Theo Todman's Web Page - Paper Abstracts

1. Material things constituted by animals¹¹
2. The clay-modelling puzzle¹²
3. The replacement puzzle and the amputation puzzle¹³
4. Thinking animals again¹⁴
5. When does constitution occur?¹⁵
6. What determines our boundaries?¹⁶

1. It is easy to suppose that we have properties incompatible with those of animals: that we are essentially capable of thinking, say, or that what it takes for us to persist is different from what it takes for an animal to persist. These claims₁₆ rule out our being animals. Yet we appear to be material things. Not only that, but each of us appears to be made of just the same matter as a certain animal. We are no larger or smaller than our animal bodies₁₇, and are located just where they are.
2. The idea that we are not animals, but are nevertheless material things made of the same matter as our animal bodies and located in the same place, may sound strange. We can increase the tension by noting the apparent truism that no two material things can be in exactly the same place at once₁₈. If we know anything about material things, we know that they compete for space and exclude one another. So we find ourselves drawn to each of four inconsistent claims:
   1. We are material things;
   2. Each of us has the same location as an animal (which is also a material thing);
   3. We are not animals; and
   4. Two material things can never be in the same place at once.
      
3. What to do? Those who think that we are partly or wholly immaterial deny the first claim. Those who think that we are temporal parts of animals deny the second: they say that our animal bodies are not located precisely where we are, but occupy larger spatio-temporal regions. We will come to these views in due course₁₉. Animalists deny the third claim. But some deny the fourth: they say that two material things can be in the same place. Not just any two material things, of course: you will never get a dog and a cat into the same place. Material things can be in the same place only if they are made of the same matter₂₀. That, the thought goes, is the truth behind the idea that material things compete for space.
4. The claim, then, is that two or more things can be made entirely of the same matter at the same time. For technical reasons it will be useful to recast this idea in slightly different terms₂₁. Let us say that some things, the xs, compose something y if and only if each of the xs is a part of y, no two of the xs share a part, and every part of y shares a part with at least one of the xs₂₂. So when a child builds a castle of Lego bricks, the castle is composed of those bricks. If each brick is itself composed of atoms, then the castle is also composed of atoms, for a part of a part of something is itself a part of that thing. So our suggestion is₂₃ that the same things can compose two different objects. In other words, different material objects can coincide materially, where x coincides materially with y if and only if x and y are material objects and some things, the zs, compose x and also compose y. Applied to ourselves, the idea is that we coincide materially with animals, even though we are not animals ourselves. One can be made of the same matter as one’s animal body while having properties that no animal could have.
5. If the relation between you and your animal body is not identity but material coincidence, we should expect to find more examples of material coincidence without identity; and indeed those who hold this view take it to be ubiquitous. The particles that compose a clay statue, they say, also compose a lump of clay; yet the lump and the statue are numerically different because the lump will ordinarily have existed before the statue did (before it was statue-shaped), and because squashing it would destroy the statue but not the lump. The blocks that compose the child's castle might also compose a material object called an “aggregate” of blocks, which predated and will outlive the castle. If we replaced one of the castle’s blocks, the castle would come to coincide with a different aggregate. Owing to metabolic turnover, an organism coincides with a different mass of matter or aggregate of atoms every fraction of a second. And so on.
6. In each of these examples, our proposal says, numerically different objects not only coincide materially, but also differ in important qualitative respects – where by 'qualitative' I mean any property that doesn't specify the identity of its bearer (being Stan Getz would be a non-qualitative property). The objects differ in kind: one is a statue and not a lump, the other is a lump and not a statue; one is a person and not an animal, the other is an animal and not a person. They differ in their modal properties: statues are essentially statue-shaped, but statue-shaped lumps are only contingently statue-shaped; people are essentially able to think, but the animals coinciding with them think at best only contingently. They have different persistence conditions: statues and people can survive things that lumps and organisms cannot, and vice versa. There is no point in saying that coinciding objects are numerically different but qualitatively identical – that we are animals that coincide with other animals exactly like us, say. The attraction in saying that statues are not lumps or that we are not human animals is that statues and people have properties that lumps and human animals lack. These are not merely historical properties: a statue and its coinciding lump are supposed to differ not only in that the lump existed before the statue did. They are supposed to differ while they coincide: for instance, while they coincide the lump but not the statue is capable of surviving squashing. A statue and its coinciding lump, or a person and her animal body, would differ qualitatively even if they coincided throughout their entire careers.
7. The view that qualitatively different things can coincide materially is called the metaphysic of constitution or constitutionalism. The name alludes to the fact that whenever two things coincide materially, one of them is supposed to “constitute” the other. Few of those who speak of constitution bother to say what they mean by it, and those who do say different things₂₄. But most agree that constitution is necessarily asymmetric and irreflexive: two things cannot constitute each other, and nothing can constitute itself. (Material coincidence, by contrast, is an equivalence relation.) And when constitutionalists take two things to coincide materially, they usually agree about which constitutes which: people are constituted by their bodies, for instance, which in turn are constituted by masses of matter, and not vice versa. (Do not confuse constitution with composition. Constitution by definition relates one thing to one thing, whereas many things can jointly compose something.)
8. Constitutionalism is a principle about material things in general, and not about ourselves in particular. It doesn’t say what you and I coincide with, or even whether we coincide with anything. For all it says we might be animals coinciding with masses of matter (Thomson 1997₂₅), or material objects that do not coincide with anything else, or even immaterial things. Most constitutionalists, though, say that we are non-animals coinciding materially with human animals. Or at least this is the usual situation. Although we are all non-animals, they say, we may not all coincide with animals. Perhaps some of us coincide with a thing made up of a human animal and a plastic knee. Someone who had a cerebrum transplant might coincide with different animals at different times, and briefly coincide with a naked cerebrum. Perhaps, by gradual replacement of parts, someone might one day come to coincide with a wholly inorganic machine. If we are lucky, we might be constituted in the next world by something glorious and indestructible but not in any recognizable sense biological. In this regard, say the constitutionalists, we are like statues, which by careful replacement of parts can coincide with different lumps of matter at different times. Coinciding with a particular human animal is supposed to be only a contingent and perhaps a temporary feature of us. But most constitutionalists say that we must always coincide with some material thing other than ourselves, be it animal, machine, or what have you: we could not become immaterial.
9. Most constitutionalists say that we have certain mental properties essentially, or that some sort of psychological continuity is necessary for us to persist through time. It follows that we come into being later than our animal bodies do: you appeared when a human animal reached a certain point in its development – perhaps when it acquired those mental features that distinguish people from non-people. Depending on what those features are, this could happen at any time between the appearance of the first mental properties five or six months after fertilization and the onset of full self-consciousness a year or two after birth. And you ordinarily cease to exist when your animal body ceases to support the relevant mental features, as it would if it lapsed into a persistent vegetative state. Here again we are like clay statues, which constitutionalists say come into being when a lump of clay is modelled in a certain way, and perish, outlived by the lump, when squashed.
10. I will call the view that you and I are non-animals coinciding with animals – as opposed to constitutionalism in general – the constitution view₂₆.

1. Whether the constitution view is right depends largely on the truth of constitutionalism in general. If constitutionalism is false, so is the constitution view. If constitutionalism is true, on the other hand, and if we are material things, it will be hard not to accept the constitution view.
2. Suppose constitutionalism is true. Then statue-shaped lumps of clay in the right circumstances constitute things that are essentially statue-shaped: statues. Likewise, lumps of matter in the appropriate organic configuration constitute things that are essentially living: organisms. These are supposed to be paradigm cases of constitution. Now it doesn’t strictly follow from this that human animals in the right state and the right circumstances constitute things that are essentially able to think. It could be that statue-shaped lumps constitute things that are essentially statue-shaped but human animals never constitute things that are essentially thinkers. But that would be surprising. We should expect there to be an explanation for this important difference. The most likely explanation, it seems, would be that no material thing of any sort could think, essentially or otherwise: thinking beings are immaterial, and are not constituted by anything. (Remember, constitutionalism in general does not imply that we are material things.) The claim that thinkers are material, but don’t stand to human animals as clay statues stand to lumps of clay, sounds rather unprincipled.
3. If human animals of the right sort constitute essential thinkers, it is but a short step to the constitution view. Those essential thinkers would not themselves be animals, for no animal is essentially able to think. And it would be hard to find any reason to suppose that we were the animals and not the essential thinkers. On the contrary, the apparently widespread and deeply held conviction that we are essentially able to think will be a reason to suppose that we are things constituted by animals.
4. So the truth of constitutionalism in general would provide a fairly strong case for our being non-animals constituted by animals. (Someone might say that we are constituted by brains or other parts of animals, rather than by whole animals. I see this as a near variant of the constitution view, and most of what I will say about the view that we are constituted by animals applies equally to the view that we are constituted by brains.) The constitution view and constitutionalism in general are likely to stand or fall together. In effect, then, any argument for constitutionalism is an argument for the constitution view. If we can find any case where qualitatively different objects coincide materially, that will be a reason to suppose that we are constituted by animals (or perhaps by brains). And there are plenty of arguments for constitutionalism.
5. Consider first the clay-modelling puzzle₂₇. Take a lump of clay of nondescript shape and knead it into the form of, let us say, Margaret Thatcher. (Iron might be a more appropriate material in this case, but clay suits our purposes better.) Then squash the lump and model it into a cube. We seem to have here a material object – a lump or mass or portion of clay – that is first shapeless, then Thatcher-shaped, then cubical. There also seems to be, for a time, a statue of Thatcher. During that time the lump and the statue coincide materially. Yet we wouldn't say that the statue starts out shapeless and ends up cubical. We wouldn't say, "See that statue? It was nothing but a shapeless lump this morning. Tomorrow it will be a cube." The statue doesn't seem to start out as a non-statue, become a statue for a while, and then revert to being a non-statue. It doesn’t merely cease to exist as a statue when we squash it. It seems to go out of existence altogether.
6. It is easy enough to make this into an argument for constitutionalism. The story invites us to accept these claims:
   • 1. There is a lump of clay that is first shapeless, then Thatcher-shaped, then cubical.
   • 2. There is a statue that is never shapeless or cubical.
   • 3. The statue coincides materially, while it exists, with the lump.
   1 and 2 imply that there are two different clay objects in the story. They have to be two because they exist at different times, and because one is shapeless for a time and the other is never shapeless. It follows, given 3, that two different things can coincide materially.
7. Now this is not yet constitutionalism – not quite. Constitutionalism says that different things can not only coincide materially, but also differ qualitatively while they coincide. If the statue and the lump were qualitatively exactly alike while they coincided – if both were statues and both were lumps, and they shared all the same dispositions, essential properties, and so on – then the clay-modelling puzzle would support only the claim that you and I coincide materially with human animals numerically different from us. That is not yet the constitution view, since it doesn’t imply that we in any way qualitatively different from those animals.
8. To make the clay-modelling puzzle into an argument for constitutionalism we need to establish that the statue and the lump in the story differ qualitatively while they coincide. But that is just what the story suggests. According to the story it is no accident that the lump persists throughout the various changes of shape while the statue does not. If we took a thousand shapeless pieces of clay and kneaded each of them first into the shape of Thatcher and then into a cube, it would always be the lump – the thing that was first shapeless – that survived the loss of its human shape and became cubical. The statue – the thing that comes into being when the lump becomes Thatcher-shaped – would always perish when it loses that shape. That is because lumps have the capacity to survive those changes and statues lack it: lumps, but not statues, have the modal property possibly surviving radical changes of shape. If that is right, then the lump and the statue in our story have different qualitative properties while they coexist. In fact they would differ in this way even if they were to coincide throughout their careers. So the story suggests a fourth claim as well:
   • 4. The lump has a qualitative property, while it coincides with the statue, that the statue then lacks.
   Grant this, and constitutionalism follows.
9. There is no easy way to avoid this conclusion. If constitutionalism is false, one of the four claims must be false; yet they all seem to be true.
10. We have already considered affirming the first three claims while denying the fourth – the view that the lump predates and outlives the statue but is qualitatively identical with it while they coincide. It is unsurprising that no one holds that view.
11. A more likely way out would be to accept the first claim but deny the second: to say that there is a lump that changes it shape twice, but there is no statue that is always Thatcher-shaped. The only large clay object in the story is the lump, which simply happens to be temporarily Thatcher-shaped. No new material object comes into being when we knead the lump into that shape, and none ceases to be when we squash it. No two things coincide materially. Call this suggestion lumpism.
12. Another possibility is that the second claim might be true and the first false: there is, if you like, a statue but no lump. Perhaps, when we knead the clay into the shape of the former Prime Minister, a new object, a statue, comes into being, and perishes when we squash it. But no clay object persists through these changes: nothing is first shapeless, then Thatcher-shaped, then cubical. (What about the clay? Isn’t it first shapeless, then Thatcher-shaped, then cubical? Maybe so; but that doesn’t obviously imply that any material object changes its shape. Perhaps the expression ‘the clay’ refers not in the singular to any one thing, but in the plural to a lot of particles₂₈ – particles that never compose anything that can survive radical changes of shape. That is, perhaps the clay is merely a lot of particles and not a large composite object.) Call the view that 1 is false and 2 is true statuism.
13. The trouble with these proposals is that they are hard to generalize (Olson 1996₂₉). Though they may sound attractive in the case of the statue and the clay, they are implausible in other cases.
14. Take lumpism, the idea that statues are merely statue-shaped lumps. If there are such things as lumps of clay, there ought to be such things as lumps of flesh and bone as well. Why should clay particles stuck together compose lumps, but not flesh particles stuck together? But although it may sound attractive to say that clay statues are just special lumps of clay, it is not plausible to say that living things are just special lumps of flesh.
15. Lumps of flesh ought to be able to survive crushing if lumps of clay can. So the lump composed of the flesh particles of a dog ought to be able to survive the same sorts of radical changes of shape as a lump of clay can survive. But the dog cannot survive that. If something analogous to squashing the statue and making the clay into a cube were to happen to the dog, the dog would not merely change its shape. It wouldn’t just cease to be a dog, and come to be a cubical piece of meat. Surely the dog would cease to be altogether. Or consider that dogs can survive wholesale changes of parts, owing to metabolic turnover. Lumps of flesh cannot survive this: if you take away half a lump’s particles and replace them with new ones – even if you do it gradually – you end up with a numerically different lump from the one you began with. So say those who believe in lumps, anyway. If there are such things as lumps of flesh, the lump now coinciding with a dog is not the one that coincided with it a year ago. A dog is not a lump of flesh. If there is only a lump of flesh in the dog story, in the way that according to lumpism there is only a lump of clay in the statue story, then there are really no dogs or other organisms at all: what appears to be a persisting organism is reality a series of numerically different lumps, each taking on organic form only briefly.
16. Those who would avoid constitutionalism by saying that statues are just lumps are likely to end up concluding that all material things are lumps – things that can survive radical changes of shape but cannot be composed of different particles at different times. They will arrive at a general “lump ontology”. Because most familiar material objects – organisms, artefacts, and ourselves as well, if we are material – are not lumps, the lump ontology implies that there are no such things. That may not be a reductio ad absurdum of the lump ontology, but it shows how tough-minded you have to be to accept it. Lumpism offers no easy way round the clay-modelling argument.
17. Now consider statuism, the idea that the only Thatcher-shaped object in the story is the statue, which has that shape throughout its career. What's wrong with that? Well, if clay particles arranged in the shape of Thatcher compose a clay statue, we should expect the organic particles that compose Thatcher herself to compose something analogous to a statue – not a statue, exactly, but something of the same metaphysical sort as a statue, with the same persistence conditions: a "statue-type object". (If you believe that Thatcher is immaterial, consider her animal body.) What principled reason could there be to suppose that clay particles arranged in the shape of Thatcher compose a statue-type object but flesh particles arranged in that way do not?
18. But Thatcher can survive things that no statue-type object could survive. She grew in size enormously in the course of her development. She could become a good deal smaller as well. She could survive the loss of her arms and legs. Given enough life-support machinery she could probably even survive as a severed head. No one thinks that a clay statue could have that sort of history. Thatcher herself is therefore no statue-type object. If the only Thatcher-sized material thing in the story is a statue-type object, then there is no such thing as Thatcher. Or if Thatcher really is a statue-type object, her history and persistence conditions are radically different from anything anyone ever thought. Those who would avoid constitutionalism by saying that lumps are really statue-type objects are likely to end up with something at least as repugnant as the lump ontology.
19. You might find my attempts to generalize lumpism and statuism too crude. Maybe the claymodelling story is disanalogous to the stories of dogs and prime ministers that I have tried to compare it with. Perhaps clay particles arranged in human or in canine form compose lumps of clay, but organic particles arranged in human or canine form don’t compose lumps of anything. Or maybe clay particles arranged statuewise compose clay statues but organic particles arranged in human form do not compose fleshy statue-type objects. That might enable us to resist the clay-modelling argument and avoid constitutionalism without going to such loony extremes as the lump ontology. The trouble with these suggestions is that they sound unprincipled. One would like to think that there was some reason why clay particles arranged in human form compose lumps or statue-type objects while organic particles arranged in that way do not. Claims like these ought to fit into some broader and more systematic picture of the ontology of material objects. Otherwise we ought to worry that they are more wishful thinking than reliable insight. And it is hard to come up with such a picture₃₀.
20. A more radical response to the clay-modelling puzzle is to reject both 1 and 2: there are neither lumps of clay nor clay statues. Of course, there is something there that sculptors work and has aesthetic value. Perhaps there are clay particles. Sculptors arrange some of these particles in special ways, with certain intentions and in special circumstances. We describe this situation loosely by saying such things as, “She has made a clay statue of Margaret Thatcher.” But really there are no clay objects, but only particles. Clay particles never compose anything: there is nothing that has many clay particles as parts and every part of which overlaps at least one of those particles. There is never any larger thing for clay particles to be parts of. The clay particles in our story start out stuck together in a nondescript fashion, then get arranged in a way that we describe as Thatcher-shaped, and end up arranged cubically. But nothing in the story is literally Thatcher-shaped or cubical. This would enable us to describe the clay-modelling case without committing ourselves to constitutionalism. We might call this proposal the sparse ontology.
21. Some philosophers have trouble understanding the sparse ontology. What is the difference, they ask, between things' being lumped together and their composing a lump? Given that some clay particles cohere together and don't cohere with any other clay particles, how can it be a further question whether there is a lump of clay there? We might as well say that there are many people gathered in the street but there is no crowd, or that there is a left shoe and a matching right shoe but no pair of shoes. How could that be a serious view? It is hard to know how to respond to this.
22. Consider the claim that any objects whatever, no matter what they are like in themselves or how they are arranged, always compose something. That is, for any things at all, there is something that has all those things as parts, and all the parts of which share a part with one or more of those things – something numerically different from any of those parts, unless there is just one of them. Call this compositional universalism. Those who don’t understand the sparse ontology appear to be assuming this principle. In fact they seem to think that it cannot intelligibly be doubted or denied: it is true solely by virtue of the meanings of the words used to state it, and in such an obvious way that anyone who understands the words ‘thing’, 'compose', and 'something' must see, on reflection, that it is true. Their view is apparently that it is a logical principle, like the principle of non-contradiction. Given that there are such things as your left leg, St Paul's Cathedral, and the planet Mars, anyone who fails to see that there is also an enormous disconnected material thing composed of those three objects is simply confused.
23. That is not how it seems to me. I find it eminently doubtful whether there is anything made up of your left leg, St Paul's Cathedral, and the planet Mars. It is not only doubtful whether those three things compose a “genuine object” or a thing with natural boundaries or anything of the sort that we have reason to pay any mind to. It is doubtful whether they compose anything at all. Universalism looks to me like a substantive metaphysical principle. It is not at all like the principle of non-contradiction. It is more like the claim that God exists. It might be true, and then again it might not be. But if we can meaningfully ask whether just any things compose a larger thing, how could it be meaningless to ask whether clay particles lumped together compose anything?
24. This is not going to satisfy those who have tried hard and failed to understand the sparse ontology. I can refer them to other sources (for instance₃₁ van Inwagen 1990b: 6-12 and 1994; Merricks 2001: 12-28). And we will see in chapter 9³³ that what we say about composition has important implications for what we are.
25. Even if we can understand the claim that there are no lumps or statues, though, we may find it hard to believe. The result of kneading some clay into the shape of Thatcher certainly appears to be a medium-sized clay object. At any rate it takes some doing to get people to take seriously the idea that there might be nothing there but particles. The proposal also raises difficult theoretical questions. For instance, if the particles in our story don't compose a lump of clay or anything else, when do particles compose something? What would it take for particles to compose something, if not their being lumped together? More to the point, if clay particles arranged in the form of a human being never compose anything, why suppose that organic particles arranged in human form compose something? Surely there is no ontologically significant distinction between clay particles and organic particles. If there are no lumps or statues, how could there be any people – unless people are immaterial? We will return to these matters in §9.5³⁴.
26. Other opponents of constitutionalism accept that the statue and lump exist and have different careers, but deny that they coincide materially. Despite appearances, they say, there are no smaller things that compose both the lump and the statue. The lump has temporal parts that do not overlap with any parts of the statue, such as the cubical part of it located later than its statue-shaped part. If a statue and a lump were to have all the same parts, including temporal parts – if the god of the philosophers were to create a clay statue out of nothing and then annihilate it without changing its shape, for instance – they would be one and the same. This proposal at least provides a systematic way of avoiding material coincidence. However, it requires the contentious assumption that statues and lumps, and presumably all persisting objects, including ourselves, are made up of temporal parts. We will return to it in Chapter 5³⁵.

1. The clay-modelling puzzle is just one of many considerations about the ontology of material objects that support constitutionalism. Here are two more.
2. The replacement puzzle, like the clay-modelling puzzle, suggests that each ordinary material object coincides materially with something of a different sort (Thomson 1998₃₅). Suppose we break off an arm of our clay statue and burn it in a very hot fire, then replace it with a new arm made of different clay. Then the argument is this: There is a clay statue that persists throughout the story, and has first one arm and then another. There is also a statue-shaped lump of clay coinciding with the statue before the replacement, and another statue-shaped lump of clay coinciding with it afterwards. I say another lump because the original lump doesn’t get smaller when we destroy part of it, as the statue does, and then regain its original size when a new part is provided. No; the lump ceases to exist when the arm is destroyed. And when the new arm is attached, a new lump comes into being. Or perhaps a previously disconnected lump – one existing in two detached pieces – comes to be a connected lump – one that is all in one piece. Either way, the new lump is not the old one. Since the statue coincides first with one lump and then with another, it cannot be identical with either lump. More generally, every clay statue has a property that no lump of clay has, namely the capacity to have different clay parts, or to be made of different clay, at different times. It follows that no clay statue is identical with any lump of clay – not even a clay statue that never has any of its parts replaced. Once more we have qualitatively different material objects coinciding materially. The alternatives to this conclusion are similar to those in the clay-modelling puzzle.
3. Then there is the ancient amputation puzzle. Consider an ordinary human organism, Peter. Presumably there is such a thing as Peter's left hand. And if there is such a thing as his left hand, there ought to be such a thing as his "left-hand complement" as well: something composed of all of Peter's particles save those that compose his left hand. Call it Pete. Pete and Peter are not the same thing: Peter is bigger. Pete would seem to be one of Peter's parts. Now imagine that Peter loses his left hand. Better, let the hand be entirely destroyed. This is surely something that Peter could survive. Suppose he does. Then he gets smaller by a hand. But what about Pete? What happens to it when Peter loses his hand? If the loss is clean and quick, Pete need not be directly affected. Only its surroundings would change. So it seems that both Peter and Pete would exist after the amputation. How would they then relate to one another? It seems that they would coincide materially: the very atoms that compose Peter would compose Pete as well. But they cannot be the same thing, for they were different things before they coincided. If these assumptions are all correct, then this is another case of material coincidence (Thomson 1983₃₆). (It will be a genuine case of constitution only if Peter and Pete differ qualitatively while coinciding; but those who accept the rest of the story are unlikely to deny this.) Again, there is no easy way of avoiding the constitutionalist conclusion. We will return to the amputation puzzle in §7.3³⁸-§7.4³⁹.

1. The constitution view promises to combine the apparent fact that we are living material things with the conviction that we have identity conditions or essential properties different from those of human animals. It claims to have all the virtues of animalism with none of its vices. And it is part of a package that appears to solve a number of vexing metaphysical puzzles. This might make the constitution view sound like a gift from the gods.
2. But it is too good to be true. For one thing, the constitution view shares some of the objections levelled against animalism. In §2.9⁴⁰ we saw that animalism was incompatible with the widely held claim that facts about mental unity determine how many of us there are at any one time. I will argue in §6.4⁴¹ that the constitution view is also incompatible with that claim. (We will consider another objection to both animalism and the constitution view in §9.3⁴².)
3. The constitution view has troubles of its own as well. The most obvious is the thinking animal problem. The constitution view says that we are not identical with our animal bodies. As we saw in §2.3⁴³, this implies that one of three things must be the case: there are no human animals at all, or human animals cannot think in the way that we do, or each of us shares all our thoughts with another being. Which of these unsavory consequences should friends of the constitution view accept?
4. They cannot deny that there are human animals. That such an animal constitutes you is part of their view. Someone might deny that there are animals and say that we coincide with lumps of flesh or masses of matter instead. (Though this is not strictly a version of the constitution view as I have characterized it, it is a close relative.) In that case there would be no thinking animal problem; but there would still be a thinking-lump problem. It might perhaps be easier to explain why a lump of flesh in human form could never think than to explain why a human animal couldn’t. But even so, advocates of this view face the considerable challenge of explaining why there are no human animals in a way that is compatible with their view of what we are. Why might there be no human animals? It might be because there are no material things at all: the physical world is an illusion. Or maybe there are no composite material things, but only elementary particles (§8.5⁴⁴). Or it might be because nothing can have different parts at different times (§7.3⁴⁵-§7.4⁴⁶). None of these claims are compatible with anything like the constitution view. Any grounds for denying the existence of animals are likely to grounds for denying the existence of any sort of “constituted” material things that we could be.
5. So according to the constitution view there are human animals coinciding materially with us. What if those animals think in the way that we do? Then there are two beings thinking your thoughts, you and the animal. And that is too many. We saw the problems this raises in §2.6⁴⁷. The human animals coinciding with us ought to count as people. Human people would then come in two kinds: animal people and the non-animal people they constitute. That makes it hard to see how we could know whether we are the non-animal people or the animal people that constitute them.
6. Baker₄₇ seems to think that these problems dissolve once we see that the animal constitutes the person (2000: 169-179, 191-204; 2002: 42). If we state the constitution view correctly, she claims, the question of how we know we are not our animal bodies does not arise. She says that I am an animal as well as a person, and that the animal constituting me – call it A – is a person as well as an animal. But A and I are animals and people in different senses. I am an animal only derivatively, she says, insofar as an animal constitutes me. A, however, is an animal non-derivatively: it is an animal independently of its constituting or being constituted by anything. Contrariwise, A is a person only insofar as it constitutes a person, whereas I am a person non-derivatively, independently of any constitution relations I enter into. How does that help? In particular, how does it enable me to know that I am not A? Well, I can know that I am a person, Baker says, because I can think first-person thoughts, and only a person can think first-person thoughts (that is her definition of 'person'). Yet A is also a person. How do I know which person I am? Baker says that although I am a person, and A is a person, and we are numerically different, we are not two people. Whenever one thing constitutes another, she says, they are one thing. And because there are not two people there, it makes no sense to ask which one I am.
7. I don’t know what Baker means when she says that that A and I, though numerically different, are one person. But whatever it means, I do not see how it could help me to know that I am not A. There is A, Baker says, and there is the person A constitutes – Olson – and they are numerically different. So I ought to be able to ask whether I identical with Olson or with A. If I am identical with Olson and not with A, as Baker claims, how can it be impossible to ask whether I am Olson or A? And if I can ask whether I am Olson or A, I can also ask what grounds I have for accepting one answer to this question rather than another – just the problem Baker’s account was supposed to do away with.
8. Friends of the constitution view will want to solve the thinking-animal problem by denying that human animals can think, or that they can think in the way that we think₄₈. But as we saw in §2.5⁵⁰, this is a hard thing for a materialist to maintain: if you say that some material things can think, you will find it hard to argue that biological organisms cannot. It is especially hard to argue that physically indistinguishable things in the same surroundings – and according to the constitution view you and I are indistinguishable in this way from our animal bodies – can nonetheless differ radically in their mental capacities. There appears to be no difference between you and your animal body that could account for any psychological difference.
9. Someone might say that what prevents human animals from thinking is not any defect in their physical structure or surroundings or history, but that they belong to the wrong metaphysical kind. The kind might be biological organism: perhaps human animals cannot think because they are organisms. We, by contrast, are able to think because we are not organisms (as well as having the right microstructure, surroundings, history, and so on).
10. Why should a thing’s being a biological organism prevent it from thinking (or from thinking in the way that we think)? Maybe organisms can’t think because they have the wrong persistence conditions. We considered this rather unlikely view in §2.5⁵¹. Even if it were true, though, it would not yet explain why we can think and human animals can’t. That is because it doesn’t explain why we are not animals ourselves. Of course, it is part of the constitution view that we are not animals. But even if we coincide materially with animals numerically different from ourselves, the question remains: what makes us non-animals? We are physically identical with human animals. We have the same developmental and evolutionary history as those animals have (we weren’t cooked up in the lab by mad scientists). How could things like that – beings that no biologist could ever distinguish from animals – not be animals?
11. Perhaps we are not animals because we lack the identity conditions of organisms: maybe our identity over time, but not that of organisms, consists in some sort of psychological continuity. But this raises a further question: what could give us different identity conditions from those of human animals? How could material things with the same physical properties (or at any rate the same microstructure) in the same surroundings differ in the sort of thing they can survive? What is it about human animals that enables them to survive in a persistent vegetative state (for instance), when we – beings otherwise exactly like them – cannot survive it?
12. I have asked why we can think but the animals coinciding with us cannot, what makes us non-organisms despite being physically indiscernible from organisms, and what could give us different identity conditions from the animals coinciding with us. We might also ask what could make one object a statue and another object, physically indiscernible from it and with the same surroundings, a mere statue-shaped lump. These are all instances of a more general question, which we might call the indiscernibility problem: how can putting the same parts together in the same way in the same circumstances give you qualitatively different wholes? If the same atoms can compose two things at once, what could make those two things qualitatively different? What could give them different mental properties, or different persistence conditions, or different modal properties? If atoms really could compose more than one object at once – if numerically different objects could coincide materially – should we not expect those objects to be qualitatively identical?
13. Constitutionalists evidently do not expect this. They are not surprised that two objects that are otherwise indiscernible should differ systematically in their mental or modal properties. Why not? Presumably it is because they take these differences to be primitive or brute: not explainable in terms of other differences. There is no saying why you would go along with your transplanted cerebrum and your animal body would not, or why you can think but the animal can’t, because there is no other difference between you and the animal that could explain it. A human animal’s inability to think is a primitive and basic feature of it. It just can’t, and that’s all there is to be said. Asking why a human animal cannot think is like asking why an electron is negatively charged: there is no more basic level of properties underlying it, in terms of which it could be explained. Or maybe animals cannot think because they have the wrong persistence conditions to think (as Sydney Shoemaker argues), and their persistence conditions are brute. Each pair of coinciding objects must have some brute difference that explains the other differences between them.
14. Constitutionalists will point out that things must have some brute properties or other: a thing cannot have every property it has by virtue of its having some other property. So why shouldn’t a thing’s persistence conditions and mental properties be brute?
15. Not many philosophers would agree that mental properties are brute. That would mean that there is no explanation of why some beings are conscious or intelligent and others aren’t. At any rate there would be no explanation in many cases. It may be that having a certain sort of brain is necessary for being intelligent or conscious. But only some of the beings with that sort of brain would be intelligent and conscious, and there would be no saying why those beings with that sort of brain that are intelligent and conscious are intelligent or conscious, and why those that are not intelligent of conscious are not. No amount of information about a being’s brain structure, history, or surroundings would suffice to explain why it is conscious or intelligent. It would not even suffice to explain why a thing is likely to be conscious or intelligent, in the way that someone’s being dealt three aces explains why she is likely to win the poker game. This would make mental properties inherently mysterious.
16. Constitutionalists may want to say that things’ mental properties are not primitive, but only their modal properties – their persistence conditions, for instance, and other essential properties – and use the difference in the modal properties of coinciding objects to explain their other differences. But it will not be easy to say how the modal properties of human animals prevent them from having mental properties.
17. And there seems to be something fishy in the claim that things’ modal properties are primitive and independent of their non-modal properties. Suppose someone said that every person located in the northern hemisphere coincides materially with a being that is essentially in the northern hemisphere: an “essential northerner”. (Perhaps Antipodeans coincide with essential southerners; perhaps not.) If you were to cross the equator from the north, the essential northerner coinciding with you would necessarily perish. Now it might occur to someone to ask why that being cannot cross the equator. You can cross the equator. And your essential northerner is otherwise just like you. What stops him from crossing? This seems to me to be a legitimate question – a question we should expect to have an answer. But if the modal properties that coinciding objects don’t share are primitive, as constitutionalists are apparently committed to saying, it is not a legitimate question. That your essential northerner cannot exist outside the northern hemisphere would be a brute property of him, not explainable in terms of his having any other properties. Why can’t he get across the equator? He just can’t. Not even God could say why. That strikes me as an absurd thing to say.
18. Constitutionalists may say that it is absurd because essential northerners are absurd: it is absurd to suppose that any material object is essentially located in the northern hemisphere. And the existence of such things in no way follows from the claim that human animals coincide with beings that are essentially able to think in a certain way. That is fair enough – but then we shall want to know why human animals coincide with essential thinkers but not with essential northerners. What’s the difference between the two cases? What makes one absurd and the other respectable? This is the topic of the next section₅₁.

1. Let me mention two further worries about the constitution view. They may appear trifling compared with the thinking-animal problem. But many philosophers say that there is no such problem, or that it can be solved, or that it is troubling but outweighed by the constitution view's advantages; so these further worries may be important. In any case they are interesting in their own right.
2. The first has to do with when one thing constitutes another. Constitutionalists say that certain qualitative properties have the following feature: necessarily, when an object acquires one of them, a new object comes into being which coincides materially with the first and has that property essentially. So when a sculptor models a lump of clay into the shape of Thatcher, the lump comes to coincide with a new object – a statue – that has that shape essentially. The lump could have any shape you like, but the statue – that particular statue – could not exist without having that shape. (Perhaps it is only an approximate shape.) Or again: when in the course of its development a human organism acquires certain a mental property (the capacity for self-awareness or for first-person thought or what you), it comes to coincide with a new being- -a person – that has that property essentially. Because most constitutionalists say that in these cases the original object constitutes the new object, we might call them constitution-inducing properties.
3. There is a complication. Constitutionalists deny that constituting objects always share the properties of the things they constitute. For instance, human organisms that come to constitute people never become self-aware themselves (or at least not in the way that the people are self-aware); otherwise human animals and human people would hardly be worth distinguishing. Rather, the organism comes to constitute a being having the capacity for self-awareness essentially when it acquires some other property: perhaps a property, the having of which suffices for a thing to be self-aware if it belongs to the right kind or has the right identity conditions or the like. This other property might be the neural substrate of self-awareness, or the conjunction of the right neural substrate and appropriate surroundings₅₂. And of course some properties that constituted objects are said to have essentially are properties that nothing could acquire: having certain persistence conditions, for instance. (Nothing can start out having one set of persistence conditions and later exchange them for a new set, incompatible with the first.) If a thing's acquiring a property P suffices for it to come to constitute a new object that has a property Q essentially, let us say that P and Q are constitutional correlates. So a property P is constitution-inducing if and only if, necessarily, whenever an object acquires P, it comes to coincide with a new object that has a constitutional correlate of P essentially. (This allows that a property might be a constitutional correlate of itself.)
4. A general question now arises: What properties are constitution-inducing? What properties are such that when a thing acquires one of them, it comes to constitute a new object that has that property or a constitutional correlate of it essentially? In other words, under what circumstances does constitution occur? What alterations in a previously existing material thing call a new material thing into being? I take it that this question must have an answer. At any rate it must have an answer if constitutionalism is true.
5. I am not asking for a definition of constitution – an account of what it would be for one thing to constitute another. That would not tell us when constitution occurs. Suppose we agree that, by definition, x constitutes y if and only if x and y coincide materially and it is possible for y to cease to exist while x endures and retains the same parts, but not possible for x to cease to exist while y endures with the same parts. Clearly we could still disagree about when these conditions obtain: you might take many properties to be constitution-inducing while I say that few are, or none. We could be like people who agree about what absolute moral rights would be but disagree about which such rights there are.
6. There are three broad sorts of answers to this question. One is that no properties are constitution-inducing: that’s what opponents of constitutionalism say. Another is that all properties are constitution-inducing: any alteration to any material object, no matter how trivial, necessarily results in its coming to constitute a new object that has the property thereby acquired, or a constitutional correlate of it, essentially. Third, it may be that some properties are constitution-inducing and others are not.
7. Call the claim that all properties are constitution-inducing the generous view₅₃. It has the important advantage of being principled. If some properties are constitution-inducing and others aren’t, it seems fair to ask why the ones that are are, and why the others aren’t – just as if some people are rich and others aren’t, we can ask why the rich ones are rich and the others aren’t. The generous view can tell us why a given property is constitution-inducing: because all properties are necessarily constitution-inducing.
8. But the generous view is hard to believe. For one thing, every persisting material object has properties that change continuously: at every moment you acquire, among other things, a new shape and a new distance from the moon. So the generous view implies that you come to constitute a new being at every moment. During the time it takes you to blink your eyes, you pass through an uncountable infinity of shapes, and each of those minute alterations brings forth a new being, coinciding with you, that has that shape essentially.
9. Now this sort of ontological extravagance is not unique to the generous view: the ontology of temporal parts has a similar consequence. But the generous view is infinitely more extravagant than the ontology of temporal parts. Suppose you stand up, thereby coming to constitute a being, S, that is essentially standing. And suppose that as you are in the process of standing up, you also begin to frown, thereby coming to constitute a being, F, that is essentially frowning. Are S and F the same object, or different? (Suppose, if you like, that S and F go out of existence at the same time.) Do the things that have their posture essentially also have their facial expression essentially? We could ask the same questions about a thing’s mass, age, distance from the moon, pH, net electric charge, and so on. Any material object acquires a vast number of new properties at any moment: perhaps an infinite number. How many new objects does it thereby come to constitute? And how are the properties a thing acquires (or their constitutional correlates) distributed essentially across the objects it then comes to constitute?
10. As far as I can see, the only principled way of answering this question is to be generous once more (Bennett 1984: 354₅₄): when you stand and frown at once, you coincide with a being that essentially stands and contingently frowns, with a second being that contingently stands and essentially frowns, and with a third being that essentially stands and essentially frowns. More generally, for every non-empty subset of the set of qualitative properties you have at a given time, there is a being coinciding with you that has the members of that subset (or constitutional correlates of them) essentially and has the other properties you have at that time (or constitutional correlates of them) contingently. So if the number of properties you have at a given moment is n, the number of material things you coincide with at that moment will be a bit less than 2ⁿ. That’s a lot.
11. Worse yet, suppose you begin to think about Vienna. The generous view implies that you thereby come to constitute a being different from yourself that thinks about Vienna essentially (you don’t think about Vienna essentially). It follows that whenever you think about Vienna, you are one of at least two beings thinking about Vienna (far more than two if the suggestion of the previous paragraph is right; but never mind that). You ought to wonder which one you are: are you the one who was absorbed with things other than Vienna a moment ago, or the one that thinks about Vienna throughout the whole of its brief career? How could you ever know? This problem arises even if constitutionalists can explain why the animals that constitute us are unable to think in the way that we do: it has nothing to do with the mental properties of animals.
12. It is not surprising that most constitutionalists want to say that only some properties are constitution-inducing. But which ones? You might think that all and only intrinsic properties are constitution-inducing. But this is hardly better than the generous view. Many mental properties appear to be intrinsic; so it would seem to follow that when you come to be conscious after a night asleep, you come to coincide with a being that is essentially conscious. What’s more, some constitutionalists say that constitution-inducing properties can be extrinsic (Baker 1997₅₅).
13. Baker proposes that a thing x constitutes something y just when “y has whole classes of causal properties that x would not have had if x had not constituted anything (Baker 2000: 41₅₆).” If we use an anvil as a doorstop, she says, it holds open the door merely by virtue of properties it would have even if it didn’t constitute anything. It acquires no new classes of causal properties. A person, by contrast, has many properties that a mere organism would not have if it didn’t constitute a person, such as the capacity to think about the future. That, she says, explains why certain organisms constitute essential people, but anvils never constitute essential doorstops.
14. Baker concedes that this is not a satisfactory account of when constitution occurs, but she thinks it is a helpful guide. As far as I can see it is no help at all. Does the human organism now sitting in your chair constitute something? According to Baker’s proposal it does if there is something now coinciding with it that has certain causal properties that the animal wouldn’t now have if it didn’t constitute anything. Well, what causal properties would the animal now have if it didn’t constitute anything? One answer is that it would have precisely the causal properties that you now have, including your mental properties. Another is that the animal would lack many of the mental properties you now have unless it constituted something that had certain mental properties essentially. How can we decide between these two answers? Only, it seems, by finding out whether the animal constitutes something. Those who think it doesn’t will take the animal to have all the causal properties that you have; those who think the animal constitutes an essential thinker will take it to lack some of your causal properties. But whether the animal constitutes something is just what we wanted to find out. Those who have no idea when constitution occurs will be none the wiser for all Baker’s proposal tells them. The same goes for other putative cases of constitution.
15. You might think that Baker’s proposal gives at least a necessary condition for constitution, even if it is no good as a sufficient condition. That is, even if it tells us nothing about when constitution does occur, it might tell us something about when it doesn’t – ruling out the generous view that all properties are constitution-inducing, if nothing else. It might imply that an anvil doesn’t come to constitute an essential doorstop when we use it to prop open a door, because it thereby acquires no new classes of causal properties.
16. But I cannot see that it helps even here. Whether the anvil constitutes an essential doorstop depends on what causal properties the anvil would have if it didn’t constitute anything. Baker says that an anvil used as a doorstop has the same causal properties whether it constitutes anything or not. But why suppose that? Someone might think that putting an anvil in front of an open door necessarily causes it to constitute an essential doorstop, and infer from this that if the anvil didn’t constitute anything it wouldn’t keep the door open. Baker’s proposal would give her no reason to change her mind. Those who deny that anvils ever constitute essential doorstops will find Baker’s proposal equally consistent with their own view. Those who have no idea whether this is a case of constitution and want to find out will get no help from Baker’s proposal.
17. Of course, it might be silly to suppose that putting an anvil in front of a door would cause it to constitute an essential doorstop, with causal powers that the anvil lacks. But then some philosophers find it silly to suppose that the normal development of a human organism causes it to constitute an essential thinker, with causal powers that the organism lacks. Baker’s proposal does nothing to help resolve these disputes. If essential doorstops are silly and essential thinkers are not, we should like to know why this is so, and Baker’s proposal doesn’t say. It implies nothing at all about when constitution occurs.
18. I have never seen a serious answer to the question of which properties are constitution-inducing, apart from ‘all’ and ‘none’. Does it matter? Every philosophical claim raises questions that no one knows how to answer. Is this a problem for constitutionalism, or merely an interesting topic for further research?
19. I think it is a problem. For one thing, if we have no idea what properties, in general, are constitution-inducing and no idea how to find out, it ought to undermine our confidence in the claim that any particular property is constitution-inducing – all the more so given that constitutionalists disagree widely among themselves about these matters. That would cast doubt on constitutionalism generally, and on the constitution view of ourselves in particular.
20. More seriously, it is hard to combine the constitution view with any acceptable answer to this question. Most friends of the constitution view say that a human animal comes to constitute a person when it acquires a certain mental capacity (or its constitutional correlate). Suppose it is the capacity for first-person thought: when an animal acquires it, it comes to constitute a person that has that capacity essentially. This has the implausible consequence that we were never foetuses, and that human beings in a persistent vegetative state are not, and do not constitute, the people by whose names we continue to call them. But never mind that. A human animal acquires all sorts of mental capacities (or their constitutional correlates) in the course of its development from a foetus into an adult. For instance, it acquires the capacity to have bodily sensations. It also acquires the capacity to have beliefs and desires, the capacity to think about the future, the capacity to do arithmetic, and many more. Friends of the constitution view say that just one of these capacities is constitution-inducing. (Otherwise each of us would coincide with a different thinking being for each constitution-inducing mental capacity.) They say that a being can acquire the capacity to have sensations without thereby coming to constitute anything that has that capacity essentially, but no being can acquire the capacity to think in the first person without coming to constitute anything that has that capacity essentially. What is it about first-person thought that makes it, alone among mental properties, constitution-inducing?
21. If there is to be any answer to this question, it will involve the claim that the capacity to think in the first person is somehow uniquely special among mental capacities. But there doesn’t seem to be anything uniquely special about it. Constitutionalists may point out that the capacity for first-person thought is a prerequisite for a wide variety of other mental capacities, such as the capacity to plan for the future. But many other mental capacities are equally special: the capacity to think at all, for instance, is a prerequisite for an even wider variety of mental capacities. The claim that the capacity for first-person thought, but no other mental property, is constitution-inducing seems arbitrary and unprincipled. It isn’t just that we don’t know what makes it constitution-inducing, but that no such account seems possible.
22. Well, maybe it isn’t the capacity for first-person thought that is constitution-inducing, but rather the capacity to think at all. (This seems to be Shoemaker’s view.) That might sound less arbitrary. But is it? We have all sorts of capacities, some mental and some not. Why should that one in particular be constitution-inducing? Why not the capacity for sense-experience? Or the capacity to move? Or the capacity to breathe? I cannot see any way of answering these questions. As far as I can see, the constitution view is inconsistent with any principled account of what properties are constitution-inducing – short of the generous view, anyway.
23. Someone might wonder whether there has to be any principled account of what properties are constitution-inducing. There must be some answer to that question if constitutionalism is true, but why suppose that the answer must be intellectually satisfying and not simply arbitrary? Could it not be that the capacity for first-person thought is constitution-inducing and no other capacity is and there is no reason why? Well, maybe. But if the facts about which properties are constitution-inducing are arbitrary, it is hard to see how anyone could ever know them.

1. Here is the second worry. The constitution view says that I coincide materially with a certain animal. Its parts are my parts and my parts are its. We share our spatial boundaries. This may be only contingently true – maybe I could come to coincide with something partly or wholly inorganic – but it is the usual situation. Why should this be so? Why is my boundary the boundary of this animal? What makes all and only the animal’s parts my parts? What is it about my feet, for instance, that makes them, but not my shoes, or your feet, parts of me?
2. I am not asking for a causal or historical explanation of how my boundaries came to lie where they do, in the way that we might ask how Texas came to be a part of the United States. Never mind history. I want to know what it is about the way I relate to my feet now that makes them parts of me now. This is analogous to asking what current geopolitical facts make it the case that Texas is a part of the US, rather than, say, a part of Mexico or an independent state. We could not understand how historical events affect a thing's current boundaries unless we understood what sort of facts fix things' boundaries in the first place₅₇.
3. I take it that this question must have an answer. Our boundaries may be indefinite – there may be things that are neither definitely parts of me nor definitely not parts of me – but even so we must have boundaries. And where they lie is no accident. According to the constitution view, they ordinarily coincide precisely with those of a particular animal; and moreover in normal circumstances they continue to coincide with those of a human animal for the whole of our lives. This fact deserves an explanation.
4. If we were animals, what fixes our boundaries would be whatever fixes the boundaries of an animal. And we know what that is: an animal extends as far as its biological life extends (more or less: see §2.2⁵⁹). My feet are parts of this organism because they are caught up in its life; my shoes are not parts of it because they are not caught up in its life: they don’t respond in the right way to the organism’s metabolic activities. If I am not an animal, though, the mere fact that something is a part of this animal cannot make it a part of me. Nor can the fact that something is caught up in a certain animal’s life, by itself, make it a part of me.
5. What might determine where our boundaries lie if the constitution view were true? It is not easy to say. Locke suggested that we extend spatially as far as our "consciousness" extends. Those particles, he said, that are "vitally united to this same thinking conscious self, so that we feel when they are touch'd, and are affected by, and conscious of good or harm that happens to them, are a part of our selves" (1975: 336₅₉). The idea is that my feet are parts of me because I can feel them, or more generally because changes in them have an immediate effect on the nature of my experience. Of course, I cannot feel any particular atom, and no single atom is such that changes in it has an immediate effect on my experience, yet on Locke’s proposal many atoms are parts of me. But an atom can be a part of a larger thing that I can feel, and any part of a part of me is itself a part of me. So we could say that our parts are those things that we can feel in this way, and the parts of those things. Most friends of the constitution view think that psychological facts fix our temporal boundaries – when we begin and end. This may lead them to think for similar reasons that psychological facts fix our spatial boundaries as well, just as Locke’s proposal has it.
6. But a moment’s reflection shows that this will not do. It implies that a limb that was completely numb would not be a part of you. An accident that left you numb and paralyzed from the neck down would literally reduce you to a head (or to something constituted by a head). An event that left you entirely numb, depriving you of all sensory input, would leave you with no parts at all. Since you could not exist without having any parts (even yourself, your “improper” part), this means that you could not survive in such a state, even for a moment. I take that to be absurd. For that matter, Locke’s proposal implies that even in ordinary circumstances our boundaries are not the boundaries of a human animal. The reason is that you can't feel the whole of your animal body. You can't feel the blood flowing through your aorta, for instance. On Locke's proposal, neither your blood cells nor the atoms composing them are parts of you: you are smaller than a human animal. It follows that no animal constitutes you. So Locke’s proposal is actually incompatible with the constitution view. Locke’s proposal may seem like a straw man. There is probably a better account of what determines our boundaries that is compatible with the constitution view. But I have no idea what it might be.
7. You may suspect that there is really no problem here. It follows from most definitions of constitution that whenever one thing constitutes another, the two things share their boundaries and their parts (or at any rate there are things that compose each of them). Thus, given that we are constituted by human animals, as the constitution view says, our boundaries will be the boundaries of those animals. Those people who come to be constituted by things other than organisms will share their boundaries with whatever it is that constitutes them then. Friends of the constitution view might therefore seem to have a perfectly good account of what determines our boundaries: they are the boundaries of whatever it is that constitutes us.
8. Our being constituted by animals would indeed imply that our boundaries are those of our animal bodies. But rather than answering the question of what determines our boundaries, this proposal merely relocates it. The question is now what makes it the case that animals constitute us. Why not things a bit larger or smaller than animals? Granted, it seems to be human animals that constitute us if anything does, and not something bigger or smaller. But that appearance is nothing more than the fact that we appear to extend all the way out to our skin and no further. It does nothing to explain why we are constituted by things that extend all the way out to our skin and no further, rather than by things with a greater or lesser extent. This is a question that friends of the constitution view need to answer. At any rate they are committed to its having an answer. And it will be no easier for them to answer than the original question of what determines our boundaries.
9. This problem might not seem unique to the constitution view. Maybe it is a good question why animals constitute us rather than bigger or smaller things. But don’t other accounts of what we are face a similar question? Take animalism. What could make it the case that we are animals, rather than things bigger or smaller than animals? Don’t animalists need to answer this question? And won’t their answer suit the constitution view equally well? Suppose animalists can say why it is that we are animals and not bigger or smaller things. Wouldn’t that also explain why it is that we are constituted by animals and not by bigger or smaller things, if indeed that is the case?
10. It is true that both animalists and constitutionalists face similar questions about what makes it the case that we are animals and constitutionalists about what makes it the case that animals constitute us. We will consider how animalists might answer their question in §9.3⁶¹. But the constitutionalists’ question will be harder to answer than the animalists’ question. There are two things animalists can say about why we should be animals rather than bigger or smaller things: they can deny that those bigger and smaller things can think; or they can deny that such things exist. (A third option would be to accept that such things exist and can think, and appeal to the “personal-pronoun revisionism” discussed in §2.6⁶² to account for our ability to know that we are not those bigger or smaller things. I take it that neither animalists nor constitutionalists will be happy with this.)
11. What might prevent something bigger or smaller than a human animal from thinking? If anything, we should expect it to be the fact that it is not an organism. That only an organism could think is sometimes said to belong to the very idea of thinking (see §4.2⁶³). But whatever merits this may have as an explanation of why you are an animal rather than something bigger or smaller than an animal, it cannot explain why you are constituted by an animal rather than by something larger or smaller than an animal. In fact it is incompatible with the constitution view, which says that non-organisms do think.
12. The second possible reason for supposing that we are animals rather than things larger or smaller than animals is that there simply are no such larger or smaller things. The only material thing that is even a candidate for thinking your thoughts – the only one that includes the neural machinery that makes your thought possible – is the animal. You are not your left-hand complement, for instance, because there is no such thing; those of your particles not located in your left hand do not compose anything. There are fewer material objects in the world than some philosophers have supposed. The animal is the only thing there that could be you.
13. Of course, if there is nothing bigger or smaller than the animal that could be you, there is also nothing bigger or smaller than the animal that could constitute you; so this story would also serve the constitutionalist as an account of what determines our boundaries. But no constitutionalist would touch it. Even if this sort of sparse ontology of material objects is consistent with the letter of constitutionalism, it goes completely against its spirit. One of constitutionalism’s main attractions is that it provides a supply of material things rich enough to make true almost any metaphysical view about material things. Are you inclined to think that clay statues and lumps of clay, organisms and masses of matter, and people and organisms all exist, yet are numerically different, owing to their different essential properties? Constitutionalism will give you those objects. It is designed precisely to squeeze more things than you would have thought possible into the same place. It is an ontological shoehorn, the metaphysical equivalent of high-rise building. Once we start economizing with material objects, and especially if we have to economize on such a scale as to deny the existence of any parts of human beings big enough to include brains, constitutionalism loses its appeal.
14. We can illustrate this conflict by recalling the metaphysical puzzles about material objects that we considered earlier in this chapter. A sparse ontology of material objects can solve those puzzles by denying the existence of one or more of the entities whose apparent coincidence was puzzling – the statue, the lump, the person’s hand complement, and so on. That solution has a high intuitive cost: the sparse ontology is hard to believe. (See §7.4⁶⁴ and §9.5⁶⁵ for more on the sparse ontology.) Constitutionalism was supposed to solve those problems without that cost. It has a cost of its own, of course: something must be done about the thinking-animal problem, not to mention the problem of when constitution occurs; but you might still prefer it to the sparse ontology. If constitutionalism requires a sparse ontology as well, this benefit is lost: there is no point in paying twice for the same thing, especially if the price is high. That would be a good reason to consider accepting the sparse ontology without constitutionalism.
15. So advocates of the constitution view cannot say what animalists say about what determines our boundaries. They need to give another answer; or at least they are committed to there being another answer. But again, I have no idea what answer they could give. This seems to be a serious problem.

Text Colour Conventions (see disclaimer)

1. Blue: Text by me; © Theo Todman, 2023
2. Mauve: Text by correspondent(s) or other author(s); © the author(s)