www.richmond-philosophy.net - © COPYRIGHT 2002-10 RUTC - Designed and maintained by Mat Carmody.

The Ontological Argument

Peter King



(A harbour-side café somewhere in the Peloponnese; Anna Kalypsas is sitting at a table outside a café with her friend, the lay cleric Theo Sevvis, and they're joined by Anna's students, Mel Etitis and Kathy Merinos. When the two arrivals have ordered their drinks, they turn to the reason for their meeting - a tutorial.)

Anna:  So, have you two been reading up on the ontological argument? Which versions did you look at?

Mel:   Just the traditional ones - mainly Anselm and Descartes.

Anna:  And what did you make of them?

Kathy:   They're odd; I mean, they obviously can't be valid, but it's difficult to see exactly why they don't work. Actually, it's difficult to see how they're supposed to work in the first place, which makes it really hard to see where they go wrong.

Anna:  Yes, that's a common complaint. Well, look, why don't we start by getting clear about the logical structure of the argument, and then look at the various ways that people have criticised it? (They all agree.)

Anna:  OK, well, who wants to start?

Mel:   Should we use Anselm's or Descartes' version?

Anna:  It doesn't matter; I always think that Descartes' is a bit clearer, but it's probably a matter of personal taste.

Mel:   All right, I'll have a go. Well, Descartes starts by explaining what he means by essences. The concept of a triangle, for example, includes the fact that its internal angles add up to 180° that means that, once we've understood what a triangle is, we can infer that its internal angles add up to 180° - and that's completely independent of whether or not there actually is a triangle in the world.

Kathy:   I didn't really get that bit; I mean, if a triangle doesn't exist, it doesn't have any angles, so how can they add up to anything?

Mel:   OK, imagine that you knew about triangles, but you'd never seen one, and someone showed you a geometrical figure and said that it was a triangle. Say you measured its internal angles, and they didn't add up to 180°...you'd say that it wasn't a triangle, wouldn't you?

Kathy:   Well, yes, I suppose so.

Mel:   OK - so why? It's not because you've seen a triangle, and you're comparing the shapes, or anything like that; surely it's because you understand the concept of a triangle, and this figure just doesn't fit that concept.

Kathy:   Er, yeees.

Mel:   So your concept of the triangle must include the fact that its internal angles add up to 180°, because otherwise why are you saying that this other shape isn't a triangle?

Kathy:   OK, I think I see that.

Mel:   Right, but the thing is, all of that would still be true is there weren't any triangles in the world - the concept of a triangle would have to involve the fact about its angles, otherwise it wouldn't be the concept of a triangle, but of something else.

Kathy:   Yes, OK, I think I'm all right on that now.

Anna:  Nicely explained, Mel. So, we've got the idea of properties being essential to conceptual entities, so what's next?

Kathy:   Well, then we take the concept of god...

Anna:  Sorry to interrupt, but I suggest that we don't mention the term "god" yet - I know that that's what we're concerned with, but it keeps the argument clearer if we just concentrate on the nature of the concept rather than what name we give it.

Theo:  Does it? Why?

Anna:  Well, all of these arguments for the existence of god are in fact arguments for a very specific sort of being; for example, the design argument tries to show that there's some kind of universe-designer, and the cosmological argument tries to show that there's a first cause or fundamental explanation of the universe. When we're asking whether the argument works, we don't want to be distracted by the ways that people think about god (being a person, being benevolent, and so on) that aren't relevant to the argument.

Theo:  I see, yes, that makes sense. Very well, go on Kathy.

Kathy:   Oh, all right then - so Descartes starts with the concept of a being that's completely perfect.

Mel:   It might be better to say: a being that has every perfection.

Anna:  Or, to use a more modern term, a maximally perfect being. Yes, fine, go on.

Kathy:   Well, just like with the triangle, we can look at the concept and infer things from it.

Anna:  Things?

Kathy:   I mean, various truths. For example, we can say that the maximally perfect being has the property of omnipotence, and of omniscience, and of existence. So a triangle must have the property of having internal angles that add up to 180°, and in the same way the maximally perfect being must have the property of existence.

Anna:  OK, fine - that's pretty well Descartes' argument. So, is everyone happy that they understand the way that the argument's supposed to work - what it's trying to do?

Theo:  I must confess that I am a little puzzled, Anna. This is not an argument with which I am very familiar, but it seems to me that Descartes is trying to define god into existence. How can the meanings of words prove anything about the world?

Anna:  Well now, Theo, you've hit on quite a common criticism in fact. Any thoughts you two?

Kathy:   I don't know; I'm afraid I thought the same as Dr Sevvis. Descartes seems to beg the question, really. He seems to be going from the definition of "god" (or this maximally perfect being) to existence, but isn't he assuming the existence of god to start with?

Mel:   But I don't really see that. I mean, for a start, Descartes isn't assuming that god exists, only that it makes sense to talk about him, that there's a genuine concept that we can grasp and infer things - I mean truths - from. And he's not interested in words and definitions - he's talking about concepts. It would be odd to go from words to the world, true, but I'm not sure there's such a problem with concepts.

Anna:  Good. Does that satisfy you Theo? Kathy? (Neither of them looks convinced.)

Anna:  Yes, you're right to look dubious. Mel's right too, of course - we are dealing with concepts not words, and that does make a difference. Still, it's not that simple. In most cases, after all, we can't go from concepts to existence. As Mel explained, it's perfectly possible to grasp the concept of a triangle whether or not there are actually any triangles in the world. I can even grasp the concept of an honest politician, so clearly concepts don't always imply existence. We need something more. OK, what else is there Mel?

Mel:   Well, the other thing that I thought was that all arguments - well, all valid arguments - have to work the same way. I mean, in every valid argument the conclusion follows from the premises, otherwise it couldn't be valid. So, for a start, in a way every argument assumes its conclusion, because if the conclusion follows from the premises, then the conclusion must somehow have been in the premises somewhere to start with.

Theo:  I am afraid that logic has never been my strong suit, so you will have to explain a little further Mel. Are you saying that to be valid an argument has to be useless, because it only tells you what you know already?

Mel:   Oh, no Dr Sevvis. It's just that an argument has at least two premises, and usually the conclusion somehow combines bits of the them; so the conclusion was there, but it wasn't obvious until you put the premises together. Of course, if you just say: "Socrates is mortal, therefore Socrates is mortal", you haven't given an argument - but if you say: "All men are mortal, Socrates is a man, therefore Socrates is mortal", then the conclusion says something new.

Kathy:   It doesn't sound new to me.

Mel:   Well neither of the premises said that Socrates was mortal, did they?

Kathy:   No - no, but it's still obvious.

Mel:   That's because it's a simple example, but the point is that the conclusion says something that wasn't explicitly in the premises, even though it was in there really. You could understand that all men are mortal, and you could understand that Socrates is a man, but if you hadn't put them together in your mind you wouldn't realise that they meant that Socrates was mortal.

Ann: We're going to come back to this point later, I think - so if Theo and Kathy are reasonably happy now, shall we move on?

Theo:  Yes, yes, Mel explained it very clearly. (Kathy nods agreement.)

Anna:  Go on then Mel.

Mel:   Oh, OK then. Well, the point is that Socrates isn't mortal because of the premises; I mean, in a way, the premises are true because the conclusion is. After all, if Socrates weren't mortal, then it wouldn't be true that all man are mortal, given that Socrates is a man.

Theo:  And if Socrates were not mortal, then he would not be a man... yes, I see,

Mel:   So if you know that all men are mortal, and you know that Socrates is a man, it follows that Socrates is mortal - but the fact that he's mortal isn't created by the premises, any more than they're created by the conclusion. And it's the same with the ontological argument. If it works, it's because the concept of god and god's existence are connected in the right way. So the premise, the concept of god, doesn't bring god into existence.

Anna:  Good, yes, that's surely right. And in fact Descartes says something very similar; wait a moment. (She pulls a book from her bag and leafs through it.) Here it is, in the fifth of the Meditations: "It is not that my thought makes it so, or imposes any necessity on anything; on the contrary, it is the necessity of the thing itself, namely the existence of God, which determines my thinking in this respect." So if all that satisfies your worries, Theo and Kathy, [they nod] let's get on to the main criticisms of the ontological argument. What are they, then?

Mel:   Well, I suppose that one main criticism is the question of whether existence is a property. Then there are the overload objections.

Kathy:   Oh yes, the perfect island and the Pegasus ones.

Theo:  It seems to me that there is another problem, though. Earlier Anna said that it was better to leave talking of god until the end, in order not to confuse matters. That, however, surely obscures an important point: can we grasp the concept of god or, indeed, of a maximally perfect being? Do not both concepts involve infinity, for example? And can we agree upon what constitutes a perfection?

Anna:  OK, those three criticisms will do to be getting on with. There's another one, in fact, but we can come to that later. So let's start with Theo's worry. Any responses to it, you two?

Mel:   Well, I think I know what to say about grasping the concepts, but I'm not sure about the business of perfections; that problem occurred to me as well, when I was reading Descartes.

Anna:  Fair enough; start with the concepts, then.

Mel:   You see, I don't really see the problem about infinity. I mean, we do understand the concept of infinity, don't we? We don't know everything about it, of course, but then we don't know everything about triangles, either, but I can still understand the concept of a triangle. I don't know, but perhaps there seems to be a problem because there's some sort of idea of the mind as a container, so that infinity can't fit into it - but the mind isn't a container like that, and we're talking about the concept of infinity, not infinity itself. After all, my idea of an elephant isn't any bigger than my idea of a mouse...

Theo:  Yes, I can see that, I suppose - but I think that the more important point concerns the question of perfection. After all, what I consider to be a perfection might not be what another person considers to be one. One person might say that omniscience is a perfection, for example, but someone else might argue that the ability to learn and grow is essential to being perfect.

Kathy:   I see what you mean, Dr Sevvis. It's got to be perfect, but what is being perfect?

Mel:   Well, I don't have an answer - but it doesn't feel right. I mean, it just seems too obvious. Wouldn't Descartes have noticed?

Anna:  Not to mention his critics... yes, that's a good point. Descartes was a great philosopher, and although he does make mistakes, he doesn't usually make silly or obvious ones. That's not enough to answer Theo's point, though; we mustn't just appeal to authority, and say that Descartes must have been right because he was so clever. In fact the problem here mainly stems from a change in our language. When we see the word "perfect" we automatically think of what's best - we think in purely evaluative terms. Well, in Descartes' day, the meaning of "perfection" was closer to its etymology; it comes from the Latin for "complete" or "finish". A perfection was a property that was complete; so power is a property, and complete power - omnipotence - is a perfection.

Theo:  Ah - I see; I should have made the connection with Latin. And it makes good sense theologically, too.

Anna:  In fact we still use the old meaning sometimes; after all, a perfect stranger is a complete stranger, not the best stranger we've ever seen, and the present perfect tense is used when some action has been completed.

Mel:   Oh, I see; in music, we talk about a cadence being perfect when it ends on the tonic chord, so that it sounds like a proper finish.

Kathy:   And in botany a plant's perfect if it has all its parts.

Anna:  That's right. There's another point to make, though. If I talk about all the Greeks who are over fifty, you understand what I mean? (They all nod.) OK - so who are they?

Theo:  But how could we know that?

Anna:  Fair enough; how many of them are there?

Kathy:   But that's no better. Well, I suppose that we could check the census.

Anna:  So you don't know how many Greeks are over fifty, or who they are; yet you understand the concept. And if I said that Kathy is one of them, you'd all know I was wrong, and if I said that every Greek who's over fifty is also over forty, you'd know that that was true... And it's the same with the concept of the maximally perfect being - the being with all perfections. We needn't know how many perfections there are, or be able to say what they all are; as long as we know what a perfection is, then we know what it means to say that a being has all of them. What's more, we can use the concept, and deduce truths from it; in fact, we can deduce truths from it just because we don't know everything about it. Well, but we'll probably come back to that later.

Theo:  I see. But that surely raises another question: even if the argument works, it proves the existence of this maximally perfect being, but it leaves open the identity of that being.

Kathy:   I'm sorry, Dr Sevvis, but I don't really understand that.

Theo:  You see, what we believers mean by god is complex, but it includes the idea that god is a being who has certain properties - essential properties. The problem with this ontological argument seems to be that it only works because, as Anna says, it is not specific about the properties that this maximally perfect being has. But why, then, should we call such a being god at all?

Anna:  Yes, some people make a similar point for a very different reason; they say that, because they think of belief as based on faith, the ontological argument (and all the other arguments for the existence of god) is just irrelevant. In other words, if the argument's sound, it proves the existence of something other than god.

Mel:   But that's a bit weird, isn't it? I mean, if it isn't god, what is it? And even if they don't accept that it is god, don't they have to account for it somehow?

Theo:  Yes, you are of course right, Mel. It is difficult to see what room there could be for both a maximally perfect being and god. I suppose that my worry must be misplaced, then.

Anna:  Well, there's an important point there, though. As I said earlier, all the arguments for the existence of god - this one, the design argument, the cosmological argument, and the rest - try to establish the existence of some specific sort of being. The ontological argument goes for a maximally perfect being, but others are concerned with a designer, a first cause, a source of morality, and so on. None of them can show that the specifically Christian, or Jewish, or Islamic god exists, and that's why people sometimes talk about the god of the philosophers as distinct from the various gods of religions. But as we've seen, the two sides can't be treated completely independently of each other. Still, we can leave that aside for the moment; we're interested in whether the argument's sound, not what believers should say if it is. So let's go on to the next criticism. Where were we again?

Mel:   We're sticking with properties, I think - the question of existence.

Anna:  Oh yes, of course. OK then, so what's the problem?

Kathy:   Is existence a predicate?

Anna:  Right, but let's be precise about it. I know that that's how it's generally expressed, but strictly speaking we should be talking about existence as a property and "exists" as a predicate - so the question is either "is existence a property?" (a question about metaphysics) or "is 'exists' a predicate?" (a question about logic). Let's stick with metaphysics, and ask about existence; so, why shouldn't it be a property?

Mel:   I've done some work on this, because I'm going to write my first essay on it, and I think that it's something like this. Say I describe a clarinet - I describe it in perfect detail, saying where and when it was made, and how and by whom; I describe the kind of ebony it's made of, the silver nickel of the keys, the corking, the padding, the tone, etc... everything. So now you all have a perfect - a complete - image of the clarinet.

Anna:  An image?

Mel:   Well, OK, we're not really talking about imagining but about conceiving - but it just makes it easier to use images.

Anna:  Fair enough; it doesn't matter here I think. Go on, Mel.

Mel:   Right, so you've all got this perfect concept, a perfect image of the clarinet, and then I say: "Oh, by the way, I didn't make up that description of the clarinet - it's not just my ideal clarinet or anything; it exists". Now, what happens to your image of the clarinet?

Kathy:   I don't know - does it become real?

Mel:   Well, I don't really know what that means; either the image was the image of a clarinet or it wasn't. What would an image of a non-existent clarinet look like? How would you tell the difference between a non-existent clarinet and a non-existent trombone?

Anna:  Good, I think that's right Mel; we can't conceive or imagine something as non-existent - we conceive it as existent, and then we judge that something in the world does or doesn't match the concept. So when I think that unicorns don't exist, I conceive of existing unicorns and then judge that nothing in the world matches that concept.

Kathy:   So existence isn't a property then?

Anna:  We can't say yet - we've just been seeing why someone might say that it wasn't. But isn't there something odd about Mel's example?

Theo:  Odd? In what sense?

Anna:  Kathy? Mel? Any ideas?

Kathy:   I don't know if this is what you mean, but it's a bit funny to say that Mel didn't tell us anything when she said the clarinet existed. I mean, surely she told us something?

Anna:  Exactly. But we've seen that she didn't tell us anything about the clarinet, because what she said didn't change our concept of that - so what did she tell us about?

Mel:   Was it... about the world?

Anna:  Yes! You changed our concept of the world, by telling us that it contained the clarinet. And another way of putting that is to say that you told us about the description of the clarinet - about its properties; you told us that that set of properties was co-instantiated in the world. In other words, just as the property of blueness is instantiated in the sky and the paint of that door, and just as the properties of brownness and roundness are co-instantiated in Mel's eyes, so the properties of blackness, being made of ebony, being such and such a length and weight and age, etc. etc., are co-instantiated in the clarinet. OK?

Mel:   So when I said that the clarinet existed, I was really saying something about the properties of the clarinet?

Ann: That's certainly one way of explaining what you were doing. Another way of putting it is to say that existence is a second-order property; that is, it's not a property of things, but a property of properties.

Mel:   Oh, I've seen something like that in one of the books - something about second-order predicates, though.

Anna:  Yes, that's right, it's the same thing but in terms of logic and language rather than metaphysics. A second-order predicate is a predicate of predicates, just as a second-order property is a property of properties, and it's easy enough to recast the ontological argument so that it uses existence as a second-order property. I'll leave you to work out how to do that. If you're happy that this objection doesn't work (they all nod) we can get on to the next one; what did we say that it was?

Kathy:   The overload objection, I think.

Anna:  Ah, right, yes...objections, in fact, because there are two different kinds. OK, what are they then?

Kathy:   Shall I start with the first one? That monk's?

Anna:  Gaunilo's, yes. Go on then.

Kathy:   Well, he's criticising St Anselm's version really, not Descartes', but it fits both. He says that, just like the ontological argument, you can imagine - (Anna looks a warning) conceive, I mean - a perfect island - but a perfect island that existed would be better than a perfect island that didn't, so to say that the perfect island doesn't exist is a contradiction. So the perfect island must exist. Is that right?

Anna:  Pretty well. Why is it called an overload objection?

Mel:   Because Gaunilo's saying that, if the ontological argument works, then so does this one, and an infinite number of others, so that the world is overloaded with perfect things.

Anna:  That's right; in other words, he's not trying to say what's wrong with the argument, he's just saying that if you don't like all those other arguments (and who would?), then you'd better not accept the ontological argument either.

Theo:  Is that not rather dangerous? After all, if we cannot find anything wrong with the ontological argument, then this Gaunilo is committed to the existence of lots of strange things.

Anna:  Strange but perfect things, yes. Don't worry, though, his argument doesn't work; can anyone see why?

Mel:   Not really - the perfect island argument seems to me to be just the same as the god argument, so if one works then the other should, shouldn't it? I must be missing something, but I don't see what.

Anna:  Well, what exactly is the essence of the perfect island? What does the concept of the perfect island have to contain?

Mel:   Oh, I see; you mean that Kathy might like palm trees, coral lagoons, and no-one else around, while Dr Sevvis, for example, might like pine trees, rock pools, and dusky maidens?

Anna (smiling at Theo's reaction): That's the first problem, yes; there isn't one concept of the perfect island - there are lots of concepts, because what's perfect for each of us will be different. And not just different, of course, but contradictory, as with being uninhabited and being inhabited by Theo's maidens...

Theo:  Yes, yes, highly amusing. But to be serious, could not Gaunilo have responded by accepting your objection, but pointing out that he meant that the island was perfect as an island. That is, not perfect for me or for you, but in itself. After all, that is more like the premise in the ontological argument, is it not? The concept of the perfect being is not of a being perfect for one person or another, but just perfect.

Anna:  Good, yes, that's half of the point. But it's only half. After all, what's an island? (They all look puzzled.)

Anna:  Don't worry, it's not a trick question.

Kathy:   Well, I suppose that it's a body of land surrounded by water.

Anna:  Right... so what's a perfect island?

Mel:   Oh, I see - something's either an island or it isn't; it doesn't make sense to say that it's perfect or not.

Anna:  Or another way of putting it would be to say that every island is a perfect island, because every island is perfectly (or completely) an island. And the same goes for anything else you care to mention: perfect scissors, a perfect olive, and so on.

Theo:  I think that I see now. In the first case, with the palm trees, and the, well, with the palm trees, et cetera, the problem was that we were thinking about what was a perfect island for us, while in this second case, the problem is that we are thinking about something being perfect as an island. In the ontological argument, however, the concept is of a perfect being; not perfect for us, or perfect as a particular thing, but simply perfect.

Anna:  Yes - and it's the maximality that does the work really. Because the concept is of a maximally perfect being, a being that has every perfection, we're not picking out which perfections we'd like, or which are most appropriate for the kind of thing we're talking about.

Mel:   All right, I can see that Gaunilo's version doesn't work, but what about the other version - John Cottingham's Pegasus one?

Kathy:   But isn't that the same as the Gaunilo, just with a different example?

Mel:   No, it works differently; it concentrates on the existence part, not the perfection.

Theo:  What exactly is this version?

Mel:   It's like Gaunilo's, because it just tries to show that there are lots of arguments with the same structure as the ontological argument, so that if you accept one, then you have to accept them all. If you take the concept of Pegasus - a winged horse - then you can see that it's necessarily true that it has wings, because being winged is part of its concept. Now, we can create a new concept, the concept of super-Pegasus, which is the concept of a winged horse that actually exists. So it's necessarily true that super-Pegasus exists, because existence is part of its concept.

Anna:  Very good, yes. Now, it's obvious that the objection to Gaunilo's overload objection won't go through, because this one doesn't talk about perfections at all. So, what's wrong with it?

Mel:   I don't know exactly, but there must be something. I mean, I don't think that Descartes' argument works, but it doesn't feel like a cheat; this super-Pegasus example does.

Anna:  I know what you mean, it feels like that to me too - and I think that that's because it is a cheat. Look, say you go up to someone in the street and ask her if she understands the concept of a being that has every perfection; do you think that she'd be able to?

Mel:   Well, I don't see why not; you might have to explain what a perfection is, but that's all.

Anna:  Right; and now you ask her whether she thinks that such a being exists. That's a reasonable question, isn't it?

Kathy:   Why not? I don't know what she'd say, though; it'd depend on her religious beliefs I suppose.

Anna:  Good, I agree. But now ask the same person if she understands the concept of super-Pegasus; what do you think she'll say?

Mel:   Well, unless she's read Cottingham, or some other philosophy book, I don't suppose she would.

Anna:  No, why would she? It sounds like a cross between classical mythology and Marvel comics.... perhaps a winged horse in blue and red lycra. So you'd have to explain the concept to her - but how would you do it?

Theo:  There is only one way, surely. One would have to say that it is a winged horse that exists.

Anna:  Exactly. And now the question "Do you think that it exists?" sounds odd, because you've just said that it does. In other words, in the first case, the ontological argument, we can understand the concept without knowing the relevant property - existence - so that it's possible to infer that property from the concept. It's just like the example of the triangle: someone can know what a triangle is without knowing that its internal angles add up to two right angles, but because she has the concept, she can infer facts about it like that.

Mel:   Yes, but it's not possible to understand the concept of super-Pegasus without knowing the relevant property, so even if it makes sense to talk about inferring existence it's trivial.

Kathy:   Oh, I think I see. You mean that the super-Pegasus example begs the question, because you can only accept the concept if you've already accepted the existence bit.

Mel:   Right, but the ontological argument doesn't beg the question, because you can understand the concept without knowing what any of the perfections is, including existence.

Anna:  I agree; if the ontological argument looks like pulling a rabbit out of a hat, this super-Pegasus argument does it after putting the rabbit into the hat in full view of the audience. I think that the problem is that the overload objections concentrate wholly on the logical structure, and it's true that arguments like the perfect island and the super-Pegasus ones have the same logical structure as the ontological argument. The problem is that there's more to an argument than just its logical structure; we have to think about the nature of the premises, and especially the concepts involved.

Theo:  Yes, I can see all that - but I must confess that I am a little bewildered. You seem to have defended the ontological argument against all of its critics. Are you saying that the argument works?

Anna:  Don't worry! I'm not suddenly going to declare my conversion and start coming to church. No, there is something wrong with the argument; in fact it's often mentioned by critics, but for some reason - perhaps because they think that it's just too obvious - they don't spend as much time on it as on all those other criticisms. Any ideas you two? (Kathy shakes her head; Mel is hesitant.)

Mel:   Is it something to do with the objection I've seen somewhere that all the ontological argument proves is that, if there's a god, then he necessarily exists?

Anna:  Yes, that's the sort of thing that I meant... but it's not really very clear, is it? We need to spell it out a bit more. The best place to start is probably with the idea of existence again. Remember that we agreed that saying that something exists is saying that the world contains it? Well it's not as simple as that. Think about a novel or a play; Aeschylus' Eumenides, for example. When Orestes kills his mother, he's pursued across Greece by the Erinyes; now, you might hear people disagreeing about the play, some saying that the Erinyes didn't really exist, that they're just in Orestes' imagination, a product of his feelings of guilt. The people who disagree, and say that the Erinyes really existed - do they believe in the real existence of such creatures?

Mel:   No, I shouldn't have thought so. They mean that the Erinyes existed in the play.

Anna:  Exactly. The point is that, whenever we talk about things existing, we relativise the existence to some specific area - what I call a domain of discourse; it might be the real world, Aeschylus' play, maths, the logically possible, etc. Usually we do this relativising implicitly, and the context tells us what's meant.

Mel:   Oh, I see. So you mean that, if the waiter here says: "Sorry, there's no more ouzo", he doesn't mean that there isn't any in the whole universe, just that there isn't any in this café.

Anna:  That's it - that's a geographical domain of discourse.

Kathy:   So what you say in one domain of discourse isn't true in another?

Mel:   Well it sometimes it must be; after all, there's no potion of eternal life in the café, and there's none in the rest of the world either.

Anna:  Or, going back to Aeschylus, the goddess Athena exists in the play, not in the real world, but the city of Athens exists in both. The point, of course, is that you can't just assume that because something's true in one it's true in the other; you have to have independent grounds for each. Now, what's going on in the ontological argument?

Mel:   Well, we start with the fact that there's the concept of a maximally perfect being.

Anna:  Right - in other words, the maximally perfect being - god - exists in the domain of discourse of concepts, or of possibilities. Now, the ontological argument proceeds: we can infer that god has the property of omnipotence, in the domain of discourse of concepts; we can infer that he has the property of omniscience, in the domain of discourse of concepts, and we can infer that he has the property of existence...

All: (in chorus) ... in the domain of discourse of concepts!

Anna:  So it looks as though the argument's valid all right, but its conclusion's just what we started with: that the concept of god exists. We wanted a bridge to take us from mere possibility or conceivability, but the argument doesn't give us one.

Theo:  But is there not a factor that you have left out? Surely, the point of the argument is that god is maximally perfect, or the greatest conceivable being; would Descartes or Anselm not argue that it is better to exist in the real world than merely in the world of concepts or of possibilities? That a perfect being cannot therefore be only a possibility or only a concept?

Anna:  Yes, that's right - but why should we accept that?

Mel:   I'm not sure, but isn't it better to have a fortune in the bank in the real world than just in your concepts?

Anna:  Ah, that's a good example; yes, of course it is - better for you. But when we were talking about the perfect-island kind of overload objection that the ontological argument can't be about what's perfect for you or for me; it has to be about what's perfect full stop. Now, imagine a character in a book who wins a fortune. Is there anything that I could buy with a fortune in the real world that she can't buy with her fortune in the book?

Mel:   Oh, I suppose there isn't.

Anna:  That's right - and surely the same's true in this case: there's nothing an omnipotent being could do in the real world that it can't do in the world of possibilities, or of concepts; there's nothing an omniscient being can know in the real world that it can't know in the world of possibility, or of concepts. So a real maximally perfect being is not more perfect than a merely possible or conceptual maximally perfect being. We might be better off with the real one, but that's just an argument from wishful thinking, isn't it? Anyway, look at the time! I don't know about you lot, but I need to be gone. I'm supposed to be giving a lecture in five minutes. What are we doing next week?

Mel:   The design argument, I think.

Anna:  OK - well, you've got the reading list, so I'll leave you to get on with it. Sorry to rush Theo; perhaps we'll see you next week?

Theo:  I hope so, if I can make it.

(Anna rushes off to the university, Theo strolls away, and the two students order more drinks...)

Peter King
Pembroke College, Oxford
peter.king@pmb.ox.ac.uk


References and Further Reading

René Descartes. Meditations

Jonathan Barnes. The Ontological Argument (London: Macmillan, 1972)

J.C.A. Gaskin. Hume's Philosophy of Religion (London: Macmillan, 1978)

Gerard J. Hughes. The Nature of God (London: Routledge, 1995)

Anthony Kenny. Reason and Religion (Oxford: Blackwell, 1987)

Alvin Planting. The Nature of Necessity (Oxford: Clarendon Press, 1978)

Alvin Plantinga [ed.] The Ontological Argument (London: Macmillan, 1965)

Bernard Williams. Descartes (London: Penguin, 1978)