Page d'accueil Metaphysics - A Very Short Introduction

Metaphysics - A Very Short Introduction

0 / 2.0
Avez-vous aimé ce livre?
Quelle est la qualité du fichier téléchargé?
Veuillez télécharger le livre pour apprécier sa qualité
Quelle est la qualité des fichiers téléchargés?
Metaphysics is traditionally one of the four main branches of philosophy, alongside ethics, logic and epistemology. It is an area that continues to attract and fascinate many people, even though it is generally thought to be highly complex and abstract. For some it is associated with the mystical or religious. For others it is known through the metaphysical poets who talk of love and spirituality. This Very Short Introduction goes right to the heart of the matter, getting to the basic and most important questions of metaphysical thought in order to understand the theory: What are objects? Do colors and shapes have some form of independent existence? Is the whole just a sum of the parts? What is it for one thing to cause another rather than just being associated with it? What is possible? Does time pass? By using simple questions to initiate thought about the basic issues around substance, properties, changes, causes, possibilities, time, personal identity, nothingness, and consciousness, Stephen Mumford provides a clear and down-to-earth path through this analytical tradition at the core of philosophical thought.
Oxford University Press
ISBN 10:
ISBN 13:
EPUB, 3,71 MB
Télécharger (epub, 3,71 MB)

Mots Clefs


To post a review, please sign in or sign up
Vous pouvez laisser un commentaire et partager votre expérience. D'autres lecteurs seront intéressés de connaitre votre opinion sur les livres lus. Qu'un livre vous plaise ou non, si vous partagez honnêtement votre opinion à son sujet, les autres pourront découvrir de nouveaux livres qui pourraient les intéresser.

Economia Brasileira Contemporanea

PDF, 97.16 MB
0 / 0

Brasil, Raizes Do Atraso; Paternalismo Versus Produtividade

PDF, 1.79 MB
0 / 0

Chapter 2
What is a circle?

The opening question of this chapter is very simple to ask. What is a circle? You would have thought that in all the centuries of human civilization, we would have answered it by now. And there is at least one very clear and simple answer, which is provided by geometry. Mathematicians have a precise definition of a circle. But we should put aside that definition for now because it is not quite the type of answer that interests us.

Around us are various circular things: a coin, a wheel, the circumference of a ball, the rim of a cup, a line-drawing on a sheet of paper. There are various individual circles here, appearing at different places and times. There seems to be a something in common to all instances. What is common to them and the other things on the list is what we call circularity. We give this feature a name and treat it as if it is a single kind of entity. The same feature can be found in lots of different places and objects. Some speak of circularity as being a One that runs through Many. In all these many different particulars, the one circularity is to be found.

Going round in circles

There seems to be something strange about this thing circularity, if it really is a thing at all. Usually when one thinks of things, they are objects: particular objects such as tables, chairs, motor cars, buildings, trees, pens, and so on. Choose one such thing at random: a particular pen. This pen may sit on a particular desk in a particular house. It will have had a beginning in existence and a particular history. Now if this pen is on a particular desk, we know that it is nowhere else. And if a certain person owns the pen, we know that no one else does. Of course, a group of people may go into partnership to buy a pen jointly but, again, if that partnership owns the pen, then no one else does. The partners perhaps own a fraction of the pen each.

Sheep, tables, and pens are, then, particular objects. We can also;  speak of a circle, as if it is a particular thing. But circularity has a very different kind of nature. Unlike the case of the pen, the fact that circularity appears in one place or time does not stop it appearing at other places and times. And at all those places it appears, it is wholly present. This is different from the case of the pen. It being wholly owned by one person stops it being owned by another, whereas one thing being circular does not stop other things being circular. And when the pen is co-owned, each of the owners owns only part of the pen. But if a number of things are circular, each is wholly circular. It is not as if circularity has to be shared out in portions, such that if exactly two things had it, for instance, each would be semi-circular. That would be absurd because then neither of those things would be circular after all. This is what is meant by speaking of circularity being wholly present in all of its instances and it being a One thing that runs through Many others.

This may all seem like a confusing tangle, so let us try to straighten things out. One view is that there are two basic kinds of entity: particulars and their properties. Tables, chairs, and sheep are examples of particulars and they seem present only at one location at a time. Circularity is an example of a property: a feature or quality of a particular. The fact that a property appears in one place, in its entirety, puts no limit on it appearing elsewhere at other places and times. Because of this feature, some like to call properties universals–they can be at any place or time–though really this term is best used for one particular theory of what a property is. Other examples of properties are redness, squareness, being hairy, soluble, explosive, tall, and so on.

Related matters

There is much to be said metaphysically about such properties: about their nature and existence. But first, it should be noted that a distinction is usually made between properties and relations, and many of the same issues are found in the case of relations.

Alan is taller than his eldest son, Bobby. We can think of this as there being a relation of taller than, which Alan holds to Bobby. And there are many other instances of taller than, all over the place. Bobby is taller than his sister Clarissa, who is in turn taller than her dog Dougal. And the Empire State Building is taller than the Chrysler building. Again, it is arguable that this relation of being taller than is wholly present in each of its instances. Perhaps even more so than in the case of properties, it seems clear that it is the very same thing that appears in both the case of Alan being taller than Bobby and Bobby being taller than Clarissa. When we say one thing is taller than another, it seems that in every case we mean the same thing.

Let us now return to the case of properties, for we have all we need here to proceed to some of the really deep philosophical issues. Suppose someone were to gather every circular thing in the world, or every red thing, or every one-hundred-sided thing. No doubt this is a practical impossibility, but it is worth entertaining just as a thought experiment. Suppose they were then to crush all the circular things, or whatever the example is, until they were destroyed or at least no longer circular. Would they thereby have destroyed circularity? Arguably not. At best, they have destroyed all the instances of it. But can we say that circularity nevertheless exists? If so, where or when does it exist?

Plato’s heaven

Plato, one of the greatest philosophers of all, had an answer to this. At first, it sounds too fanciful, but reflection may dispel the feeling. Metaphysics often works like that. Plato didn’t think you could destroy circularity. He thought that the instances of it with which we are acquainted are all imperfect copies of the true circularity. Every circular thing existing in the physical world will be defective in its circularity, to at least some degree, no matter how small. The geometrical definition of a circle is relevant here. Plato thought that only the mathematician was properly acquainted with a perfect circle, in which every point on the circumference was exactly the same distance from the centre. But all the circles we see in the world around us will have some slight deviation from this perfect or ideal circle. Now here is the fanciful part. Plato thought that the perfect circle existed in a heavenly, transcendent world: above and beyond the physical world of everyday objects that we inhabit. This heavenly realm would contain all the true versions of all the properties and relations too.

We need to take a pause at this point because the existence of a so-called Platonic realm where properties such as circularity, redness, hairiness, and relations such as taller than exist is pretty striking. This realm is not something we could ever see with our eyes or interact with physically. It has to be contemplated and understood through pure intellect, according to Plato. The idea appeals to the thought that there is more to the world than that which we humans have created.

Consider, for instance, the fact that 2 + 2 = 4. Wouldn’t this still be true even if no one had ever thought about it, or even if no humans had ever existed? Suppose the universe had been barren, filled only with lifeless rocks. 2 + 2 rocks would still equal 4 rocks, even if no one would have been around to think of it or articulate it. In that case, there should be some appeal in Platonism because it seems that some things are too great and too perfect for our mundane and everyday world and thought.

Platonism is a very strong form of realism about properties. For Plato, they were more real than their imperfect copies that we ordinarily encounter. Only the Platonic circle was a perfect circle. All the rest are flawed. He called these perfect versions of properties Forms, and he thought they were the most real things of all. We knew they existed through our intellectual grasp alone, so we didn’t have to worry about whether our senses misled us.

Not everyone likes the idea of Platonism. This may not be just a matter of personal preference. A Platonist divides existence into two realms: the one we inhabit and the heavenly one where properties exist. But whenever we divide something into two realms, we have to tell a story about how they relate and that often gets complicated. That problem afflicts this account. What is the relation supposed to be between the perfect Form of a circle and the individual circles that we see around us? Plato tried many times to answer this question, but without complete success. One circle exists in a heavenly realm, outside space and time. The other circles all exist imperfectly in space and time. So how can they possibly relate to the Form? Can they resemble the Platonic circle when they are so different in their natures?

There is a major difficulty for any such proposal. Suppose we wanted to say something like this: the worldly circles resemble the perfect Form of a circle. Resemblance is a relation. But a relation, it will be recalled, is also something that a Platonist thinks belongs in their heavenly realm. There would be a Form of resemblance, therefore. We would then have to answer the same question again: how does the Form of resemblance relate to the actual resemblance (between the worldly circle and the Form of a circle)? If we give the same answer–that it resembles it–we will have made no further progress. We will have embarked on what philosophers call an infinite regress. There will be a never-ending series of resemblances, and this would indicate that the original answer is no good: we never should try to say that the instances relate to the Form.

Platonism, however, is not the only option we have, so we need not despair if we think the theory looks doomed. There are two other main options. The first is an anti-realism about properties, certainly conceived of as universals. This needs explaining.

The starting point was an assumption that there are two basic kinds of thing in the world: particulars and properties. But not everyone accepts this. One reason for rejecting the division into two is precisely that they would need to be brought together in some way. We would have to get the roundness and the greenness, as properties, united with the physical things in the world, such as an apple. But that is when we have to start speaking of the apple instantiating those two properties, or some such account. Suppose, instead, we said that there was only one kind of thing. What if we said that everything in the world was a particular? (This is the opposite of the bundle theory of particulars considered in Chapter 1, which says that everything is a property.)

There is an attraction to this particulars-only view. I know that tables and chairs exist, and balls and screw-top jars, trees, pencils, coins, and all sorts of things. But I am slightly less sure about the idea of circularity being a thing that exists in the same way that a coin does. When the case for circularity was made earlier, it needed some work to make it sound plausible. But surely it doesn’t take any work at all to persuade someone that all these different particulars are real.

It’s all just words

What, then, do we say about the so-called properties? The view that everything is a particular is sometimes called nominalism, which means name-ism. The idea is that circularity is just a name–just a word–that we use to describe groups of particular objects. There are varieties of this theory, but one is that the name is applied to groups of particulars that resemble each other. Hence, there are particulars–a ball, a coin, a screw-top lid, a wheel, and so on–and circularity is just a name for the way in which these things resemble each other. Circularity itself is no thing. It has no existence or reality. Every single thing is a particular.

But nominalism has its own problems. It can be asked in what way the group of objects resemble each other. Suppose the group of things offered as examples–the coin, wheel, and jar lid, and so on–as well as all being circular were also all brown in colour. In that case, there seems to be more than one way or respect in which they resemble. Circularity cannot just be a resemblance between the things, therefore, for it seems it must be one individual resemblance in a certain respect.

This is potentially very damaging. It seems we have to appeal not just to the particular things but also to ways or respects in which they resemble; and a way or respect sounds like a property by another name. Our attempt to do away with properties and have only particulars looks to have run aground very quickly, therefore.

That’s not all. Here is another problem. We said that groups of particulars resembled each other, and this was all a property consisted in. But, again, what is this resemblance thing? It sounds like a relation: a universal that apparently holds between all the particulars in the group. It seems, therefore, that we are again appealing to a universal–a relation in this case. Can we avoid doing so?

Suppose we said that resemblance was not a universal or Platonic Form. Like everything else according to nominalism, it is a particular. There would then have to be a particular resemblance that held between a pair of objects; but there would also be another particular resemblance that held between a further pair of objects. In what way are those two particular resemblances both resemblances? Again, resemblance cannot be a real relation. So it seems that we have to say that these two resemblances resemble each other. And then we need an account of that further resemblance. Again, this looks like an infinite regress in the offing.

This problem afflicts another view that is worth mentioning. It might be thought of as a variety of nominalism but also stands apart from it. The idea is that while the world is made up of particulars only, these particulars should be thought of not as particular objects but as particular qualities. This is a rejection of the view of properties as a genuine One running through Many. Rather, this redness is an entirely distinct thing to that other redness. Various patches of red can be found all over the place, just as there can be various circles. These particular attributes should be thought of as entirely separate existences from each other. The redness of one snooker ball is, after all, separate from the redness of another. One could exist even if the other didn’t.

The technical term for these particularized qualities is tropes. But the same kind of difficulty arises for them. In virtue of what are all these tropes red tropes, for instance? What gives them their red nature? One may just say that it is a primitive fact about them that permits no further explanation. But that then starts to look like a realism about properties after all. Or one might say that they are all red because they resemble each other. But we have already seen the difficulties resemblance gets us into. Are there resemblance tropes? And do they resemble each other?

Back down to earth

Is there an alternative to Platonism and nominalism? We have seen the difficulties of both and might wonder whether there is a third way. Fortunately, there is. One of the most famous images in the history of philosophy is of the School of Athens painted by Raphael (Figure 2). In the centre, Plato and Aristotle debate. Plato is seen pointing up to the heavens. All that really matters is up there. But Aristotle has a different view. He gestures down to earth. No, he is insisting, it is all down here.

[image: image]

2. Detail of The School of Athens, 1510, by Raphael

This Aristotelian view is one to consider. Plato’s theory was described as a realism about properties but it is not the only form that such realism could take. The worry about it concerned its transcendent nature: properties residing in the Platonic heaven. But perhaps the properties could be real and exist down here, in the regular world of which we feel part. Such was Aristotle’s view, which we can call an immanent realism, because properties are here with us. Circularity would be a real feature of the world but exist only in its instances: in circular things. One would have to accept that some such circles were imperfect. Perhaps that means that only imperfect circularity is a real property. Why not accept that? The mathematician’s circle is really nothing more than a stipulated definition of something, which doesn’t mean that it thereby exists. To exist, on this view, is for something to be it; and if nothing is perfectly circular, mathematically defined, then perfect circularity is not a property of our world.

But what of the point that if I crush every circular object, I cannot thereby have destroyed circularity? It doesn’t seem that properties can go out of existence or come into it. Although this point may tempt one towards Platonism, there is another response. Suppose I say that a property exists in nothing more than its instances, but I mean in all the instances that ever have and ever will be? We can treat all times as equal rather than privileging the present. So if something somewhere is circular, be it only once, at any time, then that property exists and is real.

This doesn’t end the matter. We still need to consider what it is to be an instance, and whether there are any tricky instantiation relations lurking around to thwart our theory. But we have seen that there is a possible position in which properties and relations can be understood as real but also down to earth. Undoubtedly, this view would need more development and defence, but it looks worth a try.

Chapter 10
What is metaphysics?

Now that we have looked at some of the main questions discussed in metaphysics, we are in a better informed position to answer the question of what metaphysics is. Metaphysics is the activity we have been doing in the previous nine chapters. Having engaged in it, we are in a better position to understand it.

Many of the questions will have sounded simple, silly, or childish, and they are often dismissed as such. Once we grow up, we are not expected to ask what a circle is, whether time passes, or whether nothing is something. It is almost as if the natural sense of wonder with which we are born is disciplined out of us.

Perhaps metaphysics is thought of as a useless waste of time or, even worse, a dangerous distraction. We shouldn’t forget the story (probably only half true) that Socrates was put to death for being such an annoyance. The reader will have seen, however, that these very simple questions can lead to complicated answers. Just from asking what a circle is, we quickly got into some deep issues about the nature of the world, reality, and what exists. Pondering on such issues may have helped develop our minds. We have had to think quite hard and perform some mental acrobatics. But is metaphysics any more use than that? We have gained understanding, perhaps, but it looks like there is no use to which we can put that understanding. The charge of pointlessness seems to stand.

What have we been doing in these past nine chapters, then? One answer is that we have been trying to understand the fundamental nature of reality. But it is only one aspect of reality that we have been interested in, and only one kind of understanding that we have sought. Science also seeks to understand the nature of reality, but it does so in a different way. Science looks for some general truths, but they are also concrete, whereas the truths of metaphysics are very general and abstract.

When we consider what exists, the philosopher’s answer will be at the highest levels of generality. They may say there are particulars that fall into natural kinds, there are properties, changes, causes, laws of nature, and so on. The job of science, however, is to say what specific things exist under each of those categories. There are electrons, for instance, or tigers, or chemical elements. There are properties of spin, charge, and mass, there are processes such as dissolution, there are laws of nature such as the law of gravitational attraction.

Metaphysics seeks to organize and systematize all these specific truths that science discovers and to describe their general features. Although, in explaining metaphysics, I have tried to use plenty of examples by way of illustration, the reader will also have seen that the choice was somewhat arbitrary. I asked what is a circle and what is a table. I could just as well have asked what is redness or what is a molecule. These were just ways of getting into the issues of properties in general and particulars in general.

Physics and metaphysics

There is a difference between metaphysics and science in the level of generality, but there is also a difference in approach. Although the disciplines have the same subject matter as their focus–the nature of the world–they seek to understand it from different directions. Science is based on observation, which is often its starting point and the ultimate arbiter of the truth of a theory. Metaphysics, while it’s concerned with the world, is not so much concerned with that part of it that can be observed. What we can see with our eyes is of little help in metaphysics, or philosophy in general. The evidence of the senses is not what decides whether a philosophical theory is to be accepted or rejected.

We considered, for example, whether a table was just a bundle of properties or was a substance underlying and holding together all those properties. We should note that we cannot decide between these two theories on the basis of observation. The world would look the same whichever of them is true. It is not as if we could actually remove the properties of a real object and find a propertyless substratum. What would one look like, given that it was propertyless? Our questions are not, therefore, scientific ones. A difficulty students often have in starting metaphysics is that they cannot distinguish it from science, especially physics. They think that if the world is our subject matter, then we should look at the world and do so scientifically. We may use it for examples to get us started, but we cannot expect it to answer our metaphysical questions.

It may have been a historical accident that our discipline got its name. Metaphysics was the book of Aristotle’s that came after the Physics. But, intended or not, there is another way of seeing the name that does describe the activity quite well. Meta can be interpreted to mean ‘above’ or ‘beyond’, and what we do in metaphysics is indeed above and beyond physics. It is above in its level of generality; and it is beyond the observational investigation of the world, thinking about the features that rationally the world should or could have. In that case, the discipline has a very suitable name indeed, because the problems discussed in Aristotle’s book were precisely of this nature. It would also follow, however, that calling a practitioner of metaphysics a metaphysician is wrong. A physician is one who practises medicine. Our subject is not above and beyond medicine. A practitioner of physics is a physicist, and thus those who do metaphysics are metaphysicists.

Metaphysics has been challenged not just on the grounds that it is childish or annoying, however. When we concede that it is non-empirical–it’s not about what can be observed with our senses–then there are many sceptics willing to attack it. If metaphysics is about something that cannot be observed, then how do we know anything about it and answer its questions? How do we know it is not all empty words or pure nonsense? Hume, and those who followed him, dismissed metaphysics on these grounds. Our ideas had to originate in our experience in order to have any meaning, the empiricist philosophers said. And if the terms used in metaphysics cannot be traced back to some original observation, then they are literally meaningless. Because of this, Hume recommended consigning metaphysics to the flames. He didn’t see himself as a metaphysician, even though he has influenced many of the subsequent debates. Few have followed his suggestion of burning books, but Hume’s critique has frequently resurfaced in more modern guises, such as in logical positivism.

One response to Hume was Kant’s version of metaphysics in his Critique of Pure Reason (1781). This is depicted as a defence of metaphysics, but it is also possible to interpret Kant as somewhat lowering the ambitions of the subject. Kant scholars disagree on how best to understand him, but one way is to see his metaphysics as a description of the structure of our thinking about the world, rather than being about the world itself. Because of the limits of our thinking and experience, we must understand it in a certain way. Running with this idea, one might say that we have to think of the world in terms of particulars and their properties, or as being located in time and space, or as involving causes. Or one might say that metaphysics is just about the concepts that we use in describing the world and how they relate to each other. Perhaps metaphysics of this ilk is easier to defend, but it has to be acknowledged that it changes its nature fundamentally. Metaphysicists want to discover what the world is like, not the facts about our concepts, or about our psychology, or about only that part of the world that our psychology accommodates. Whether Kant’s own position warrants this response is something I wouldn’t dare comment upon. The point to make, however, is that it is tempting when metaphysics is under attack to retreat to some ground that is easier to defend. In doing so, however, it may no longer be metaphysics that is being defended.

Can we describe the practice of metaphysics in the pure original form and defend it against the attack that it is meaningless, useless, and pointless? The reader might consider whether the problems they have encountered over the past nine chapters have been meaningless. Perhaps you have the same intuition as me.

[image: image]

11. Immanuel Kant (1724–1804)

Although they are not settled through observation, it can still be claimed that they are genuine and meaningful questions. We can also contend that they are about the nature of the world, rather than just concepts or the way we think. Let us see if we can defend this view.

Seeing is believing?

To begin with, let us acknowledge that the powers of observation have been greatly exaggerated. Although scientists may use their senses, observation does not settle everything. What we believe in science is theoretical to a large extent. In that case, perhaps metaphysics is continuous with science. What metaphysicists do is not outlandishly different from what everyone else does who wants to understand the world. Metaphysics is at the more abstract and theoretical end of the scale, but it looks like more of a sliding scale than there being a very sharp boundary between the two disciplines.

In metaphysics, we have our thinking, our reasoning, as our guide to how the world could be or how it should be. We can reject accounts of some feature of the world if they are absurd, either by being counterintuitive or outright contradictory. The latter is preferable. A theory may entail that the world is some way but also not that way, in which case we would have grounds to reject that theory as incoherent. It may be rare that things are so clear cut, however. And even if they were, the defender of the theory might try to explain away the contradiction. More often, we would challenge a theory for saying or entailing something implausible. If one says that time flows, for instance, then it seems that one ought to be able to state the rate at which it flows. But we saw that this could not be meaningfully answered: it would trivially have to flow at one second per second. We also considered the theory that absences could be causes, but this also produced an absurd consequence. We would have to allow that a nothing could nevertheless have causal powers and, as we saw, that among the causes of any event would be the absence of anything that could have prevented it.

When a consequence is so counterintuitive, we say that we have reduced the theory to absurdity. In the case of substance, it looked like we had reduced to absurdity the theory that a particular was a bundle of properties. There couldn’t be, within that theory, two particulars with all the same properties, and there seems no good reason why we should accept this. There are, of course, philosophers who think the theory can avoid this reductio ad absurdum, so it is never the absolute end of the matter. And others might deny that such a consequence really is absurd. Nor should we assume that the truth of the matter always will be intuitive. We sometimes have to follow the argument and accept all it entails, even if it’s surprising. There is always room for further philosophical debate, therefore.

Nothing here really hinged on the use of the senses. Certainly, you need to be a thinking thing to even get started in philosophy, and experiencing the world seems a precondition of that. Your senses probably also showed you that particularity was a feature of the world that needed explaining. It looks like there are particular things. But after that, it was down to your thinking in the abstract what a particular must be, or a cause, or time. You can explore the various theories, as fully as needs be. Some of them do not stand up to scrutiny.

Provisionally, we can reject some of the theories if it looks as though they have an intractable problem: if they seem to involve a contradiction or other absurdity. We accept that the problem may be overcome at some future point. So far, it doesn’t look as if what the metaphysicist is doing is that much different from what some of the cleverest scientists do. A theoretical physicist might reject theories on similar grounds. There is one difference. Metaphysicists reject theories on the basis of reasoning alone: where the problem of the theory is one of internal coherence or contradiction with some other set of theories that we hold dear. In the case of sciences, the conflict that is the basis for rejecting a theory may be with some observational evidence. Science wants the theory to fit the observational facts, whereas the data of metaphysics are non-observational.

So far, this looks defensible. Metaphysics does not look conspicuously any less defensible than science. But, it might be argued, this is because we have only looked at the negative case in which a theory is rejected. Of course, a theory should be rejected if it is self-contradictory or leads to absurdity, whether it is in science or philosophy. But that only gives us grounds for ruling out theories, not for accepting them. And here it might be thought that science has the upper hand because it can find empirical confirmation of its theories. In philosophy, it is conceivable and very likely to be the case that there is more than one coherent theory about some issue. Multiple theories could be true individually even if they couldn’t all be true at once. So how do we decide which one is right? Perhaps there is no truth in metaphysics, just a bunch of theories between which we cannot choose.

This again ignores the way that we think in both philosophy and in science. As we know from the philosophy of science, it is problematic to state simply that we know which theory of the world is true, simply by looking. Doing so may provide useful data, that rules some theories out, but more than one theory could be consistent with the data. How do we decide which theory is true, then? How do we know that anything is true? This is not an easy question to answer. Some observations may even be partly determined by the theories that we believe, so it is far from a simple matter in the scientific case. Truth in metaphysics is also hard work, but the point is that we shouldn’t subject truth in metaphysics to any more stringent standards than we do in other cases. And what we probably then find is that our theories are provisional and fallible. We accept that they may have to be revised but that, nevertheless, it may be rational to work with them.

The relationship between metaphysics and science is likely to be a lot more complicated than the story just told. We often like the two to cohere even if they are not in outright conflict. Ideally, we would like a metaphysics that is fit for the world described by science, and science that is metaphysically sound. We have already seen a case that illustrates this. Questions were raised about simultaneity in Chapter 6: a notion that is challenged by relativity theory in science. Relativity theory does not refute certain philosophical accounts of absolute space and time, it could be maintained. But we may nevertheless think it preferable to develop a metaphysics for the way the world is, as described in the best available scientific theory. What we might then achieve is a scientifically informed metaphysics. This might give us a coherent account of the world that works on both a concrete and abstract level.

Theoretical virtues

In metaphysics, when deciding which account to hold, we have to look for a range of theoretical virtues. We look at how much of what really counts the theory can explain. Does it gel with our other theories to provide a unified account of the world? And does it explain much with very little, or are we assuming so much to get the theory started that its explanatory power is an illusion?

The theory that there are many other worlds, for example, has been recommended because it explains so much. It tells us what is possible, for instance. As we saw, however, for the theory to even get started, we had to assume the existence of each possibility at some real world. In that case, it looks like we only get out of the theory as explanation what we have put into it as assumption. We are thus often performing a balancing act whereby the number of assumptions is compared to explanatory power, like some sort of cosmic cost–benefit analysis.

Again, when we compare metaphysics to the search for knowledge in the sciences, it does not seem worlds apart. We want to find theories that explain the phenomena in a relatively simple way, requiring no outlandish assumptions or ad hoc hypotheses. For the sciences, the phenomena to be explained can include observational phenomena, such as the unexpected presence of a particle or movement of a planet. Such a movement would have to be entailed by the theory if that is to count as an explanatory success of it.

In metaphysics, the phenomena are not observational in the same way. We accept that some very abstract things have to be explained: there appears to be a plurality of particulars, for instance, and causation seems to be a feature of the world. How best to explain these general features? Once we explain what causation must be like in our world, we will then leave it to science to tell us what causes what. We just want to know what it is for one thing to cause another. And we adopt the same theoretical virtues as anyone else. We want to discover the best possible explanation of some feature of the world. It is just that the features that interest metaphysicists are abstract and very general. They cannot be observed in the way that you can observe a table or a cat, but perhaps they have been abstracted from what we have observed. We have a notion of particularity, from these things, and this is what we want to explain.

The value of metaphysics

It might seem that our critics have an argument that we cannot answer. We may have grounds for rejecting some theories and accepting others, provisionally, and in this respect, we may be offending no rational principle that anyone else would adopt, but the sheer uselessness of metaphysics shows why we shouldn’t do it. Because science has at least some connection with the empirical world that we observe and with which we interact, we can use our scientific theories to get what we want. Science can be applied usefully. We can manipulate the world and use science to our own ends. Good science produces good results. Because metaphysics is so theoretical, so abstract, and so non-empirical, it seems that it has no pay-off at all. It doesn’t allow us to manipulate the world around us. It is literally useless.

We need have no fear of this argument. In the first place, its basis is contestable. Causation, for instance, matters to everything. Our very manipulation of the world depends on causation. Nothing would make sense without it. And it turns out that we cannot know what causes what unless we have some theory–an inevitably metaphysical one–of what it is for something to cause something else. A philosophical misunderstanding could lead to a practical error, if we assumed causation was nothing more than correlation, for instance.

But suppose the basis of the argument is correct. Suppose metaphysics really is useful for nothing. Does this mean it is worthless? No. The value of many things is instrumental: they get you something else that you want. But some things have intrinsic value. They are valuable just for what they are, in their own right. Even if metaphysics is useless, its insights may be so deep and so profound that it could have the highest intrinsic value to us. It would give us a useless but deep understanding of the nature of reality. Indeed, Plato’s idea makes sense.

A metaphysical understanding of what the world is, how it works, and how it all fits together, in general and abstract terms, could be the most real and important thing there is. In that case, we don’t do metaphysics so that we can stay healthy and wealthy: we want to stay healthy and wealthy so that we can do metaphysics.

What is an introduction?

Metaphysics is one of the traditional four main branches of philosophy, alongside ethics, logic, and epistemology. It is an ancient subject but one that continues to arouse curiosity. It holds an attraction for many who have only a basic inkling of what it is but are keen to know more.

For some, it is associated with the mystical or religious. For others, it is known through the metaphysical poets who talk of love and spirituality. This book will aim to introduce the uninitiated to how metaphysics is understood and practised by philosophers. Many introductions to the topic begin with a consideration of what metaphysics is and how its truths can be known. But this itself is one of the most difficult and contentious questions, and the reader could quickly become bogged down and lose interest. This book is therefore written back to front. The question of what metaphysics is and how it is justified will be left to the very last. The best way to understand an activity is often through doing it rather than theorizing about it. In that case, we start by doing some metaphysics: considering some seemingly simple little questions but which concern the fundamental nature of reality.

We will go through a variety of issues with only a few technical concepts and terms. By the end, there should be a fair grasp of the problems around substance, properties, changes, causes, possibilities, time, personal identity, nothingness, and emergence. It is hoped that the book will not intimidate its readers in a way that many philosophy books–particularly in metaphysics–can.

Often, the ideas, concepts, and questions of metaphysics sound easy–childish even. What are objects? Do colours and shapes have some form of existence? What is it for one thing to cause another rather than just being associated with it? What is possible? Does time pass? Do absences, holes, lackings, and nothingnesses have any form of positive existence at all? To some, these seem like silly questions, but for others they are at the core of what philosophy is all about. And those who see it that way often get a sense that the issues these questions raise are the most fundamental and profound about which humans can think. Metaphysics is the subject among all others that inspires the sense of wonder in us, and for that reason some think that doing metaphysics is the most valuable use we could make of our time.

If you have made it this far, perhaps metaphysics has already captured your imagination and your curiosity. In that case, we should begin forthwith on our little tour of the metaphysical furniture of the world. But where to begin? Philosophers never really know. The things they worry about are often interconnected. To understand one issue, you need first to understand another. Yet we have to say the same about the second issue as well: to understand it, you need to understand a third, and so on. And this seems to be true no matter where we start. Sometimes an understanding of the world comes only by grasping the whole, which makes it hard to explain the problems of philosophy in a neat sequence, as books must inevitably try to do. Where we begin is thus to an extent arbitrary.

Chapter 6
How does time pass?

We have been considering what there is in terms of some of the most general categories. There are particulars and properties, but there are also things like wholes, parts, changes, and causes. The latter two could not exist unless there was also something else: time. For a change to occur, there must at least be something at one time that is no longer at a later time, or vice versa. Philosophers argue about whether there could be time without change, but it surely looks certain that there could not be change without time.

What is this thing we call time? This is perhaps a question we have all thought about, and in doing so we will have been engaging in metaphysics. One thought is that time is some thing itself and acts as a background in which events are situated. We think of this time as flowing and having a direction. It can pass you by. You need to keep up with it and not waste it. It has reached a certain point. Perhaps it’s a finite resource. This image of time may resemble the flow of a river. It moves at a certain pace, passing certain points along the river bank. At your birth, you jump on a raft and the river takes you downstream, passing all the signposted years on the bank along the way. At your death, you jump off the raft, and the flow of time carries on along its way without you.

[image: image]

7. The direction of time

Some of these thoughts do seem puzzling, however. We speak of time flowing but, while we can measure the speed of a river, can we really do anything similar for time? If time passes, how quickly does it pass? Presumably one second per second? Does that make sense? Could there be any other answer? If I get a paddle on my raft, could I go faster or even slower than time itself? And we say that time has a direction. We say it goes forwards rather than backwards, but what are we asserting with this? Perhaps time is after all going backwards. What would it look like if it did?

Treating time as a thing in itself, having some kind of existence independently of the events situated within it, also suggests that there could be a period of time in which nothing happens. Is this really a possibility? How do I know that there hasn’t just been a gap of a year in which everything stood still and then resumed again? Or perhaps it was a gap of two years.

When we discuss time, we find it so hard to grasp that we have to resort to metaphors. These may be misleading, however. I might say that time has passed me by, or that a long time has passed, but these can’t literally be true. Length is something attributable to space and passing is attributable to motion, as in a dog running past me. Maybe the thought is that a certain event or process–your teenage years, for example–come from the future, into the present, but are then gone into your past. You can wave them goodbye, just as when you see the dog running up, for a time he is by your side, but then he is gone again. Does time pass like this?

How soon is now?

There are two models of time that have been debated by metaphysicians over the past century. I will spend some time on the first.

Think of an event such as the assassination of President Abraham Lincoln. For those living before 1865, the event was in the future. For us, the event is in the past. And on 14 April 1865 at about 10.15pm Washington DC time, it was a present event. Should we understand this case in terms of the spatial analogy? Did the event creep up on people, briefly attain presentness, and then go off into the past? Or is there some other way we should understand it?

There is a view that still has some respectability in which events have temporal properties of a sort. The assassination of Lincoln has a property of being past. A number of events have the property of being present, such as the event of you reading this sentence (and think of all the other events going on while you are reading). Many events have a property of being future, a property that might be called futurity. The Qatar football World Cup, the next UK general election, the solar eclipse of 21 September 2025, and the Earth’s human population reaching eight billion are all examples, as far as can be told in 2012.

And here is where we might get some sense of the direction of time. Events always are first future, then present, then past. It never goes in the opposite direction, as far as we know. If backwards time travel is possible, that might complicate it, but it seems the three temporal properties are always possessed in that order. I am, of course, treating events as particulars here rather than what metaphysicians call types. The Olympics occurs every four years, but by that we mean a type of event. Each particular Games is a one-off, and it is such events-as-particulars to which I am referring. These events ‘flow’ from the future, through the present, and into the past.

Temporal properties would have some strange features. It seems they would be able to come and go in different combinations. What was future may now be past. The 2025 eclipse is in the future as I write this, but eventually it will be in the past. Perhaps you are reading this after it occurred. This shows me from my 2012 position that it has a property of being future past; that is, a future event will eventually become past (by October 2025, for instance). And there is also a past future. Lincoln’s assassination was the future in 1860 but it isn’t any longer. His assassination hasn’t been future since 1865. It might then be wondered how things can have qualities such as these and what is happening when they undergo a change in respect of them. Is it that there are lots of things standing around somewhere with the property of futurity, waiting until they attain the property of presentness? Are there future people who are longing to be present, thinking to themselves ‘How soon is now?’? And where do they go when they attain a property of pastness? Does anything really have that property, or is it merely that they go out of existence?

No time like the present

There is a view that only the present is real; appropriately, it’s called presentism. This could be thought of as a response to some of the questions just posed. For isn’t it absurd to say there are things with the properties of futurity and pastness? To exist seems a condition of bearing properties, but one could argue that future and past things have no existence at all. Barack Obama was born in 1961. Wouldn’t it be misleading to suggest that he existed in 1959 though at that time with the property of futurity? And Julius Caesar did exist for a time but he doesn’t now. It would again be wrong-headed to say that he exists now but with the property of pastness. It seems an option, therefore, to say that instead of there being three temporal properties, we should instead substitute a simple notion of existence and allow that things come into and go out of existence. When present, they are real. After that, they are not.

That seems a sensible view, but here are some issues to be considered. First, how long does the present last? Is it today, or this minute, or just a second? At 20:50 in the evening, midday today is surely past. Indeed, even 20:49 is past, and two seconds ago also. The present seems like a tiny sliver. We can wait for its existence, but it is too quickly gone. Indeed, if there is a smallest unit of time–some micro-micro-second, which we might call an instant–then the present seems only to be as long as that instant. If we deny that, and argue instead that the present has some extension, then how long should we allow it to be? Two minutes? That looks an arbitrary figure. And yet if we don’t allow the present to have some temporal extension, it seems almost to vanish to nothingness.

Here is a second problem for presentism. The notion of the present is challenged by relativity theory. I may think that the sun is now shining, and it thus seems to be part of the present. But I’m also told that it takes 8 minutes and 19 seconds for the light of the Sun to reach Earth. Absolute simultaneity has been challenged in physics, and we are told it’s illegitimate to speak of two spatially separated events being simultaneous. You could view two stars collapsing in distant galaxies and it might look as if they are collapsing at the same time. But if one is much closer to your telescope than the other, then those events are not really simultaneous at all. There is a problem then of what exactly we mean by the present when it seems always relative to a position or a viewpoint. We could settle for a purely subjective account of the present–it’s what appears to be now, for some viewer–but many of us don’t want our metaphysics to be so dependent on one’s point of view. We like to feel that we are dealing in objective, eternal, and immutable truths, unaffected by our human perspective on things.

Speaking of which, there is a further problem for presentism. Although Caesar is not alive, there is a strong sense in which he is nevertheless real, even now. There are facts about him–he crossed the Rubicon–and there must be something in virtue of which those facts are true. If only the present exists, what makes it true that there was a Second World War or an assassination of Lincoln? Wouldn’t it be wise to say that those past occurrences and things are a part of our reality even if they are not present? Given the above considerations from relativity theory, there are even some past facts that I can still see: for example, what the Sun looked like eight minutes ago. What is there to stop someone rewriting history if we deny any reality to what happened in the past?

Getting pasturized

There is a view, therefore, that treats the past and future differently. It’s one thing to call absurd the idea of future people standing around waiting to be born. But the past is not quite the same. It did exist. It was present. And in this sense, it should be counted part of the totality of reality. Being a part of reality but not in the present could account for our property of pastness, then.

This view is often likened to a growing block. One could think of the present as a thin layer on top of a big solid cuboid. New layers keep being added to it all the time on its top surface. Caesar, and all that he did, is there in the block, some little way down. When we speak of what exists, there are two things we could mean. What exists now is only the top surface of our block, which is in that position only fleetingly. Perhaps it is just a few molecules of the block thick. But we could also mean by what exists the entire block, which is the whole of existence from its start until present. The past is now part of this. But as new layers are added on to the growing block, former present events recede into the distance. We could say they become pasturized, just in virtue of having a new future built upon them.

We have moved, therefore, from a view that privileges the present to one that privileges both the present and the past, though not the future. This second view still has to face the problem of what counts as present: of how thin it is and of the problem of absolute simultaneity. To an extent, the problem remains of treating presentness and pastness as properties of events or things. The growing block picture has merely dispensed with the property of futurity.

I said earlier that there were two models of time philosophers had debated. The first tries to explain the passage of time in terms of events and things having a property of presentness, pastness, or possibly futurity. But we have seen that this leads us into saying strange things at every turn. Perhaps the problem is that we started by looking for a theory that would satisfy an image we had in which time flowed: it passed like the water in a river. There is a different way of understanding the temporal sequence, however. In this view, there is no property of presentness, nor pastness, nor futurity. Instead, we can only say that the events and things in our world stand in relations of order to each other. They are temporally related and to that extent can stand in a sequence.

Early, late, or on time

The basic relations out of which such a sequence could be built are being earlier than, being later than, and being simultaneous with. Obama’s birth is certainly earlier than his death, but it was later than the assassination of Lincoln, which itself was earlier than the assassination of Kennedy. The notion of simultaneity has been challenged, as we have seen, though that applies only to events at different places. I might thus legitimately still say something like Obama’s birth was simultaneous with his first breath, given that those events occurred at the same place.

The flow or passage of time could be seen, on this view, as a misleading metaphor created to accommodate the earlier than and later than relations that hold between things and events. There is no change of properties, from presentness to pastness. The temporal relations between events in this new series hold for all times. It is at any time the case that Obama’s birth is later than Lincoln’s death. Nothing has to pass from one state to another. Nor do we need to see time as thing-like, such as a medium within which events occur. So perhaps there need be no worry about whether there could be time without change. Instead, we could just think of all the world’s events being placed in an order–what was before what–and then we have the sequence of time.

This last idea gets us to the heart of a very important matter: one on which there is another Platonist–Aristotelian divide. The divide has lurked in the background throughout this chapter. Do we treat time as an objectively real thing, existing in its own right, whether or not any events are happening within it? Or do we think that time is nothing more than the ordered sequence of events?

At the start of the chapter, I almost suggested that we needed the reality of time as a background against which changes could occur. But an Aristotelian way of looking at it would be to start with change–maybe all of the world’s changes–and see time as some sort of construction from them. If the thought of everything standing still for a year–and then resuming unnoticeably where it left off–seems absurd, then the Aristotelian view is probably more appealing. Time would be judged to have started with the first event: the Big Bang, if you like. The idea of there being anything ‘before the Big Bang’ would be absurd for an Aristotelian but not necessarily for a Platonist. The latter might also countenance a serious answer to the question of at what time the Big Bang occurred, as if there were some kind of godly astronomical clock that dated everything. For an Aristotelian, the first event was the point at which the clock started ticking.

One may have some attraction towards coupling this Aristotelian view with what is called eternalism about events and things. We considered privileging the present, or the present and past, but the eternalist takes all events as equally real even if from one perspective they are future. I don’t know whether the 2020 Olympic Games will pass off successfully or not, but, if they do, an eternalist takes them just to be as much a part of reality as anything. This may sound confusing. Trading on the image of the block again, the eternalist takes reality to be one huge block of everything that ever was and will be. We are located some place in the middle, able to look back at what occurred earlier than us but unable to see what is later than our perspective. But it’s all just as real.

When we consider what exists, we are sometimes tempted to think of the question only three-dimensionally: of what exists in the whole of space. But shouldn’t we be thinking four-dimensionally instead: about what exists in the whole of space and time? Obama’s birth is earlier than his death. We cannot maintain this if one of those events is not yet real (as I write this in 2012). Why do I say this? The thought is that a relation is real only if its relata–the things it relates–are real. Obama’s birth could not bear a relation to something non-existent. We must grant reality to Obama’s death, therefore, while of course hoping there will be some time to go before it.

This may sound like a tempting account as it does away with the idea of time as a flowing medium. But there could also be a worry that it ignores something fundamental about time. It certainly does seem like there is a present that has a special quality about it. We may allow all times to be equally real in so far as they are all existent, but couldn’t one also argue that the present has something that neither the past nor future has? What is that? Well, it’s what’s happening now–at one place and point of view at the very least. And does the view of time as nothing more than a relative ordering of events have the resources to explain what any kind of now is?

A further issue in the philosophy of time is worthy of mention. There is the question of its topography. We sometimes think of time as a single straight line. It has a beginning, it runs its course, and has an end. But there are other ways of picturing it. Perhaps the line continues indefinitely. Time might not be a finite resource. And it might continue infinitely in both directions. On the other hand, it could branch out as it progresses. There might be two separate timelines branching out from a single source, based on some significant difference. A more radical idea would be that time goes round in a circle. What caused the first event in the history of the universe? Perhaps it was the last moment in the history of the universe. These debates remain live, and perhaps the reader can see how our stances on some of the issues discussed earlier might inform our decisions on these latest options.

Chapter 8
What is possible?

‘I coulda been a contender’ is one of the most famous lines in movie history (Marlon Brando in On the Waterfront, 1954). Could he really? We accept lots of things to be possible that are not actual. You could have been late to your appointment if there had been an incident on the way; Michael Foot could have been prime minister; the Eiffel Tower could have been disassembled in 1909, and it could have been 350 metres high, instead of its actual height. Yet we also accept that many things are not possible. You couldn’t jump to the Moon; lead can’t transmute into gold; a leopard can’t turn into a chicken; and so on. There are some things that might be possible or might not, and we don’t know which. We seek a cure for cancer, for example, and we don’t yet quite know whether one is possible, though we hope so. Who would have thought when sticky black oil was first found in the ground that it would make motor-car propulsion possible? Science and technology often progress by discovering hitherto unknown possibilities within things.

What are these possibilities? Are they a part of the reality that we have been cataloguing? We have found particulars, their properties, changes, causes, so what about possibilities? Are they things? Do they have any kind of being? Or are they a mere fabrication: things we can think about but which are not really a part of the furniture of the world?

Before addressing those questions, we need a clarification. The possibilities considered here are those that are possible but without being actual. It is necessary to say this so that the answer to our question is not trivial. All that is actual is of course possible: for how else could it be actual? So those possibilities are real enough. No dispute there. That London is the capital city of England is both possible and actual, so certainly a part of reality. That Winchester was the capital until 1066 may be a part of reality depending on your view of facts about the past (see Chapter 6). But some really interesting metaphysical questions arise when we consider those possibilities that are not also actual. We can distinguish such non-actual possibilities by calling them ‘mere’ possibilities. Lincoln escaping assassination in 1865 is a mere possibility, as is Wayne Rooney succeeding David Cameron as prime minister and Beeston being the current capital of England. You are likely to get tired of me using the word ‘mere’, however, so I will just say now that when I discuss possibilities, I mean those that are non-actual.

Coulda, woulda, shoulda

The key question, then, is whether such things are real or not. In a way, they are clearly not. If you see a sign by the road saying ‘Possible queues ahead’, and you know that it is a mere possible queue, then you have nothing to fear, for only an actual queue can delay you. The idea that Beeston could now be the capital city matters not at all. And although the Eiffel Tower could have been 350 metres high, an air balloon passing over it needn’t consider this but only its actual height.

Then again, it seems foolish to ignore possibilities if they could become actual. There’s a possibility of skin cancer if you stay in the sun regularly and for too long, and a possibility of lung cancer if you smoke. Surely you shouldn’t ignore these possibilities. And when a glass is fragile, it means it could break relatively easily. This should determine how you handle it, assuming you don’t want it breaking. Every time you drive a car down the motorway, there is any number of possible crashes awaiting you if you were to be careless. You should take care precisely to avoid making them actual. The possibilities that surround us do indeed seem to shape the way we interact with the world and, on that basis, seem to have some reality.

Maybe there are two types of possibility, then. That the Eiffel Tower could have been 350 metres high–or 450 metres high even–seems of no real consequence. Only its actual height affects the way we should behave towards it: if trying to get over it in a hot-air balloon, for instance. But other possibilities are of consequence, where our behaviour could easily bring them about. And although we want to avoid bringing about car crashes through recklessness, there are many other possibilities we aim to achieve. It’s possible to be rich, or learned, or athletically excellent, and many people strive to make these things so.

We have already encountered an idea (in Chapter 5) that many people find useful in understanding what possibilities are. When we considered causation, we saw the theory that in some other possible world an event that occurs in our world does not occur in that one. The notion of a possible world has a broader use than just accounting for causation, however. There is also an idea that when we think of a possibility, we are thinking of something that does occur or is a fact at another world. Hence, there is another world in which Michael Foot was prime minister and another, not necessarily the same one, in which the Eiffel Tower is 350 metres tall. In another, Lincoln survived assassination and died of old age. On this account, every possibility of our world is actual at some other world; so there are as many worlds as there are possibilities, which sounds like an awful lot.

[image: image]

9. What is possible here?

But what are these other worlds? Are they a part of reality? First of all, by a possible world, we don’t mean something like a planet. In old science-fiction stories, we may refer to other planets as worlds, but the idea here is that a world is a whole universe. Our world–the actual one–encompasses everything there is. That would be the whole of space and time complete with its occupants: the properties, particulars, changes, and causes that make up the entirety of which we are a part. And the other worlds are supposed to be just like that too, complete with their planets and stars, people, trivial facts, tables and chairs, prime ministers, and capital cities. Some possible worlds will have no life in them at all, however, and while some will be vast universes, others will be very small. There’s even a possible world that contains just two objects sitting in a spatially symmetrical relationship (we came across this world already in Chapter 1).

In other worlds

Opinion divides from this point on. It may be a surprise to hear that there is a view in which all these worlds are just as real as ours. They exist in the fullest sense of the word, in exactly the same way that our world exists. Just as the table in front of you is a physical object that makes a noise when you bang it, there are physical tables in other possible worlds that can make a noise when they are banged too. Some worlds are even inhabited in places by people: not necessarily human beings but things capable of thinking and acting (in still other worlds, remarkably, these persons are computers). These people can feel sensations just like we do, and they can think about whether there are other possible worlds like theirs. You can even have counterparts in some of these worlds: people who look very like you, with a similar history, and they may even have the same name as you.

Although these worlds are real, however, we can never visit them. They are within space and time, with physical occupants, which is why they are sometimes called concrete possible worlds. But they are not within our space and time. Every world is spatiotemporally separated from every other world. Indeed, this could be part of what we mean by a world. If something is spatiotemporally connected to our world, it is thereby a part of our world; for you could in theory get there. And in this metaphysic, it should be clear that while we use the term actual to refer to our world, occupants of other worlds use the term to refer to their own. Actual thus becomes an indexical term, like I, we, here, and now. The reference of such terms varies according to who uses it. When I say here, it refers to Nottingham, but, when you use it, it refers to Oslo, Istanbul, St Louis, Antigonish, or wherever you are.

When David Lewis (in Counterfactuals, 1973, and then in On the Plurality of Worlds, 1986) first proposed this realism about possible worlds, his account was met with incredulous stares. Yes, he really did mean that reality was packed with an infinite number of concrete worlds of which ours was just one. But an incredulous stare is not an argument. One reaction was to say something along the following lines. When we speak of the possibility that the Eiffel Tower is 350 metres high, we could think of that as meaning it has that height in another possible world, but we shouldn’t treat such a world as real. It’s a merely possible world, not a real one. So perhaps we can use possible worlds as a conceptual aid in articulating our thoughts about possibility. Such worlds would be abstract only. By saying something is possible, we can say that there’s a possible world in which it’s true, but we don’t mean this literally. We could accept the language of possible worlds without being committed to them as a metaphysical reality.

Lewis was dismissive of such ‘ersatz’ realism, however. One thing he was trying to do was analyse what it is for something to be possible. His answer was that there was a concrete world in which that thing was true. Such worlds were as real as ours. If we try to resist this last step, then we cannot do what Lewis was trying to do. We cannot analyse what a possibility is because all we can say about these other worlds is that they are possible, rather than Lewis’s which are all actual worlds. To say that for something to be possible is for it to be true at a possible world, is to provide a circular, uninformative account. We should not, therefore, look to downgrade such worlds; if we do, then they may be unable to perform the substantial task for which they were intended. Such a line of argument might not be the end of the matter, however.

The motivation to back away from Lewis’s strong realism about possibilities is understandable, however. He makes them as much a part of reality as anything. They are just not a part of the little corner of reality that we call home. They are someone else’s home. Reality becomes a much bigger thing than we might have thought, therefore. We could think of this as rather metaphysically cumbersome. There must be a world for every single possibility, so our own world is an infinitesimally small part of the totality. Do we really need this much just to account for possibility? Philosophers are not very economical, are they?

Again, though, this is not an argument. There is no reason why the most economical theory should be true. The world (or all of them) could be a messy place where the truth of the matter is complex and involves a great many things. But here is a consideration that might be a bit more serious (opinions on its significance vary greatly). The statement with which we began was ‘I coulda been a contender’. Note the emphasis on I. Many of the claims we make about possibility refer to particulars. The Eiffel Tower could have been 350 metres tall, Lincoln could have survived assassination, and I could have been a professional footballer. But if realism about possible worlds is true, these claims about possibility, strictly speaking, are not.

According to the theory, what it is for it to be true that I could have been a footballer is that there is another world containing someone who is a lot like me, my counterpart, who is a footballer. But what do I care about this other guy? He’s not me; indeed, he’s not part of my world at all. Nothing he says or does can affect me because worlds cannot causally interact. So, in a sense, I couldn’t have been a footballer, after all, if I follow the implications of the theory. Only some other guy was. Similarly, the Eiffel Tower of our world is 324 metres tall. We think it could have been 350 metres tall, but the theory tells us only that there is another world containing a counterpart of the Eiffel Tower that is 350 metres tall. What does this tell us about our world and the tower within it? It couldn’t have been 350 metres, given that it is in fact 324 metres. All the theory says is that it’s like some other tower that is 350 metres tall. This looks like quite a big letdown of the theory. In establishing realism about worlds, in order to tell us what possibility is, it seems to have had the opposite effect because it restricts us to differences occurring to counterparts of our world’s inhabitants. Brando wanted to know whether he could have been a contender, and we wonder whether our Eiffel Tower could have been dismantled in 1909. The realist about possible worlds has some work to do to answer this criticism.

Combinations and recombinations

The possible worlds account is not the only game in town, however. It is worth looking at another. We might think of this second account as more Aristotelian than Platonic. Let us suppose that mere possibilities have no existence whatsoever. Nevertheless, there seems to be something to them. How can we account for them? It has been suggested that we could think of possibilities as recombinations of all the existing elements in reality.

Here’s how it could work. Suppose you are fairly new to the world with only a limited experience. One minute a white dog walks by. The next minute you see a black cat. You now know that there are these two kinds of particulars: a dog and a cat. But you also know that there are two properties as well: whiteness and blackness. Although you’ve seen only a white dog and a black cat, you now see that a black dog and a white cat are also possible. All you have done is rearrange in your head the existing elements. You have seen which particulars exist and which properties; and you have just redistributed or rearranged them. Similarly, you know that there are various buildings in the world, and you know that there are various height properties. 350 metres is not the height of the Eiffel Tower, but given that you know that the Eiffel Tower exists and that 350 metres is a height, then you can put those two elements together in your mind and think of a possible 350-metre-high Eiffel Tower.

The possibilities have some kind of existence: but only in so far as the elements exist out of which they are built. The recombinations that are merely possible have no concrete existence. The 350-metre Eiffel Tower is a mere fiction. The tower exists, and the height, but not their combination. That is the thing that is merely imagined.

We need not require that someone does the actual imagining, however, or write the fiction. For something to be possible, it can suffice just that the particular and the property exist that would, if they were combined, constitute it. They could form a recombination that no one ever thinks of. Abraham Lincoln is a particular, for instance, and there is such an activity (or property) as scuba diving. So Abraham Lincoln scuba diving is a possibility even if no one ever mentions it. It has now been mentioned here, but there are lots of other possibilities that no one has ever mentioned nor thought of.

This theory of possibility was developed by David Armstrong (A Combinatorial Theory of Possibility, 1989). We can think of all the world’s elements in the form of a grid. Arrange all the particulars on one axis and all the properties on the other. A particular apple would be one particular, and greenness one of the properties, but there will be many more besides. We could think of some of the grid spaces where a certain property and a certain particular intersect as filled in (with a tick, perhaps). This would indicate that the property was held by that particular. It would be a fact, for instance, that this apple is green. But many of the spaces would be left blank. Abraham Lincoln was not green, so there would be no need to tick that box. If we discover all the spaces that are ticked on our grid, then we would have discovered all the facts about our world. We could have total knowledge. But the grid also tells us, on this theory, what all the mere possibilities are. They are all the blank spaces: all the boxes that are not ticked. Lincoln could have been green. This just means that Lincoln is a particular and greenness is a property.

This is a simple idea. As always, the devil’s in the detail. This combinatorial account only works if you allow that the elements are recombinable. The principle of recombination–that any particular can be matched with any property–is doing all the work, one might think. And one could then wonder whether the combinatorial grid really tells us what the possibilities are or just assumes them by adopting the principle of free recombination.

One might wonder whether this principle should in any case be supported. Could Lincoln really have been green? Surely that is not a possible colour for human beings. Maybe there are at least some restrictions on what is possible. It was remarked earlier that you couldn’t jump to the Moon, but on this account it seems this might count as a possibility. You exist, and there is a property of jumping to the Moon, so that seems to be possible unless we impose some restrictions.

Distinctions are often made between types of possibility. There is that which is logically possible, meaning simply that it would not involve a contradiction or violate the laws of logic. And there is natural possibility: what the laws of nature permit. The latter include laws of physics, biology, chemistry, sociology, optics, and so on. One could say that it is logically possible that Lincoln be green even if it is not a biological possibility. The natural possibilities would be some subset of the many more things that are logically possible. Hence we could allow unrestricted recombination to account for the logical possibilities and impose restrictions if we are talking about the natural ones.

The combinatorial theory of possibility can be criticized along these lines for allowing too much to be possible. But it also faces an objection from the opposite direction. It is not always able to deliver enough possibilities. The theory constructs possibilities out of recombinations of all the existing elements. But there is a thought that there might have been a bit more than there actually is. It seems a possibility that there might have been one more particular or one more property. Kennedy could have had an extra child, for instance, instead of the four he actually had. It seems a possibility but not a recombination of what there is. The extra child I’m imagining is a whole other extra particular: another element there would have to be in reality. The theory can protect a little bit against this objection but not completely. One could couple it with a four-dimensional view, which would give you all the particulars and properties there ever was and ever will be and leave them all available for recombination. But even if the number of elements is huge, it is still finite. There could still, we feel, have been more: such as that unborn child.

One who adopts the possible worlds account, on the other hand, can say simply that there are worlds that contain more than ours does. There would be a world in which Kennedy indeed has a fifth child. We should not forget the problem with that account, however. The man with five children in the other world is not Kennedy, and the fifth child is not an unborn one of this world. We have two main theories of possibility, therefore, and both have their weaknesses. You may have noticed a pattern of inconclusiveness by now. Such is often the way in our discipline, but it does show that there is plenty of work left to do for all the possible future philosophers.



List of illustrations

What is an introduction?

1 What is a table?

2 What is a circle?

3 Are wholes just sums of parts?

4 What is a change?

5 What is a cause?

6 How does time pass?

7 What is a person?

8 What is possible?

9 Is nothing something?

10 What is metaphysics?

Further reading


Chapter 1
What is a table?

When I look at the world around me, I see that I am surrounded by all sorts of things. I see a table and two chairs, buildings, an aeroplane, a box of paper clips, pens, a dog, people, and a wide variety of other kinds of things. But this is a book about metaphysics, and in metaphysics we are concerned with the nature of things in very general terms. I am tempted to say, as a metaphysician, that all of these things I have listed are particular things, or groups or kinds of them. The notion of a particular is very important to us. I want to know that the pen on the table is my particular one rather than someone else’s, or that the woman in the room really is my wife rather than her identical twin sister. To understand the importance of these issues, we need to probe them more deeply.

In front of me stands a table that I can see, feel, and hear if I rap my knuckles on it. I have no doubt that it–the table–exists. But now I will start the philosophical questions. What is this thing? What is the nature of its existence? Is the table something I know through experience or do my senses reveal to me something else? After all, when I look at it, I see its colour: the brownness of the wood. And when I feel it, I feel its hardness. Brownness, hardness, four-leggedness, and so on, are the qualities or properties of the table. One might then be tempted to say that I do not know the table itself but only its properties. Does that then mean that the table is an underlying something about which I know nothing? Its properties seem wrapped around it and impossible to strip away.

What goes for tables, goes for other particular things too. There is nothing special in the choice of a table as my example. In the cases of coins, motor cars, books, cats, and trees, I know them only through knowing their qualities. I see their shape, their colour, I can feel their texture, smell their fragrance, and so on. The nature of these properties of things–redness, roundness, hardness, smelliness, and so on–will be the topic of the next chapter. But we really cannot avoid mentioning properties as soon as we mention the particulars to which they attach.

The more things change, the more they stay the same

Now why would I suggest that the table is something other than the brownness, hardness, and four-leggedness that I can see in front of me? One reason is that I could imagine these properties changing while the table remains the same particular that it was. I could paint the table white, for instance, because it fits in better with the decor of my office. If I did that, then it would still be one and the same table, it would simply have changed its appearance. Something will have changed, while something has remained the same.

In philosophy, we see that all sorts of confusion can reign if we speak loosely of it being the same table, so we employ an important distinction. We can say that something has changed qualitatively even though it has remained numerically the same. So the table can be different in its qualities–it was brown and now it is white–but it remains one and the same thing. The table that was brown is now the table that is white. Imagine if a visitor comes into my room and asks what’s happened to my old brown table. It’s perfectly acceptable for me to respond that it’s still here: it’s just that they didn’t recognize it because I had painted it. Being one and the same, despite such changes in qualities, is what we mean by numerical sameness (the topic of change will be explored more in Chapter 4).

It is this consideration that leads me to think that the table itself cannot be the same thing as its properties. At least some of them could change and yet it would still be the same table. So when I look at and feel the properties of the table, I am observing just that–its properties–and not the table itself. But what, then, is the table, if it is not its properties?

Here is a suggestion. The table is something that underlies the properties and holds them all together in one place. It is something I cannot see or touch, because all I experience is a thing’s properties, but I know it is there through my rational thinking. When I move the table across the room, for instance, all of its properties move with it. They are clustered together in a semi-permanent way. It is not as if the brownness and hardness of the table can move but the four-leggedness can get left behind. I say that the properties are clustered only semi-permanently, though. As we have seen, some properties can be shed from the cluster and new ones take their place, so we cannot be absolutely strict and say that the properties are bound together inseparably. The brownness can be shed and replaced by whiteness.

Such a view of particulars may be best understood through the metaphor of a pin cushion that is used to hold pins together in one place. The pins represent the properties of an object and the cushion represents the particular itself. Some call this a substratum view of particulars, where the pin cushion is the substratum that underlies all the properties on view. One pin stands for the brownness of the table, another stands for its hardness, and a third stands for its weight, another its height, and so on for every single property the table has. And if we could strip these away–mentally, through a process of abstraction–we would come to understand that the thing itself is separate from them and is that in which they all inhere. Of course, when you remove all the pins from a real pin cushion, you are still left with something that you can see and touch. But remember that our metaphorical pin cushion, when all its pins have been removed, is a particular that has been stripped of all its properties so that we can think of what the table itself is. And without properties, it couldn’t therefore look or feel like anything.

Consider, for instance, a cat. We can think of it without its blackness; for that is a property and we want to know what the thing is that underlies all its properties. But removing its blackness isn’t like skinning a cat. As well as removing its colour, we also have to take away its shape, as that is just another property like the rest, and so is its four-leggedness, smelliness, and furriness. Take all those away and we could well wonder what this underlying substratum really is. It would have to be invisible. It would have no length, breadth, or height, and no colour or solidity. There would be a bareness to it that may really make us start to wonder whether we have anything at all.

Philosophers are notorious for working out all the implications of an idea. But they don’t necessarily always accept those implications. Sometimes a consequence is so ridiculous that it can be taken as good grounds for rejecting the initial supposition. Such a counterintuitive consequence will have reduced the supposition from which it sprang to absurdity. Perhaps we can say that’s happened in this case. It was suggested that the particular had to be something other than its properties. But once we started to abstract away the properties of the cat from the cat itself, we realized that it would leave hardly anything. Our substratum-cat seems to be nothing at all. It has no weight, no colour, no extension in space, and so on. And this starts to look like a non-thing. Isn’t it the case that everything that exists has properties? It is not as if ‘bare’ particulars could exist and that some of them were just fortunate enough to accidentally acquire properties. Certainly every physical thing that ever has and ever will exist has some shape or weight or feature. And to talk as if the thing can in some way exist independently of those properties was perhaps the mistake that led us to absurdity.

Bundles of properties

Let us, in that case, consider a different approach. If there can be no ‘bare’ particulars, existing without having properties, then we might want to think again of the cluster or bundle of properties with which we began. When in our minds we stripped away those properties, in a process of abstraction, the fear was that we were left with nothing at all. So shouldn’t we then just countenance the possibility that there is nothing more to a particular than that bundle of properties? If there really is no remainder once all the properties have been removed, then we know that our particular cannot be more than them. The bundle view is that particulars can be accounted for in terms only of properties. How plausible is this view?

There are a couple of problems associated with it, which come from the problem of change that we already discussed. If a thing were just a collection of properties, it couldn’t survive any change. If one property were lost and another gained, we would have a different collection: for I am assuming that what makes a collection the same thing at different times is that it is composed of the same component things. Consequently, two collections are different if the things collected within them are different. And clearly, the particulars that interest us change all the time while remaining (numerically) the same. A cat changes its shape frequently. Sometimes it is lying out flat, other times it is rolled up in a ball, and then it might be running around, changing its shape continuously. How can the cat be just a collection of properties when they change all the time?

It may be possible to answer this objection, though. Perhaps we should think of a thing as a series of bundles of properties, united by a degree of continuity. So while the table can be changed and painted white, it keeps roughly the same weight, height, and physical position. I am assuming the physical position of an object is one of its properties, and clearly it is a pretty important one in this context. I am confident the white table is the same thing as the previous brown table in no small part because I find it in the same room. And if it has moved, I expect that it did so gradually by passing through a series of locations between where it started out and where it ended up. While the cat changes shape rapidly, it keeps the same colour, furriness, smell, and, importantly, it is in the same place; or if it has changed its position, it has done so through a series of locations. We could say, therefore, that while the bundles of properties come and go, a particular thing is a succession of such bundles with an appropriate continuity running throughout.

There are a number of other difficulties to be faced, but before going on to consider one of them, it is worth mentioning what might be a big advantage of this bundle view. The first account we considered was one in which particulars were underlying substrata that held the properties of a thing together. To account for particular objects such as a table, a chair, a dog, and a tree, we had two kinds of ingredients. We had a thing’s properties and its substratum. But with this new bundle theory, it seems that we need only one kind of thing. We just have the properties and, when they come in a bundle or a continuous sequence of such bundles, we say that we thereby have a particular object. So where we previously needed two elements, we now have only one. Another way of looking at this is to say that the notion of substratum has been reduced away entirely in other terms. Objects would just be nothing more than bundles of properties, appropriately arranged.

The second theory is thus a simpler one in so far as it invokes fewer kinds of entity. The unknowable formless substratum seemed to give us nothing extra: if the bundle theory is correct, then the substratum is dispensable. Now there is no particular reason why a simpler and more economical theory is more likely to be true than a complex and uneconomical one, but philosophers prefer the simple ones. Certainly, there seems no reason to tolerate redundancy in one’s theory of the world because any redundant elements are clearly not needed for the account to work. They serve no purpose.

Identical twins

The bundle theory looks simpler than a substratum view, therefore. But is it too simple? Would it have enough resources to deliver all we want of a particular thing? There is one consideration that suggests not. A particular, we are told by this theory, is just a collection of properties. A snooker ball, for instance, is just a bundle of the properties red, spherical, shiny, 52.5 millimetres in diameter, and so on. The problem for the theory, however, is that there could be another object with exactly those properties. Indeed, for the game of snooker to be fair, there should be many red balls with those same properties: they are standardized. The theory has a difficulty here, however. It tells us that a particular just is the bundle. But then, if we have the same bundle, it implies that we have the same object. In other words, there could not, on this theory, be more than one object that is the same bundle of properties.

It might be said that this objection is a mere technicality that doesn’t really matter. Couldn’t it just be that, as a matter of fact, two distinct objects never really do share all the same properties? Even tables that are mass-manufactured will have some very slight difference in weight, colour, or even just the pattern of fine, microscopic scratches on the surface. Our snooker balls need only be close enough in their properties for the game to be playable fairly so they too can have some slight differences. This response misses the point of a philosophical theory, however. This was supposed to be an account of what it is to be a particular thing. The truth of that theory should not have to rely on luck working out for it, such that every particular thing just happens to be a different bundle. It does seem at least a possibility that two things could share all their properties. And if, as the theory states, particulars are only and nothing more than bundles of properties, then it is inconsistent with that possibility. Two particulars with the same properties collapse into one.

There are two possible ways out for the bundle theorist but both have problems. The first apparent solution is to say that there is a reason in principle why two particulars could not share all their properties. If one allows relational properties, then these arguably must differ because they allow spatiotemporal location to come into the equation. The following example illustrates what is meant by a relational property. Even if all the red snooker balls are indistinguishable when you inspect them, perhaps one is just 20 centimetres from the bottom-right pocket of the snooker table, while the other is 30 centimetres from it. One ball has the relational property of being 20 centimetres from the pocket, while the other has the relational property of being 30 centimetres from the same pocket. Assuming no two entirely distinct particulars can occupy the same space at the same time, then it seems that all things will bear a unique set of relational properties.

Here is the problem with this proposal. There is no guarantee that distinct things really will have different relational properties unless we reintroduce particulars into our metaphysics. This is why. Should we think of position in space (and time) as an absolute or relative matter? If it were absolute, it would suggest that there is some kind of particularity to spatial positions. A position would be a particular. The notion of a particular–one that is not defined as a bundle of qualities–will have come back into the theory. That’s no good because we were looking to eliminate particulars in terms of bundles of properties.

So do we instead define spatial positions in relation to each other? The problem with doing so is that there is at least the possibility that the space of a universe has a line of symmetry; and thus places in corresponding positions either side of the line of symmetry would bear an identical set of relations to all the other places within the whole of that space. If we then position two of our snooker balls at those corresponding points within our symmetrical universe, then it remains a theoretical possibility that two distinct particulars nevertheless are identical in all their non-relational and relational properties. (This sounds a bit complicated, but Figure 1 shows what’s meant.) On the bundle theory, they again collapse into each other.

This is a complicated argument. A short summary might help. We tried to separate indistinguishable particulars on the basis of them having different locations. But either those locations are themselves particulars, in which case we have not succeeded in eliminating particulars, or locations are just distinguished by their relations to each other. And in the latter case, the possibility of a symmetrical structure means that we could have two particulars that were not distinguishable even on the basis of location.

[image: image]

1. A symmetrical universe

What we just had was a proposed first way for the bundle theorist to avoid the implication that particulars with all the same properties collapse into one. As that didn’t seem to work, here is a second proposal. The objection, that the theory entails bundles with all the same properties must be one and the same, strikes only if the properties are to be understood in a certain way: as nothing like particulars. But there are other conceptions, as we will see in Chapter 2. Perhaps those properties are particularized in some way. Hence, the red in this bundle might be a different thing or instance from the red in another bundle. There might then be the possibility that there are distinct particulars with all the same properties. They consist of the same types of property but different instances of them. Isn’t this what we think of all the red snooker balls? The red of this ball is not the same as the red of that ball. They are two different instances of red.

But there is again a problem with this apparent solution. We have saved the bundle theory but at a cost. An advantage of the bundle theory, it was noted, was that it accounted for particulars entirely in terms of properties. Particularity was reduced away in terms of properties. But it now seems that we are able to salvage the bundle theory from the objection that two identical bundles would collapse into one only if we understand properties in some way as particulars. We spoke of having two distinct instances of red and a property instance looks like some kind of particular. So to make our bundles behave more like the particulars that we take objects to be, we have had to make our properties like particulars. Particularity has managed to sneak back into the theory.

There are countless mistakes that we may have made along the way. But it looks like we might be forced to conclude that particularity is an irreducible feature of reality, for there could, in theory, be two distinct particulars whose distinctness did not consist in them having different properties.

So what, then, is a table? After the considerations in this chapter, it seems that we have to say it is a particular that bears certain properties but is not identical with, nor reducible to, those properties. The table was chosen arbitrarily as the object we examined, and it thus seems safe to generalize from it. We should then give the same answer for any other object.

The properties of particulars have been mentioned throughout this chapter. We need next to consider what these things are supposed to be, if indeed they are things at all. We move on, therefore, to this topic.

Very Short Introductions available now:

ADVERTISING • Winston Fletcher

AFRICAN HISTORY • John Parker and Richard Rathbone

AGNOSTICISM • Robin Le Poidevin



ANARCHISM • Colin Ward



ANCIENT WARFARE • Harry Sidebottom

ANGLICANISM • Mark Chapman


ANIMAL RIGHTS • David DeGrazia

ANTISEMITISM • Steven Beller



ARCHITECTURE • Andrew Ballantyne

ARISTOCRACY • William Doyle

ARISTOTLE • Jonathan Barnes

ART HISTORY • Dana Arnold

ART THEORY • Cynthia Freeland

ATHEISM • Julian Baggini

AUGUSTINE • Henry Chadwick

AUTISM • Uta Frith

BARTHES • Jonathan Culler

BESTSELLERS • John Sutherland

THE BIBLE • John Riches


BIOGRAPHY • Hermione Lee

THE BOOK OF MORMON • Terryl Givens

THE BRAIN • Michael O’Shea


BUDDHA • Michael Carrithers

BUDDHISM • Damien Keown


CAPITALISM • James Fulcher

CATHOLICISM • Gerald O’Collins

THE CELTS • Barry Cunliffe

CHAOS • Leonard Smith

CHOICE THEORY • Michael Allingham

CHRISTIAN ART • Beth Williamson


CHRISTIANITY • Linda Woodhead

CITIZENSHIP • Richard Bellamy


CLASSICS • Mary Beard and John Henderson

CLAUSEWITZ • Michael Howard

THE COLD WAR • Robert McMahon

COMMUNISM • Leslie Holmes

CONSCIOUSNESS • Susan Blackmore

CONTEMPORARY ART • Julian Stallabrass


COSMOLOGY • Peter Coles

THE CRUSADES • Christopher Tyerman

CRYPTOGRAPHY • Fred Piper and Sean Murphy


DARWIN • Jonathan Howard


DEMOCRACY • Bernard Crick

DESCARTES • Tom Sorell

DESERTS • Nick Middleton

DESIGN • John Heskett

DINOSAURS • David Norman

DIPLOMACY • Joseph M. Siracusa

DOCUMENTARY FILM • Patricia Aufderheide

DREAMING • J. Allan Hobson

DRUGS • Leslie Iversen

DRUIDS • Barry Cunliffe

THE EARTH • Martin Redfern

ECONOMICS • Partha Dasgupta

EGYPTIAN MYTH • Geraldine Pinch


THE ELEMENTS • Philip Ball

EMOTION • Dylan Evans

EMPIRE • Stephen Howe

ENGELS • Terrell Carver


EPIDEMIOLOGY • Roldolfo Saracci

ETHICS • Simon Blackburn

THE EUROPEAN UNION • John Pinder and Simon Usherwood

EVOLUTION • Brian and Deborah Charlesworth


FASCISM • Kevin Passmore

FASHION • Rebecca Arnold

FEMINISM • Margaret Walters

FILM MUSIC • Kathryn Kalinak

THE FIRST WORLD WAR • Michael Howard



FOSSILS • Keith Thomson

FOUCAULT • Gary Gutting

FREE SPEECH • Nigel Warburton

FREE WILL • Thomas Pink



FREUD • Anthony Storr


GALAXIES • John Gribbin

GALILEO • Stillman Drake

GAME THEORY • Ken Binmore

GANDHI • Bhikhu Parekh

GEOGRAPHY • John Matthews and David Herbert