Categories of Being and Universals metaphysical
We human beings sort things into various classes. And we often suppose that the classes into which we sort things enjoy a kind of internal unity. In this respect they differ from sets in the strict sense of the word. (And no doubt in others. It would seem, for example, that we think of the classes we sort things into—biological species, say—as comprising different members at different times.) The classes into which we sort things are in most cases “natural” classes, classes whose membership is in some important sense uniform—“kinds”. We shall not attempt an account or definition of ‘natural class’ here. Examples must suffice. There are certainly sets whose members do not make up natural classes: a set that contains all dogs but one, and a set that contains all dogs and exactly one cat do not correspond to natural classes in anyone’s view. And it is tempting to suppose that there is a sense of “natural” in which dogs make up a natural class, to suppose that in dividing the world into dogs and non-dogs, we “cut nature at the joints”. It is, however, a respectable philosophical thesis that the idea of a natural class cannot survive philosophical scrutiny. If that respectable thesis is true, the topic “the categories of being” is a pseudo-topic. Let us simply assume that the respectable thesis is false and that things fall into various natural classes—hereinafter, simply classes.
Some of the classes into which we sort things are more comprehensive than others: all dogs are animals, but not all animals are dogs; all animals are living organisms, but not all living organisms are animals …. Now the very expression “sort things into classes” suggests that there is a most comprehensive class: the class of things, the class of things that can be sorted into classes. But is this so?—and if it is so, are there classes that are “just less comprehensive” than this universal class? If there are, can we identify them?—and are there a vast (perhaps even an infinite) number of them, or some largish, messy number like forty-nine, or some small, neat number like seven or four? Let us call any such less comprehensive classes the ‘categories of being’ or the ‘ontological categories’. (The former term, if not the latter, presupposes a particular position on one question about the nature of being: that everything is, that the universal class is the class of beings, the class of things that are. It thus presupposes that Meinong was wrong to say that “there are things of which it is true that there are no such things”.)
The topic “the categories of being” is intermediate between the topic “the nature of being” and the topics that fall under the post-Medieval conception of metaphysics for a reason that can be illustrated by considering the problem of universals. Universals, if they indeed exist, are, in the first instance, properties or qualities or attributes (i.e., “ductility” or “whiteness”) that are supposedly universally “present in” the members of classes of things and relations (i.e., “being to the north of”) that are supposedly universally present in the members of classes of sequences of things. “In the first instance”: it may be that things other than qualities and relations are universals, although qualities and relations are the items most commonly put forward as examples of universals. It may be that the novel War and Peace is a universal, a thing that is in some mode present in each of the many tangible copies of the novel. It may be that the word “horse” is a universal, a thing that is present in each of the many audible utterances of the word. And it may be that natural classes or kinds are themselves universals—it may be that there is such a thing as “the horse” or the species Equus caballus, distinct from its defining attribute “being a horse” or “equinity”, and in some sense “present in” each horse. (Perhaps some difference between the attribute “being a horse” and the attribute “being either a horse or a kitten” explains why the former is the defining attribute of a kind and the latter is not. Perhaps the former attribute exists and the latter does not; perhaps the former has the second-order attribute “naturalness” and the latter does not; perhaps the former is more easily apprehended by the intellect than the latter.)
The thesis that universals exist—or at any rate “subsist” or “have being”—is variously called ‘realism’ or ‘Platonic realism’ or ‘platonism’. All three terms are objectionable. Aristotle believed in the reality of universals, but it would be at best an oxymoron to call him a platonist or a Platonic realist. And ‘realism’ tout court has served as a name for a variety of philosophical theses. The thesis that universals do not exist—do not so much as subsist; have no being of any sort—is generally called ‘nominalism’. This term, too is objectionable. At one time, those who denied the existence of universals were fond of saying things like:
Present-day nominalists, however, are aware, if earlier nominalists were not, that if the phrase ‘the name “horse” ’ designated an object, the object it designated would itself be a universal or something very like one. It would not be a mere puff of sound but would rather be what was common to the many puffs of sound that were its tokens.
The old debate between the nominalists and the realists continues to the present day. Most realists suppose that universals constitute one of the categories of being. This supposition could certainly be disputed without absurdity. Perhaps there is a natural class of things to which all universals belong but which contains other things as well (and is not the class of all things). Perhaps, for example, numbers and propositions are not universals, and perhaps numbers and propositions and universals are all members of a class of “abstract objects”, a class that some things do not belong to. Or perhaps there is such a thing as “the whiteness of the Taj Mahal” and perhaps this object and the universal “whiteness”—but not the Taj Mahal itself—both belong to the class of “properties”. Let us call such a class—a proper subclass of an ontological category, a natural class that is neither the class of all things nor one of the ontological categories—an ‘ontological sub-category’. It may indeed be that universals make up a sub-category of being and are members of the category of being “abstract object”. But few if any philosophers would suppose that universals were members of forty-nine sub-categories—much less of a vast number or an infinity of sub-categories. Most philosophers who believe in the reality of universals would want to say that universals, if they do not constitute an ontological category, at least constitute one of the “higher” sub-categories. If dogs form a natural class, this class is—by the terms of our definition—an ontological sub-category. And this class will no doubt be a subclass of many sub-categories: the genus canis, the class (in the biological sense) mammalia, …, and so through a chain of sub-categories that eventually reaches some very general sub-category like “substance” or “material object”. Thus, although dogs may compose an ontological sub-category, this sub-category—unlike the category “universal”—is one of the “lower” ones. These reflections suggest that the topic “the categories of being” should be understood to comprehend both the categories of being sensu stricto and their immediate sub-categories.
Does the topic “the categories of being” belong to metaphysics in the “old” sense? A case can be made for saying that it does, based on the fact that Plato’s theory of forms (universals, attributes) is a recurrent theme in Aristotle’s Metaphysics. In Metaphysics, two of Plato’s central theses about the forms come in for vigorous criticism: (i) that things that would, if they existed, be “inactive” (the forms) could be the primary beings, the “most real” things, and (ii) that the attributes of things exist “separately” from the things whose attributes they are. We shall be concerned only with (ii). In the terminology of the Schools, that criticism can be put this way: Plato wrongly believed that universals existed ante res (prior to objects); the correct view is that universals exist in rebus (in objects). It is because this aspect of the problem of universals—whether universals exist ante res or in rebus—is discussed at length in Metaphysics, that a strong case can be made for saying that the problem of universals falls under the old conception of metaphysics. (And the question whether universals, given that they exist at all, exist ante res or in rebus is as controversial in the twenty-first century as it was in the thirteenth century and the fourth century B.C.E.) If we do decide that the problem of universals belongs to metaphysics on the old conception, then, since we have liberalized the old conception by applying to it the contemporary rule that the denial of a metaphysical position is to be regarded as a metaphysical position, we shall have to say that the question whether universals exist at all is a metaphysical question under the old conception—and that nominalism is therefore a metaphysical thesis.
There is, however, also a case to made against classifying the problem of universals as a problem of metaphysics in the (liberalized) old sense. For there is more to the problem of universals than the question whether universals exist and the question whether, if they do exist, their existence is ante res or in rebus. For example, the problem of universals also includes questions about the relation between universals (if such there be) and the things that are not universals, the things usually called particulars. Aristotle did not consider these questions in the Metaphysics. One might therefore plausibly contend that only one part of the problem of universals (the part that pertains to the existence and nature of universals) belongs to metaphysics in the old sense. At one time, a philosopher might have said,
The universal “doghood” is a thing that does not change. Therefore, questions about its nature belong to metaphysics, the science of things that do not change. But dogs are things that change. Therefore, questions concerning the relation of dogs to doghood do not belong to metaphysics.
But no contemporary philosopher would divide the topics that way—not even if he or she believed that doghood existed and was a thing that did not change. A contemporary philosopher—if that philosopher concedes that there is any problem that can properly be called “the problem of universals”—will see the problem of universals as a problem properly so called, as a problem having the kind of internal unity that leads philosophers to speak of a philosophical problem. And the same point applies to the topic “the categories of being”: every philosopher who is willing to say that “What are the categories of being?” is a meaningful question will assign every aspect of that question to metaphysics
Let us consider some aspects of the problem of universals that concern changing things. (That is, that concern particulars—for even if there are particulars that do not change, most of the particulars that figure in discussions of the problem of universals as examples are things that change.) Consider two white particulars—the Taj Mahal, say, and the Washington Monument. And suppose that both these particulars are white in virtue of (i.e., their being white consists in) their bearing some one, identifiable relation to the universal “whiteness”. Suppose further that we are able to single out this relation by some sort of act of intellectual attention or abstraction, and that (having done so) we have given it the name “falling under”. All white things and only white things fall under whiteness, and falling under whiteness is what it is to be white. (We pass over many questions that would have to be addressed if we were discussing the problem of universals for its own sake. For example, both blueness and redness are spectral color-properties, and whiteness is not. Does this fact imply that “being a spectral color-property” is, as one might say, a second-order universal? If so, does blueness “fall under” this universal in the same sense as the sense in which a copy of Philosophical Studiesfalls under blueness?)
Now what can we say about this relation, this “falling under”? What is it about the two objects whiteness and the Taj Mahal that is responsible for the fact that the latter falls under the former? Is the Taj perhaps a “bundle” of universalia ante res, and does it fall under whiteness in virtue of the fact that whiteness is one of the universals that is a constituent of the bundle that it is? Or might it be that a particular like the Taj, although it indeed has universals as constituents, is something more than its universal constituents? Might it be that the Taj has a constituent that is not a universal, a “substrate”, a particular that is in some sense property-less and that holds the universal constituents of the Taj together—that “bundles” them? (If we take that position, then we may want to say, with Armstrong (1989: 94–96), that the Taj is a ‘thick particular’ and its substrate a ‘thin particular’: a thick particular being a thin particular taken together with the properties it bundles.) Or might the Taj have constituents that are neither universals nor substrates? Might we have been too hasty when we defined ‘particulars’ as things that are not universals? Could there perhaps be two kinds of non-universals, concrete non-universals or concrete individuals (those would be the particulars, thick or thin), and abstract non-universals or abstract individuals (‘accidents’ or ‘tropes’ or ‘property instances’), things that are properties or qualities (and relations as well), things like “the (individual) whiteness of the Taj Mahal”? Is the Taj perhaps a bundle not of universals but of accidents? Or is it composed of a substrate and a bundle of accidents? And we cannot neglect the possibility that Aristotle was right and that universals exist only in rebus. If that is so, we must ask what the relation is between the matter that composes a particular and the universals that inhere in it—that inhere simultaneously in “this” matter and in “that” matter.
The series of questions that was set out in the preceding paragraph was introduced by observing that the problem of universals includes both questions about the existence and nature of universals and questions about how universals are related to the particulars that fall under them. Many of the theories that were alluded to in that series of questions could be described as theories of the “ontological structure” of non-universals. We can contrast ontological structure with mereological structure. A philosophical question concerns the mereological structure of an object if it is a question about the relation between that object and those of its constituents that belong to the same ontological category as the object. For example, the philosopher who asks whether the Taj Mahal has a certain block of marble among its constituents essentially or only accidentally is asking a question about the mereological structure of the Taj, since the block and the building belong to the same ontological category. But the philosopher who asks whether the Taj has “whiteness” as a constituent and the philosopher who supposes that the Taj does have this property-constituent and asks, “What is the nature of this relation ‘constituent of’ that ‘whiteness’ bears to the Taj?” are asking questions about its ontological structure.
Many philosophers have supposed that particulars fall under universals by somehow incorporating them into their ontological structure. And other philosophers have supposed that the ontological structure of a particular incorporates individual properties or accidents—and that an accident is an accident of a certain particular just in virtue of being a constituent of that particular.
Advocates of the existence of ante res universals, and particularly those who deny that these universals are constituents of particulars, tend to suppose that universals abound—that there is not only such a universal as whiteness but such a universal as “being both white and round and either shiny or not made of silver”. Advocates of other theories of universals are almost always less liberal in the range of universals whose existence they will allow. The advocate of in rebus universals is unlikely to grant the existence of “being both white and round and either shiny or not made of silver”, even in the case in which there is an object that is both white and round and either shiny or not made of silver (such as a non-shiny white plastic ball).
The two topics “the categories of being” and “the ontological structure of objects” are intimately related to each other and to the problem of universals. It is not possible to propose a solution to the problem of universals that does not have implications for the topic “the categories of being”. (Even nominalism implies that at least one popular candidate for the office “ontological category” is non-existent or empty.) It is certainly possible to maintain that there are ontological categories that are not directly related to the problem of universals (“proposition”, “state of affairs”, “event”, “merepossibile”), but any philosopher who maintains this will nevertheless maintain that if there are universals they make up at least one of the higher ontological sub-categories. And it seems that it is possible to speak of ontological structure only if one supposes that there are objects of different ontological categories. So whatever metaphysics comprehends, it must comprehend every aspect of the problem of universals and every aspect of the topics “the categories of being” and “the ontological structure of objects”. For a recent investigation of the problems that have been discussed in this section, see Lowe (2006).
We turn now to a topic that strictly speaking belongs to “the categories of being”, but which is important enough to be treated separately.
Some things (if they exist at all) are present only “in” other things: a smile, a haircut (product, not process), a hole …. Such things may be opposed to things that exist “in their own right”. Metaphysicians call the things that exist in their own right ‘substances’. Aristotle called them ‘protai ousiai’ or “primary beings”. They make up the most important of his ontological categories. Several features define protai ousiai: they are subjects of predication that cannot themselves be predicated of things (they are not universals); things exist “in” them, but they do not exist “in” things (they are not accidents like Socrates’ wisdom or his ironic smile); they have determinate identities (essences). This last feature could be put this way in contemporary terms: if the prote ousia x exists at a certain time and the prote ousia y exists at some other time, it makes sense to ask whether x and y are the same, are numerically identical (and the question must have a determinate answer); and the question whether a given prote ousia would exist in some set of counterfactual circumstances must likewise have an answer (at least if the circumstances are sufficiently determinate—if, for example, they constitute a possible world. More on this in the next section). It is difficult to suppose that smiles or holes have this sort of determinate identity. To ask whether the smile Socrates smiled today is the smile he smiled yesterday (or is the smile he would have smiled if Crito had asked one of his charmingly naïve questions) can only be a question about descriptive identity.
Aristotle uses ‘(prote) ousia’ not only as a count-noun but as a mass term. (He generally writes ‘ousia’ without qualification when he believes that the context will make it clear that he means ‘prote ousia’.) For example, he not only asks questions like “Is Socrates a (prote) ousia?” and “What is a (prote) ousia”?, but questions like “What is the (prote) ousia of Socrates?” and “What is (prote) ousia?” (Which question he is asking sometimes has to be inferred from the context, since there is no indefinite article in Greek.) In the count-noun sense of the term, Aristotle identifies at least some (protai) ousiai with ta hupokeimena or “underlying things”. Socrates, for example, is a hupokeimenon in that he “lies under” the in rebus universals under which he falls and the accidents that inhere in him. ‘To hupokeimenon’ has an approximate Latin equivalent in ‘substantia’, “that which stands under”. (Apparently, “to stand under” and “to lie under” are equally good metaphorical descriptions of the relations a thing bears to its qualities and accidents.) Owing both to the close association of (protai) ousiai and hupokeimena in Aristotle’s philosophy and to the absence a suitable Latin equivalent of ‘ousia’ ‘substantia’ became the customary Latin translation of the count-noun ‘(prote) ousia’.
The question whether there in fact are substances continues to be one of the central questions of metaphysics. Several closely related questions are: How, precisely, should the concept of substance be understood?; Which of the items (if any of them) among those we encounter in everyday life are substances?; If there are substances at all, how many of them are there?—is there only one as Spinoza contended, or are there many as most of the rationalists supposed?; What kinds of substances are there?—are there immaterial substances, eternal substances, necessarily existent substances?
It must be emphasized that there is no universally accepted and precise definition of ‘substance’. Depending on how one understood the word (or the concept) one might say either that Hume denied that there were any substances or that he held that the only substances (or the only substances of which we have any knowledge) were impressions and ideas. It would seem, however, that most philosophers who are willing to use the word ‘substance’ at all would deny that any of the following (if they exist) are substances:
- Universals and other abstract objects. (It should be noted that Aristotle criticized Plato for supposing that the protai ousiai were ante res universals.)
- Events, processes, or changes. (But some metaphysicians contend that substance/event is a false dichotomy.)
- Stuffs, such as flesh or iron or butter. (Unfortunately for beginning students of metaphysics, the usual meaning of ‘substance’ outside philosophy is stuff. Aristotle criticized “the natural philosophers” for supposing that the prote ousia could be a stuff—water or air or fire or matter.)
The nature of being, the problem of universals, and the nature of substance have been recognized as topics that belong to “metaphysics” by almost everyone who has used the word. We now turn to topics that belong to metaphysics only in the post-Medieval sense.