.

(10) This list is what 'we' consider to be substances (Metaph 1028b8). Presumably Aristotle means this in a generic way, though the candidates are in some way still live possibilities. He goes on to mention some other candidates, drawn from other thinkers, mainly Plato and Speusippus: the limits of body, bodies, solids, forms, the objects of mathematics, and the one (Metaph 1028b16-26).

(11) I accept soros over orros on the ms. strength with both Ross and Jaeger. I note, however, that the comparison of the elements with a heap calls for some explanation. See [section] III below for an explanation. Cf. Metaph 1041b12, 1044a4.

(12) Perhaps not all of them, though the matter is disputed. At de Caelo 285a29-30, the heavens are ensouled and have a source of self-motion; cf. Cael 291a27 on the individual stars, with Cael 289a11-19, 292a21-2.

(13) Reading, with J, as in Ross 1924 II: kata phusin kai phusei sunestekasi.

(14) Cf. Metaph 1043b19-23: 'Whether there are separable substances among destructible things is not yet clear, although it is clear that in some cases this cannot be, namely those which cannot exist apart from certain things, for example a house or an implement. Presumably then these are not substances--neither these things themselves nor anything else which has not been formed by nature. For one might well take nature alone among destructible things to be substance.'

(15) On the connection between having a nature and an impulse (horme) to change, see also: APo 95a1, Metaph 1023a9-23, and EE 1224a18-b9). Aristotle will allow that fire has an impulse to go upward, something it does neither by force nor voluntarily. Aristotle remarks that the opposite of 'by force' (bia(i)) in the case of fire is nameless (EE 1224a19-20).

(16) The opening lines of Physics II 1 suggest that the distinction between those things which do and those which do not exist by nature is to be made in terms of their causal histories. When something arises by a cause other than nature, presumably by craft or purposive decision, or by luck, it dos not qualify as existing by nature. The distinction as developed later in the chapter, however, points instead to the presence or absence of an internal source of change and rest (Ph 192b13-19). The second distinction is not obviously equivalent to the first, not even extensionally. Since the bulk of the chapter relies on the second, but not the first, distinction, we should probably understand the opening lines as a general orientation.

(17) I thank M. Scharle for impressing this point and its importance upon me. It is noteworthy too that, as Ross 1936, 500, observes, Aristotle evidently refers to inanimate compounds as having an internal arche at Ph 192b11.

(18) On teleology and life, see Shields 1999, Ch. 7.

(19) For an explication of the Aristotle's notion of core-dependent homonymy, see Shields 1999, Ch. 4. For a treatment of life as a core-dependent homonym, see Ch. 7 of that same work.

(20) Van Inwagen (1990, 87) asserts easily that 'not just any self-maintaining event is a life. A flame or a wave is a self-maintaining event, but flames and waves are not lives. It is ... the business of biology to tell us what lives are.' He agrees with Aristotle about the extension of life in this respect. The difference is that Aristotle thinks that it falls to the philosopher to investigate the nature of life.

(21) This also comes out clearly in the Physics, where Aristotle asserts an important connection between being a teleological system and having a nature. He denies that seeds germinate and grow 'by chance'. In fact, 'anyone who asserts this destroys what is "by nature" and "nature" itself. For whatever arrives at some goal by a continuous movement from an internal source happens "by nature". The result is not the same for each, but yet it is not a matter of chance; but what occurs always tends toward the same goal, if nothing intervenes' (199b14-18). I agree with Charles who, in commenting on this passage, observes: 'If one removes talk of goals, one thereby removes talk of natures and natural processes' (1991, 117).

(22) On fusion (or 'concoction') see Lloyd 1996, 70, who rightly remarks that this is a 'remarkable use of the notion of concoction'. See all Lloyd 1996, Ch. 4, for a full treatment.

(23) Earlier versions of this paper were read at Dartmouth College, New York University, the University of Virginia. I am grateful to the members of the audiences on those occasions for their consideration. I also had the good fortune of discussing this paper at length in a seminar on Aristotelian hylomorphism conducted at the University of Alaska in the summer of 2007. That discussion, which was helpfully initiated by critical comments offered by John Mouracade, proved especially instructive. I also thank Paul Studtmann for discussing the themes of this paper with me in depth on several different occasions.