CHAPTER XI

Things and Conditions

§1. The Incompatibility of Necessary Connections with an Ontology of Conditions. The concept of an entity that has components but is not itself a component has played a role at numerous stages in our discussion. The natures required by necessity were components of entities of such a kind, as were the actions required by time and the elements of fine structure required by capacities. But so far, little has been said that would tie the concept of an entity that has components but is not itself a component together with the concept of a thing. Since individuals and conditions are the same particulars as their components, components are like individuals and conditions in being particulars. But individuals and conditions are “complex” particulars since they have components, whereas properties, parts, and actions are “simple” particulars. Conditions, unlike individuals, are themselves the havings of components. Thus an “individual” is a particular that has components though it is not itself the having of any component. In this catalogue of kinds of entities, things fall under the heading of individuals. Traditionally, the concept of a thing has contained the concept of an enduring individual. But since we have carried the idea of time back to that of action, it will turn out to be more natural for us to analyze the concept so that a “thing” is an individual with a capacity to act. Things will, then, be different from simple particulars, which have no components and hence no capacities. They will also be different from non-physical individuals such as numbers, in that non-physical individuals cannot act. Finally, things are different from conditions both by being individuals, which have conditions but are not conditions, and by having capacities, which, as we shall see, conditions have only derivatively.

The question to be raised here is whether there is any call to add a further dimension–that of the ability to act–to the concept of an individual. Do we really need things in the required ontology? I shall argue that unless there are entities with capacities to act there can be no necessities across time. An ontology coherent with the practice of acting on prior experience must contain things. It was pointed out at the beginning of Chapter I that denials of necessary connections do not come from ontologies that admit things. In this section, I shall attempt to show that ontologies that admit conditions yet reject not just things but any kind of physical individual inevitably deny real necessities of any kind. In this attempt, and in the remainder of the chapter, the arguments will rely almost exclusively on principles established earlier. The point will be to draw consequences from heretofore isolated claims.

By an ontology of conditions I shall mean an ontology that admits conditions but not individuals and hence no things. As we noted in Chapter III,§1, conditions are sometimes called facts-of and sometimes events. For us a condition, fact of, or event is the having of a component by a particular that need not be a condition itself. Within an ontology of conditions a different view of conditions is required since there are no individuals, and hence no things, to have components. Nonetheless, the competing ontologies will refer to the same entities when they employ expressions such as ‘doing running’ and ‘having tallness’. The difference lies in the analysis of the entities referred to.

The ontology of conditions has received support from rather different quarters. On the one hand, the emphasis on the unceasing transformation of institutions and ideas in the dialectical conception of history has led some to suppose that belief in enduring particulars is incompatible with this conception of history.¹ Since things are one kind of enduring particular, they seem too static to merit inclusion in an ontology that gives history its due. If things are conceived of as inert and as making no contribution to their own endurance, then indeed things are incompatible with the universal striving toward quantitative and, ultimately, qualitative difference recognized by the dialectical conception of history. Once it is pointed out, however, that things are centers of action and that actions are components of things rather than entities distinct from things, there is no longer even an appearance of incompatibility between the relative permanence of things and their striving toward novelty. For then things endure precisely because they are active.

On the other hand, the substitution of space-time for space and time by special relativity theory has been thought to render an ontology of conditions “more natural” than an ontology including things.² Somehow the fact that, in the space-time view, spatial points are no longer enduring is supposed to confirm the theory that there are no enduring particulars. At best, however, there is only an analogy between that fact and this theory. From the perspective of our action view of time and hence, through simultaneity, of space, the analogy is used to argue in precisely the wrong way. It is used to argue from space to the content of space, rather than from the content to space. When we start with a consideration of enduring particulars, we find no incompatibility between them and the rejection of enduring spatial points.

The ontology of conditions is thus associated with two of our strongest cultural traditions. So to argue for things easily admits of misinterpretation as a manifestation of both an antihistorical and an antiscientific attitude. But I have just indicated, if only briefly, the faulty logic involved in attempts to reject an ontology including things by reference to the dialectical theory of history and the theory of special relativity. I wish now to show, first, that conditions without individuals can account for no necessities and, second, that conditions without things can account for no necessities across time. Thus, attempts like the above to reject things are incoherent with the practice of action on prior experience.

For the ontologist of conditions, the natures of individuals cannot be the bases of necessities, since individuals are rejected. The fundamental role played by individuals having components is taken over by conditions obtaining or having existence. This suggests that necessities will be necessities of conditions rather than of individuals. And so it will be the “natures” of conditions that base necessities. The view that a feature had by an entity is the nature of the condition that is the having of that feature is the only reasonable one to adopt, if conditions are to have natures at all. For a condition is nothing but the having of a component. If the having itself is its nature, conditions would all have similar natures.

Moreover, it seems clear that distinct conditions of having similar features will have similar natures. This requirement is easily satisfied if the features are the natures. It is important that conditions do not generally have the features of which they are the having. Otherwise, a problem arises about what grounds the necessity with which a condition has the feature that is its nature. (I speak here about features rather than components since there are now on individuals to have components.)

The counterpart of the claim in an ontology including individuals that □  a(Fa) will be the claim in the ontology of conditions that □  (a’s Fing)[a’s Fing obtains]. This is not to say that the claims are synonymous but only that they are paired between the ontologies. The counterpart of □  a(Fa → Gb) will likewise be □  (a’s fing)[a’s fing obtains→b’sGingobtains]. Just as, when □a(Fa), Fa follows from a’s having a certain nature, N, so too, when □  (a’s Fing)[a’s Fing obtains], the obtaining of a’s Fing follows from a’s Fing having a certain nature, F. Now given these interpretations, there are three important failures of the attempt to account for necessity in the ontology of conditions, each more general than the preceding one.

First, there are cases where a is necessarily F, though b hasF with a different modality. In the ontology that includes individuals, this happens because of dissimilarities in the natures of the individuals a and b. But in the ontology of conditions, the natures involved must be similar, since the conditions are havings of similar՛ features. Thus if □  (a’s Fing)[a’s Fing obtains], then it cannot be contingently true or impossible that b ’s Fing obtains, since the nature of a’s Fing and of b’s Fing is simply F. If a sodium atom necessarily combines with a chlorine atom to form a molecule, then rather than its being impossible (as it in fact is), it would have to be necessary that a fluorine atom combines with a chlorine atom to form a molecule. Moreover, if some atoms are naturally unstable, then all would have to be, even though we know some break apart only when subjected to high energies. So the ontology of conditions is committed to the view that similar features are had with the same modality. And this conflicts with evident modal differences among instances of similar features.

In view of this criticism, the ontologist of conditions will retreat to a view of necessity that requires universality. He will reject the view that there are necessities of some or many particulars that are not equally necessities of all particulars. A necessity that holds for a single particular will hold for all. This view has been widely held since Locke claimed that necessities are relative to classifications. But what are the consequences of joining it to the only account we found possible to give of induction? It undercuts the desired effect of principle (II) of Chapter V, §1, the principle of necessary basic properties. That principle leads to the conclusion that we must regard some basic properties of individuals as necessary if induction is to be a reasonable practice. Thus the antecedents of supportable hypotheses must stand a chance of signifying a property that is necessary for some individual. But suppose we impose the condition that all singular necessities can be universalized. Then the antecedent of any supportable hypothesis must stand a chance of signifying a property with universal application. Otherwise, it would not stand a chance of signifying a property that is necessary for some individual. The imposed requirement would have the effect of limiting supportable hypotheses to ones that are mostly irrelevant to the practice of action on prior experience, considered as a fundamental practice. For practice deals largely with connections among properties known to have limited application. Thus inductive practice is incoherent with the theory of necessity based on universalizability. But let us see if the ontologist of conditions can even make sense of his own universalizable necessities.

Second, there is a difficulty for necessities across time. The action view of time was possible since things had progressive actions that did not themselves have durations. By going back to progressive actions, it was possible to ground the time of and between conditions (Chapter VIII, §3). Yet conditions, as opposed to progressive actions, are of themselves temporally extended. In an ontology of conditions, a progressive action is not had by an individual, nor is it had by the condition of having it. Yet apart from being had by some entity it is not a source of time, since if it is not had by some entity it is in no way particularized. Thus it cannot be the source of the time of any particular condition or of the time between any two particular conditions. However, in an ontology containing physical individuals, progressive actions are components of individ־ uals and can thus be the basis for the extended conditions of having those actions as well as for times between conditions. In short, an ontology of conditions will have to assume temporal and spatial relations as entities. Conditions will have positions in a network of such relations, as will their stages. But the result of having temporal relations as entities was seen to be that necessities across time were excluded (Chapter VII, §1). Thus the ontology of conditions can admit no necessary connections across time, universalizable or otherwise; it is a natural basis for a Humean denial of necessities across time. But we saw that denying such connections is incoherent with the practice of action on prior experience.

Finally, it is even difficult to account for any necessity within the ontology of conditions. What is the status of the features of which conditions are the having? It is customary to say that they are the “constituents” or the “components” of conditions.³ But this is usually explained by saying merely that F is a constituent of condition e when e is, at least in part, the having of F. In any event, for the ontology of conditions, these features are not, in our sense, components of individuals or even of conditions. They are not the same particulars as individuals–there being no individuals–and they are not the same particulars as conditions–conditions not, in general, having the features of which they are the having. So for the ontology of conditions, F is related to the having of F in either of two ways. Either F is no entity at all or F is an entity distinct from the having of F. If it is no entity at all, it cannot be used to account for the necessity of a condition by being its nature. If, however, F is distinct from, say, a ’s Fing, the ontology of conditions accepts features as a category of entities distinct from conditions. This is not its first admission of entities distinct from conditions. This ontology must also countenance temporal relations as entities distinct from conditions. We thus arrive at a physical ontology similar to that of the Timaeus, in which Plato treated physical appearances as conditions–the havings of distinct entities called forms–in the relations collectively called the receptacle.⁴

The difficulty now encountered by the ontology of conditions is that only an entity that a necessity is of can account for this necessity. For, as we saw in Chapter III, §2, if F is the nature associated with, but distinct from, as having F, then it will be a necessity of F itself that if the nature F does support an implication to the feature G, then, indeed, the condition of a’s having F will imply b’s having G. Otherwise, a’s condition of having F could fail to have connections its nature, F, requires it to have, which would be absurd. When, however, the nature is not distinct from an individual that has it, the association between the nature and the individual is a necessity of the individual, and does not depend on the nature of the nature of the individual. But here the condition and its nature are distinct, and, thus, to ground the association of the two, a nature of this nature is needed, and so on endlessly. We saw that this was a harmful regress.

So unless necessities of physical conditions can be based on the natures of complex particulars other than conditions–hence, on the natures of individuals–there can be no necessities of physical conditions. This requires abandoning the ontology of conditions. The whole idea of natures of conditions is seen to be useless, for all necessities are ultimately grounded on the natures of individuals, that is, on the natures of things or of other individuals if there are any. Moreover, since the ontology of conditions fails to allow even for necessities that can be universalized, it has no basis –other than its general inability to handle necessities–for rejecting restricted necessities of particulars.

As we have seen, our version of the limited-independent-variety account of necessities needed by induction requires non-universalizable singular necessities for some basic properties. Conditions belong to the surfaces of individuals, and necessities utilize the depths of the natures of individuals. It is no surprise, then, that when one tries to have the surfaces without the depths one loses necessities, and thus settles for an ontology that is incoherent with the presuppositions of any human practice.

§2. Things as Individuals with the Capacity to Act. Conditions without individuals yield an impoverished ontology. There must then be individuals in order to provide conditions with a locus for occurrence. But a further step is needed to require that there be individuals with a capacity to act, that is, that there be things.⁵

As we saw in Chapter VIII, there must be entities that act if there are to be necessities across time. Since conditions and nonphysical individuals do not act, it is clear that the required entities must be physical individuals. However, it does not follow that an individual has a capacity to act simply because it acts. According to Chapter X, two requirements must be met if there is to be a capacity. There must be an appropriate fine structure, the having of which is a causal factor in realizing the behavior the capacity is for. There must also be a necessary connection involved in the causal relationship from the fine structure to the behavior. If either of these requirements fails, the behavior in question is not the realization of a capacity but is a caprice. However, the actions involved in necessities across time cannot be mere caprices of nature, so the individuals having these actions will have capacities to act and, hence, will be things.

Consider the consequence of saying that, though individuals act, they have no capacity to act. Can it be necessary that if Fa then at a subsequent time Gb? If, where F itself is not an action, neither a nor b nor a third individual has the capacity to act, then it is a contingent matter that, when Fa holds, either a or b or a third individual will act. For if it were necessary, then some component of a or b or a third individual would insure that, when Fa holds, one of the three individuals would act. And this would amount to one of them having the capacity to act. But if it is contingent that, when Fa holds, either a or b or a third individual will act, then it cannot be necessary that if Fa then at a subsequent time Gb> for there can be no subsequent time unless one of the relevant entities acts.

Where F is itself an action of a, it cannot be supposed automatically that something has the capacity to F. What needs to be shown is that, whether or not F is the actualization of a capacity, there is some action involved that is. For there to be a necessity of the above kind, there must be a condition that is a result. Otherwise, there would be a time subsequent to as being F only contingently. The result need not be b’s being G, but it will be like b’s being G in being after as being F. A condition that is a result comes about as a change in the way things are, and a change is an action, in our broad sense. There must be such a change when a F’s or else what follows a’s being F follows it only contingently. So there will be a component, a nature, that functions as fine structure in bringing about the change. Either a or some other involved individual will, then, have a capacity to act. If one rejects such a capacity, one must reject the necessity across time.⁶

In §1, I showed that the necessities required for induction require individuals. Given our account of necessity, the required individuals will have natures, since the necessities are necessities of the individuals. I have now shown that necessities across time require individuals with capacities to act. There is the possibility that some of the individuals required by the former argument lack capacities. They may have natures but no natural actions. If these individuals had natural actions, their natures would provide a basis for their having capacities for these actions. (There is no obstacle to expanding the notion of fine structure to include natures.) We cannot then say that the individuals required by induetion are all things. But we have just shown that at least some of them must be things and hence that the required ontology will contain things.

The differentia in the definition of things is the having of a capacity to act. It turns out that having the capacity to act is not characteristic of any entities other than things. Thus, having the capacity to act implies being an individual, which was the genus in our definition. And so the seemingly more general predicates ‘particular with the capacity to act’ and ‘entity with the capacity to act’ have in fact the same extension as the definition ‘individual with the capacity to act’. To justify this claim, I shall examine the relation of non-individuals, such as components and conditions, to capacities. I wish to establish the general position that the claim that a component or a condition has a capacity is true only because an individual has a capacity.

§3. Capacities, Conditions, and Components. First, consider the question of whether conditions have capacities. The condition of being sharp, we might say, is capable of causing a cut. But to what entity does the required fine structure belong? The fine structure will be, say, the component of sharpness belonging to the knife. And the basis of the causal necessity required for the capacity will be the nature of this sharp individual. The fine-structure factor and the factor behind the causal necessity are not then factors of the condition of being sharp but of the individual with this condition.

Admittedly, the fine-structure factor is a “constituent” of the condition of being sharp, but this means only that sharpness is a component of the individual with this condition. Moreover, the causal necessity is not of the condition since, in view of □˜1, the nature of the condition would be distinct from the condition and could not ground the causal necessity. So the causal necessity is of the individual, and hence the nature needed to support this necessity is the nature of the sharp individual. In brief, conditions do not themselves have capacities; whenever a statement that a condition has a capacity is accepted, it is accepted only because individuals with these conditions have the corresponding capacities.

Second, consider the question of whether components have capacities. The attribution of capacities to components demands reinterpretation to avoid sheer absurdity. The component redness that an individual has does not have the capacity to be or to become some other component. A component is always precisely what it is and does not change, since we employ components to explain just what change is. In addition, a component does not have the capacity to make something happen or to be the object of some act, such as seeing. Yet why not?

A component by itself is not a cause, though an individual’s having a component can be a cause. Otherwise, two causes might be different components without being distinct particulars; the heat of a fire that causes a burn and the brightness of the same fire that causes a shadow would not be distinct causes, being components of the same fire. Causal claims could then change truth value by the replacement, at the point of the term for the cause, of a reference to one component with a reference to a different one, even though the reference does not change to a distinct particular. On the surface, at least, this is a difficulty.

But what if one allows the distinction between the transparent and the opaque to apply to the context of the term for the cause as well as to that of the term for the effect (cf. Chapter X, §4)? Thus:

(1) A crisis is caused by Wilson’s speaking about war might be true, when:

(2) A crisis is caused by Wilson’s speaking 200 words

is false, since, though Wilson’s speaking about war and his speaking 200 words were the same event, the context is opaque for terms for causes. However:

(3) A crisis is caused by the condition that is Wilson’s speaking 200 words

would be true, given the truth of (1) and the sameness of the conditions, since the context is transparent for terms for causes.

Still, the above-mentioned difficulty remains. The heat and the brightness of the fire have distinct effects. Thus the replacement of a term for heat by a term for brightness in a transparent context like (3) would affect a change of truth value. To allow for this, one would be forced to the conclusion that the heat and the brightness are, in fact, distinct particulars. Then the terms for them could not be interchanged in a transparent context. But since we have strong reasons for thinking that components of this kind are not distinct, we should rather conclude that components are simply not causes at all. Then all attempts to insert terms for them in the place for causal terms in affirmative causal claims are misguided. (Similarly, since conditions, not components, are extended, we avoid the difficulty that if ‘the sharpness of a lasts one year’ is true then, since components of a are all the same, ‘the hotness of a lasts one year’ would be true.)

But perhaps components can be objects of acts. If a component were the object of an act, a different component of the same individual should not have to be an object of such an act. I do not see the red of an object’s back when I inspect the green of its front. But normally, if one sees an entity, then one sees any entity that is the same particular it is. Here I am taking the object of the act to be a real and not just an intentional object. This assumption is perfectly natural where we are speaking of real components and not of the contents of perceivers’ beliefs about what they perceive. Thus ‘Hugo sees green’ implies ‘There is green that Hugo sees’. So we arrive at the objectionable result that if the green individual is red in back, Hugo’s seeing red is implied by his seeing green, even though Hugo may not have noticed the individual’s back. It is not open to us to avoid this absurdity by claiming that the green and the red of the same individual are distinct. For then we would be faced with the vicious regress of Chapter VII, §1.

Now if components do not change and if they neither act nor are acted on, they lack capacities. Ascribing capacities to them is at best a way of saying that individuals with these components have certain capacities, and that these components are themselves parts of the fine-structure factor in these capacities.

It suffices, then, to characterize things as entities with capacities to act in order to distinguish them from conditions and components. But this characterization would also distinguish things from physical individuals that, despite our definition of them, are simple and from any kind of momentary entity. Simple physical individuals have no components and are thus propertyless foci for external relations to properties. Now perhaps the properties externally related to physical simples could do the work of those components of complex individuals that ground the capacities of these individuals. But externally related natures will not, without the unpalatable regress discussed in Chapter III, §2, do the work of grounding the necessity involved in the causal connection required by capacities. Thus simple physical individuals, sometimes referred to as bare particulars, have no capacities and a fortiori no capacities to act.⁷ They cannot do the job that things do in the required ontology.

Momentary entities, conceived of not as instantaneous events but as instantaneous slices of enduring individuals, can have capacities, though not capacities to act. A momentary entity could still have the capacity to be a factor in influencing other entities. Its momentary presence could, in the context of the action of some other entity, be a causal factor in influencing the outcome of that action. But beyond this, it lacks even the capacity to be in certain ways corresponding to non-relational features. It does not have the capacity to be red just by being red, since its having red would then have to cause it to be red, which it clearly does not do. And when it is red it does not have the capacity to be red because of a component other than red. For if it has the capacity because of another component, then either this other component would have to have functioned earlier as the cause of this red now, or it would have to function now to cause a later and hence a different redness in this individual. But in neither case is the individual momentary; it becomes an enduring individual. Finally, a momentary entity does not have the capacity to be what it is not, for then some feature of its fine structure would have to have, in conjunction with some act, a certain outcome that makes a difference to the momentary entity. Such an outcome would, in view of the intervening act, occur later on, and hence too late to be a condition of the momentary entity.

How does this relate to having the capacity to act? Since actions are components, they belong to a kind of being. So if momentary entities are barred from having capacities to be certain ways, they are equally barred from having the capacity to act. In other words, the above argument can be transferred point by point to capacities to act. If a momentary entity is acting, it can have a result only in another entity, for otherwise it could have a condition–the resuit of the action–that is later than this action. But momentary entities do not, in view of the above argument, have capacities for actions that have results in other entities. Since this is the only kind of action possible for them, they do not have the capacity to act in any way. This action with results in other entities could well establish momentary entities in time, but, without capacities to act, momentary entities would stand only in contingent connections.⁸

The futility of upholding the physical necessity needed for human practice while maintaining an ontology of conditions, an ontology of simple individuals, or an ontology of momentary individuals should be apparent. Conversely, it is not surprising that devastating critiques of necessities across time should have come from philosophers for whom one or another of these ontologies was ascendent. Since momentary individuals could have natures, there can be necessities of these individuals, though none of these necessities would be across time. However, since conditions and simple individuals could have no natures, they are the subjects of no necessities at all. Given the strong reasons we have for necessities across time, it becomes impossible for us to allow either events, simple individuals, or momentary entities to have the exelusive title of physical particulars. Of course, entities of any one of these kinds may exist; we are only denying that without things they can constitute an adequate ontology.

Consider, finally, the relation of capacities to time and space. From where I am seated, I am able with my hands to move individuals on my desk but not individuals on the moon. Capacities to act on individuals thus seem conditioned by their proximity. Perhaps then spatial and temporal relations must be assumed as entities if there are to be capacities to act on distinct individuals. However, capacities for actions, but not for actions on other individuals, do not seem to suppose a temporal and spatial network. I have the capacity to run even when denied the space to run in.

According to the view described in Chapter IX, §4, actions between distinct individuals are possible either because they ultimately occur and then provide their own temporal basis or because other actions occur between those individuals and provide what we called a temporal triangle. Independently of actual actions, there are no possible actions between distinct individuals. There is a perfectly analogous situation for potential actions between distinct individuals. Independently of actual actions there are no capacities for actions between distinct individuals. This means that actions must in some way be included among the components of fine structure when our analysis of capacities is applied to capacities to act on distinct individuals. Part of the fine structure will be the non-specific feature of there being some action, of the entity with the capacity, on the distinct entity that is potentially acted on.

Thus a’s having the capacity to act in the manner F on the distinct entity b requires, first, that some action occur from a to b, either directly or through intermediaries, and, second, that F be causally explained by other features of the fine structure of a. (The only action there is between a and b may turn out to be F itself.) This is not to say that actions for which there are capacities will end up requiring actions for which there are no capacities. This account allows that, for every action between a and b, there can be a capacity. There is neither regress nor circularity, and no need for spatial or temporal relations.

§4. Sameness and Existence for Conditions. By establishing that things are indispensable we have in no way undercut the need for conditions. In fact, there must be conditions, for things must have components in order to have capacities. Even if there were only one individual in the physical universe–and nothing I have said implies there must be many–there would still have to be many particulars–at least one individual and at least one condition. So monism, as the view that there is only one particular, is incompatible with the required ontology. Recall that, unless conditions are particulars distinct from things, they too will be components, and there will be a regress of conditions required for the having of any property (Chapter VIII, §4). But if things are distinct from conditions, something more needs to be said about their relation to one another. I shall argue that, though conditions are unlike components in that they are distinct from things, they are like components in having the status of secondary entities, whereas things are primary entities.

By a “primary entity” I shall mean an entity that satisfies two requirements. The first is that it is “independent as regards sameness.” An entity satisfies this requirement when it is not a component of any other entity. Otherwise, what entities it is the same as would depend on what entity it is a component of. Whether this redness is the same particular as this circularity depends on whether there is an individual both are components of. But whether this penny is the same particular as this redness does not depend on what this penny is a component of, since it is not a component at all.

The second requirement is that it is “independent as regards existence.” An entity satisfies this requirement when the associated component called existence does not depend on any entity other than this entity. In other words, there being the component of existence associated with an entity depends on nothing other than there being this entity. I say the component of existence “associated with” an entity rather than the entity’s component of existence since, for example, the dependent existence of a com- ponent like redness is not a component in redness. Since components do not have components, the dependent existence of the color is ultimately the existence of the individual with the color.

It is important to emphasize that independence as regards the feature existence need not imply independence as regards the condition of having existence. For a primary entity will have existential independence even when its coming into existence and its having existence are dependent conditions.⁹ A thing, for exampie, might be brought to exist or caused to be in the condition of having existence by some other thing’s condition. Still, once its having existence is caused, it has a component of existence and this component is not thereby caused; indeed this component, as the existence of a thing, does not depend on any condition.

I have had to rely on the notion of component here; an entity is independent as regards sameness when what it is the same as does not depend on what it is a component of, and an entity is independent as regards existence when the associated component of existence does not depend on any entity other than it. But with some sacrifice in definiteness, the reliance on the notion of com- ponent can be avoided. Then we say an entity is independent as regards sameness when what it is the same as does not depend on its belonging to any whole; an entity is independent as regards existence when, over and above any dependency it might have for its coming to exist and for its being sustained in existence, there is no other way that involves existence in which the entity is dependent.

Thus the phenomenalist treats impressions as primary and things as secondary. For the existence of things depends on impressions in a way that implies neither that things come to be nor that they are sustained because of impressions. In short, the dependency involves existence but is not causal. The historical materialist treats economic roles as primary and a juridical system, say, as secondary. For the existence of a juridical system depends on roles in the process of production in a way that is compatible with its being a cause of the coming into existence of some modifications of economic roles. The scientific realist treats microentities as primary, and trees and birds as secondary. For him, the existence of trees and birds depends on microentities in a way that does not imply that they are caused by microentities to come to be and to endure. .

Here then are ontologies in which the notion of component is either rejected—because of a commitment to simples—or not well worked out. Yet in regard to them the notion of ontologicai inde- pendence, that is, of independence as regards existence, has appli- cation. (The same point could be made for independence as re- gards sameness.) But in what follows I shall keep to my characterization in terms of components. In any ontology the connections between entities will include dependencies as regards sameness and existence. The justification of these dependencies will be one of its main problems. In my ontology, this justification will quite naturally rely on the notions of component and individual.

The general notion of dependency was explained in Chapter IX, §3, with the help of the notion of the factive conditional. Thus, to say that the existence of an entity a depends on a but not on any other entity is to say that:

(4) There is the entity a > there is the component of existence that is associated with a

is true, but the following is false for any b other than a:

(5) There is an entity b > there is the component of existence that is associated with a.

Of course, the component existence is a sufficient condition for there being the entity associated with it; conversely, there being the entity associated with it will be a sufficient condition for this component. So the general notion of sufficient condition will not explain the asymmetrical notion of dependency needed here. This notion of dependency is, however, illuminated by the notion of making something true which was built into the notion of the factive conditional. When a’s existence depends on a, there being an entity a makes it true or is the ground for the fact that there is the component of existence that is associated with a. But there being the component of existence does not make it true and does not ground the fact that there is the entity a. In short, the part does not imply the whole. There being the component is at best a sign of there being the entity.

Further, if properties and conditions are dependent, as regards existence, on things with them, then that there are these things factively implies that there are the components of existence that are associated with their properties and conditions.¹⁰ That there is existence that is associated with the yellowness of a thing a is made true or grounded by there being the thing a, but not conversely. Of course, a might change color, but as was emphasized earlier, the factive conditional is contextual.

Which entities in the required ontology are primary, if any are? Things, but neither components nor conditions, qualify as primary. (Moreover, if there are physical individuals that are not things, they too would be primary, but I shall limit attention to things.) Things are independent as regards sameness since they do not have the status of components, and for the same reason conditions are independent as regards sameness. But though things are independent as regards existence, conditions of things depend on things for their existence. Similarly, components are not independent as regards existence; they are dependent as regards their existence on the things that have them. Let us look now at the justification for the claims that conditions and components are dependent as regards existence.

As for the primacy of things, I shall consider only the question of whether things, if they have independence as regards existence, are necessary beings. If existence is a component that an entity has only because it has the nature it does, then the entity is said to be a necessary being. It does not logically follow from the fact that an entity is a necessary being that it always exists or that it exists in all possible circumstances. All that logically follows is that, when it exists, its having the component of existence will not be due to the condition of any entity playing the role of its cause, but will flow from its having the nature it does.

Even so, primary entities need not be necessary beings. What follows from a primary entity’s having the nature it does is not its having existence, but at most its having the conditional property that it has when it is true that there being this entity factively implies there being its existence. Thus, ontological independence means that there being the component of existence depends only on there being the entity with it. It does not mean that having the component of existence depends on either there being the entity or on its having the nature it does. The necessity of an entity concerns its condition of having existence, whereas its primacy concerns its component of existence.

Let us inspect this distinction simply from the perspective of causation. A necessary being could neither be caused to come to have existence nor sustained causally in its having existence. Otherwise, its having existence would not be due simply to its having the nature it does. The situation is different when the entity is merely a primary entity. Even though the entity’s component existence depends only on this entity, there may be a cause of its coming to have this component or of its being sustained in having this component. In a similar manner, the component yellow an old newspaper has is dependent on the molecular structure of the newspaper and hence on the newspaper, even though it is not the newspaper that causes its having yellowness but rather the sunlight to which it was exposed. The component existence of an entity may depend on the entity, even though for the entity to exist–that is, for it to have this component existence–a cause may be needed. In sum, things can be independent as regards existence without having to be necessary beings.

The theological argument from contingency depends on a blurring of this distinction. One goes from the premiss that physical things do not exist by nature to the conclusion that some being exists by nature by way of the premisses that contingent beings are dependent as regards existence and that dependent existence is grounded in independent existence. But, in view of the above, it is simply false that contingent beings must have dependent existence and, hence, it is simply false that independent beings must be necessary beings.

Why are conditions secondary entities? We can approach this question by examining the foundations of the relation condition of. Assume we are dealing with a condition of a thing. The component the condition is the condition of having is the foundation on the part of the thing. But it is not so clear what the foundation is on the part of the condition itself. If the condition is one of having redness, then it is not the case that redness is itself a component of the condition, for then the condition, as well as the thing, would be red. Redness will not, then, be a foundation at both ends of the relation. My proposal is that the unity of the condition is all there is to the foundation of the relation condition of in the condition itself. The reason for this is imply that nothing more is required of the condition than that it be, and hence be a unity, for it to stand in this relation to a thing. But a thing must have the relevant component to stand in this relation to a condition.

Now the unity of a thing is the sameness of the thing both with itself and its components (Chapter VII, §5). The unity of the thing is the binding together of the components, including unity, with one another and of the thing with each of its components. That is, if a has unity, then any two components of a are the same, and a is the same as any component of a. The unity of a condition is also the binding together of the condition with its components and of these components among themselves.

But the important fact is that the unity of a condition of a thing has its roots elsewhere. This is understandable since the condition concerns a thing and one of its components. The unity binds together since this unity is itself a complexity whose elements are elsewhere. The unity is the complexity there is when there is, on the one side, a thing and, on the other side, a component of that thing. This is not the complexity there is just among the components of the thing, for, since a thing is different from any of its components, the thing is not reflected in this complexity. It is, rather, the complexity there is as a result of there being a thing and one of its components. This complexity does not contain the thing or its component as a component, for, as the unity of a condition, this complexity is itself a component, and thus has no components.

Moreover, even though the condition has this complexity, the thing and its component are not had by the condition as components. If it had them as components, it would be the same particular as the thing and be qualified in the way the thing is qualified by having that component. Since the unity of the condition is this complexity, it cannot be the case that the condition is the ontological ground for either the thing or the component of it. For the complexity of there being a thing and one of its components requires there being a thing and one of its components. Since the existence of a condition requires its unity and since its unity is the complexity of there being a thing and its component, the existence of the condition depends on a thing and its component. Once it is shown that components depend on things for their existence, we can say simply that conditions depend on things as regards existence if they are conditions of things.

To show that a component is existentially dependent on a thing or whatever has the component, it suffices to point out that a component depends on a thing or whatever has the component for sameness. If a component were not dependent on, say, a thing with it for its existence, it would have its own existence or depend on some other entity for its existence. Thus, its existence could not be one shared with the thing. So the component’s existence would not be the same particular as the thing. But if its existence is not the same particular as the thing, the component itself is not the same particular as the thing. So a component would not be dependent as regards sameness on the thing with it. In sum, the existence of a component is dependent since it is not a component of this component, but is the existence of the thing with this component.

§5. The Dialectic of Sameness. The doctrine that relations are entities was seen to lead to the result that things are really isolated from one another. By contrast, if in place of relations there are relational properties and foundations for them, then there is a togetherness of things (Chapter VII, §4). This togetherness resuits from the fact that a thing has a relational property, not by it alone having a foundation, but in addition by another thing having a matching foundation. Given the foundations in distinct things, these things will have properties by which they internalize one another, that is, they will have relational properties.

Nonetheless, this togetherness seems to be associated with a rigid distinctness of the things participating in it. And such a rigid distinctness is incompatible with the fact that these things often exist in communities in which their distinctness is modified. The togetherness we have offered, while a step beyond the separateness imposed by relations as in-between entities, still leaves the things participating in it as distinct entities. But at every level–the atomic, the chemical, the biological, and the social–we are faced with phenomena for which the assumption of the distinctness of the key entities provides an inadequate account.

One way to attempt to solve this problem is to call a retreat on the distinctness of things. Things are then reduced to the level of components in order that they might be members of communities, which in turn acquire the role of primary entities.¹¹ But this solution fails to provide for the fact that things come into and go out of communities. On the one hand, they may become members of a community and before they cease to exist the community dissolves. On the other hand, they may cease to be components in one community and shortly thereafter begin to be components of a distinct community by ceasing to participate actively in the former and beginning to participate actively in the latter. This attempted solution to the problem of community by reducing things to components is a static solution in that it ignores these changes of sameness and traps things once and for all in given communities.

The concept of part elaborated in Chapter X, §2, provides the basis for another approach to reconciling the distinctness of things with their roles in communities. A part is the kind of component that can become a distinct entity by being separated from the entity of which it is a component. So there is a change of sameness involved in the separation. Conversely, if an entity becomes a part, it is no longer the same particular. It becomes the same as the entity of which it is now the component.

Of course, a part is in some sense the same entity as the corresponding separate entity. But we can reconstruct this sense in terms of the materials at hand. First, the part and the corresponding separate entity are similar, in the way the green of this thing is similar to the green of that thing, but both distinct and different. This alone is not enough even to approach closely the intuitive sense of sameness as applied to a part and a corresponding separate entity. What more is involved? Second, a part and a corresponding separate entity are related by becoming. The separate entity becomes a part, or the part becomes the separate entity. So a part and a separate entity are ‘the same entity’ in the sense that they are similar and the one becomes the other.

If, on the other hand, things were simples, there would be nothing describable as a part becoming a separate entity, since in this view parts must already be separate entities. An ontology of simples does not countenance a dialectic of sameness whereby, in time, the many become a complex one. In the ontology of simples, communities must be simples if they exist at all, in which case there are no entities that are components of them. Conversely, in the same ontology, if communities have components playing roles in them, then the communities themselves do not exist. Every community that has component members is a fiction, and the distinctness of things is an inflexible law. Individualism is then an ontological necessity and not just an ideology if there are component members of communities.

One might seem to resolve this dilemma by saying that communities are sets to which individuals stand in the membership relation. In this view, both the community and the individuals in it can be distinct simple individuals. But if communities are sets, then it becomes impossible to explain their changes when their membership remains unaltered, or their sameness when their membership changes. Communities change their form and even dissolve without, in the process, changing members. The only basis for understanding communities is a component ontology, in which the individual-component contrast is radically different from the set-member contrast.

The ontology of components allows us to say simultaneously that things are entities and that communities of things, or of entities that when separated are things, are entities. In addition, given the theory of parts we have associated with the ontology of components, there can be processes whereby entities no longer remain the same since they become components of communities, and whereby, conversely, communities dissolve or lose some of their component members, which then revert to distinct entities.

Such processes are not limited to the physical, chemical, and biological levels, where it seems obvious that bits of matter are appropriated by systems that can, in turn, generate distinct systems from their parts. People are also entities that can exist either as distinct entities or, through genuine participation, as entities that are components of various sorts of communities. Mere formal membership without active participation is never sufficient for a change of sameness from personal to communal sameness. Conversely, mere formal community that depends exclusively on ritual is not an entity for whose sameness members exchange their personal sameness.

Since people are genuine entities, even as components of communities, and since the communities of which they are components are genuine entities, the dichotomy between individualism and holism breaks down. Communities do not, as they do for individualism, contain individuals in the sense of entities that are rigidly distinct. But neither are communiâtes, as they are for holism, wholes in which the members have lost their entitative status. Communities are distinct entities in which members are components that can still become distinct entities. Insofar as communities are distinct entities, they must have components other than their parts. They will have properties, actions, and natures. Thus it is clear that not all explanations of interaction among communities will be explanations by means of their parts, as the individualist would contend.¹² Of course, it does not follow that the extreme holist would be right in thinking that the parts never serve an explanatory role in regard to the action of the community. He would, nonetheless, be led to say this because of his denial of entitative status to the parts. But if parts are entities, then like properties or natures, they can be appealed to in accounting for what the community does. Further, as parts they may have come to the community from having been entities with the status of distinct entities. And they are not trapped by the community since it is possible for them to return to the status of distinct entities or to participate in distinct communities at distinct times.