Ontology. The second assumption of ontological philosophy is about ontology. Ontology is, literally, the study of the nature of being (or existence), and what we shall assume is that ontology is a kind of explanation.

Ontological philosophy takes ontology to be a kind of explanation in which the causes are basic substances (along with their basic relationships to one another), and the effects are what is found in the world, or all the phenomena. Given the existence of certain kinds of basic substances and basic relationships, it explains the things found in the world by showing how their existence is constituted by such substances and relations among them.

If ontology is a valid kind of explanation, an adequate ontology should explain everything found in the world, for it is a theory about the nature of existence and what we mean by "the world" is everything that exists. To assume that ontology is a valid kind of explanation is to assume, therefore, that everything found in the world can be explained by showing how its existence is constituted by basic substances, given how they exist together as a world — including all the objects in the world, all their properties, all their relations to one another, and every way that they can change. It holds, in other words, that nothing exists, ultimately, but the basic substances.

The other way of doing philosophy is based on epistemology, and for epistemological philosophy, ontology is something quite different. Ontology is simply a thesis about what exists. Epistemologists base their claims about certain truths being necessary relative to our ordinary ways of knowing on a theory about how we know. Thus, they find themselves committed to the existence of entities of all the kinds that are known, including the entities presupposed by their foundation as well as all the additional entities entailed by their conclusions (assuming that they succeed in defending those conclusion). Since it is committed to the reality of additional entities of some kind, its ontology is called "realism." It is the belief in the "reality" of those additional entities. But since, as it turns out, they never fit together intelligibly with the entities constituting the epistemological foundation, realism is a form of ontological (or metaphysical) dualism that engenders skepticism. Hence, realists have always had to do battle with so-called anti-realists, who accept only the entities presupposed by their epistemological foundation. To mark how this view of ontology differs from ontological philosophy, let us call it "ontology as realism."

Ontology as a form of explanation. For ontological philosophy, ontology is explanatory. We assume that a certain kind of explanation is valid, which is to believe that there are causes and effects of certain kinds. In this case, the causes are the basic substances and their basic relationship to one another, and their effects are what they can constitute, which includes, if adequate, everything that can be found in the world, including all the objects, their properties and relations, and how they change over time.

Ontological causes. To see how such effects are produced ontologically, let us consider, first, the nature of the causes, both the substances and their relations, and, then, their effects.

Substances. Substances are one part of every ontological cause, and in order to explain how they help produce effects, we must consider both the nature of substance itself and a relevant difference among the kinds of basic substances that may be postulated by an ontology.

Nature of substance. Substance, we shall assume, has a nature that includes to two basic aspects. For something to be a substance, it must not only have a certain determinate nature, but must also be self-subsistent. That is, a substance must have, as a substance, both an essential aspect and an existential aspect to its nature.

Essential aspect of substance. A substance must have an essential aspect to its nature as substance, because in order to exist at all, it must exist in a determinate way. It is not possible for anything to exist without existing in a determinate way; indeterminate existence would be tantamount to nothing existing. The essential aspect of a substance includes all its kind-differentiating properties that do not change as time passes.

To assume that substance as substance has essential properties is not to assume that properties exist in addition to the substances that have them. We can and shall assume that properties are simply aspects of the substances themselves. Thus, essential properties are simply how substances exist, implying that substances can exist in different ways, as in substances being of different kinds. Beings like us can think about aspects of substances and distinguish their aspects from one another, and when we do, we are thinking about their properties. But ontological philosophy cannot answer questions about how rational beings have the ability to think about the aspects of substances as distinct from the substances themselves until it has explained the nature of what exist and the existence of beings in the world, like us, who can think at all. (See, for example, Change:AbstractObjects, or for a briefer statement of the entire theory about the nature of reason, Relations: Ontological theory of mathematical knowledge.)

We will take up the kinds of basic substances after explaining the nature of substance as substance.

Existential aspect of substance. Substances also have an existential aspect to their nature as substance. They must, because, in an ontological explanation of the world, it is the existence of substances (in certain relations) that explains the existence of what is found in the world. Substances are, in other words, self-subsistent.

Existence is, therefore, a property of substance as substance, just as having an essential aspect is. But in both cases, these aspects of substances have to do with their having aspects. The essential aspect is that they have an aspect of the kind we will call their "essential nature," and the existential aspect is that what has such an essential aspect exists independently of the rational being who know about them. That there are aspects of substances that have to with their having aspects is no more puzzling than that they have aspects at all and is answered in the same way, as we shall see, by the ontological explanation of the nature of reason. (See Stage 9, Rational Spiritual Animals, under Reproductive Global Regularities under Change.)

There are, however, two aspects to the existential aspect of the nature of substance as substance: particularity and temporality.

Particularity. First, substances are self-subsistent in the sense each substance has an existence that cannot be reduced to the existence of any other substance or substances in the world. Each substance exists on its own. That is not to say that substances must be able to exist even if all the other substances were to drop out of existence. (For example, it may not be possible for material substances, given their essential nature, to exist without having spatial relations to other material substances.) It is merely to say that there is something in the world whose existence would not be accounted for if only all the other substances in the world were assumed to exist. In short, each particular substance has an existence that is distinct from every other substance in the world.

We must accept that substances are related to one another in one way, at least, since we are assuming that there is more than one substance in the world. By "the world," we mean everything that exists, and thus, if there is more than one substance in the world, the world is a whole composed of parts. Since every substance is, by virtue of the existential aspect of its nature as a substance, something that exists, each substance is a "particular" substance in the further sense of "being part of the world." Each substance has a relationship to the world as a whole, and since it has an existence that is distinct from every other substance in the world, it also has a relationship to the other substances as a different part of one and the same world with them. In other words, when we postulate basic substances, we assume that they are parts of one and the same world.

There is another relationship that all substances have, namely, being identical to themselves. Relationships, like properties, are not something in addition to what has them, but merely an aspect of the substances that have them. And we continue to put off discussing how beings like us know about relationships until we explain the nature of reason ontologically. Although identity is a relationship, it is a relationship that something has to itself, and thus, it may be considered another aspect of each substance taken separately, like its properties. That is, each substance is identical to itself.

By the way, this is to assign ontological meaning to each of three basic senses of "is." "Is" can be used to say that something exists, and in that sense it refers to the property of existence, or the existential aspect of substance as substance. "Is" can also be used as a copula, to attach a predicate to a grammatical subject. In this case, it is referring to the relationship between a substance and some aspect of it, either a property that characterizes its essential nature or one that characterizes a changeable aspect of it (such as the roundness of a piece of wet clay). Finally, "is" can be used to assert identity. When identity is asserted of two substances, it says that the two substances have the same relation to one another as each has to itself, that is, that they are identical. But when identity is asserted of aspects of substances, that is, of properties, it has a different meaning, because different substances can have the same aspects and be of the same kind under each aspect. For example, all substances have the existential aspect, and "being" is the same property in each case. Likewise, substances of the same kind have the same essential properties. It will be possible to keep track of which properties are identical and which are different, because one thing an ontology provides by explaining everything in the world is an inventory of all the aspects of substances.

Temporality. Second, we assume that substances are self-subsistent in a temporal sense. Substances do not go out of existence over time, nor do they come into existence. Thus, a substance that exists at one moment must have existed at the previous moment. And it will continue to exist the next moment. Thus, if a substance exists at all, it exists at every moment in the history of world. It is permanent. The substances that exist at any one moment are the same substances that exist at every other moment in the history of the world.

This is a strong assumption to make about the nature of substance as substance, and it is not one that has always been made, even by naturalists.

According to Aristotle, for example, substances come into existence and go out of existence over time in a process of generation and corruption, though he did assume that they also had "material causes," or matter, that endures through change.

Other naturalistic ontologists do not postulate substances at all, but only "tropes," or properties considered as particular entities. Though tropes are supposed to explain everything in the world, they are not substances in our since, for they are supposed to come into existence and go out of existence at determinate locations in space from moment to moment. See Williams.

Though ontological philosophy makes this strong assumption about the temporal aspect of the existential aspect of substance as substance, there is an issue about the temporal aspect that we will leave open for the time being. To hold that substances never come into existence nor ever go out of existence over time is to presuppose that they are in time. That is, time is built into the nature of substance, as part of the existential aspect of the nature of substance as substance. But there are two different views about the nature of time and how it is related to existence. One is the "endurance" theory and the other is the "perdurance" theory.

Endurance theory of time. The first view holds that substances endure through time. This theory assumes that existence itself is in time. That is, only the present exists. The past and the future do not exist. Thus, for a substance to exist at all is for it to exist at the present moment. This view is also called "presentism." But since substances never come into existence, every substance must have existed at every past moment in the history of the world. And since they never go out of existence, every substance will still exist at every future moment in the world’s career. In other words, substances are identical through time: each substances that exists now is identical to some substance that existed or will exist at every other moment in the history of the world.

Since endurance theory assumes that the past and the future do not exist, they must explain the sense in which statements about the past and the future are true. It holds that such statements are true of substances that exist now, though the properties being ascribed to them have to do either with what has happened or with what will happen to them. That is, the aspects of substances which exist now include the states they had in the past and the states they will have in the future.

Perdurance theory of time. The other view is that substances perdure across time (or over time). Instead of assuming that existence is in time, this theory holds that time is a relation that holds among parts of substances. On this view, the past and the future exist in the same sense as the present. Though perdurance theorists can agree that substances never come into existence nor go out of existence over time, what they mean is that each substance is made up of a continuous series of moments stretching all the way back and all the way forward in the temporal dimension. Thus, instead of seeing substances as identical through time, they see substances as involving a part-whole relation: each substance is a whole whose parts include its state at every moment in its history. Thus, corresponding to the part of each substance that exists at any one moment, there is another part at every other moment in the history of the world.

The distinction between the endurance and perdurance theories about the existential aspects of substance as substance can be traced to McTaggart, who argued around the turn of the twentieth century that it is self-contradictory to hold that only the present exists. But recently, it has been resurrected by analytic philosophers defending the so-called "tenseless theory of time," as opposed to the "tensed theory of time". (The tenseless theory holds that statements about the past, present and future can all be translated, without any loss of content, into sentences about the relations of moments in time that hold eternally, whereas the tensed theory insists that some content is lost, namely, what they imply about which moment is actually present, that is, not just present relative to some particular time of utterance. See Oaklander and Smith.) And even more recently, the perdurance theory has been defended, albeit without admitting it, as what is called "four-dimensionalism" against "three dimensionalism." (See Sider.) But the reason I leave the issue open here is because a similar view is currently accepted by naturalists who are trying to be realists about the notion of spacetime introduced by Einstein’s special and general theories of relativity. Spacetime taken ontologically entails the perdurance theory.

Since purdurance theory assumes that all moments in the history of the world are ontologically equivalent, it holds that statements about the past and the future are true in exactly the same sense as statements about the present. There is no need to hold that statements about the past and the future are really about substances that exist now.

Whatever the relationship between time and existence, the temporal aspect of the existential aspect of substance involves a relationship between moments in time. Everyone agrees that moments occur in a continuous series, though endurance theorists think of time as flowing from the past into the future, and perdurance theorists think of time as just an order about the moments that all exist. But since endurance theorists take existence itself to be in time, they take time to be as ontologically basic as existence and substance, and thus, they take temporal relations to be a measure of the separation between different moments in the existence of a substance that is identical over time. Perdurance theorists, on the other hand, take all the moments in the history of a substance to exist in the same way, and thus they explain time, in effect, as how these moments exist together as a substance in the world.

The difference between these theories can be seen in what they imply about change. In a world where substance is permanent, what changes are the properties or relations of substances, or aspects of them. Endurance theory holds that change involves properties or relations coming into existence or going out of existence over time, because if the future and the past do not exist, there is no "place" for them to come from or to go to. On the other hand, perdurance theory holds that properties and relations never come into existence and never go out of existence, because if the future exists, the properties and relations already exist before the change takes place. And if the past exists, the properties and relations continue to exist after the change is long over.

Kinds of substances. The substances that an ontology postulates are the causes by which it explains the world. But in order to explain completely what is found in the world, those substances must be the most elementary substances that constitute the existence of things in the world. Let us call such ultimate parts of the world "basic substances."

All substances have, as substances, the same kind of existential aspect, but the essential aspects of their natures may be different. Thus, there may be different kinds of basic substances making up the world. But it is important to recognize at the outset that the essential natures that distinguish kinds of basic substances from one another may be either temporally simple or temporally complex.

We are assuming that the properties that characterize basic substances are simply aspects of them. The properties that characterize the essential nature of a substance are aspects of the essential aspect of their nature as substance, and they distinguish one kind of basic substance from another. Such essential properties do not change over time.

Temporally simple. Now, a substance that exhibits its full nature at each moment is a simple substance. That is, a substance will be said to have a "temporally simple essential nature" insofar as its essential properties are aspects of it that exist complete at each moment in the history of its existence. The contrast to complex substances will make this clear.

Temporally complex. The essential nature of a substance may also be defined by how its properties change over time. Properties that can change over time are contingent (or "accidents"), but if contingent properties always change in the same way, the way in which they change may be an essential property. For example, the properties a substance exhibits at one moment may depend on the properties it had the previous moment (together with its relations to other substances), and since the regularity about how they change would be a property that the substance has at every moment, it would be an essential property of the substance. But its essential nature would be dispositional. Insofar as the essential aspect of the nature of a substance is defined by a regularity about how its contingent properties (or relations) change over time, it will be said to have a "temporally complex essential nature."

Relations. Substances are only one part of every ontological cause. The other part is the relationship that holds among the basic substances. Relations are necessary for ontological explanation, because substances have nontrivial ontological effects only by working together, that is, by combining with one another in some way to constitute the existence of things found in the world. What makes ontological explanation explanatory is that substances can work together in different ways to produce different effects.

We must assume, therefore, that there is more than one substance in the world. Though it is conceivable that the world is made up of a single substance, nothing in such a world could be explained ontologically, in our sense, for everything found in such a world would be the same as what is assumed by the ontology in postulating that single substance.

Spinoza was not, therefore, giving an ontological explanation of the world in our sense, because according to his Ethics, he assumed that a single substance makes up the entire world.

If there is more than one substance in the world, they must have, as we have noted, at least one basic relationship to one another, for they are parts of the same world. Since their combination causes the world to exist, that relationship together with the substances might be said to explain the world. But if having such a relationship did account for everything in the world, it would be trivial, for nothing that is contained in any one of the ontological causes is really explained. It is merely assumed.

Finally, if the substances in the world had no further relationship to one another, beyond being different parts of the same world, they could not combine to constitute anything, except for the world as a whole. Though each substance might be said to cause itself ontologically (because it would still constitute its own existence), that would explain nothing, for its existence is precisely what is assumed in postulating the substance. It too would be trivial.

Nature of relations. We must assume, therefore, that basic substances have relationships of some kind to one another (beyond simply being parts of the same world). That is not to assume that relationships are something that exist in addition to the substances that have them. We can and will assume that the basic relationships are simply how basic substances exist together as a world. For example, bits of matter may be assumed to have spatial relations to one another as how they exist together as a world; or bits of matter may be assumed to exist together with space as a substance by coinciding with some part of space or other; and parts of space may be assumed to exist together as a world by having unchanging geometrical relations to one another. Such basic relationship are like properties, which, as we have assumed, are simply aspects of substances. But instead of being aspects of substances taken separately, the relationships we are assuming are aspects of the world, or how substances exist together as a world.

The basic relationships among substances being postulated as part of the ontological causes to be used in explaining everything in the world should be distinguished from the two relations, already mentioned, which substances have to themselves or among their parts: the identity relation and temporal relations. We are considering the relationships that an ontology must postulate along with substances in order to explain things ontologically, whereas the identity relation and temporal relations are aspects of how each substance exists on its own and do not depend on how they exist together as a world.

Since naturalism is the belief that what exists is just what is in space and time, one kind of basic relationship that any naturalism will require among substances is spatial. It is hard to see how any substance could be in space and time without having spatial relations to other substances. By spatial relations, I mean the distances that can hold between substances in three independent dimensions, and I assume that such distances are continuously variable.

Though spatial relations are found in the natural world, that does not mean that a naturalistic ontology must assume that having spatial relations is how substances exist together as a world. There is another way of existing together that would entail their having spatial relations: if space is a substance, bits of matter could have spatial relations by coinciding with parts of space. The real nature of spatial relations is another issue that we will leave open for the time being, until we are in a better position to decide what to believe. (See Space under Spatiomaterialism.)

Kinds of relations. As in the case of substances, there is an important difference to be recognized between kinds of basic relations that might be assumed to hold among the substances postulated. Though such basic relationships are just how the basic substances exist together as a world, they can, like the essential aspects of substances, be either temporally simple or temporally complex.

Temporally simple. Relations that exhibit their full nature at the moment that the substances exist together in that way are temporally simple. That is, relations are "temporally simple" to the extent that they are how substances exist together at a single moment in the history of the world. In a world constituted by space and matter, for example, the basic relationship between the two basic substances would be simple in this sense, for it would be true at every moment that each bit of matter coincides with some part of space or another.

Temporally complex. The relations that exist fully at any one moment may, however, change the next moment. That is, some relations may go out of existence over time and other relations come into existence. Such relations would be contingent, and the only way to define the basic relations by which substances exist together as a world may be the way in which contingent relations change over time. If change in contingent relations were regular, the way that substances exist together as a world might be defined by how their contingent relations change, for that would be a relationship that does not change over time. That is, the relations among substances might be dispositional. To the extent that the relationship by which substances exist together as a world have a nature that is defined by how contingent relations change over time, it will be said to be a "temporally complex relation."

For example, an ontology may assume that the way that substances exist together as a world is by having spatial relations. Particular spatial relations change over time, for example, as objects move, and the possibility of such change could be built into the the meaning of "having spatial relations." "Having spatial relations" might accordingly be defined as meaning that substances have spatial relations of some kind or other at each moment, but that they can change from one moment to the next as long as they are all geometrically consistent as a whole. "Having spatial relations" would then be a temporally complex relation among substances, and the substances themselves could have a relatively simple, inert nature.

It is possible to hold that spatial relations are temporally simple without postulating space as a substance. The change in spatial relations could be explained by the temporally complex essential natures of the substances, such as material substances defined as substances that move and interact according to the basic laws of physics. That is, everything that happens in the world, including all the spatial relations that come to exist, might be explained as what is required because material objects obey the laws of physics. What must be assumed is that those material objects had certain spatial relations at the beginning, say at the Big Bang or when God created the world. The spatial relations assumed by such an ontology could be temporally simple, for they could all exist fully at a single moment, at the very beginning. (It might be mentioned, however, that this view would not even be possible, given the Heisenberg uncertainty principle of quantum mechanics, unless there is a so-called hidden variable that makes the indeterminism of quantum theory a mere appearance of the incompleteness of its explanation.)

Ontological effects. Ontological explanations use substances as causes to explain things in the world as their effects. Such causes produce their effects by constituting the things being explained. Since there are relations among substances, different effects can be produced when basic substances are combined in different ways.

It should be emphasized, however, that insofar as the phenomenon being explained is the same as the substance that constitutes it, the explanation is trivial and, thus, not a genuine explanation at all. The explanatory power of ontology comes from showing how the substances cited as ontological causes work together so that jointly they constitute what is being explained. Thus, even if the existence of some object is explained by showing how it is constituted by the combination of various particular substances, the object's properties are still not explained if they are simply the essential properties of the basic substances constituting it. For example, it does not explain why something is moving in a certain direction to say that all its parts are moving that way. The "explanation" in ontological explanations comes from showing how ontological causes work together to produce something that may seem different from them. Anything that is entailed by the essential natures of substances taken separately is not explained, but just assumed.

What is explained by ontological causes includes both the objects found in space and how they change over time.

Objects. The existence of particular objects can be explained by the substances constituting them. Substances have, as substances, an existential aspect to their nature, that is, they are self-subsistent, and the relations by which they exist together as a world permit them to work together in constituting objects. How they do so depends on the specific ontology.

Likewise the natures of objects found in the world, or their properties, can be explained by the substances constituting them because of the essential aspects of their natures as substances, that is, their essential properties, and the relations by which they exist together as a world permit substances to be combined in different ways.

Thus, it is possible to explain a diversity of things in the world. Things may be different in kind because they are constituted by different kinds of basic substances combined in the same way, or because they are constituted of the same kinds of basic substances combined in different ways, or because of some combination of both factors.

If only out of respect for the Pre-Socratic philosophers, it should be noted that the attempt to explain the world ontologically was first attempted about 600 BC, before epistemological philosophy began. These first philosophers were naturalists looking for the "first principle" (or arche) by which to explain the natural world, and they assumed that it must be a "stuff" of some kind that constitutes the existence of everything in the world. Thales thought it was water. His student, Anaximander, insisted it was an inchoate stuff ("apeiron") without properties of its own. And Anaximander's student, Anaximines, argued for it being air. Though these so-called "Ionian" Pre-Socratics disagreed about its essential nature, they all agreed that the world is constituted by only one basic kind of material substance. Their ontologies were forms of monistic materialism. Spatial relations were taken for granted.

As the Pre-Socratics soon discovered, however, none of these ontologies offered an adequate explanation of the natural world, for they could explain neither the diversity of the objects in nature nor the change that occurs in them. The only properties postulated by any of them were those that characterize the essential nature of the single kind of material substance making up the world, and that left unexplained all the properties that distinguish one kind of object from other kinds, not to mention how such properties could come or go from existence as time passes.

Parmenides can be read as making this point. What Parmenides was referring to by his famous dictum. "What is, must be, and what is not, must not be," was a basic aspect of the nature of substance (the temporal aspect of its existential aspect). Substance cannot go out of existence, nor can it come into existence. But since Parmenides agreed that the "first principle" for explaining the world is a single kind of substance (with a temporally simple essential nature), he argued that there cannot be any real change or diversity in the world. Thus, he insisted that change and diversity are an illusion.

Heraclitus drew the opposite conclusion from the assumption that there is only one first principle for explaining the natural world. But he took, as the first principle, change and diversity itself. That was, in effect, to deny that there is any such thing as substance underlying change or diversity. Since the essential natures of substances are defined by their properties, to take the change of properties as basic was to deny that properties are aspects of substances, for otherwise substances would have to be coming into and going out of existence as time passes. Though Heraclitus did assume that change and diversity are guided in a regular way by Logos (which is something like laws of nature), this is to read Heraclitus' famous claim that you cannot step in the same river twice as saying that what exists in the natural world is nothing but properties that change over time.

Between them, therefore, Heraclitus and Parmenides posed a dilemma for any explanatory ontology that would postulate only one basic principle to explain the world: either the first principle is a material substance of some kind and there is no change nor diversity, or else change and diversity themselves are the first principle, and there is no substance. The former fails to explain the natural world, and the latter abandons ontological explanation altogether.

Pre-Socratic philosophy was a process of posing hypotheses, criticizing them, and posing new hypotheses, and it discovered two ways of solving this dilemma.

Pluralists held that the world is constituted by more than one kind of material substance. That made it possible to explain diversity and change by the mixture and separation of different kinds of material substances each with a simple essential nature.

Empedocles postulated four basic substances, earth, air, fire, and water, and he explained the diversity and change of things in the world by their mixture and separation (according to the forces of "love" and "strife"). Anaxagoras gave the same kind of explanation, except that he postulated infinitely many different basic substances (or "seeds," as he called them). In both cases, the essential natures of the basic substances were defined in terms of their qualitative properties, such as hot and cold, wet and dry, and their mixture was supposed to account for all the other sensible qualities of objects. (It was probably the limited range of objects that could be explained by only four basic substances that led Anaxagoras to insist on infinitely many "seeds.")

The other solution to this dilemma was offered by the atomists, Leucippus and Democritus. They are said to have explained diversity and change "quantitatively", rather than "qualitatively," because they took spatial relations into account. They assumed that the material substances are atoms whose natures differ from one another only by their size and shape, and they explained the differences in kinds of objects not only by the shapes and sizes of their constituent atoms, but also by the spatial relations that hold among them. That forced the ancient atomists to believe, however, that the sensible qualities that objects seem to have are actually subjective, a view that was not generally accepted until the beginning of the modern era.

Change. In order to explain change, an ontology must not only assume that substances have a temporal aspect to their existential nature, but also that they can be combined in different ways at different times. In that case, as time passes, an object may change because some of the kinds of basic substances constituting it are exchanged, or because the relations by which the same basic substances are related in constituting it change, or because of some combination of such factors. But that is to assume that, in addition to having relations, the relations among basic substances are capable of change over time.

This is clearly what Empedocles was assuming in holding that the objects perceived in nature change because of the mixture and separation of elements, such as earth, air, fire and water. He took it for granted that they can move, explaining one kind of change by assuming the possibility of another, namely, motion.

The atomists, however, believed that it was necessary to explain how motion itself is possible. That is why they postulated the void as well as all the atoms. They are traditionally understood as having argued that bits of matter would not be able to move, if there were no void, because there would always be other bits of matter in the way. But if there were a void as well as the atoms, atoms would be able to move without obstruction, at least, until they collided with other atoms. However, since the void exists only where atoms do not exist, the void can be understood as a very subtle kind of material substances that atoms can displace more easily than other atoms. On that interpretation, atoms move through the void like fish through water, displacing a fluid-like substance which offers no resistance. We will return to their explanation of the possibility of change.

There is, it should be emphasized, no ontological explanation of change, if the change being explained is the same kind of change that the substances undergoing that change are postulated as having as part of their essential nature. Whether we are explaining objects and their properties or change in them, when cause and effect are the same, there is no ontological explanation, but only ontological assumption.

Explanatory ontology is, in sum, the attempt to reduce everything in the world to the various kinds of basic substances constituting them and the relations by which those substances exist together as a world. But that is explanatory only to the extent that the substances and their relations are more elementary than what they explain and produce those effects by how they are combined. But if it were successful, an ontological explanation of the world would be a simple and complete explanation of the world, for it would show how everything in the world is identical to certain basic kinds of substances and certain basic kinds of relations among them. Everything in the world would be explained in the same way.

Ontology as realism. For traditional, epistemological philosophy, ontology is realism (or, more precisely, its ontology is determined by the position it takes on realism). The foundation of epistemological philosophy is a theory about how we know (or a theory about the nature of reason) which is based on reflecting on our mental processes. From this foundation, it attempts to justify certain conclusions about the world, which would be necessary relative to our ordinary ways of knowing about it. Thus, success generally means that it is committed to the existence of certain entities beyond those assumed at the beginning. "Realism" is the name for belief in their reality. But realism is usually a form of dualism. Epistemologists are already committed to the existence of the subject whose way of knowing is the foundation for their epistemological argument, and realism commits them to the existence of entities of a fundamentally different kind. Hence, they wind up defending some form of ontological dualism, and that typically leads to anti-realism, since the two kinds of substances do not fit together intelligibly as a world. This pattern can be found in every era of the history of Western philosophy. I will suggest how, very briefly, in order to make clear what I mean.

Ancient. Reflecting on the difference between the objects of perception and the objects that seem to be present to us in reasoning about kinds of things, Plato argued that, in addition to all the visible objects in the realm of Becoming, there is a realm of Being where such objects of rational intuition exist as unchanging Forms. He called the latter realm "Being" because the Forms were supposed to be permanent and unchanging. It was supposed to be outside space and time, beyond the natural world of changing, visible objects. Thus, his realism committed him to believing in the existence of both Being and Becoming, and since they are so fundamentally different in their natures, his ontology is clearly a kind of dualism.

Plato’s was a very problematic dualism, because it is hard to explain how entities that are not supposed to be in space and time are related to visible objects which are, much less to show how such Forms could cause visible objects to have the natures they seem to have. That makes it easy to be skeptical about the transcendent realm of Being, and naturalists are already inclined to be anti-realists about abstract entities of any kind, because they assume that everything is located in space.

Aristotle tried to avoid these problems by postulating, instead, substances in the natural world that are compounds of two elements, matter and form. This was not, however, to abandon Plato’s epistemological foundation, for Aristotle continued to assume that the "material cause" is an object of perception and that the "formal cause" is an object of rational intuition. Though essential forms were located in space, they had to have a peculiar nature to play their role, because each had to be located in many different particular substances at the same time and yet be one and the same thing. That earned them the name "universals." Though Aristotle could claim to be a naturalist, he was still a realist about essential forms as something beyond what is known by perception. That landed him with his own ontological dualism because, even though neither matter nor form can exist without the other, the existence of one is distinct from and cannot be reduced to the existence of the other. Realism about universals invited a type of skepticism called "nominalism."

Attempts to avoid matter-from dualism characterize Aristotle’s later work on the nature of substance as substance. Though there is much dispute about it, Aristotle seems to argue in Metaphysics, Books VII and VIII, that substances are basically just essential forms. He apparently reduces the material cause to the fact that forms exist only as particular substances despite being entities that exist as many different particular instances of the same form (that is, as universals). That position seems to reduce matter to a principle of individuation. This later notion of essential form and matter is closer to the distinction between essence and existence assumed here (see Substances above). In any case, Aristotle's conception of being as being (that is, substance as substance) poses so many problems that many traditional philosophers have been inclined to avoid ontology altogether.

Medieval. In the Medieval period, realism took the form of belief in the existence of God, rather than a realm of Being, outside space and time. Theists believed that it was possible to prove the existence of God on the basis of what can be observed in the natural world. For example, they argued from the natural belief that every event has a cause to the existence of God as the first cause, or cause of nature as a whole. And they argued from natural teleology to God, both as the designer of the natural order and as the ultimate final cause of natural things. Realism about God, or theism, committed them, therefore, to believing in the existence of God as well as nature. After Augustine, this ontological dualism was modeled on Plato’s, and it was no less problematic. The fundamental difference in their natures makes it difficult to explain how God and the natural world are related as parts of a single world. It was ultimately left as a mystery that could not be fathomed by finite rational minds. Denial of this kind of realism is generally considered atheism, though mere skepticism about it is often distinguished as agnosticism.

Modern. With the rise of modern science, it was recognized that our perceptual experience of the natural world is something distinct from the natural world itself (as the ancient atomists first held), and the foundation of epistemological philosophy shifted from reflection on how we know in which we are living bodies in the natural world to reflection on how we know in which we are minds where ideas have an appearance. Mind is the epistemological foundation from which Descartes tried to prove the existence of the body and the external world of which it is part. The success of Cartesian philosophy would entail realism about the natural world, and thus ontological dualism. But mind and body are substances with such radically different natures that it is, once again, a very problematic ontology, namely, mind-body dualism.

There were, of course, skeptics about its success, notably, the British Empiricists, and they are interesting for their views about substance. Locke argued that realism about material objects involves belief in a substratum, or substance as nothing but a support of the properties that perception reveals objects to have. Since that was to believe that substances have no properties of their own, it was, in effect, to reduce substance as substance to its existential aspect, and thus, Locke could plausibly hold that substratum is an incoherent idea. But even the existential aspect was denied by Berkeley and Hume. They accepted the "bundle theory" of substances, that is, the view that substances are just the bundle of properties that we seem to perceive in them. In any case, since the foundation of modern philosophy was mind, they were implicitly committed to one kind of substance, and the only ontological position open to skeptics was idealism of some kind or other, though only Berkeley embraced it explicitly.

Later attempts to justify science from the epistemological foundation of modern philosophy led to other forms of realism, though they were not called that. Kant tired to avoid the problems of Cartesian philosophy by holding that space and time are merely forms of intuition in the mind. But since he continued to believed that there are things in themselves, he was implicitly committed to entities that are not in space and time. That landed him with the same kind of problematic ontological dualism as Plato, and like Augustine, he simply denied that it is possible to explain the relationship between the natural world and the things in themselves which are outside time and space.

Contemporary. Early in the twentieth century, developments in logic by Russell and Frege offered a new foundation for epistemological philosophy. Reflecting on our use of language, so-called Anglo-American analytic philosophy took as their epistemological foundation what we all know about the meanings and references of the terms and sentences we use. This foundation has been used in various way, leading to different forms of realism.

Analytic philosophy was able to reformulate empiricism as a justification of science at the expense of modern metaphysics. Logical positivists took the observation of objects in the natural world as the epistemological foundation of science, and they tried to show how scientific conclusions were supported by it. Though their original purpose was to show that whatever is not based on observation is meaningless metaphysics, it was soon noticed that even theories in physics mention unobservable entities, such as electrons, quarks, and force fields. Thus, those who believed in their existence came to called "realists about theoretical entities."

More recently, the recognition that such unobservable entities are not very different from the observable objects on which science bases its theories has led to calling the defenders of science "scientific realists." Scientific realism is taken to involve a commitment to the existence of both the observable and unobservable objects recognized by science. Or in the words of Wilfred Sellars, "science is the measure of all things, of what is that it is, and of what is not that it is not" (p. 173). But disputes still rage in the professional literature about the significance of calling it "realism."

Most recently, philosophers of science have tried to avoid problems about realism by simply abandoning traditional epistemology all together. They often call themselves "naturalized epistemologists," for they hold that the only foundation for justifying science is science itself (that is, the conclusions that science draws about how we know). Though they say that they believe that philosophy is continuous with science, to ontological philosophy, they seem to be giving up philosophy altogether in favor of being cheerleaders for science. See Kitcher and Rosenberg.

Giving up epistemological philosophy does not necessarily mean, however, taking up ontological philosophy. The habit of epistemology makes it seem that ontology is purely descriptive. The job of ontology seems to be just to discover the kinds of entities to which one is committed by holding certain beliefs to be true.

With regard to natural science, for example, ontology is just realism about the conclusions of science.

In the philosophy of mathematics, realism is defended by so-called Platonists, who hold that numbers and other mathematical entities exist independently of the subjects who know about them (in opposition to logicists, who argue that mathematics can be reduced to logical truth, and to constructivists, who argue that mathematical objects are simply constructs of the imagination).

Even language is taken as a foundation for descriptive ontology. Quine (1953, 1960) has argued that talk of classes implies the existence of at least that one kind of abstract entity. Some analytic philosophers now argue that to believe in the truth of descriptive statements is to be committed to the existence of properties as well as the substances that have them, or what might be called substance-property dualism.

Scientific realism leads some analytic philosophers of science to take laws of nature to be real, which entails a dualism of laws and the objects that that obey them.

In any case, realism is not explanatory ontology, but just ontology as realism. It does not use the entities it postulates to explain anything beyond the phenomena on which their existence was defended. That leaves plenty of room for philosophical argument, because descriptive ontologists generally take a skeptical attitude and are inclined to deny the existence of any kinds of entities whose existence is not forced on them by their epistemological foundation. But that is a different issue entirely from explanatory ontology.