Excursus on Natural Theology (Part 26): The Ontological Argument

May 18, 2016

A Posteriori Defense of the Ontological Argument

For today’s lesson, we want to attempt to conclude our discussion of the ontological argument for God’s existence. We saw that the crucial premise in this argument is the first one – that it is possible that God (or a maximally great being) exists. If it is possible that a maximally great being exists then it follows that a maximally great being (or God) actually does exist. So the key premise is that first one – what warrant or justification is there for thinking that it is possible that God exists.

We had looked at a priori arguments, or evidence or warrant, for that first premise. As I said, the word a priori (if it is not part of your vocabulary, you can remember what it means by the word “prior” in it) means prior to experience. This is not based upon empirical studies of the world. It is not based upon your experience through the senses. This is prior to experience. We saw that when we reflect upon the idea of a being which has omnipotence, omniscience, and moral perfection in every possible world that this seems to be a coherent idea. It seems to be perfectly coherent, and that gives good reason for thinking that it is possible that such a being exists. Then we looked at objections to those a priori considerations and found them to be inconclusive.

But that is not the end of the story, for although the ontological argument is usually presented as the paradigm example of an a priori argument, Plantinga suggests that there might actually be some a posteriori considerations that would support that key premise. You can remember what this means by seeing the word “posterior” in this Latin phrase. These are Latin expressions. In it you see the word “posterior.” That connotes posterior to experience or based on experience. There are certain experiential knowledge that we have that might lead us to think that that key premise is true. So, for example, Plantinga says that if we “carefully ponder” the key premise and the alleged objections to it, and if we “consider its connections with other propositions we accept or reject and still find it compelling, we are within our rights in accepting” this premise.[1] So there may be some a posteriori considerations that would reinforce these a priori intuitions we have of the premise’s truth.

What might these be? It seems to me that the other arguments of natural theology for the existence of God can go some distance toward making us think that it is possible that such a being exists. For example, Leibniz’s argument from contingency, certain versions of the moral argument for God’s existence, and a so-called conceptualist argument for God’s existence could lead us to think that it is plausible that maximal greatness is possibly exemplified.

For example, remember the Leibnizian argument from contingency that we’ve already talked about. It goes like this:

1. Everything that exists has an explanation of its existence. This explanation might either be a necessity of its own nature (in which case it is a metaphysically necessary being) or in some external cause.

2. If the universe has an explanation of its existence, that explanation is God.

3. The universe is an existing thing.

From these three premises it follows logically that:

4. Therefore the explanation of the existence of the universe is God who is a being that exists by a necessity of his own nature.

Any being that exists by a necessity of his own nature is going to be a metaphysically necessary being. So Leibniz’s argument gives us good grounds for thinking that there is a metaphysically necessary being upon which everything else depends for its existence.[2] So there are no limits to the power of this being. There are no non-logical limits to the power of this metaphysically necessary being because everything that exists depends upon it for its existence.

Or consider the moral argument for God’s existence. Again, we’ve already talked about the moral argument. It could be summarized like this:

1. If God does not exist then objective moral values and duties do not exist. If there is no God then there is no absolute standard of right and wrong, good and evil. Everything becomes relative.

2. Objective moral values and duties do exist. Certain things are really right or really wrong. Certain things are really good or really evil.

3. Therefore God exists.

Now, I’m not interested in redefending the moral argument now. We’ve already done that. But what I want to point out is that if moral values and duties depend in this way upon the existence of God then, since certain moral principles are necessarily true, it follows that the God who grounds them must be a metaphysically necessary being. That is, a being that exists in all possible worlds. For example, one naturalist philosopher of science, Michael Ruse, has written, “The man who says that it is morally acceptable to rape little children is just as mistaken as the man who says 2+2=5.”[3] Think of that statement. Here Ruse equates the truth of these moral principles with the elementary truths of arithmetic like 2+2=4. These are necessary truths. These are logical truths. Therefore these moral principles, to be grounded in God, require that God be a metaphysically necessary being who exemplifies these moral values necessarily. Therefore the moral argument also leads to the existence of a metaphysically necessary being who is not merely perfectly good but who is the actual standard of goodness, who is the paradigm of goodness from which all moral value derives.

Thirdly, in addition to this, some people might be persuaded by a sort of conceptualist argument for God’s existence as the best grounding for abstract objects like mathematical objects such as numbers and sets and functions and geometrical shapes. An argument might go along these lines:

1. Abstract objects like numbers and other mathematical objects are either independently existing realities (that is Platonism that says they just exist on their own) or else they are concepts in some mind. This premise would involve a denial of anti-realism with respect to these objects, and therefore I myself am not persuaded by this argument as it stands because I think it is defensible to say there just aren’t any mathematical objects – they just don’t exist at all. But if you are a realist about them then it would seem that you would think they are either independently existing objects (abstract entities of some sort) or else they are concepts in a mind.

2. Abstract objects are not independently existing realities.

This would involve rejecting Platonism about these abstract objects. You would probably do so on the grounds of their causal isolation. If these objects exist, they don’t exist in space and time, they have absolutely no causal influence upon the world, and therefore they are irrelevant. How could we know anything about them anymore than I have knowledge of some distant village in Nepal in the Himalayas, and why would they be applicable to the physical world? Why would they be so useful in science if they are causally isolated from the world?[4] So it would be plausible to think that abstract objects are not independently existing realities.

3. If abstract objects are concepts in some mind then an omniscient, metaphysically necessary being exists.

If mathematical entities are concepts in a mind, they can’t be concepts in some human mind because there are too many of them to be thought about by any finite mind. They would be necessary in their existence. If 2 and 4 exist, they don’t happen to just exist in the actual world. 2+2=4 is a necessary truth, so 2 and 4 would exist in every possible world. Therefore if these mathematical entities exist in some mind as the concepts of that mind, it must be an infinite, metaphysically necessary person who grounds these abstract objects. Thus you are brought to the conclusion of the existence of an omniscient, metaphysically necessary mind as the foundation of abstract objects.

Look at what these three arguments give us. Each one of them gives us a metaphysically necessary being – a being that exists in every possible world. The Leibnizian argument, the moral argument, the conceptualist argument all yield a metaphysically necessary being. The Leibnizian argument gives you a being which is the source of all reality outside itself and therefore plausibly all-powerful. It has no non-logical limits on its power. The moral argument gives you a being which is morally perfect. The conceptualist argument gives you an omniscient being which grounds these mathematical truths. In other words, you’ve got here the elements that make up maximal greatness, don’t you? You’ve got metaphysical necessity, omnipotence, moral perfection, and omniscience.

So these other arguments of natural theology can provide support for the idea that it is possible that a maximally great being exists. I think here as well considerations of simplicity also come into play. For example, it is simpler – isn’t it? - to postulate one metaphysically necessary, infinite, omniscient, morally perfect being than to think that there are three separate beings that have these properties, that there are three metaphysically necessary beings – one of which is omniscient, one is morally perfect, one is omnipotent, and the source of all reality. It is simpler to say that all of these arguments are leading to the one metaphysically necessary being who is omnipotent, omniscient, and morally perfect.

Moreover, as Richard Swinburne has pointed out, it seems plausible that it is simpler or less ad hoc (less contrived) to say that a property that comes in degrees has either zero or infinity as its value rather than some arbitrarily selected finite value in between. It would be simpler to just say it is infinite or it is zero rather than having an inexplicable finite value somewhere on the measurement. So it would be more plausible to think that maximal greatness is possibly instantiated than quasi-maximal greatness – a being which has infinite power, intelligence, and perfection morally.

So on the basis of these sorts of considerations which are a posteriori I think that we could well consider ourselves to be warranted in believing that it is possible that a maximally great being exists.


Student: The only hesitation I have with using the other arguments for the existence of God with the ontological argument is because I feel it makes the ontological argument redundant.[5] I always like it to stand by itself. Even when you bring this up to some atheist friends they point out the same thing. Then they will just say, “What do you even need the ontological argument for at that point?”

Dr. Craig: That is the next point that I am going to make. So let me hang on to that question. I will get to it. But don’t forget that we still have the a priori warrant for the key premise. These a posteriori arguments confirm that or are a second line of defense, but it is not as though the ontological argument depends wholly on these a posteriori considerations.

Student: I was curious regarding the moral aspect of your talking about . . . you used the example of numbers to say that they exist independently of any one person essentially. Their existence is dependent on a maximally great being. What about moral virtues such as things like courage or honesty or integrity? You could also say that those exist independently of any one given individual person. Could that, in a sense, add fuel to the argument as well? To what degree does it relate?

Dr. Craig: Yes, that is what I was trying to say. I think that you’ve captured correctly what I was trying to say. Imagine a world in which no human beings exist. The moral principle “It is wrong to torture a child for fun” would still be true in such a world. That principle doesn’t depend for its truth upon any actual children that exist. Or imagine a world in which there are actual children but none are actually tortured. That moral principle would still be true in that world that it is wrong to torture a child for fun. Plausibly, as you say, these moral principles are true independent of whether any people exist or these actions actually occur. Therefore they need a grounding that will be metaphysically necessary – that will exist in every possible world.

Student: When you said a metaphysically necessary being is simpler than any other explanation . . . or less ad hoc I think you said . . . what about people who posit physical explanations for the universe like multiverse hypotheses? And they would sometimes consider those more simpler than going to the metaphysical – staying in the physical.

Dr. Craig: What you are doing there is recurring to those arguments again and I’d rather not go back and rehearse those arguments once again. Look at our discussion there where I interact with multiverse hypotheses as a possible explanation of the origin of the universe. What I simply want to do here is say let’s suppose that you are convinced that these arguments are good. We’ve already defended them. Suppose you agree with me that these three (or at least two of the three) are good arguments. Would these lend a posteriori support to the ontological argument? That is all I am saying here now. Again, it would take us too far afield to go back and rehearse the arguments once more. Look at the discussion of those. What I want to draw out here is the remarkable coalescence, I think, of the conclusions of these arguments with the key premise in the ontological argument.


Now the question which arises at this point is the one that was just broached. Doesn’t the ontological argument now become question begging? What does it mean for an argument to be question begging? What it means is that your only reason for accepting a premise in the argument is that you already believe in the conclusion. You accept the premise because you already believe the conclusion is true. So you are begging the question or you are arguing in a circle. Let me give an example.

Consider this as an argument for God’s existence.

1. Either God exists or the moon is made of green cheese.

2. The moon is not made of green cheese.

3. Therefore God exists.

That is a sound argument for God’s existence. Both of its premises are true. Since God does exist it is true that either God exists or the moon is made of green cheese. In order for that disjunction to be true, one of the disjuncts needs to be true. Since God does exist that is true that either God exists or the moon is made of green cheese. The second premise is true – the moon is not made of green cheese. Therefore God exists.[6] So this is a logically valid argument with true premises which implies the existence of God. Yet no one would think this is a successful piece of natural theology. Why not? Because the only reason you would have for thinking that the first premise is true (that either God exists or the moon is made of green cheese) is because you already believe the argument’s conclusion that God exists. So you are reasoning in a circle. You are begging the question.

The question now is: if the reason that you think it is possible that a maximally great being exists is because you believe that a maximally great being does exist then aren’t you similarly begging the question? If on the basis of the Leibnizian, moral, and conceptualist arguments for God’s existence you come to believe that a maximally great being exists then of course it is possible that he exists. You’ve proven that he already does! So doesn’t the ontological argument then become question-begging or redundant?

I think that this misgiving may result from thinking of the project of natural theology in too linear a fashion. We should not think of the arguments of natural theology as being like links in a chain where the chain is only as strong as the weakest link. If the weak link is broken then the whole chain is broken and becomes unable to bear the weight that you would want. We shouldn't think of the arguments of natural theology in that linear fashion, as like a chain of links which is only as strong as the weakest link. Rather the arguments of natural theology are more aptly compared to a coat of chain mail like a knight used to wear where the links all reinforce each other so that the coat of mail is not as weak as the weakest link. In a case like that the ontological argument can play, I think, an important role in a cumulative case for the existence of God in which a whole host of independent factors lead us to the overall conclusion “therefore God exists.”

If that is right then Anselm was wrong in thinking that he had discovered a single argument which standing alone would prove God’s existence in all of his greatness. I don’t think the ontological argument would succeed in that way. But it does seem to me that the ontological argument does have a role to play as one of the links in the coat of mail – one of the arguments in a cumulative case for God’s existence. So the arguments taken together show that God, a maximally great being, does in fact exist.


Student: Just an observation with this. It seems like these natural theology arguments would be interrogatory in nature. You get out of this begging of the question or circularity. Why do I think I have a concept of a God that’s omnipotent? Why do I have an experience of salvation, why would we have such experiences?

Dr. Craig: It seems to me what you are saying there – correct me if I’m wrong – another link in the chain mail could be religious experience. Religious experience might be one of the links in this coat of chain mail leading to the conclusion that God exists. I think that would be certainly justifiable. There are many other arguments for God’s existence that we haven’t surveyed in this class.[7] I am participating in a book dedicated to Alvin Plantinga that is based upon an essay Plantinga wrote back in 1982 called “Two Dozen or So Arguments For God’s Existence.” The folks over at Baylor University have decided to do a volume honoring Alvin Plantinga by assigning each one of these two dozen or so arguments to a contemporary philosopher and writing a chapter on where that argument stands today. I was assigned the kalam cosmological argument, as you might anticipate. Big surprise! But just think of that – two dozen or so arguments in Plantinga’s coat of chain mail supporting theism. We’ve just looked at a few. Certainly that could be a part of it.

Student: I was just thinking we could frame these types of arguments like ontology, personal experience, properly basic things, as questions as to why do we have the dimension to ask such a question rhetorically rather than say “I am using this as a given.”

Dr. Craig: Right. Don’t forget about what we said right at the beginning that I think it is perfectly defensible to say you don’t need arguments in order to know that God exists, that this can be a properly basic belief in the same way that my belief in the reality of the world around me is properly basic or in my belief that there are other minds besides my own as a properly basic belief not grounded in argument.

Student: I’ve actually seen that kind of linking in many different kinds of interactions like where if you talk about the ontological argument it goes to the contingency argument or something else. I’ve seen it also with the moral argument jumps into the argument from desire. There always seems to be some kind of something, and they kind of back each other with people having questions that lead into something else. Or the cosmological argument going into the design argument. My question about the ontological argument – because you were saying last week that the way that Plantinga presents it that it represents God as a necessary being – I was curious if it is kind of a blend into the contingency argument. Doesn’t that basically just say that God is necessary without needing . . .?

Dr. Craig: All three of the arguments that I mentioned today (contingency, moral, conceptualist) lead to a metaphysically necessary being. So this is not just a feature of the ontological argument or even of Leibniz’s argument. Several of the arguments for the existence of God lead to not just God’s existence in the actual world but God’s necessary existence. In fact, I think it is arguable that even the kalam cosmological argument could lead to the existence of a metaphysically necessary being since it would be metaphysically impossible for the world to be past-infinite, past-eternal, and metaphysically impossible for something to come out of nothing. So in every possible world there would need to be a cause of the origin of the universe. I haven’t pressed that point, but I think that is defensible.

Student: My question was more of: providing that God is necessary a priori.

Dr. Craig: The Leibnizian argument doesn’t assume that a priori. I think it gives an argument based on an explanation – the nature of explanation. The only adequate explanation for the existence of contingent beings is a metaphysically necessary being. So this is still an argument. It is an inference. It is not just an assumption.

Student: I think the ancients used to think if anything exists then it is either contingent or it has to be necessary. Ultimately if anything exists there has to be a necessary existence that has always been. If you say there could be more than one then they would have no contact with each other. For this existence there always is a maximal supreme being.

Dr. Craig: Let me just say a couple of things. You are certainly right that ancient Greek philosophers did think there was a necessary being, but the way they understood necessity was different than the way it is discussed today. For Aristotle, to say something exists necessarily basically meant it exists eternally and wasn’t subject to being generated or corrupted.[8] It would just always exist. But it wasn’t the sort of logically necessary existence that we are talking about today. That concept of God evolved among medieval Islamic theologians such as al-Faradi and Ibn Sina and then was inherited in the West by our medieval Latin-speaking theologians like Anselm and Thomas Aquinas and others. That was a deeper more robust understanding of necessity than the ancients had. So while your point is correct, they understood the word differently than it came to be understood.

Student: I just want to say I recently read a blog post that was talking about the ontological argument and the title was very provocative. It was about Richard Dawkins going around in his bus that said “God probably doesn’t exist so relax and enjoy your life.” This is going back to the a priori saying “If God probably doesn’t exist then God exists.” He is saying that if it is even possible that God exists then therefore God exists. If Richard Dawkins is right that God even probably doesn't exist then that means that God exists.

Dr. Craig: I get the point. If there is any possibility that he exists then the ontological argument goes through. I think that you are making a good point. The atheist has to maintain not simply that God does not exist. He has to maintain that it is impossible that God exists because once you grant the possibility then the argument goes through. The atheist is taking a pretty radical line here – not simply that there is no God, but that it is impossible that God exists. That is a helpful perspective to keep in mind.


With the ontological argument we draw to a close our discussion of the arguments of natural theology. Let me ask if there is any final comment that anyone has on this section of the course which is so, I think, critical toward establishing the coherence of a Christian worldview in a secular society, namely that God exists. We’ve looked at the proper basicality of belief in God’s existence. We’ve looked at the contingency argument, the cosmological argument, the fine-tuning argument, the moral argument, and the ontological argument for God’s existence. These give us not simply the existence of a vaguely characterized being that could be the Flying Spaghetti Monster. These give us a being replete with certain specific attributes like moral perfection, omnipotence, omniscience, metaphysical necessity, and so forth.


Student: Do you know of any prominent non-theist philosophers of religion who accept that if God even possibly exists he must exist?

Dr. Craig: Yeah, I actually do. One of the most famous proponents of the ontological argument is the great (even legendary) mathematician Kurt Gödel who developed Gödel’s Theorem that you may have heard of. Gödel wrote an essay defending the ontological argument. There was a gentleman at the University of Birmingham in England when I was doing my doctoral thesis on the cosmological argument for God’s existence. One of my fellow grad students was doing his thesis on the ontological argument. He told me that there was a man in the department of philosophy at the University of Birmingham who believed the ontological argument proved God’s existence. I said, “Is he a theist then?” He said, “Well, not really. He just thinks that as an argument this is a sound argument. Whether it has any religious significance is immaterial to him.”

Student: So there might have been a psychological element in that case?

Dr. Craig: I couldn’t judge. But what I am saying is there have been philosophers who think that the argument is sound, but may not find religious significance in it. In fact, some people have said probably no one has ever come to believe in God through the ontological argument because it is so abstract. In fact, that is false because we get emails at ReasonableFaith.org – I’ve seen them – from people who say the ontological argument is what convinced them that God exists and has led them to become a theist. Even this argument as abstract and a priori as it may seem has been used by the Lord to being people to belief in him.[9]

Student: You said at the beginning of this section of your class – I think you repeated it earlier today – that argumentation is not necessary. I think you are talking that you can rely on the internal witness of the Holy Spirit. You can have this private, inaccessible-to-other-people experience and believe God exists on that basis – God is properly basic. What do you do with the response that says every time in Scripture where you have skeptical people who are asked to believe in God or accept the resurrection of Christ they are given some evidence or argumentation? One case is in the Old Testament, 1 Kings where Elijah demonstrated the power of God and the people were skeptical. It says at the beginning of his test the people were hesitating between two opinions. After he called down fire then they believed.

Dr. Craig: This is the contest with the prophets of Baal.

Student: That’s right. One other example is in Acts 17 at Mars Hill where Paul confronts people who are skeptics. At one point he says I give you proof that this unknown God you are searching for exists because he raised Jesus from the dead. So he gave them evidence. What is your response to that?

Dr. Craig: I draw a fundamental distinction between knowing Christianity to be true and showing Christianity to be true. We can know that Christianity is true in a properly basic way through the inner witness of the Holy Spirit. But when it comes to showing skeptics that our faith is true then we need to give apologetic arguments and evidence. So Elijah didn’t need the fire to come down from heaven to convince him that Yahweh is the true God. He was already convinced. He knew that God exists. But to show the people and demonstrate the imposture of the prophets of Baal he set up a contest where there would be argument and evidence. So it seems to me that this fundamental distinction is very helpful in showing the proper role for both of these. Both proper basicality but then also showing – the use of apologetics. That is the way I would understand it.


The arguments of natural theology for God’s existence are only in one sense one side of the coin because we also would need to consider what are the arguments on the other side against the existence of God. Maybe there are such powerful arguments for atheism that they simply outweigh these arguments for God’s existence and tip the weight of the scale in the other direction. What we want to now take up beginning next week will be arguments for atheism, and especially the problem of suffering and evil in the world. If there is an omnipotent, all-loving God then isn’t that incompatible or improbable with the truly horrendous and innocent suffering in the world? How are we going to deal with that problem? That will be the question that we will take up next time.[10]



[1] Alvin Plantinga, The Nature of Necessity, Clarendon Library of Logic and Philosophy (Oxford: Clarendon Press, 1974), p. 221.

[2] 5:10

[3] Michael Ruse, Darwinism Defended (London: Addison-Wesley, 1982), p. 275.

[4] 10:10

[5] 15:00

[6] 20:11

[7] 25:00

[8] 30:02

[9] 35:14

[10] Total Running Time: 39:45 (Copyright © 2016 William Lane Craig)