By Steven W. Halady

Steven W. Halady, University at Buffalo, SUNY

Our world is populated by a variety of things with which we interact. I am sitting on a chair, typing on a computer, which rests upon a desk. A book lies just to the left of my hand. Photographs hang on pins which are stuck into my office wall. At least, that is how it all seems.

Trenton Merricks is not so sure that this picture of the world, one overpopulated by myriad objects, is the correct understanding of how things actually are. In his book, Objects and Persons, he argues strongly against what he calls “folk ontology,” the common sense position that the objects of our everyday world-view actually exist. He argues for eliminativism, an ontological position that denies real metaphysical status to the ordinary objects which we incorrectly believe we encounter: chairs, desks, computers, books, and baseballs, just to name a few. Instead, he claims that instead of a baseball, what really exists is a collection of “particles arranges baseball-wise,” referring to the elementary particles of subatomic physics, remaining silent on which particles are the most elementary and what they are like.

Merricks’ motivation for this claim gets its support from a number of arguments, the strongest of which he calls the Overdetermination Argument (OA). The heart of this argument is the claim that the ordinary objects of folk ontology do not do any causal work, and therefore do not deserve the status of actually existing. The causal powers of folk-ontological objects lie entirely in the particles that would compose those objects, if those objects existed. Assuming, for the sake of reduction, that baseballs exist, any causal work that we might ascribe to, say, a baseball when it shatters a window is accounted for by the particles that make up the baseball. Once we recognize that the particles (arranged baseball-wise) are adequate for getting things done, there is nothing left for a baseball to do. Thus, baseballs are extraneous and unwarranted objects and should be eliminated from our ontology. The same holds for other ordinary objects. When we say that baseballs exist, what we are saying is strictly false, although close enough to true to be useful for practical purposes. Merricks argues that his position, eliminativism, is able to avoid many of the traditional problems that have plagued ontologists, such as the problem of the statue and the clay.

There is an important problem with OA, an exposition of which undermines the strength of the argument. In his book, Merricks never gives any account of causation. In this essay I intend to show how this oversight weakens OA by giving an account of causation based on power properties and arguing that on this account of causation, some ordinary objects of folk ontology can plausibly said to possess non-redundant causal powers, and so to avoid OA. 1

First, I shall present OA. I follow this by giving a slightly less brief account of causation based on power properties, and then show how on this understanding of causation, ordinary objects can avoid OA. That is, I show that ordinary objects have non-redundant causal powers. I do this be means of a reductio ad absurdum, in which I argue that when we interpret one of the premises of OA in terms of a powers account of causation, we find that it results in the failure of OA. My purpose here is not to argue that a powers account of causation should be accepted, but merely to show that if an ontologist accepts such an account, then she has a means for avoiding Merricks’ argument for eliminativism. At most this means that in order for OA to remain strong, Merricks owes us an account of causation which is independently motivated and which complements his eliminativism. Finally, I defend a key premise of my argument, as well as the claim that my account of causation does not beg the question against Merricks on the issue of the existence of ordinary objects.

I. Merricks’ Overdetermination Argument

Using the example of a baseball shattering a window, where “shattering a window” is shorthand for the scattering of the particles that were arranged window-wise, which would have composed a window if windows existed. Merricks presents OA as follows:

(1) The baseball – if it exists – is causally irrelevant to whether its constituent atoms, acting in concert, cause the shattering of a window.
(2) The shattering of a window is caused by those atoms acting in concert.
(3) The shattering of the window is not overdetermined.

(4) If the baseball exists, it does not cause the shattering of the window. 2

I will focus on (1) and (2) here and in the sections that follow. If either of these premises can be shown to be false, then any claims of the truth of (4) are unsupported, and Merricks’ motivation for eliminativism is found to rest on shakier grounds than he claims.

II. Power Properties Proposed

A power property is a way that an object is in virtue of which that object, in conjunction with the reciprocal powers of other objects, causes the manifestation of a particular event under a particular set of conditions and arrangement. 3  I will use “object” in a broad sense, to include both particles and any macro-sized objects that exist. I will use “object” in a restricted sense, to refer to the ordinary objects of folk ontology. The tendency of an object to yield a particular outcome under certain conditions is its disposition to behave that way under those conditions. The power property, or power for short, is the way that an object is, given in terms of what the object does; the disposition is how that object will behave under a particular set of conditions.

On an account of causation that is grounded in a theory of powers, x causes y under conditions Cbecause x exists in such a way that given C, it cannot avoid producing y. The ability to cause y under C is simply apart of what it is to be the way that x is. This is to say that the connections between powers, their dispositions, and the various manifestations of their dispositions, is a brute fact of the world. This is similar to Merricks’ apparent position that causation is brute – that things cause things. However, my position differs from his in that on my account, the causal ability is grounded in the relevant objects, whereas he does not seem to take a stand on what grounds the causal power––it could just as easily be grounded in laws of nature that are external to the objects as in the way the objects are. 4

To give an example of a powers analysis of an event, a window, assuming that windows exist, is disposed to be fragile if and only if it would shatter when struck in an appropriate way by an appropriate object (or collection of objects). It has this disposition in virtue of the power properties that the window possesses. Note that the same power property yields different dispositions under different conditions: under conditions where the window is not struck in an appropriate way by an appropriate object, the window is disposed to remain intact, unshattered. This disposition to resist shattering is grounded in the same properties that the disposition to shatter is – it is the reciprocal dispositions of the other objects involved along with the external conditions that make the difference in which dispositions are manifested at a given time. This reading is silent on the question of whether there are laws of nature and what those laws are like.

This fact points to an important factor in a powers account of causation: every disposition which manifests does so in conjunction with a reciprocal disposition. The disposition of the window to shatter manifests only in conjunction with a baseball manifesting the disposition to smash windows. This can be translated into eliminativist jargon: the particles arranged window-wise will manifest their disposition to be scattered only when the particles arranged baseball-wise

manifest their disposition to scatter particles that are arranged window-wise. Under either description, the dispositions being ascribed to the objects involved are grounded in the nature of the objects themselves, in the way that those objects are. The relevant objects are such that under certain circumstances, they will mutually manifest particular reciprocal dispositions.

The important point is that on a power properties account, all powers are internal to objects, and dispositions to behave in certain ways are the result of the internal powers of other objects. There are no free-floating, external causal powers. Consequently, any causal power observed is the power of some object or set of objects. It is for this reason – to rule out any causal powers that are grounded outside of objects – that I will assume a version of a power properties account of causation.

On such an account, powers are not conditions for objecthood but marks of objecthood – they are symptoms by which we can recognize that a genuine object exists. If there is a real power which seems to be possessed by a proposed object and which cannot be accounted for by reference to the constituent parts of that object, then we have a good indication that the proposed object actually exists, and, all other things being equal, ought not be eliminated from our ontology. Notice that this view is fully compatible with Merricks’ position; indeed, OA depends largely on this notion, and it argues that in the case of ordinary objects, there are no such non-reducible causal powers, while with persons, there are.

Contra Merricks, I doubt that (1) and (2) can both be true at the same time, and I do so on the basis of a powers account of causation, as sketched above. Once we unpack what exactly (1) and (2) are saying, it is not clear that they are compatible. The key problem in OA rests in the necessary but fatal phrase, “acting in concert.” In Section III, I will examine the way in which Merricks believes that “acting in concert” should be understood. In Sections IV and V, I argue that there is an ambiguity in the phrase, and that no matter how we resolve the ambiguity, the end result is a weakening of OA, and thus of Merricks’ general eliminativism.

III. “Acting in Concert”: Some Analogies

Merricks says that it is the particles “acting in concert” that causes the window to shatter. 5 He takes “acting in concert” to mean that each relevant unit, in this case, each relevant baseball particle, makes its own causal contribution to the scattering of the relevant window particles, and that the sum of all of these separate individual contributing causes are the event we call the window’s shattering. 6 According to Merricks, any case of “acting in concert” is analogous to the lifting of a couch. I, puny as I am, cannot by myself lift a heavy couch. Neither, say, can you.

However, with each of us strategically placed at either end of the couch, we put our individual modest strengths towards lifting the couch, and succeed in raising it off the ground by exerting our causal influence upon the same couch at the same time. Similarly, a loud crowd is making a great din. Each individual in the crowd is contributing her own small part to the general racket. The individual contributions of each separate unit in the crowd, acting in concert, sum up to create what we take to be the noise of the crowd. 7 The sum total of our unique and separate strengths is what causes the overall effect of a lifting. The members of the crowd do not combine their noise-making ability into one single great and powerful noise-making ability; the more modest volume of each crowd member makes a separate and distinct contribution.

Merricks is drawing an analogy between teams and crowds on one hand, and baseballs and tables on the other. On this understanding, the baseball particles do not form one cohesive whole, they do not compose an object, which shatters the window; each particle has its own power which causes its own modest effect, perhaps of causing a single window particle to fly away, but when these effects are summed, we get the overall effect which we call “the window’s shattering.” Merricks’ analogies suggest that there is no internal unity to the baseball particles; they are simply summing up their separate causal powers, not in virtue of being a unified whole, but in virtue of being together in roughly the same place at roughly the same time, such that the summing of their moderate individual effects seem to add up to one grand effect. This seems to be all there is to “being arranged x-wise”: if the particles are arranged such that they act in concert, then they are arranged in such a way that we are likely to refer to them as an x. The summed effects lead us, or mislead us, as the eliminativist would say, to assume that there is one large object where there are in fact many smaller ones. In what follows, I will use the locutions “particles acting in concert” and “particles arranged x-wise” somewhat interchangeably to refer to a collection of particles that are arranged such as to warrant our paying attention to the sum of their effects as a whole.

Notice that in all of these examples of “acting in concert,” folk ontology attributes different kinds of unity to the causal units which contribute to the overall effect. According to folk ontology, you and I do not combine to be one object with one strength which lifts the heavy couch, even if you take us to be a team. Couch-lifting teams and crowds are not commonly thought of as being a single unit in the same way that a baseball is thought of as a single unit. Thus, for the folk ontologist, there is a dis-analogy here between the couch-lifting team and the crowd, on one hand, and baseballs, on the other. These alleged analogies risk begging the question against the defender of ordinary objects, in that they suggest that, contra common sense understanding, the baseball is a collection of things working together, rather than a single, unified whole. All that this establishes is that we should not take the analogies as an argument for eliminativism, but rather as a heuristic for understanding the point that Merricks is arguing for. The analogy’s force is in explaining the position, not in defending it.

IV. Loose and Unified Arrangements of Particles

On closer observation, however, it seems clear that Merricks’ analogy is not correct. I will argue that there is an ambiguity in the phrases “particles arranged x-wise” and “acting in concert” and that there are two ways of interpreting them. These two interpretations place Merricks between the horns of a dilemma, for the first interpretation does not get us the causal powers we want from particles acting in concert, while the second allows ordinary objects to get a foot into the ontological door, and thus weakens OA.

Merricks claims that it is the particles hitting the window at the same time and in roughly the same place at the medium-paced speed of a thrown baseball that causes the shattering. 8 There are two ways to understand what is meant by this arrangement, and as stated, “acting in concert” is ambiguous between them. On the first reading, the particles are completely independent of each other but located together in time and space in such a way as to warrant being called an “x.” Call this interpretation of “arranged x-wise” the “Loose” interpretation. Alternatively, the particles may be thought of as not completely independent of each other and interacting with each other in a way such that there is some degree of unity to them. Call this interpretation of “arranged x-wise” the “Unified” interpretation. Since Unified covers any arrangement of particles that has any degree of unity beyond complete independence, there is no third option. First, I shall argue that Merricks’ eliminativism is committed to Loose. Second, I shall argue that Loose cannot adequately account for the causal interaction of events like baseballs shattering windows, and so we need some version of the Unified interpretation to explain what “arranged x-wise” means.

Merricks does not ever explain exactly what he means by “objects.” Instead, he assumes a common understanding of what exactly the extension of the word is. This leaves an ambiguity which I believe is detrimental to his eliminativism. Indeed, I believe that this common oversight, discussing the metaphysics of objects while giving neither criteria for recognizing what kinds of objects there are nor criteria for recognizing when one in fact has a genuine object on one’s hands, mars many a metaphysical argument. 9While a full account of the necessary and sufficient conditions, or some other kind of definition, for the application of “object” is beyond the scope of this little work, I would like to restate something I mentioned at the end of section III: one of the marks of being an object, one of the signs that we have a real object on our hands, is the fact that real objects possess properties, and among those properties are power properties, which cannot be accounted for by reference to constituent parts of that object. If we find something which possesses a real, non-reducible causal power, a power property, then we have good reason to acknowledge the existence of the object which bears that property. Now, if for Merricks the objects of folk ontology are simply particles arranged x-wise, then he is faced with two possibilities. Either objects are particles arranged Loosely or in a Unified way. Both of these interpretations are problematic for the eliminativist, for Loose is untenable and a Unified understanding of “arranged x-wise” severely weakens eliminativism, both in terms of its scope and overall plausibility. Unified allows at least some ordinary objects into our ontology that Loose does not. Let me illustrate.

On a Unified interpretation, particles arranged x-wise are intertwined in ways that that Loose particles are not. There are interactions among those particles in addition to merely being found together, and networks of those interactions serve to individuate that collection of particles from other collections of particles in their environment. There does not seem to be any one principle which “acting in concert” requires. Rather, there are many different principles of unity which will account for different causal powers, and that what the principle of unity is for a given arrangement seems to depend on what, exactly, the x in “particles arranged x-wise” compose. Particles arranged baseball-wise may have a different principle of unity than particles arranged puddle-wise, or particles arranges team-wise. 10

From this, we can see that different arrangements of particles yield different powers – if you take the same particles and arrange them differently, you get different powers. Particles arranged baseball-wise have the power to shatter windows while particles arranged puddle-wise have the power to splatter when stepped upon, but not vice versa. In addition, if you arrange a different particular set of particles, you are also likely to get different powers. For example, protons and neutrons are made up of different combinations of quarks. Thus, it seems that it is the specific combination of particular particles and their arrangement that yields the specific power of the particles arranged baseball-wise to shatter a window under certain conditions. From this, we can see that the particles plus their interconnections have powers that the independent particles alone do not have. And on Merricks’ own principle – that objects with non-redundant causal powers should be given a place in our ontology – we get an object that is composed of the interconnected particles.

Note that this conclusion follows from a powers account of causation, but need not follow from other accounts. On the account sketched above, the causal power necessarily rests with the objects, and so the existence of powers is a clear sign that there is an object there. On some other account, the causal work need not be grounded completely in any object or set of objects, and so the existence of a causal power does not necessarily point to the existence of any object or set of objects. It is thus possible, in theory, to avoid the conclusion that there are any objects present over and above the constituent particles.

One way of describing the relationship between an x and the particles arranged x-wise that compose it is by way of what Frank Jackson calls the entry by entailment thesis. According to this principle, “the one and only way of having a place in an account told in some set of preferred terms is by being entailed by that account.” 11 What this means is that if we can tell an ontological story regarding an object in which the existence of that object is entailed by the story that we tell about more fundamental objects, we should accept that, according to our ontological story, the entailed objects do in fact exist. But this opens the door to objects of folk ontology, for what are the objects of folk ontology if not individual objects composed of elementary particles arranged in such a way so that there are networks of causal interactions among their composite particles that do not include the particles in their environment? Many folk-ontological objects are simply particles that are interconnected in ways that particles that do not compose those objects are not. All of this is to say that what it is to be a baseball is, at least in part, to be particles arranged in an appropriately unified way so as to give rise to the properties normally associated with baseballs. On a Unified interpretation, folk-ontological objects like baseballs are entailed by the existence of particles arranged x-wise, and thus it is proper to say that they exist.

I am not arguing that we must accept a principle like the entry-by-entailment thesis. For example, we can invoke Gilbert Ryle’s insight that we are simply mistaken if after touring classrooms, deans’ offices, and so forth, we say that those buildings are lovely, but ask where the university is. 12I claim only that if we do, then folk ontologists have the tools to defend the claim that on a Unified interpretation, “particles arranged x-wise” is not a radical rejection of folk ontology but a creative restatement of it, at least for some ordinary objects. Thus, on a Unified interpretation, the scope of eliminativism is limited – it does not apply to all of the objects of folk ontology, although it may still apply to some or many of them – and thus eliminativism becomes less plausible as an overall ontological theory. Loose, however, by not positing any kind of distinguishing interconnectedness among the particles arranged x-wise, is further from the folk-ontological theory, and does not entail any kind of folk-ontological object. Thus, Loose avoids this criticism, and so, if successful, allows for a stronger eliminativism, which is the eliminativism that Merricks wants.

So, it seems that Merricks is committed to the Loose interpretation of “arranged x-wise,” on pain of otherwise having a weaker and more limited eliminativism than he argues for. However, the Loose interpretation does not fit well with what we observe about the causal interactions of the world and should be rejected as an explanation of the nature of ordinary objects. Take the following thought experiment. Start with a single particle of any of the kinds that could, in theory, compose a baseball if baseballs existed, and hurl it projectile-wise at a window. The window will not shatter. Add a second particle of the same general kind, that is, of a more specific kind that would compose a baseball if baseballs existed, and hurl both of them, arranged relative to each other as they would be if they were part of a baseball-wise arranged set of particles, at the same window at the same time. Again, the window will not shatter. Repeat this process of adding particles and hurling them at the window until you have as many particles as you would need to compose a baseball, if baseballs existed. This collection of loose particles, however large in number, will not cause the window to shatter. Take as support of this point the fact that the window is constantly being bombarded by huge numbers of loose particles in the form of light and particles arranged air-wise, yet shows no signs of shattering.

One may argue in response to this that at the sub-micro level, each individual particle that hits the window causes only a tiny scattering of the window particles, but that we are not able to detect those scatterings without the aid of sophisticated detection equipment. So, each individual particle causes a tiny scattering, but these are detectable by the naked eye only when a large number occur at any one moment. But this response will not do. If it were true that individual particles had the power to cause an individual particle to scatter, then over a long enough period of time, all of the window particles would disperse one-by-one; the window would eventually appear to dissolve. But we know that windows do not do this – the window particles maintain their cohesion despite the constant bombardment of other particles. They have the power to persist in their arrangement in the face of particle bombardment. To put this in another way, the individual particles lack the power to cause the shattering of the window, when this is understood as the large-scale scattering of the window particles, even over extended periods of time.

The difference between the baseball particles’ ability to shatter windows and the ambient air particles’ inability to shatter windows is not to be found in what kind of particles they are. Remember that when we speak of particles, we are referring to the most elementary physical particles, and there is a relatively small, finite number of these particles. So, at the most fundamental level, baseball particles and air particles are arrangements of the same basic kinds.

The particles’ velocities are also not a factor here. Light travels faster than baseballs, and even if we slowed the speed of light down to the speed of baseballs, it seems that light would not thereby get the power to shatter the window – if the higher velocity does not do it, it seems correct to say that the lower velocity certainly will not. This is an extension of what we usually experience: objects moving at greater velocities have more force and do more damage than those moving at lesser velocities.

Let me illustrate this in another way. Say that you are a soldier on a battlefield and find yourself face to face with an enemy soldier. You are out of bullets, and the closest thing to a weapon that you have is a bag of metal flakes. You reach into the bag and pull out a single metal flake, which you hurl at your opponent at the speed of a bullet fired from a gun. To nobody’s surprise, she survives your onslaught. You repeat this with two metal flakes, to no avail. Again with three, and so on, until you are throwing an amount of metal flakes which has the same mass as a bullet. The difference between the collection of flakes and the bullet is that while the bullet might do some damage to the enemy soldier, the mass of metal flakes will not.

From this, it seems that we are justified in saying that the particles alone lack the power to shatter the window, even when those particles are taken en masse and are hurtling at a window or an enemy soldier at great velocities. We need more than a loose collection of particles, independent particles located together in time and space, to get the causal power we want. What we need is a causal whole, and it seems that there must be some principle that unifies the particles into that causal whole, a principle that distinguishes particles arranged baseball-wise from the same particles in a non-arranged state. Loose spatial-temporal coincidence is not enough to bring about the causal power we need for the baseball to shatter the window. We should reject Loose and accept Unified.

V. From Unified Particles to Ordinary Objects

At this point in the argument, we seem to be in a curious position which friends of Merrick’s style of eliminativism will not like. Premise (2) of OA states, “The shattering of a window is caused by those atoms acting in concert.” It seems that embracing (2) of OA does not support eliminativism as strongly as Merricks would like. If we accept a Loose interpretation of “acting in concert,” then we do not get the causal powers that we observe. But, the powers of particles that we get from “acting in concert” on the Unified interpretation will in some cases suggest that there exists an object over and above the particles whose existence is entailed by that of those particles in their interconnected arrangement, and which is needed to account for the powers that we observe. From this, we now appear to have good reason to accept the existence of something which Merricks would rather not acknowledge: a middle-sized object, of the kind folk ontologists love so much.

This leaves us with the question: what exactly is this object that causes the window to shatter? Attheveryleast,theobjectishard;itshardnessisperhapsthemostimportantfactorinits causing the window to shatter. It is about the size of my clenched fist, roughly spherical, and generally white with red laces stitched in a distinctive loop pattern around it. That is, it looks like a normal baseball. We know that this is what the object looks like, and Merricks says as much: particles arranged x-wise are empirically indistinguishable from x’s; it is only metaphysically that eliminativist and folk-ontological world-views differ. It weighs as much as a standard baseball, makes a distinctive basebally clunkwhen struck by a baseball bat, and is in all ways indistinguishable from what a baseball would be like, if baseballs existed. The object is and does everything that we would expect a baseball to be and to do. The moral of this story is clear: the real object which causes the window to shatter is a baseball, and I can see no good reason why we ought to deny this conclusion.

What is more, contrary to Premise (1) of OA, given the above arguments and a power property account of causation, the baseball is causally relevant to whether its constituent particles, acting in concert, cause the windows to shatter. Premise (1) states that the baseball – if it exists – is causally irrelevant to whether its constituent atoms, acting in concert, cause the shattering of a window. As I have argued, from the perspective of a powers account of causation, to be the constituent particles acting in concert is simply what it is to be the object we call the baseball. From that, it is clearly nonsense to claim that the baseball is causally irrelevant to whether the baseball causes the shattering of the window. So, the baseball must be causally relevant to the shattering of the window.

Thus, as a result of unpacking the ambiguity in Premise (2) of OA within the framework of a powers account of causation, we can see that the baseball both exists and is causally relevant to the window’s shattering. This directly contradicts (1). Thus, we cannot accept both of the first two premises of OA at the same time, thus OA fails to establish the case for eliminativism. In undermining OA, I have also, I believe, given some indirect support for accepting the existence of the objects of folk ontology.

VI. Response to Some Objections

In claiming that Merricks’ account of “acting in concert” needs what I call a principle of unity in order to be empirically accurate, I have not begged the question against him on the existence of ordinary objects. Rather, I am arguing that on the account of causation that I have laid out, we have good reason to believe that ordinary objects like baseballs have non-redundant causal powers, and the best conclusion we should draw from this is that (at least some) ordinary objects exist. This point is perfectly consistent with the claims about what objects exist that Merricks himself makes. 13

As I have stated above, there is a distinction between criteria of objecthood, the necessary and sufficient conditions for being an object, and marks of objecthood, features of our experience that clue us in to the fact that we are dealing with a genuine object. I have also claimed, with Merricks, that an irreducible power property is one such mark of objecthood: if there is a real power which seems to be possessed by a proposed object and which cannot be accounted for by reference to the constituent parts of that object, then we have a good indication that the proposed object actually exists, and, all other things being equal, ought not be eliminated from our ontology.

Merricks himself states, “Composite objects that cause things that their parts do not redundantly cause can resist the eliminative sweep of the Overdetermination Argument […] to be is to have non- redundant causal powers.” 14 I have argued that there is a causal power, the power to shatter the window, which, on a power-property account of causation, cannot be accounted for by Merricks’ understanding of what it means for objects to be “acting in concert.” Thus, there must be something which is not just the simple sum of the individual causal powers of the relevant particles. I call this factor a principle of unity.

The principle of unity is a mark of objecthood. Recognizing a need for a principle of unity does not entail that there is an object when we think there is one. We may be wrong about the relevant causal story. Further investigation may reveal that the best account of an event is not what we thought it was – that the causal powers lay somewhere else than we thought they did – and we may be best off concluding that nothing like a principle of unity is required in a particular case. A hurricane is an example of such an object. The units we had thought make up a genuine object may in fact account for the overall final effect, in which case OA will apply to that putative object, and we should eliminate it.

This is to say that Merricks’ understanding of “acting in concert” may be the correct way of viewing some, but not all, cases. However, I have argued that, given a particular account of causation, the baseball’s shattering the window is one causal story that does not fit this picture. If my argument is correct, then the baseball has at least one non-redundant causal power, and so it is able to avoid the claims of the overdetermination argument, not because I assume that there is an object there, but because Merricks’ eliminativism does not properly characterize the causal story in that case. The conclusion that there is a baseball that shatters the window comes as an a posteriori conclusion, not an a priori assumption. No questions have been begged in the making of this argument.

While I have not directly argued for it, it seems to me that similar conclusions can be drawn for a great many, although probably not all, ordinary objects of folk ontology. It seems that the folk assumption that this is a genuine object is not immediately undermined by OA, and we should lay the burden of proof on the eliminativist to show that a particular alleged object falls prey to OA.

VII. Conclusion

I have argued that Merricks’ OA does not work for many of the ordinary objects of folk ontology, when viewed through the lens of a power-property theory of causation. Under this account of causation, the first two premises of OA are incompatible, and the argument fails. This is not necessarily the case for all accounts of causation, hence I have not showed that OA necessarily fails. What I have argued is that the case for OA, and with it much of the support Merricks rallies for his eliminativism, is not airtight. A clearer account of causation, supported by arguments which are independent of, and complementary to, his ontology of material objects, will help Merricks’ case for eliminativism. Without such an account, the success of his case for eliminativism is an open question.


Armstrong, David M. Universals: An Opinionated Introduction. Boulder, CO: Westview P, 1989.

Heil, John. From an Ontological Point ofView. Oxford: Clarendon P, 2003.

Jackson, Frank. From Metaphysics to Ethics: A Defense of Conceptual Analysis. New York: Oxford UP, 1998.

Lowe, E.J. The FourCategory Ontology. Oxford: Oxford UP, 2006.
Lowe, E.J. The Possibility ofMetaphysics: Substance, Identity, and Time. Oxford: Oxford UP, 2001. Merricks, Trenton. Objects and Persons. Oxford: Clarendon P, 2001.

Prior, Elizabeth. Dispositions. Aberdeen: Aberdeen UP, 1985.
Prior, Elizabeth, Pargetter, Robert, and Jackson, Frank. 1982. ‘Three Theses About Dispositions’.

American Philosophical Quarterly19 (1982): 251-57.

Olson, Eric. ‘The Ontology of Material Objects’. Philosophical Books 43.4 (2002): 292-99.

Ryle, Gilbert. The Concept of Mind. London: Hutchinson, 1949.

Seargent, D.A.J. Plurality and Continuity: An Essay on G.F. Stout’s Theory of Universals. Dordrecht: Martinus Nijhoff Publishers, 1985.

Thomasson, Amie. Ordinary Objects. Oxford: Oxford UP, 2007. van Inwagen, Peter. Material Beings. Ithaca, NY: Cornell UP, 1990.


  1. Eric Olson makes a similar argument using a different account of causation in “The Ontology of Material Objects,” Philosophical Books 43.4 (2002): 292-99.
  2. Trenton Merricks, Objects and Persons (Oxford: Clarendon P, 2001) 56.
  3. Properties as “ways” is defended by David M. Armstrong in Universals: An Opinionated Introduction (Boulder, CO: Westview P, 1989) 96-98, as well as throughout John Heil’s From an Ontological Point of View (Oxford: Clarendon P, 2003). It is also suggested in the work of E.J. Lowe in his discussion of what he calls modes – see E.J. Lowe, The FourCategory Ontology (Oxford: Oxford UP, 2006). See also, D.A.J. Seargent, Plurality and Continuity: An Essay on G.F. Stout’s Theory of Universals (Dordrecht: Martinus Nijhoff Publishers, 1985) Chapter 4.
  4. I have in mind here functionalism about dispositions, which holds that ontologically, the tendency of objects to behave in a certain way is determined by the qualitative properties of those objects along with the laws of nature under which the object finds itself. For clear statements of dispositional functionalism see Elizabeth Prior, Dispositions (Aberdeen: Aberdeen UP, 1985) and Elizabeth Prior, Robert Pargetter, and Frank Jackson “Three Theses about Dispositions,” American Philosophical Quarterly 19 (1982): 251-57.
  5. In the baseball example, Merricks makes reference to atoms. Physically speaking, atoms themselves are made up of more elementary particles. Speaking in mereological terms, an atom is absolutely simple. To avoid ambiguity, in what follows I will use “particle” to refer to those most basic objects whose non-redundant causal abilities warrants its status as an existing object.
  6. I admit that eliminativism makes for awkward phraseology. To avoid such awkwardness, unless otherwise clear from the context, let my reference to baseballs and other ordinary objects be taken to be neutral with respect to ontological status.
  7. Merricks used these examples in personal correspondence.
  8. An anonymous commentator has noted that in order for there to be a shattering, there must be an object, like a window, to shatter. Otherwise, there is only a “scattering.” I agree. However, Merricks has an easy way around this criticism. While claims that there is a shattering are, strictly speaking, false, they are close enough to true to be useful shorthand for practical purposes. See Merricks, Objects and Persons, Chapter 7.
  9. See E.J. Lowe, The Possibility of Metaphysics: Substance, Identity, and Time (Oxford: Oxford UP, 2001), Chapters 2 & 3, for a refreshing escape from the widely shared assumption that “object” is unambiguous.
  10. Peter van Inwagen calls this type of position a Series-style answer to his “Special Composition Question.” See Peter van Inwagen, Material Beings (Ithaca, NY: Cornell UP, 1990) 64-71. Van Inwagen argues that this kind of answer is unprincipled and should be rejected. However, this criticism lacks force, for while non-Series answers to the Special Composition Question are more principled and thereby more elegant answers, the ability of Series answers do far more to explain the wide variety objects that are found in folk ontology, and this more than makes up for their inelegance.
  11. Frank Jackson, From Metaphysics to Ethics: A Defense of Conceptual Analysis (New York: Oxford UP, 1998) 5.
  12. Gilbert Ryle, The Concept of Mind (London: Hutchinson, 1949). This is also the approach taken by Amie Thomasson, Ordinary Objects (Oxford: Oxford UP, 2007), Chapter 1.
  13. See, for example, Merricks, Objects and Persons, Chapter 4, Section VI.
  14. Merricks, Objects and Persons, 114-15, emphasis in original.  

Steven W. Halady