A Primer for Understanding Ontological Arguments for the Existence of God

Here’s a very brief history of the Ontological Argument for the existence of God: Italian Anselm originated it around the 11th century. Italian Thomas Aquinas rejected it in the 13th century. Frenchman Rene Descartes resurrected it in the early 17th century. Prussian/German Immanuel Kant refuted it in the latter half of the 17th century. Then in the past several decades Americans Norman Malcolm, Charles Hartshorne, and Alvin Plantinga have all defended it. Criticisms and debate about it abound in almost every philosophy textbook.

It is generally agreed that the Ontological Argument never converted anyone (even though Bertrand Russell once thought it was correct, but later changed his mind). It is an amazing argument—a philosopher’s delight! This is the “most famous, the most mystifying, the most outrageous and irritating philosophical argument of all time.” “It remains as one of the most controversial arguments in all of philosophy.” Yet, “whenever I read the Ontological Argument, I have the same feeling that comes over me when I watch a really good magician. Nothing up this sleeve, nothing up the other sleeve; nothing in the hat; presto! A big fat rabbit. How can Anselm pull God out of an idea?” [Robert Paul Wolff in About Philosophy, (pp.284ff)].

Anselm’s Ontological Argument for the Existence of God:

1) On the assumption that that than which nothing greater can be conceived is only in a mind, something greater can be conceived, because
2) Something greater can be thought to exist in reality as well.
3) The assumption is therefore contradictory: either there is no such thing even in the intellect, or it exists also in reality;
4) But it does exist in the mind of the doubter;
5) Therefore that than which nothing greater can be conceived exists in reality as well as in the mind.

Anselm argued that even those who doubt the existence of God would have to have some understanding of what they were doubting: Namely, they would understand God to be a being than which nothing greater can be thought. Given that it is greater to exist outside the mind rather than just in the mind, a doubter who denied God's existence would be caught in a contradiction, because he or she would be saying that it is possible to think of something greater than a being than which nothing greater can be thought. Hence, God exists necessarily.

The basic Kantian criticism of Anselm's argument is that someone cannot infer the extra mental existence of anything by analyzing its definition. Yet defenders reply that Anselm is not defining God into existence. He’s asking whether we can reasonably suppose that something than which nothing greater can be conceived exists only in the intellect. Consider these statements: “No square circles exist,” and “an infinite set of prime numbers exists.” If we can move from concepts to statements about reality here with math, then why not with God?

However, if we asked an Easterner what he conceives to be the greatest conceivable being, his conception will start off being different than those of westerners from the get-go. I think Anselmian arguments, including those of Hartshorne, Plantinga and Malcolm's all begin with Occidental not Oriental conceptions of God, and their western conceptions of God are theirs by virtue of the prevalence of the Christian gospel in the west. If ontological argumentation is sound, then the eastern conceptions of God will entail that their God (or the One) also exists. Since these two conceptions of God produce two mutually exclusive conclusions about which kind of God exists, then the ontological argument itself does not lead us to believe in the Christian God alone.

John Hick says this about Plantinga’s formulation: “the reasoning looks suspiciously like an attempt to prove divine existence by definitional fiat. I believe that the suspicion is justified. Plantinga’s argument for a maximally excellent being, if valid, would also work for a maximally evil being.” [Which Hick offers in An Interpretation of Religion (Yale Univ. Press, 1992), pp. 78-79]. But there cannot be two omnipotent beings, one being good and one being evil, even though the ontological argument could prove that they both exist. Since the ontological argument can be used to prove that two mutually exclusive beings both exist, the reasoning itself is faulty.


Edward T. Babinski said...


"The Ontological Argument and the Sin of Hubris" by Toni Vogel Carey from Philosophy Now (Dec. 2005).


It's professor Carey's answer to the most argued-over argument for the existence of God.


And speaking of the existence of God, what about the existence of yawns? Huh?

Research Provokes Yawns

While the act of yawning is common to a wide range of creatures, it has long been supposed that contagious yawning is confined to human and great apes alone, since only they were believed to be able understand how others of their species feel. Recently published research by a University of Stirling team in the journal Biology Letters suggests otherwise. Stumptail macaque monkeys yawn too, but it had been thought that they did not have the empathy to ‘catch’ a yawn. Yet video footage of natural yawns by macaques contrasted with footage of control mouth movements showed that the yawns triggered a statistically significant increase in the behaviour. It is proposed that this contagious yawning, recognising that a peer feels tired, may date back to a common ancestor of the monkeys and apes more than twenty million years ago.

Both articles I cited are from the same website, Philosophy Now.

Cool site.

JustinOther said...

Perhaps I don't get the whole concept here, but what I hear is that if I can think of a being , entity or whatever, then it exists?

That, I'm sure, is oversimplification. However, if I am on the right track, then I can imagine an omnipotent flying spaghetti monster who controls this christian god to make humans think that god is the ultimate of beings.

I know, kind of convoluted thinking. I should brush up on my philosophy.

Steven Carr said...

I can conceive of a potato which weighs more than any other potato.

Which weighs more, any potato somebody has grown in the ground or the potato I have conceived of which weighs more than that potato?

It is a foolish question. The atrributes of things I have conceived are not real attributes.

I can conceive of a perfect being, but the perfectness of that conceived being does not exist and cannot be compared to the perfectness of things which exist outstide my conception.

Anonymous said...

The potato in your head would be some real potato, somewhere, somebody has actually grown. Of course, you wouldn't know where, or what it was, or how much bigger, or when it was grown. In short, all you'd know about it, by definition, is that it is bigger than any other potato. Whoop-dee-do. Sure, it exists, but you can't say anything about it. This arguement, applied to gods, would mean god is... something great. It doesn't give it any particular qualities. In fact, by ways mere mortals can't understand, my shoelace could be god, being greater is some measure than all other things.

Ebonmuse said...

"But there cannot be two omnipotent beings, one being good and one being evil, even though the ontological argument could prove that they both exist. Since the ontological argument can be used to prove that two mutually exclusive beings both exist, the reasoning itself is faulty."

An excellent point. I made the same argument in my essay "Unmoved Mover" (http://www.ebonmusings.org/atheism/unmovedmover.html), where I dubbed it the Manichaean Reductio.

Something else to point out is that Anselm's argument is really just a circular argument, though cleverly disguised. He smuggles in the crucial presupposition by including existence in the list of possible perfections and then defining God as a being with all perfections. But that can only be true if God actually exists! One cannot simply assume that to be the case and use it to prove itself. In essence, the argument is nothing more than, "By definition, God exists. Therefore, God exists."

UberKuh said...

The concept of God itself is internally contradictory no matter how you slice it, so using that concept to extend its property to cover whatever someone subjectively defines as "greatest" presupposes otherwise and is, therefore, useless in debate. The number of worlds you have is irrelevant and this is what Plantinga and others overlook or ignore.

Francois Tremblay said...

You made a grave omission... you forgot the atheist ontological argument. See :

James said...

The Ontological Argument is playing around with infinity and words that imply infinite concepts, such as "perfect" and "greatest". Infinity is not a point, it is a direction. It is not objective, and therefore, any concepts dealing with infinity (i.e. greatest, etc) are not objective, but conceptual. Thus, god (is the greatest) is conceptual only. With infinity, there is always something greater.

In addition, you can not, in fact, conceive of a perfect being, unless you pinpoint what characteristic said being is perfect at. A being with multiple perfect characteristics (i.e. perfectly just and perfectly forgiving) will run you right into the non-cog snag.

CalvinDude said...

Consider these statements: “No square circles exist,” and “an infinite set of prime numbers exists.” If we can move from concepts to statements about reality here with math, then why not with God?

Math brings up all kinds of different logical conundrums. And while this doesn't deal with the ontological argument, I think it's still too cool to not pass on.

An infinite list of numbers does not contain all numbers (in fact, it is missing an infinite number of numbers). How is this possible?

Consider all the numbers between 0 and 1. Let's just list a few for example:

a(1) = .000000....
a(2) = .175999....
a(3) = .374111....

list continues through infinity

Now let me highlight a diagonal through this infinite list:

a(1) = .000000....
a(2) = .175999....
a(3) = .374111....

The diagonal likewise extends through infinity (every single number on the infinite list), but we'll only deal with the first part in this example, which is 074...

Add 1 to each digit of the numbers of the diagonal, with the stipulation that 9 + 1 = 0 (this is modular math so you ignore the carried 1). This becomes your new number:

x = .185

Now, we know this:

x does not equal A(1) because the first digit in x is 1 and the first digit in A1 is 0.

x does not equal A(2) because the second digit in x is 8, and the second digit in A2 is 7.

x does not equal A(3) because the third digit in x is 5, and the third digit in A3 is 4


Therefore, x does not equal a(n) because the nth digit is one more in x than it is in a(n).

Thus, the number that we built by adding 1 to each number of the diagonal is not in the list of infinite numbers.

Now this would only give us a single number not in the infinite list of numbers. But remember that we could add 2 to each digit of the diagonal instead of 1, or we could add 3, or 4, etc. Also, we could create a pattern where we add 1 to the first digit, subtract 2 from the second, add -1 to the third, etc. Since the numbers are infinitely long, there are an infinite number of patterns that you could choose to create, thus resulting in an infinite number of numbers not included in the infinite list of numbers.

What's interesting about this is two-fold. First, an infinite list doesn't have an infinite number of numbers! That very fact should be causing your brains to melt already :-)

Secondly, though, WHICH numbers are not included in the infinite list of numbers depend completely upon the ORDER in which the numbers on the list are listed (a different sequence results in different diagonals). Thus, changing the order of the list (not even the content) will change the individual numbers that do not exist on that infinite list. No single number is, therefore, necessarily absent from the list, yet it is necessary that there are an infinite number of numbers not in the infinite list of numbers.

Needless to say, I would hesitate before saying that our logic, reasoning, and rationality are actually sufficient to describe anything, let alone everything, about reality.

Anonymous said...


"An infinite list of numbers does not contain all numbers (in fact, it is missing an infinite number of numbers). How is this possible?"

Did you mean to say "A FINITE list of numbers does not contain all numbers (in fact, it is missing an infinite number of numbers..."

exbeliever said...


You wrote: "Needless to say, I would hesitate before saying that our logic, reasoning, and rationality are actually sufficient to describe anything, let alone everything, about reality."

Are you suggesting that the logic your god imposed on the creation is not sufficient "to describe anything, let alone everything, about reality"?

Or are you saying that "our" logic is different than god's logic? And if we are using a logic different than god's is this not an atheistic logic?

JustinOther said...


That was a fun post, man. I'm still trying to wrap my brain around it, but I do get the point.

Even something as simple as 0.33333333333333333333333333333333333333333333333333333333333333333333333333333333333333 etc. to infinity never ends, therefore never reaches a point at which it can be accurately described. It can be desccribed conceptually as 0.333 going on to infinity, but it cannot actually be written...

For another example, If I say write the number 2 five times, you would get:

If I say write the number 2 an infinate number of times, you cannot do it. You would never finish writing it.

Bringing this back to the topic at hand, god cannot be perfect because perfection, like infinity, can never be reached.

Phewwww, I could use a drink now.


Steven Carr said...

Calvindude is correct that an infinite list of numbers can miss off an infinite list of numbers.

Perhaps he should call himself Cantordude

exbeliever said...

I guess I don't see the mind-numbing part of it all. True, numerical infinities are odd birds.

An infinite set of even numbers will have the same number of numerals as an infinte set of all integers.

A library with an infinite number of dictions will have just as many pages as it does books.

I still don't think calvindude's criticism of tautologies is that impressive.

Is he saying that square circles can exist within the language-game of Western logic?

A circle is a round by definition. If one finds something that is square, then that same thing cannot be defined as a circle according to our normal language conventions.

In other words, within our particular language-game, a square circle cannot exist.

If calvindude is merely imagining a different language game in which "square" and "circle" are defined in such a way that a square circle can exist, that doesn't seem to undermine logical statements made within the traditional, Western, logical convention.

exbeliever said...

What does any of this have to do with the Ontological Argument?

James said...

exbeliever, did you read my thoughts in my post above about infinity and the Ont-Arg?

Daniel said...

I wish I'd known this post existed before replying to the original ontological argument.

Derek said...

My version of the ontological argument:

(1) The concept of God is identical to the concept of the greatest conceivable being possible.
(2) The concept of the greatest conceivable being (henceforth GCB) contains the concept of necessarily existing.
Argument for (2): (along Anselmian lines)
(2a) If GCB failed to exist necessarily then it would be lacking a necessary prerequisite to be such a being, for surly a GCB who does not exist necessarily is not as great as a GCB who actually does, in which case the GCB that does not exist necessarily would not in fact be the GCB, therefore:
(3) To be GCB is to necessarily exist (entailed by (1) and (2)).
(4) To necessarily exist means nothing more that GCB must exist in all possible worlds.
(5) This world is a possible world.
(6) God exists (entailed be (4) and (5)).

I must concede that this argument is different than Anselm’s in an important respect, for it introduces the modal notion of ‘necessary’ as the type of existence had by the GCB in question. But we still haven’t left the notion of the GCB, so it’s not entirely distinct from Anselm’s either. And in introducing the notion of necessary (existence) we’re able to narrow the arguments scope, which in turn might reveal what St. Anselm was trying to get at as well as point out why Gaunilo's Island and Chris’ Teleportation Device- counterarguments (by why of analogy) have no effect.

Gaunilo’s Island argument (and Chris’ too) serve to counter St. Anselm by using the same argument in order to render a ridiculous conclusion. But the modal version above can easily block such a move, which we can begin to see by replacing the concept of God with the concept of ‘Teleportation Device’ in premise (1).

(1’) The concept of (the Teleportation Device) is identical to the concept of the greatest conceivable being possible.

Which is clearly false; the concept of a teleportation device could never be identified with the greatest conceivable being, for at least two reasons:

-(1’)a The greatest possible being is too big of a thing to be reduced to a teleportation device, despite what a wonderful thing it could be (if it existed), because the GCB must also be able to be conscious, have the capacity to love, create universes, etc.

-(1’)b a teleportation device is clearly a contingent thing (it’s existence is in no way necessary), which is evidenced by the fact that it doesn’t happen to exist, and therefore can never be a necessary being.

But perhaps I’m cheating. By way of the first premise I attempted to weed out Chris’ analogy, and to be fair I need to change some things. In order to carry out Chris’ (and Gaunilo’s) response appropriately, the argument should be reconstructed like so:

(1’’) The concept of the greatest conceivable Teleportation Device (henceforth GCTD) contains the concept of necessarily existing.
Argument for (1’’): (along Gauniloian lines)
(1’’)a If GCTD failed to exist necessarily then it would be lacking a necessary prerequisite to be such a being, for surly a GCTD that does not exist necessarily is not as great as a GCTD that actually does, in which case the GCTD that does not exist necessarily would not in fact be the GCTD, therefore:
(3’) To be GCTD is to necessarily exist (entailed by (1)).
(4’) To necessarily exist means nothing more that GCTD must exist in all possible worlds.
(5’) This world is a possible world.
(6’) GTCD exists (entailed be (4) and (5)).

At first glance this argument seems to fly just as well as the original, but perhaps we should take a closer look. It’s obvious the subject of this argument (the GCTB argument) has changed from the GCB one, but that’s to be expected since that’s how you argue analogically. But the thing to notice is that because the subject of the argument has changed, so has the sense in which the modal term ‘necessity’ is read. Compare, for instance,

(3) To be GCB is to necessarily exist.
(3’) To be GCTD is to necessarily exist.

Despite the identity in syntax between these premises, the way in which these two premises predicate ‘necessarily exist’ to their subjects is different. We can’t read (3’) in the same way we read (3) because the way in which we ascribe the necessity to the greatest conceivable teleportation device is dissimilar to how we ascribe necessary existence the greatest conceivable being, and the way to show this to is to go back to argument that entailed this premise, which would be (2)a (for GCB) and (1’’)a (for GCTD).

(2) said that for the GCB to be such it must necessarily exist, because it wouldn’t be (as) great if it didn’t exist. (1’’)a said that GCTD must necessarily exist for the same reason (that it wouldn’t be as great if it didn’t exist). But the reasons why we say that both need to necessarily exist is different: in the case of GCB the reason why it must exist necessarily is because it’s apparent to our reason that such a being would posses such a property necessarily; that is, we perceive the necessary existence of the GCB to be an intrinsic property of that thing. But when we said that the GCTD must exist necessarily to be such, the reason why we would say that is because it would be greater to us that it actually existed, for if it didn’t, we couldn’t use it, and hence it wouldn’t be so great. But this means that when it comes to TDs, the greatness they would have by existing or wouldn’t have by not existing is external to the thing itself. And this simply isn’t the case when you think of the greatest conceivable being, for in conceiving such a thing we realize that to be such a thing, we think that it must not be great merely because it would make a difference to us, but also because it seems that such thing must exist necessarily to be such a thing in the first place.

Simply put, necessary existence is an essential intrinsic quality a GCB must have to be such a thing, but on the other hand the greatness of a TD might only imply existence when it would be great for us, in which case in terms of thinking about the greatest conceivable TD, it’s greatness would be a matter of an external property of the thing in question.

Some clarifying objections:

Kant’s rejection of the ontological argument, ironically, helps to point out what I’m trying to get at. Basically Kant’s problem is that existence isn’t a predicate (or if it is, it’s never an essential property of a thing). For instance, suppose two people are arguing over whether unicorns existed. Assuming the debate is a semiformal one, the members of the debate decide to define their terms. The one arguing against the existence of unicorns says, “so let me get this straight; when you say that there are such things as ‘unicorns’, you mean an animal that more or less resembles a horse and it has a horn on its forehead?’ The one arguing pro says, “Why yes, the qualities which a unicorn has are as you described, but you mustn’t forget that unicorns exist, that is, they have the essential quality of existing.’ “I thought we’re arguing over whether unicorns exist or not, so you mustn’t assume they exist until you’ve proved it to us!”, responds the opponent. “No no no,” says the man arguing for the existence of unicorns, “The unicorns I’m talking about do exist; they have such a quality! So you’re not thinking of the same thing I am.” “Well,” says the opponent, “if you can’t separate the idea of a unicorn from it’s existence, then it seem you’ve already won the debate, in which case it seems there is nothing to debate about.” “Hmmm, I guess we’re at an impasse.”

Kant says that if existence is a property, it must be assumed to be irrelevant to the object of a thing or else you would never be able to prove to someone who doesn’t already believe in the existence of a certain thing in the first place. This is so, says Kant, because whether a thing exists or not it makes no difference because existence is never an essential attribute of an object; hence why we’re able to debate whether certain things exist.

I think perhaps Kant is right, that one can rightfully separate the existence of certain objects without changing or altering the object in question. This is obviously the case when it comes to things like Islands and teleportation devices. The difference between the greatest possible island that does exist and greatest island that does not is nothing. They’re identical intrinsically speaking since they have all qualities in common. The only way existence would become relevant to perfect islands is when we equivocate between the senses of greatness. Only in the observer-relative sense of ‘greatness’ does existence become relevant. This is to say, again, that a perfect island can be intrinsically great without needing to exist.

But is Kant’s observation true of all objects, whatever? What if we introduce the notion of a type existence such as ‘necessary existence’? Think of GC islands or teleportation devices: examining them with our mind’s eye we see that if they necessarily existed or not it would make no difference; even worse both objects (all created things, actually) could never possess such a property, even if they did exist, because all such objects are contingent things. It’s necessarily false that a contingent thing posseses the property of necessary existence.

But what of the very idea of greatest conceivable being, the very idea of God? If we look to see with our mind’s eye at such a being, wouldn’t we see that to be such a being it couldn’t help to exist necessarily? Isn’t it the case that unlike all other objects, if we tried to remove existence from God’s essence we would cease to be contemplating the same thing? Isn’t the difference between a being who necessarily existed and one who needn’t not to exist to be such being an infinite difference? Doesn’t it seem to be the case that the notion of the GCB includes within itself necessary existence, such that if we took it away it wouldn’t be what it is?

Identity of essence and existence in God.

God’s essence cannot be other than His existence.
In any being whose essence is distinct from its
existence, what it is must be distinct form that
whereby it is. For in virtue of its essence we say
what it is. This is why a definition that signifies
an essence manifests what a thing is. In God,
however, there is no distinction between what He
is and whereby He is… therefore God’s essence is
nothing else than His existence.

St. Thomas Aquinas.
Aquinas’s Shorter Summa, 11