The absence of an Aristotelian work specifically dedicated to artefacts or well-developed argumentation about artificial objects forces us to address this topic in a piecemeal fashion, drawing on a variety of works and passages that are often not clearly related to one another. Although Aristotle’s ontology of artefacts could be reconstructed within the boundaries of the Metaphysics, here too the reconstruction needs to skip between different and, at times, quite distant passages. The reason for this is that Aristotle uses artefacts as a means of engaging in metaphysical debates, drawing on problems present in Plato. Aristotle’s theory of artefacts is related to Plato in three different ways. First, Aristotle uses artefacts against Plato, in the sense that artefacts provide counter-examples to or central elements of a refutation of Plato’s theories. I will dedicate the entirety of Chapter 2 to this first usage. Second, since artefacts play a crucial role in the critique of Plato’s theory of Ideas, even when Aristotle intends to build on Plato’s insights, whether correct or not, he does so in such a way as to accommodate artefacts. Third, Aristotle inherits from Plato a series of open philosophical questions that he sometimes addresses using artefacts. On this basis, he eventually derives conclusions that hold true for artefacts within the framework of his own metaphysics. As a result of this complex engagement with the Platonic tradition, however, Aristotle’s theory of artefacts must inevitably be reconstructed in a piecemeal fashion. Indeed, Aristotle’s primary goal is not to design a theory of artefacts, but rather to build on Plato without making the same mistakes (i.e. overlooking the advantages of having an account of artefacts), as means both to pre-empt counter-arguments and to effectively develop our sound intuitions.
It might seem surprising that Aristotle does not always use artefacts in argumentative contexts where Plato explicitly talks about them. It is, however, less surprising if we consider that the assumption underlying Aristotle’s criticism and appropriation of Plato is that he overlooked the importance of artefacts – as a source of both potential objections and philosophical insight. Hence, Plato’s own ontology of artefacts does not coincide with those philosophical issues raised by Plato that Aristotle thought could be effectively addressed or critiqued by appealing to artefacts; and these issues, in turn, do not directly provide a comprehensive picture of Aristotle’s own position on artefacts.
In Section 1.1, I outline Plato’s ontology of artefacts on the basis of his explicit references to Ideas of artefacts, his discussion of the eidetic cosmos and his introduction of divine art in the Timaeus. Thus, Section 1.1 considers Ideas of artefacts as they are mentioned in the Cratylus and Republic X, the discussion of the extent of the eidetic cosmos presented in the Parmenides and the description of the material world as a product of the divine artisan in the Timaeus. A consideration of these different loci reveals several inconsistencies that make it difficult to know what Plato really thinks. Later interpreters sketched the position that Plato might have – or should have – favoured.
In Section 1.2, I present the shortcomings of Plato’s account detected by Aristotle and addressed through artefacts: these are Plato’s failure to recognise final causes and the concept of imitation. In Section 1.3, I outline the metaphysical intuitions upon which Aristotle builds by taking artefacts into account. The roots of certain metaphysical problems are not explicitly identified as Platonic, but they are arguably to be found in the Platonic corpus. For this reason, the enquiry begins with one or several Platonic passages, which are then compared with certain Aristotelian arguments. These problems concern the range of things that have a form, the separation between axiology and metaphysics, the concept of real kinds and the relation between parts and whole.
1.1 Plato’s Metaphysics of Artefacts
1.1.1 Plato’s Ideas of Artefacts in the Cratylus and Republic X
In his oeuvre, Plato explicitly mentions Ideas of artefacts twice.Footnote 1 In both cases, he is neither arguing for their existence nor even dealing with the problem of the scope of the eidetic cosmos. I shall not argue here that Plato denies the existence of Ideas of artefacts, as Aristotle would have it,Footnote 2 but rather I shall point to some inconsistencies that may have prompted part of the Aristotelian discussion.
In a sense, Plato’s notion that there are Ideas of art objects goes back to the Gorgias.Footnote 3 At 503D–E, Socrates invites Callicles to address the definition of the good man in light of the way in which artisans tackle their work. Artisans keep the finished product in view, selecting what they need not randomly but with a view to the form that they intend to bestow. Socrates mentions artisans such as housebuilders, shipwrights and painters, and ascribes to them the ability to organise the matter in an orderly way by keeping the object as a whole in mind. In the Gorgias, in other words, artisans work with a view to an idea of what they ought to produce.
The first explicit occurrence of Ideas of artefacts is, however, to be found in the Cratylus, where Plato speaks of the shuttle per se (389B5). Socrates and Hermogenes are developing the analogy between naming and weaving. Names are tools for establishing relations, differences and therefore order. The production of a given tool must take into account certain formal features (i.e. the functions the object is supposed to fulfil). It is in this sense that Socrates says that the carpenter looks to the shuttle itself, while producing a particular shuttle. The material constitution must correspond to the formal features of the object: the material is chosen with a view to the function that the shuttle is naturally predisposed to fulfil. The same holds in the case of names: names are naturally predisposed to fulfil a pedagogical and discriminating function. Hence, their material constitution (i.e. sounds and syllables) must be apt to fulfil that function. In general, however, this passage does not necessarily introduce an opposition between models (i.e. Ideas) and particular images or likenesses, or argue for the existence of an Idea of shuttle that stands in opposition to particular shuttles.Footnote 4 Thus, it cannot be immediately considered a piece of evidence for the Platonic acknowledgement of Ideas of artefacts. For instance, Reference FineFine (1993) distinguishes between Socratic forms (which are not separate) and Platonic forms (which are perfect, non-sensible, eternal and separate). In the cases of the Gorgias and the Cratylus, the Ideas of artefacts are, on her account, Socratic rather than Platonic.Footnote 5 Fine takes the forms of artefacts introduced by Plato in most passages to be Socratic forms. The only exception, she claims, is Rep. X, where the form of couch and the form of table are taken as perfect and non-sensible, and hence as Platonic forms. She additionally suggests that although these forms are presented as Platonic forms, they need not also be everlasting and separate, as Aristotle takes them to be.
Let us now turn to the Idea of Couch in Rep. X, which represents the second occurrence of an Idea of an artefact. In this passage, Socrates and Glaucon agree to discuss the meaning of imitation. They opt for the customary method of defining the one that stands over and above a set of things that have the same name. ‘There are many couches and tables’, Socrates says, without providing any further justification for this quite spontaneous example. In 596B–C, Socrates introduces three kinds of couches: the couch that exists by nature, which is made by God himself, the couch made by the carpenter and the couch portrayed by a painter, which is created by the painter. The third couch is notoriously twice removed from the Couch itself, in that it is a likeness of a likeness. The couch produced by the carpenter is once removed from the couch made by God, but it still deserves the name of couch. The painter, by contrast, only produces imitations, because he reproduces things as they appear to us, and not as they really are. Socrates and Glaucon seem to agree that the carpenter produces the couch by drawing on the couch made by God and that God makes the ‘truly real couch’ (597D). The God is the real maker of the one true couch ‘and of everything else’ (597D). Plato thus introduces here the notion of the true couch, which is numerically one and which exists prior to the many couches that the many carpenters are capable of producing.
Whether forms are being posited depends on just how the ‘customary method’ (Armstrong 1978) or ‘regular practice’ (Reference Sedley, Chiaradonna and GalluzzoSedley 2013) advocated in Rep. X 596A is understood. This method consists in positing a form corresponding to a set of things with the same name. Sedley argues that the sameness of the name does not concern the plurality of items taken on their own, but rather relates these items to the form. That said, the claim is not that for any set of things with the same name one can posit a form, but rather that the particulars are named after the form. In other words, ‘Socrates is not asserting the principle of one form per plurality, just the more modest principle of one plurality per form’ (Reference Sedley, Chiaradonna and GalluzzoSedley 2013, 128). If this is true, there are important consequences for Plato’s ontology of artefacts. The Three Couches Argument would establish, first and foremost, that there is precisely one Idea of Couch.Footnote 6 Corresponding to the unitary Idea of Couch we find a plurality of couches (i.e. the couches made by human artisans). Again, it is not so much the case that for any set of things that are all called ‘couch’ one can posit a corresponding Idea of Couch, but rather that the particular couches are named after the Idea of Couch. If this is the case, Plato would be providing a strong defence of the existence of Ideas of artefacts. Furthermore, not only is there an Idea of Couch standing over a plurality of things called ‘couches’ but also, in making couches, human artisans look to this form.
Nonetheless, there are several reasons for thinking that this is not sufficient to justify ascribing to Plato a belief in the existence of Ideas of artefacts. One could say that the aim of Book X of the Republic is to assess the kind of poetry permitted in the Kallipolis and that the question of the extent of the eidetic cosmos is in no way a concern. More generally, however, one could point to the apparent inconsistencies with other Platonic dialogues. Indeed, if the God of the Republic is the Demiurge of the Timaeus, the two theories are different and incompatible. In the Timaeus, the Demiurge is not the cause of the existence of Ideas, which are given. Instead, he looks to the Ideas in order to create the cosmos. In the Republic, the God himself creates the one true form of Couch – he is the only artisan capable of making the form itself – while the other artisans create a plurality of particular couches by looking at the couch created by the God. In describing the God, Socrates adds that ‘this same artisan is able to make, not only all kinds of furniture, but all plants that grow from the earth, all animals (including himself), the earth itself, the heavens, the gods, all the things in the heavens and in Hades beneath the earth’ (596C–D). The concept of a God creating both natural and artificial beings raises concerns about the consistency of the Platonic position on artefacts, if we compare it with both the Timaeus and Socrates’ doubts in the Parmenides. As regards the Timaeus, although in this dialogue too the Demiurge creates the entire cosmos, there is no evidence that he creates artificial beings. In fact, artefacts are not even mentioned at all. As regards the Parmenides, if God creates Ideas of artefacts, as well as Ideas of all living beings, either the doubts of the young Socrates about the existence of Ideas of animals are incomprehensible or the passage from the Cratylus does not, in fact, imply the existence of such Ideas. Either way, within the Platonic oeuvre there are inconsistencies regarding Ideas of artefacts and their ontological status. Perhaps it is precisely these inconsistencies and difficulties that lie at the core of Aristotle’s enterprise.
1.1.2 The Doubts of the Young Socrates in the Parmenides
While it is true that every Idea is an eternal model for certain particular likenesses, it is not true that every particular kind of thing has a corresponding Idea. The first objection that Parmenides raises against the theory of Ideas advocated by Socrates concerns the extent of the eidetic cosmos. In fact, the only Platonic dialogue that explicitly deals with the extent of the eidetic cosmos is the Parmenides.Footnote 7 The question at issue is: of what things are there Ideas? Socrates is eager to accept Ideas of similarity and Ideas of values, such as justice, beauty and good. However, faced with Parmenides’ question about whether there is an Idea of man, fire or water, Socrates does not conceal his embarrassment: he finds it hard to answer to this question. Parmenides presses on, asking Socrates whether there are Ideas of things devoid of value, such as hair, dirt and mud (130C6–7). Socrates is more confident here, denying that such Ideas exist, yet he once again states that he feels somewhat uncomfortable with the question, since he is afraid of ending up admitting to the existence of too many Ideas. Parmenides blames Socrates’ youth and inexperience and seems to suggest that Socrates is still too concerned with the opinions of others to be open to all sorts of things, both more and less valuable. In a context in which the mention of artefacts would be extremely natural, Plato unfortunately makes no reference to them.Footnote 8 However, artefacts might be thought to belong among those things that are not particularly valuable, such as dirt, mud and hair.Footnote 9 This would be odd, given that the whole cosmos is an artefact. Perhaps Plato would not even list them among the non-valuable things, but Socrates’ doubts about the existence of the Idea of man seem to be reasonably applicable to artificial things, such as couches and tables.Footnote 10 A definitive answer is hardly possible, particularly within the framework of the Parmenides. Is there any way of evaluating the problem of the existence of Ideas of artefacts elsewhere without contradicting what is said in the Parmenides? Later interpreters elaborated the position that Plato would have – or should have – favoured, namely that there are no Ideas of artefacts.
Although the Parmenides does not provide any clues to understanding whether Plato denies or accepts Ideas of artefacts, the Neoplatonic tradition shows considerable agreement in denying the existence of Ideas of artefacts.Footnote 11 One highly interesting piece of testimony is that of Alcinous,Footnote 12 who states that most Platonists deny that there are Ideas of artefacts. This compact agreement is probably due to the Xenocratean definition of an Idea as a ‘cause exemplifying things always constituted by nature’.Footnote 13 The Neoplatonic tradition denies the existence of Ideas of artefacts, but it is difficult to understand who, within the first Academy, was the first to adopt this position. Generally, the scholarship has identified four possible answers: Plato,Footnote 14 some Academics,Footnote 15 Aristotle misinterpreting PlatoFootnote 16 or Alexander misinterpreting Aristotle.Footnote 17 Setting aside this rich debate, it seems that in Peri Ideôn, as well as in some passages of the Metaphysics, Aristotle ascribes the denial of separate artificial forms to the Academy without excluding Plato, either partially or wholly. There are some reasons for at least thinking that Plato would not willingly allow for the existence of Ideas of artefacts. It is true that Plato mentions Ideas of certain artefacts, such as that of a shuttle in Cratylus 389B7–8, or the Ideas of Couch and Table in Rep. 596B. There are also passages in which Plato seems to suggest a universally applicable version of the one-over-many argument.Footnote 18 However, careful contextualisation suggests rather that Plato is not expressly arguing for the existence of Ideas of all kinds of thingsFootnote 19 and, more importantly, that he is not even dealing with the question of the existence of such Ideas.
An explanation for why Plato might deny the existence of Ideas of artefacts comes from the Republic. At 601D–E, Plato ascribes right opinion to the artisan and science to the user of artefacts. Arts do not have the same epistemological status as the sciences. If the artisan looks at the Ideas while producing something like a table, their productive activity would have the same dignity as the theoretical one, meaning that the artisan would be a philosopher. On a more general level, in order to produce a statue, the artisan does not require Ideas, but merely the thing that they are imitating. If the sculptor wants to create a statue of Socrates, they do not need to look to the Idea of Socrates, but only to have Socrates in front of them or present in their mind. Of course, this reasoning applies better to imitative arts. However, we can say that health too needs only to be present in the mind of the doctor. In order for the doctor to produce health, they merely need the notion of what health is in their mind. As I shall argue in Section 2.2.1, this is precisely Aristotle’s main counter-argument against the Platonic theory of Ideas. Therefore, this last line of reasoning is crucial for Aristotle, but there is no evidence for it in the Platonic text. Perhaps, this is a point that Aristotle uses against Plato, while Plato did not himself think of it as a reason to deny the existence of Ideas of artefacts.
A major reason why Plato might have denied the existence of Ideas of artefacts has to do with their dependency on humans. Humans are needed for there to be particular chairs since humans make chairs and there would be no chairs if there were no humans. So, there have not always been chairs: the kind chair is not eternal – or the form of chair is not an eternal model. I shall come back to the issue of eternity in Chapter 5, where I also discuss passages in which Aristotle ascribes the denial of Ideas of artefacts to Plato and the Platonists.Footnote 20 One plausible option is that Plato did not directly address this problem, but that subsequent Platonists did start to see a potential problem, leading them to take a clearer position. Thus, at this stage, let it suffice to show that Plato left the problem open to several different interpretations and that not only Aristotle, but also later Neoplatonic interpreters, established that there should not be Ideas of artefacts.
1.1.3 Divine Craftsmanship in the Timaeus
The Timaeus represents an important application of the concept of technê. The divine Demiurge is the artisan par excellence: everything in the world is a product of his infallible art.Footnote 21 The Demiurge’s craftsmanship is the model for human production. He understands the nature of his subject-matter and could therefore give an account of his method and reasoning. Through discrete phases of design and step-by-step construction, the divine artisan creates the cosmos wholly free from deficiency.Footnote 22 The product of his activity is good in its very constitution. The Demiurge is not, however, the creator of everything. The material he works on (i.e. the physical elements) is given to him. He does not create the material out of which he creates the cosmos. This means that his power does not overcome the limitations of his material. Just like a human artisan, the Demiurge too is constrained to work on already existing material, which dictates the limits of his creative power. Furthermore, the Demiurge shapes his material with a view to a model. The models for his creation are the eternal forms.Footnote 23 Although the Demiurge does not create, for instance, the sun by copying it from an already existing sun – as a painter would – the models (i.e. the Ideas) are already given.Footnote 24 The divine artisan does not create them. Not only are there models for the particulars populating the material world, but there is also a model for the cosmos as a whole: an ideal animal containing in itself all kinds of living beings. In other words, not only each particular but also the material world as a whole is a product of art and only resembles an ideal animal, which is the one true animal.
Although the Demiurge creates neither his model nor his material, he does create the soul through the mixture of being, sameness and difference: the soul is intermediate between the likenesses and the models and places the likenesses in contact with what is truly natural and real. After creating the soul, the Demiurge directly produces the following particular entities: the planets, the sun, the moon, night and day, and finally the gods. As for animals, the Demiurge makes the individual souls, whereas the gods are in charge of their bodies: first, the animals living in the air, then those living in water and finally those living on earth. If human bodies are created with a view to the fact that they ought to host a soul, other animals are created as degenerations from the human archetype. Human beings are therefore the best likenesses that are created. If other animals represent degenerations from the human archetype, the question of whether there are Ideas of those other animals on which animals are modelled is also raised in the context of the Timaeus.Footnote 25 In 41D–42D, the process of degradation occurs if the soul has been enslaved by passions. The first degradation is from a male human being to a female human being, but then the process goes on: from female human beings to birds, from birds to quadrupeds, from quadrupeds to reptiles and from reptiles to aquatic animals (90E–92C). It seems that each kind of animal is a condition for the generation of another. If the existence of, for example, reptiles is a sufficient condition for the generation of aquatic animals, the existence of Ideas of aquatic animals is ruled out. According to Reference Rashed and ChiaradonnaRashed (2012, 70), in the Timaeus, contingency consists in the occurrence of the various acts of metempsychosis, whereas necessity consists in the fact that there are precisely four classes of animals. Viewed in this way, the existence of reptiles is contingently sufficient for there to be aquatic animals and what is additionally necessary is the general rule that there ought to be four classes. Again, there seems to be no reason to posit Ideas of animals here; if at all, one would posit Ideas as mathematical objects (in this case, the pentadic structure). The scala naturae presented in the Timaeus speaks against the existence of Ideas of animals. Ideas seem to entail mathematical structures and objects, while the sensible world is ontologically hierarchised according to ethical rules.Footnote 26 If there is no need for Ideas of species or more general kinds of animals, it seems that there is even less need for Ideas of artefacts.
Whether considering the cosmos as a whole (i.e. as one likeness), or in its parts (i.e. as many likenesses of many models), the material world seems to consist in products of art that are ontologically inferior to the Demiurge’s models. Likenesses are only insofar as they have been modelled after the Ideas, which truly are. The distinction between particulars and universals reproduces the distinction between artificial and natural. Plato provides a distinction between particulars and universals in terms of the distinction between artefacts and natural beings, but this introduces inconsistencies and unresolved issues across the different dialogues.
1.2 Building on Plato’s Theoretical Shortcomings
Aristotle’s theory of artefacts is related to Plato in different ways. First, Aristotle uses artefacts against Plato, in the sense that artefacts provide counter-examples to or central elements of a refutation of Plato’s theories. I dedicate the entirety of Chapter 2 to this first way, while I concentrate here on the second way. Since artefacts play a crucial role in the critique of Plato’s theory of Ideas, even when Aristotle intends to build on Plato’s insights, whether correct or not, he does so in such a way as to accommodate artefacts. He makes two criticisms of Plato without referring to artefacts. Rather, these criticisms point at shortcomings, which Aristotle overcomes by incorporating artefacts. The two criticisms concern Plato’s failure to recognise final causes and his deficient notion of imitation. Chapter 3 discusses in detail the upshots of this approach.
1.2.1 Final Causation
Aristotle criticises Plato’s theory of causation in terms of final causation, as well. He complains that Plato fails to recognise final causes. In Met. A 7 Aristotle states that:
That for the sake of which actions, changes and movements take place, they assert to be a cause in a way, but not in this way, i.e. they do not assert [to be a cause] in the way in which it is its nature to be cause. … so that in a sense both say and do not say the good is a cause; for they do not call it a cause qua good but only coincidentally.
His target is two different groups of philosophers who are guilty of overlooking the final cause: Anaxagoras and Empedocles on one side, Plato and the Pythagoreans on the other. In Aristotle’s view, Plato fails to recognise the good for the sake of which something occurs.Footnote 27 Whether this is a fair criticism falls outside the scope of my discussion. What I want to suggest, instead, is that Aristotle’s discussion of the art analogy serves the purpose of showing that there is a final cause in nature, which one can recognise by the fact that art imitates nature and that there is a good for the sake of which an art is exercised. Two premises ought to be accepted by a Platonist and are taken as endoxastic. The first is that arts work for the sake of the good. Aristotle has good reasons to believe that a Platonist would readily agree with this premise: Plato’s account of technê and its use as a model for ethical behaviour takes it to have a teleological structure aiming at some good. The second premise is that art imitates nature, which is also endoxastic. Not only would Plato agree that likenesses imitate models, but in Rep. X Plato endorses the view that human artisans imitate the products of the Demiurge. There is an important sense in which Aristotle employs artefacts in order to prove not only that there is a final cause qua good in nature, but also that this is easily recognisable from the fact that art imitates nature. Given that everyone agrees that art imitates nature, once one sees that there is a good for the sake of which an artefact comes-to-be, one ought also to see that there is a good for the sake of which a natural being comes-to-be. Aristotle is not using artefacts against Plato (i.e. he is not postulating that there is a final cause in artefacts); rather, he is addressing a shortcoming in his theory by arguing for the existence of a final cause on the basis of the plausible postulate that art imitates nature. I shall explore the art analogy in more detail in Section 3.1 (Chapter 3).
1.2.2 Models, Likenesses and the Notion of Imitation
In Met. A 6, Aristotle criticises Socrates for engaging with only ethical issues, and once he turns to Plato he says:
Things of this sort, then, he called Ideas, and perceptible things were apart from these, and all called after these; for the multitude of things which have the same name as the forms exists by participation in them. He only renamed it ‘participation’; for the Pythagoreans say that things exist by imitation of numbers, whereas Plato says they exist by participation, changing the name. But what the participation or the imitation of the forms could be they left an open question.
Plato left unclear the way in which things come to exist, since he calls participation ‘imitation’ but does not further explain what it entails. Aristotle is explicitly dissatisfied with the concept of participation, even if it is understood as imitation. In Met. A 9, 991a20–2, Aristotle questions the meaning of the claim that Ideas are paradigms and wonders whether calling them paradigms is only a metaphorical and ultimately empty way of talking about them. In fact, it is not clear how paradigms are supposed to fashion particulars (i.e. to be the efficient causes of their existence). Anything, says Aristotle, can be or become like something else without imitating it (as particulars do with Ideas in Plato’s account). Even if Socrates were something eternal, the becoming-like-Socrates of a given man does not necessitate that Socrates is a paradigm. If imitation is supposed to explain the relation between models and likenesses, this is a dissatisfactory answer, for something can exist without imitating anything.
The first problem is the metaphysical hierarchy implicit in the Platonic notion of imitation. According to Aristotle, imitation does not hierarchise; rather, it is grounded in a metaphysical trait that is equally shared by model and image. As we anticipated in Section 1.1.3, Timaeus’ description of the whole cosmos as being the result of the Demiurge’s activity makes all particulars, irrespective of whether they are ostensibly natural or artificial, (artificial) imitations of the models (forms), which are the truest (natural) things. Timaeus’ picture introduces the problem of the metaphysical difference between artefacts and living beings, for according to the dialogue there is a metaphysical difference only between particulars and models, but not among particulars. The relation between the likeness and the model is one of imitation. The presence of the Demiurge and the Timaeus’ general framework introduce the concept of imitation as a relationship between likenesses and models, and therefore between particulars and Ideas.
If we leave the Timaeus aside and refrain from introducing the divine Demiurge, the relation that cries out of explanation is that of participation. In the Phaedo (74A, 74C–D, 76A) and in the Parmenides (132A–133A), imitation – or resemblance – is presented as the correct account to give of the relationship between Ideas and particulars. In the Parmenides, after having shown the difficulties involved in a mereological interpretation of participation, Parmenides raises the further difficulties arising from the attempt to interpret participation as the imitation of a model by a likeness. The metaphysical debate at issue concerns the ontology entailed by the Platonic conception of imitation: the likeness (i.e. the particular) lies on a different ontological level than the model (i.e. the form, which is supposedly universal). It seems that, just as in the Timaeus, there is a metaphysical difference between the Ideas and the particulars, but not between particulars. However, this conclusion is inconsistent with the Three Couches Argument from Rep. X: the products of representative arts are imitations in that they are twice removed from the model and they do not imitate how things are, but merely how they appear. The same can be said about the Sophist (236C), where the imitative art (mimêtikê) is divided by the Eleatic Stranger into copy-making (eikastikê) and appearance-making (phantastikê) branches, with sophistry falling under the appearance-making. In this way, the sophist imitates the appearance of the wise man. Plato, in other words, seems to introduce a metaphysical distinction between particulars as well, but fails to provide an argument or an explanation for how the two types of metaphysical difference (i.e. between Ideas and particulars and between different kinds of particulars) hang together. Even if we grant Plato the existence of these two types of metaphysical difference and assert that the Timaeus does not necessarily deny the existence of differences between the likenesses, there is still a potential inconsistency that Aristotle might be exploiting. How do the differences between particulars manifest themselves? According to the model found in the Timaeus, the differences occur as a continuum in which one kind of being (e.g. one animal species) is the degradation of another one. By contrast, according to the model laid out in Rep. X, the differences between being a model and appearing, as well as between being a likeness and being an imitation (in the sense that the term is used here) are presented as discrete differences.
On the basis of these passages, we come to see that Aristotle is unhappy with the notion of imitation, although his use of artefacts to oppose Platonic imitation is not entirely straightforward. Still, Aristotle challenges the importance of the concept of imitation by means of the art analogy. To start with, Aristotle denies that the relationship between particulars and universals is one of imitation. In general, as we shall see in Section 1.3.4, the relationship between particulars and universals cannot be explained in this metaphysically impoverished vocabulary. Furthermore, imitation is a relation between things that exist on the same ontological level (i.e. two particulars that have forms). The Aristotelian art analogy establishes a relation of imitation between art and nature, as well as between artificial and natural beings. Despite the ontological differences between artefacts and natural beings, they are still situated on the same ontological level with regard to their forms: none of these particulars, whether a dog or a couch, is an imitation of an ontologically superior form, but a particular artefact is endowed with a form and is an imitation of a particular natural being, which is also endowed with a form. As we shall see more clearly in Chapter 3 (Section 3.1), the Aristotelian art analogy establishes the presence of intrinsic ends in artefacts, as well as natural beings. Intrinsic ends constitute the forms of these objects. Artificial beings are therefore just as metaphysically ‘true’ as natural beings.Footnote 28 The concept of imitation cannot explain the relation of forms to particulars. The example of artefacts makes this plain: it shows that imitation is a relationship that exists among particulars and that it does not necessarily produce a hierarchy between different ontological levels. As a doctor, one might well imitate health; yet, the health restored by a doctor is health just as much as the health which comes-to-be spontaneously.
Another problem with the conception of imitation as the relation between models and likenesses is that it portrays the relation between universals and particulars in such a way that every particular – whether natural or artificial – becomes a mere artificial copy of the only true things (i.e. models). To be sure, the prelude of the Timaeus (27D–29D) establishes that the story that will be told is only a likely story. The Demiurge is ‘hard to find and impossible to declare to all men’, and Timaeus seems to make use of an image that cannot be taken literally. The reason why Timaeus tells only a likely story has to do with Plato’s metaphysics: any account of the physical world can be no better than a likely story, because the world itself is only an image of an unchanging reality. The physical world and its components are mere images of the eternal world of Ideas. Thus, even on a metaphorical reading of the Timaeus, if everything in the physical world – including, first and foremost, Timaeus’ story – is a likeness of eternal and universal models, particulars are mere likenesses (i.e. artificial beings) that imitate Ideas that are in the true sense (i.e. natural beings). Plato, therefore, can be interpreted as offering a distinction between universal Ideas and particulars by drawing on the distinction between what is natural and what is artificial. Aristotle, however, appears to take the Timaeus literally. On a literal interpretation of the Timaeus, if everything created is created by an artisan, what we commonly define as artificial and natural beings are ultimately all artificial beings. Thus, Aristotle defends a sharp distinction between natural beings and artefacts by fashioning conceptual tools that allow for the possibility that something might come-to-be by art or by an artisan without being an artefact. In particular, Aristotle’s distinction between artificially or naturally caused and artefacts or natural beings enables animals and plants to be defined as natural beings, even within a Platonic framework in which the world is created by a Demiurge. Chapter 3 (Section 3.2) focuses on Aristotle’s manoeuvre in response to Plato’s notion of imitation and its consequences.
1.3 Platonic Intuitions as the Source of Aristotle’s Account of Artefacts
Aristotle deploys artefacts against Plato. Mindful of the risks of not dealing properly with artefacts, as his teacher did, Aristotle incorporates artefacts into his theories in a positive way, as he builds on Plato’s metaphysical intuitions. Faced with a genuine metaphysical issue, Aristotle goes beyond Plato in addressing it by devoting attention to artefacts. Most of this book will focus on the account of artefacts that Aristotle ends up constructing as a result of this approach.
1.3.1 Ideas and Forms
There are several inconsistencies in the Platonic corpus that concern the eidetic cosmos. In the Parmenides, Socrates voices his doubts about the range of things for which there are Ideas. Unfortunately, Ideas of artefacts are not mentioned here, but Socrates does appear to be dubious about the existence of Ideas of animals, a stance that is inconsistent with the mention of the Idea of Bee in Meno 72A8–B2. The Parmenides also rules out the existence of Ideas of relatives, although they are reintroduced in Theaet. 176E. This state of affairs suggests that Plato is not proposing an ontology of artefacts, but that he certainly lays out the problem of what things possess a formal cause. The Parmenides leaves open the question of what things have corresponding Ideas, a philosophical challenge that Aristotle inherits from Plato and that leads him to attempt to determine what things have forms. Ideas and Aristotelian forms are certainly different in many respects, the most obvious being their separateness. However, they are nonetheless comparable in at least one respect. Both Plato and Aristotle want to account for what it is that gives things common features. Aristotle notoriously rejects the separation of this common factor. However, the question of what makes something what it is shared by both philosophers, although Plato emphasises its transcendence, while Aristotle focuses on its substantiality. The final difficulty raised by Parmenides concerns the substance of particular objects and is formulated in a way that an Aristotelian would find quite familiar. Parmenides says with respect to the particulars ‘down here’ that – like the Ideas – they also have their substance and name in relation to each other and not to the Ideas. For instance, he argues, the slave is the slave of the master ‘down here’ prior to being the slave of the master-in-itself.Footnote 29 Plato might be introducing here the concept of an immanent character that is close to Aristotle’s immanent forms.Footnote 30 He might also be speaking of particulars imitating Ideas.Footnote 31 Precisely this ambiguity seems to be at the core of Aristotle’s approach of taking up issues and tensions already present in the Platonic oeuvre. In other words, the seventh difficulty raised by Plato produces an ambiguity in the concept of immanent character that oscillates between the Platonic notion of separate Ideas and the Aristotelian notion of immanent forms. As we have seen, Plato is not clear on the status of Ideas of artefacts. However, Aristotle unhesitatingly ascribes to the Platonists the rejection of such Ideas. As we shall see more extensively in Chapter 2, in Peri Ideôn, he claims that the Platonists deny that artefacts have Ideas and throughout the Metaphysics repeatedly alludes to this denial. In Peri Ideôn, in the context of the arguments from the sciences, Aristotle twice affirms that the Platonists do not want to admit the existence of Ideas of artefacts. In the Metaphysics, he repeats, on four occasions, the common opinion according to which there is no separate Idea of things such as rings or houses.
On these grounds, the question of what things have Ideas – which is raised by Parmenides in the eponymous dialogue – can be translated into the Aristotelian question of what things have forms. Aristotle does not raise the question explicitly, but he does provide an explicit answer with respect both to the entities mentioned in the Parmenides and to those that Plato should have mentioned in this dialogue. As for the first group, Aristotle, in his catalogues of substances, mentions elements and parts of natural beings as recognised substances. By the end of Z 16, however, we learn that elements and parts of natural beings are not substances after all, because they do not possess a form. This topic will be extensively addressed in Chapter 7. Eventually, Aristotle implicitly takes a stance on the questionable entities identified by Parmenides. Mud, hair and dirt do not possess a form in the strict sense, qualifying instead as matter or parts. As for artefacts – those cases that Plato does not mention, but should have – although Aristotle agrees with the Platonists that there are no separate Ideas of artefacts, he might nonetheless hold that artefacts possess forms of another kind. In conclusion, Plato’s valuable intuition was to raise the question of what things have forms, but in order to answer this question satisfactorily, we must take artefacts into account in a way that Plato failed to do. We will see how Aristotle does so in Chapter 4.
1.3.2 Axiology and Metaphysics
Another Platonic intuition that shapes, if not motivates, Aristotle’s approach to artefacts can be found in a range of different dialogues. Plato often oscillates between metaphysical and axiological arguments. Despite spotting and elaborating genuinely metaphysical problems, Plato fails to provide metaphysical solutions. However, he consistently expresses doubts that axiological considerations can play any useful role in metaphysical discussions, a stance that Aristotle builds on. Notwithstanding his commitment to an axiological structure according to which some things are better than others, Aristotle does not introduce this structure in order to solve metaphysical problems. Thus, I shall begin by identifying what, to my mind, are the most significant instances of Plato’s doubts in this regard, before illustrating how Aristotle takes them further by appealing to the case of artefacts. Moreover, the very fact that Aristotle gives an account of artefacts, despite their inability to participate in eternity, can be considered an outcome of his Platonic heritage.
In the Sophist, Theaetetus and the Stranger employ the dihairetic method to distinguish the sophist from the philosopher. The search for a definition and the dialectical exercise concern transcendent eidê. What one calls ‘sophist’ must correspond to a species that can be grasped by a dialectical procedure – or, at least, the latter must have it as its aim. Theaetetus and the Stranger call ‘cleansing’ the kind of differentiation that keeps what is better and throws away what is worse. In 226A8–C6, cleansing is subdivided into the cleansing of the internal part of living bodies (which is done by gymnastics and medicine) and the cleansing of the external part (which is done by bathing and such things). The Stranger calls the external part ‘base’ (phaula) and all kinds of furbishing the external part ‘trivial’ (smikra) and possessing ridiculous (geloia) names. After this dismissive treatment, the Stranger contrasts our judgement with the dialectical method. The dialectical method, in fact, does not bother to distinguish between cleansing by taking medicine and cleansing by means of a sponge – as far as the (real) kind of activity is concerned, axiological considerations play no role. The dialectical method ‘tries to understand how all kinds of expertise (pasôn technôn) belong to the same kind or not’ (B1–2) and ‘so for that it values them all equally without thinking that some of them are more ridiculous (geloiotera) than others, as far as their similarity is concerned’ (B3–5). Similarly, the dialectical method does not care whether hunting performed with military expertise is more impressive (semnoteron) than that which is conducted by means of the art of louse-catching. In the same way, in order to perform a correct division of cleansing, one need not – and should not – consider what sounds finest (euprepestaton). This passage not only argues that the dihairetic method does not follow axiological principles, but it also suggests that even contemptible things can be led back to Ideas – without assuming the existence of corresponding Ideas. One lesson that one might take from the Sophist is that, as far as method is concerned, we should not distinguish between dignified and undignified things. And indeed, no such distinction is made. This lesson is repeated in different terms later on (247E2ff.), where Plato states that a thing really is if it has any capacity at all to either do something to something else by nature or to have something done to it even at the smallest scale and by the basest thing (hupo tou phaulotatou), and even if it only happens once.Footnote 32
Plato appears to point to the same problem in the Statesman. The dialogue explicitly adopts the dihairetic method from the Sophist in order to define the statesman. At 266D4–9, the Stranger reminds Socrates the Younger that the lesson of the Sophist should be kept in mind, namely: ‘That such a method of argument as ours is not more concerned with what is more dignified (semnoteron) than with what is not, and neither does it at all despise the smaller more than the greater, but always reaches the truest conclusion by itself.’ Therefore, the division might happen to group together human beings with pigs, but this is of no concern, since ‘noblest’ (gennaiotaton) and ‘most easy-going’ (eucherestaton) are not categories that one uses in our ‘method of inquiry’. Once again – just like in the Sophist – the dialectical method and our hierarchy of values do not mirror each other, and baser animals can be referred back to the Ideas no less than the noblest ones can.
This lesson is taken seriously throughout the Statesman, and also in the Parmenides. As has already been mentioned, Socrates raises doubts about the existence of Ideas of things like mud, hair and dirt – on the grounds that these things are wholly undignified (atimotaton) and basest (phaulotaton). In this case too, Plato seems to suggest that the worthlessness of a particular group of beings cannot be taken as a metaphysical argument for the exclusion of certain Ideas from the eidetic cosmos. Once again, axiological considerations cannot play a decisive role in metaphysical matters: indeed, the young Socrates does not provide any rational argument for his denial of Ideas of base things. Hence, although Plato clearly sees the problem with mixing axiology and metaphysics, in Aristotle’s eyes he fails to consolidate this intuition by offering a metaphysics able to account for the differences between base and noble things.
To be sure, considerations of value and worth are at the core of the Timaeus, where the Demiurge is a good artisan and all that he does is valuable and dignified. Although Aristotle does not recognise final causes in the Timaeus, he agrees with Plato in rejecting the principles of nous, friendship and strife, for such principles are not final causes because they are not causes qua good.Footnote 33 There is indeed a strong sense in which one can say that Aristotle makes axiology part of his metaphysics: final causes are – and must be – good. There are also certain things that are better than others, and certain animals that are superior to others. However, their lesser worth does not mean that they are not worthy of close examination. In PA 1.5, 644b22–645a36, Aristotle states that natural things that do not undergo generation and corruption are of highest dignity (timias) and divine (theias). Nevertheless, the study of nature should still be directed towards animals and plants, which we can come to know better and which can also improve our understanding of ourselves. The study of animals is not an undignified pursuit (atimon). Moreover, the philosopher ought also to take an interest in the less dignified animals (atimoterôn) because their lesser worth does not imply that they are not marvellous works of nature. Just as in Plato base things, too, can be led back up to the Ideas, in Aristotle undignified living beings can also be shown to exemplify the beauty of the workings of nature. They might appear less attractive to the senses, but they might still be beautiful. It would be childish (paidikôs) to refrain from the investigation of certain animals on the basis of their unattractive looks, an admonition that echoes Parmenides’ explanation of Socrates’ denial of Ideas of base things in terms of the latter’s inexperience.
Aristotle develops this intuition further. The fact that some things are worse than others does not mean that they should be disregarded as they do not contribute to our understanding of reality. What is more, the fact that some things are better than others does not constitute an argument for their metaphysical superiority. This also applies when he introduces artefacts as inferior to living beings. In GA 2.1, he says:
Now some things are eternal and divine, others admit of both being and not being. But that which is noble and divine is always, in virtue of its own nature, the cause of the better in such things which admit of it, while what is not eternal admits of being and not-being, and of partaking in the better and the worse. And soul is better than body, and the living, because of the soul, is better than the lifeless, and being is better than not being, living than not living. (731b23–30)
Eternity is a trait of the noblest things, while things that partake of non-existence to a greater or lesser degree are more or less noble. The value ascribed to things depends upon their degree of eternity. While heavenly bodies are fully eternal, living beings only partake in eternity thanks to their ability to reproduce themselves. Non-living things are therefore ascribed less value, because they do not partake in eternity due to their inability to reproduce (i.e. to be eternal at least in species).Footnote 34 In DA 2.1, Aristotle refers to the common sense (dokousi) and states that the recognised substances are most of all (malista) bodies – specifically pointing to natural ones, because they are principles of the others (412a11–13).Footnote 35 Although the passage from DA does not introduce axiological considerations, it connects with the passage from GA in providing a scala naturae extending down from higher entities that are eternal or partake in eternity. Non-living things, ostensibly including artefacts, are situated at the bottom of the scale on account of their inability to produce, i.e. to be a principle, while being a principle is the way in which living beings partake in eternity. The highest worth is commonly ascribed to eternal things. Aristotle presents the common-sense view and, as is often the case, shows that there is some truth in it: living beings are, in fact, better than artefacts. If it seems that the heavenly bodies are more valuable than mortal beings, this is because eternity is the basis for this distinction. However, no matter how legitimate, axiology is not the reason why or the basis upon which living beings are said to be ontologically superior to artefacts. If anything, as we shall see in Chapter 5 (Section 5.4), beings that are ontologically superior might also turn out to be better according to common sense or to be eternal. As in PA, eternal beings have the highest worth. Eternal beings are substances not because they are better; rather, they are better because they are the highest causes. However, the question of whether both living beings and artefacts are substances and whether they are so to the same degree is kept separate by Aristotle from the question of their eternity or worth. These observations indicate that when Aristotle states that natural beings are better than artefacts, he is not using the remark as an argument about their respective metaphysical status. On the contrary, he is taking it as the result of their different metaphysical (and causal) status. The rest of the book will shed light on the other side of the coin: that when Aristotle does, in fact, make a metaphysical point about artefacts as opposed to natural substances, he does not rely on any axiological principles. Indeed, the very fact that Aristotle conducts a metaphysical analysis of artefacts is the result of his acceptance of Plato’s insight. If value judgements should not guide metaphysics, then one ought to go a step further and elaborate a metaphysics of base beings. Aristotle’s assessment of matter and artefacts responds to this exigency.
1.3.3 Carving Nature at the Joints
In the Statesman, Socrates and the Eleatic Stranger embark upon an investigation of the nature of the statesman, particularly as he stands in opposition to the sophist and the philosopher. Socrates the Younger and the Eleatic Stranger aim to elaborate a definition of the statesman through the method of division promoted in the Sophist. The Eleatic Stranger advises Socrates the Younger not to make rough divisions that do not capture ‘real classes’ (i.e. real eidê). One ought not to rush the dividing up of reality through broad divisions that do correspond to real classes along with them (262B). The advice is somewhat vague, and Socrates does not fail to draw attention to this, leading the Stranger to agree that ‘it is impossible to show what I mean with absolute completeness’. However, he goes on to provide an illuminating example. Dividing the human race into the Greeks, on the one hand, and all of the other races together, on the other, and then calling the latter by a single name (i.e. ‘barbarians’) does not distinguish one real kind or class. Notwithstanding the single designation ‘barbarians’, this class is not a real one since it does not represent a natural unit: the other races are, in fact, unlimited in number and do not even share the same language. The passage thus denies the existence of an eidos of ‘barbarians’. Another example proposed by the Stranger concerns numbers: dividing numbers into two classes – on the one hand, the class composed of the number 10,000 and, on the other, the class consisting in all other numbers taken together – would be a mistake, even if the latter class had a name, like ‘barbarians’. A more promising way of marking off the statesman is to look for real kinds via twofold divisions: in the case of numbers, a better division would be between even and odd; in the case of the human race, a better division would be between male and female. Of course, one might easily be at a loss in looking for real kinds. The temptation to divide the human race into Lydians, on the one hand, and everyone else, on the other – or Phrygians, on the one hand, and everyone else, on the other – may be indulged if the division into real kinds (genos) and parts turns out to be too arduous. Plato is laying out a methodological as well as ontological problem. He establishes that, if a division does not correspond to nature, it is not a correct division. What such a division carves off is not a real unit, but only a multiplicity lacking an eidos.
The discussion in the Statesman has an antecedent in the Phaedrus. There, Socrates speaks of the nature of two kinds of things. The first is seeing things that are scattered as belonging together and being able to collect them into a single kind. The second is managing to cut up each kind ‘according to its species along its natural joints’ (266A). In other words, the dialectician should bear in mind two questions. The first is at what point a multiplicity is, in fact, a unity; the second is whether a unity is the same as a natural unit. Working with these questions in mind is the dialectician’s task, and the dialectician should refrain from being like the bad butcher who splinters the parts of natural beings. ‘The beast is complex’, says the Stranger in the Sophist (226A8–C6) and, in order not to splinter its parts like a bad butcher, one needs to grasp the beast with both hands (i.e. through careful division). The famous Platonic metaphor of the butcher warns the speakers that dialectic concerns the transcendent eidê and that one needs to look for divisions that respect and correspond to reality and are therefore natural. A division is natural when it marks off a real kind (i.e. a real eidos). As Rep. X also establishes, there exists an eidos of every multiplicity that is a real unit.
Aristotle sees in this discussion something metaphysically promising but does not find in Plato a complete or satisfying formulation of – and hence solution to – what is actually a metaphysical problem. Aristotle builds on the problem of real kinds by accommodating artefacts. What is a real kind and how can one establish that a certain division corresponds to reality and respects nature? Plato does not provide an answer and the way in which he lays down the problem is vague enough to accommodate at least two different readings. First, Plato might consider a division ‘natural’ when it marks off whatever has a real existence. Second, Plato might consider a division ‘natural’ when it marks off whatever is given by nature and not by human beings. Regardless of whether he meant the first or the second option, one might still complain that Plato did not give any instructions as to how one can perform divisions in such a way as to be sure to mark off something natural. Aristotle, by contrast, demands instructions and takes Plato’s formulation to be unsatisfactory precisely because of its vagueness. Aristotle can be shown to engage with both possibilities, by means of a consideration of the status of artefacts. With regard to taking ‘natural’ as ‘real’, Aristotle discusses cases of conventional or merely linguistic phenomena and links them to considerations of unity. In Met. Z 4, proper essences belong absolutely (haplôs) and primarily (prôtôs) to things that are one and determinate, like substances, but only in some sense (pôs) and non-absolutely (ouch haplôs), to things like properties, which are not one, for they always imply something more (i.e. the subject they inhere in). The very possibility of a definition – as opposed to a formula – depends upon the unity of the object being defined. If the object represents a unity, it is amenable to proper definition. If one defines ‘pale man’ as pale, something is being omitted. But the attempt at defining ‘pale man’ as a unitary thing fails primarily because such a being has no essence. An indication of this is that one can never come up with an essential definition of ‘pale man’, but only with an account signifying it. The situation is no different if one replaces the label ‘pale man’ with a single name (e.g. ‘cloak’). We would not be defining something that is one and determinate, just because it has a single, determinate name. The attribution of a single name does not guarantee that the object is something determinate. Aristotle famously takes the Iliad not to be one and to exemplify the case of a single name that might mislead us into thinking that there is some one thing that the name captures. Because the Iliad does not constitute a single, unified thing, it has no proper essence and, hence, definition. Just as in Plato’s Statesman, the existence of the name ‘barbarian’ does not guarantee that the multiplicity of non-Greeks constitutes a real unit of which there exists a corresponding Idea, in Aristotle’s Metaphysics the existence of the name ‘Iliad’ does not guarantee that the collection of words to which it refers constitutes a real unit of which there exists a corresponding definition.
With regard to understanding ‘natural’ as referring to what is by nature in opposition to what is by art, Aristotle does have grounds for identifying what is ‘not natural’ with whatever comes-to-be through other principles (i.e. technê), but also chance and luck. This is what he does when he identifies principles in the Physics. What is natural is whatever has an inner principle of motion and rest. Whatever is not natural lacks an inner principle of motion and rest, while having an external one. A multiplicity constitutes a natural kind when it makes up a unity of things that have an inner principle governing their behaviour. A multiplicity constitutes an artificial kind when it makes up a unity of things that have an external principle governing their behaviour. Both sorts of kinds – natural and artificial – have objective existence in that they are equally real.
If what I am arguing is correct, Aristotle accepts the twofold advice of the Phaedrus: to ask oneself the questions: (i) ‘When is a multiplicity in fact a unity?’ and (ii) ‘Is a unity the same as a natural unit?’ To answer the first question, Aristotle draws on his theory of definition; to answer the second question, he elaborates distinctions that are fundamental for his physics and metaphysics. Chapter 3 engages with ‘natural’ in the sense of ‘by nature’, as opposed to artificial, while Chapter 6 (Section 6.1) deepens our understanding of real kinds as opposed to conventional kinds.
1.3.4 Parts and Whole
The second difficulty raised by Parmenides regarding Socrates’ support of the theory of Ideas concerns the concept of participation. What does it mean for particulars to participate in one Idea? Parmenides shows the difficulties involved in a mereological interpretation, according to which particulars have parts that somewhat overlap with a part of the whole (i.e. the Idea). If particulars are beautiful, because they possess a part that overlaps with a part of the Idea of Beauty, the Idea of Beauty would lose unity and would itself be many instead of one.Footnote 36 Moreover, understanding participation mereologically leads to contradictory results. For instance, if some particulars are small in virtue of their partaking, as parts, in the whole Idea of Small, the Idea of Small itself will turn out to be bigger (i.e. less small) than the particulars. In other words, smallness will better apply to the particulars than to the Idea of Small in itself, whose parts overlap with small particulars, but whose wholeness ought to be something more than the parts. Parmenides urges Socrates to come up with a definition of participation (i.e. the relation between particulars and the Idea) that does not become entangled in such contradictions. Socrates appears to agree with Parmenides that the mereological interpretation is unsuccessful and suggests understanding participation as imitation instead. Although the mereological interpretation fails, the theory of imitation also runs into difficulties – such as the Third Man Argument – and does not quite persuade the interlocutors: speaking about particulars down here, Socrates at 133D indeed adds ‘whether these are imitations or however one wants to call them’. The mereological interpretation thus opens a debate about the relationship between parts and whole, which the imitation theory cannot. In her book from 2002, Harte argues that Plato arrives at the negative result that wholes are not identical with their parts. In other words, Plato engages with the metaphysical problem of identity and composition, and leans towards a rejection of composition theory, according to which the whole just is its parts. However, Plato’s positive theory of what the whole is identical with (if it is not identical with its parts) turns out to be underdeveloped. The Parmenides (together with passages from the Sophist, Philebus and Timaeus) suggests that wholes are structures, with the parts depending for their identity conditions on the wholes that they belong to. Plato’s positive theory is not ‘yet a fully developed theory of composition’, but ‘an attempt to say what a whole of parts must be like’ (273). The positive mereology remains underdeveloped.
Despite the lack of explicit references, it is plausible that Aristotle is building on the problem of identity and composition and the relation between parts and whole as sketched by Plato. His teacher’s underdeveloped mereology left much to be desired.Footnote 37 There is, however, one specific point that Aristotle seems to address by means of an analysis of artefacts and that I would like to focus on. As has been mentioned, parts depend for their identity conditions on the wholes to which they belong. As Reference ShieldsHarte (2002, 161) puts it, parts ‘are not identifiable independently of the structure of the whole they compose’. The example given is that of a dinner party (whole), in which the identity of the guests (parts) essentially depends on the party they compose. Unfortunately, Plato does not provide further insights into the way in which parts are essentially dependent on the whole. There might be parts that also have an identity outside of the whole or parts that are not essentially dependent on the whole they compose. Precisely this difficulty lies at the core of Aristotle’s account of artefacts. To be sure, he makes sense of the relations between parts and wholes by means of his philosophical concepts of actuality and potentiality (as we shall see in Chapter 7). However, the most crucial ontological distinction between living beings and artefacts has to do with the identity of the parts within the whole, thus taking up and responding to this Platonic difficulty. As I shall show in Chapter 7, Plato fails to account for the fact that some parts are not essentially dependent on the wholes they compose, while Aristotle takes care not only to better qualify the notions of whole and parts, as well as their relationship, but also to clarify which parts are essentially dependent on the whole – and which are not – by appealing to the case of artefacts.
Open access