ABSTRACT
In 1078, Anselm of Canterbury finished his work Proslogion, which features his famous ontological argument for God’s existence. Most readers will find the notoriously controversial argument rather unconvincing, and it is these intuitions that motivate this paper, whose project is to disprove the argument by means of denying Anselm’s hidden premise that asserts different levels of existence. It will then interact with Kant’s, Plantinga’s, and Oppy’s responses to the argument, and discuss how Anselm’s argument can, at least, be reduced to a probabilistic one.
INTRODUCTION
Anselm of Canterbury (1033-1109) finished Proslogion in 1078, and it is perhaps his most well-known work. Beginning and ending with prayer, the pious text argues for every aspect of God that Anselm believed could be reached by reason alone, beginning with God’s existence. Anselm composed Proslogion not out of spiritual doubt, but rather out of strong faith. In his prayer in Chapter One, he writes:
I am not trying to scale your heights, Lord; my understanding is in no way equal to that. But I do long to understand your truth in some way, your truth which my heart believes and loves. For I do not seek to understand in order to believe; I believe in order to understand. For I also believe that “Unless I believe, I shall not understand.”[1]
Anselm is starting from Christian premises—that God exists and is good—and thus, his project is to further strengthen his faith, out of piety. Because of this, his ontological arguments, which I will outline next, ought not to be viewed exclusively as trying to convince those who do not already believe in God, those Anselm calls the ‘fools’. Alvin Plantinga calls Anselm’s argument, “remarkably unconvincing, if not downright irritating; it looks too much like a parlor puzzle or word magic.”[2] And yet, whether or not Anselm intended those outside the faith to read his writing, the atheistic reader should be at least a little insulted to be called a fool, and so the argument at least must be considered partially in terms of its ability to convince. But ultimately, a deductive logical argument is a deductive logical argument, regardless of its discoverer’s intentions, and Anselm’s ontological argument will be considered as such.
ANSELM’S FIRST ONTOLOGICAL ARGUMENT
Anselm presents two slightly different ontological arguments for God’s existence, though they are incredibly similar. I will present, and attempt to disprove, each argument one at a time, starting with the First. Anselm does not present his argument in a strict premises-and-conclusion format, but I will do my best to convert it into a more digestible and logical form, using his language.
(1.i) We understand God to be that than which no greater can be thought.
(1.ii) When a man hears “that than which nothing greater can be thought”, he understands what he hears.
(1.iii) If a man understands something, then it exists at least in his understanding.
(1.iv) Therefore, the man, even if he does not believe such a being exists, acknowledges that it at least exists in his understanding. (1.ii-iii)
(1.v) Existing in reality is better than existing only in the understanding (in the sense that to exist in reality would be a quality possessed by that than which no greater can be thought.)
(1.vi) If the man thinks of this being existing only in the understanding, but not in reality, then one could think of a being that existed in reality as well, which would be greater.
(1.vii) Therefore, that than which no greater can be thought, in the man’s understanding, is actually that than which a greater can be thought, or, not that than which no greater can be thought. (1.v-vi)
(1.viii) And (1.vii) is self-contradictory, since it states that the being both is and is not that than which no greater can be thought.
(1.ix) Therefore, that than which no greater can be thought must exist in the understanding, because if it did not, it could not be understood, which it can be. (1.i-viii)
(1.x) Therefore, if that than which no greater can be thought exists in the understanding, then it must also exist in reality, because it is better to exist in reality, than only in the understanding. (1.v, 1.ix)
DESTABILIZING THE FIRST
Anselm’s argument seems sound—his premises do not appear unreasonable, nor his conclusions invalid. However, I will argue that (1.iii) is actually a nonsensical statement. To say that something “exists in the understanding” seems like a reasonable inference from the idea that one understands something, but this jump actually carries a lot with it. There is essentially a hidden premise in his argument: that there are different levels of existence. Now this itself is not incredibly controversial, but it is something Anselm assumes the reader will agree with; this premise is further implied when he states that it is better to exist in reality, than to exist only in the understanding. Anselm sees existence as a spectrum, perhaps where the lowest level of existence is a thought, or a word in a book, and the highest is God. This is fairly intuitive—however, it is an assumption that I happen to disagree with.
I say that there is no difference between something existing “in the understanding”, and “existing in reality”, or simply that it is a nonsensical comparison to make. At first, this claim may sound ridiculous, so let me clarify exactly what I mean. Let’s say I have a thought of an apple. Is there a difference between this thought, and God? Yes, of course. And is God a greater type of thing than a thought? Absolutely. Is it true that God has always existed, and will always exist, whereas my thought has only just come into existence, and may go out of existence soon? Indubitably. However, is it also true that at this present moment, both God and my thought exist in reality, and in the same manner? I would argue yes.
Let me offer an example to help explain this claim. Anselm’s jump from “a man understands x” to “x exists in the man’s understanding” is essentially no different than jumping from “I have painted a portrait of a dog”, to “the dog exists in my painting.” Though some readers may also accept this as true, I imagine most will see how this seems unlikely. Of course, the painting represents the dog, and resembles it in many ways: shape, color, perhaps even size—anyone who saw the painting would exclaim “that is a dog!” And yet, it is still just a painting. I do not deny the human intentionality that was present when the painting was created, but it seems wrong to say that the dog is infused into the painting in any way.
The concept of identity might be useful here. It seems that the real animate dog that was referenced for the painting has some kind of personal identity; call it Fido. This Fido-identity seems present, of course, in Fido, and in his paws, tail, and fur that he leaves around the house. But I would argue that Fido’s identity is not present either in the painting, or in our thoughts. That is not to say that we do not have thoughts pertaining to or about Fido-identity, or that we do not see the painting, which inspires in our minds the true fact that it resembles Fido-identity. But the identity itself, whatever kind of thing that is, cannot be in anything but Fido himself.
This seems unshakably true, because if Fido-identity were in the painting, or in our minds, would it not be harmful to the dog if the painting were destroyed, or if everyone who thought about Fido were to die? The painting of Fido, the thought of Fido, and Fido himself, all exist in the same manner. They are three completely separate things, that do indeed resemble each other. In the same way, I would argue that insofar as everything that has ever been exists, at the moment that my thought of God exists, it exists in the same manner that God does, because 'is' has only one 'manner'. The difference between them lies in what they are, not in how they are. Once again, I do not mean to impiously suggest that God and my thought are equals in some way, which is how it may sound if the reader is used to language pertaining to the notion of different levels of existence.
Having shown that there are not different levels of existence, I will explain why this invalidates Anselm’s argument. As a reminder to the reader, Anselm is not dealing explicitly with our Christian conception of God in this argument, but rather, “that than which no greater can be thought”, a type of theology known as “perfect-being” theology. (1.v) states that it is better to exist in reality than to exist only in the understanding, but we now see that this statement may not make sense. What I will grant, is that that than which no greater can be thought (which I will now refer to as ‘ThatThan’) is a better type of thing than a thought, and I would even be willing to admit that the fact ThatThan has always existed, while the thought has not, makes it better. But this is not a different level of existence—it is only talking about existence with regard to time, and so in comparing the existence of a thought, which exists in the present moment in which the argument is being conceived, with ThatThan, which also exists in the same moment, there is no difference.
What this means, is that just as Fido-identity is not present in the painting, ThatThan-identity is not actually present in my thought of it. We cannot imagine a thought of ThatThan, and the real, tangible ThatThan, as being on different ends of a spectrum of existence; there is no actual connection between the two. My thought of ThatThan, of course, pertains to, mirrors, and describes ThatThan (if it is real), but it does not contain any ThatThan-identity, and at that moment, it exists in the same way that ThatThan does (once again, if, in fact, ThatThan is real). The idea that 'God is present in all things' may remain true; it is irrelevant to my point, since this is an argument about whether such a being exists in the first place.
We cannot make a claim about the real ThatThan’s existence using the thought in our minds which pertains to it, because they are unconnected. Thus, we must reject (1.x), because it requires we agree that there is a difference between existing in the understanding, and existing in reality.
ANSELM’S SECOND ONTOLOGICAL ARGUMENT
As mentioned, Anselm’s Second Argument for God’s existence appears, at first glance, very similar to his First. In logical form, it proceeds as follows:
(2.i) A man can understand the notion of a thing that cannot be thought not to exist, so it exists in his understanding.
(2.ii) Such a thing would be greater than that which can be thought not to exist.
(2.iii) Therefore, if that than which no greater can be thought could be thought not to exist, then that than which no greater can be thought would not be that than which no greater can be thought. (2.i-ii)
(2.iv) And (2.iii) is self-contradictory since it states that the being both is and is not that than which no greater can be thought.
(2.v) Therefore, that than which no greater can be thought cannot be thought not to exist. (2.iv)
DESTABILIZING THE SECOND
All these steps may appear sound, but they require some unpacking. Essentially what Anselm is doing here is showing that not only does ThatThan exist, but ThatThan must necessarily exist. That is to say, in every possible world, ThatThan exists, and cannot even be thought not to exist, and this is because necessary existence is the greatest possible form of existence. It may not be clear how this is an argument for God’s existence, but the proof lies in what it means to necessarily exist. For example, if I were to say, “Imagine a world where unicorns are real, and one their qualities is that they necessarily exist.” Assuming that in our world, unicorns do not, in fact, exist, this is itself not only a false statement, but a nonsensical one. One cannot even imagine a world where something that is not necessary in our world (but rather, contingent) is necessary in another world, because if something is necessary, it is so in all worlds, by definition. It is like imagining a square circle: it is inherently self-contradictory. And, in the same way, necessary things (for example, the reader may grant that every effect has a cause) cannot even be imagined to not be necessary.
Thus, when Anselm says that ThatThan cannot be thought not to exist, he is saying that it is a necessary being—and he is either correct, or being nonsensical. Assuming the latter, where might he go wrong in his Second Argument? It turns out, we can disprove it in a similar manner as the First. (2.i) asserts that a man can understand the notion of a thing that cannot be thought not to exist, and so it exists in his understanding. This premise is borrowing from one in the First Argument: namely, that ThatThan can be understood, and thus can exist in a man’s mind. But if indeed there is no necessary connection between the thought of something, and the thing itself, then we cannot safely assert (2.i), because a man can only understand that a thing possesses the quality of being necessary, if it actually does. Otherwise, it is nonsensical. Now we see that Anselm was hiding his conclusion in his premise: since one can only imagine necessity to be a thing’s quality if it actually has that quality in reality, and is otherwise nonsensical, then Anselm must assume from the outset that ThatThan possesses necessity, and thus exists, to be able to assert that a person is imagining it. If ThatThan is real, then it surely is a necessary being—but we have yet to prove it is real at all, because of the failure of the First Argument.
Thus, if ThatThan indeed exists, the Second Argument is sound. But it cannot prove God’s existence on its own because of its hidden circularity—it is begging the question.
KANT’S CRITIQUE OF PURE REASON
Immanuel Kant had similar feelings about Anselm’s argument. In his section, “On the Impossibility of an Ontological Proof of the Existence of God” in the Critique of Pure Reason, Kant argues that ontological proofs at large are not possible because they rely on the predication of existence; but what does this mean? The Stanford Encyclopedia of Philosophy defines an ontological argument as follows:
Ontological arguments are arguments, for the conclusion that God exists, from premises which are supposed to derive from some source other than observation of the world—e.g., from reason alone. In other words, ontological arguments are arguments from what are typically alleged to be none but analytic, a priori and necessary premises to the conclusion that God exists.[3]
Kant, however, argues:
You have already committed a contradiction, if, in offering the concept of a thing that you wanted to think merely as regards its possibility, you have already brought into this concept–no matter under what covert name–the concept of the thing’s existence. If this move is granted to you, then you have seemingly won your point; but in fact, you have said nothing, because you have committed a tautology…Being is obviously not a real predicate, i.e., it is not a concept of anything that can be added to the concept of the thing. It is merely the positing of a thing in itself or of certain determinations in themselves…hence, all effort and labor is lost if expended on the famous ontological proof of the existence of a supreme being from mirror concepts; and, I suppose, a human being could not from mirror ideas become richer in insights any more than a merchant could become richer in assets if he tried to improve his situation by adding a few zeros to his cash balance.[4]
Kant’s writing is famously difficult to understand, but in essence, he is arguing here that ‘being’ or ‘existence’ is not a legitimate predicate. What this means, is that claiming a thing exists does not add anything useful to its description, because just by talking about it, you are already asserting that it exists, even if it is still unknown whether it is a tangible thing in reality. This does not mean, of course, that when someone asks, “do unicorns exist?” the polite response would be, “that is a meaningless statement—you just said yourself they do.” That would be to misunderstand the intent behind the question. But Kant indeed sees existence as binary. Do unicorns exist? They must exist in some way, because you are talking about them. Are there actually physical horse-like horned beasts in forests? Probably not—but this is a different question for Kant.
Thus, Kant would surely agree that Anselm’s ontological argument is impossible because of the jump from “existing in the mind” which is a nonsensical statement, to “existing in reality” which is a useless and redundant statement. The ThatThan we understand in our brains is a thought; a concept. It fully exists. ThatThan in reality (if indeed it is) is a being, and it also fully exists. One cannot connect these two.
Further, specifically with regards to Anselm’s Second Argument, Kant writes:
To posit a triangle, and yet to reject its three angles, is self-contradictory; but there is no contradiction in rejecting the triangle together with its three angles. The same holds true of the concept of an absolutely necessary being. If its existence is rejected, we reject the thing itself with all its predicates; and no question of contradiction can then arise.[5]
As noted in the last section, defeating the Second Argument is as simple as recognizing that Anselm’s conclusion is contained in his premise, which is essentially what Kant is explaining here.
PLANTINGA’S GOD, FREEDOM, & EVIL
Plantinga also interacts with Anselm in his book, God, Freedom, & Evil. As noted, he does not like Anselm’s argument, but disagrees with Kant’s objections, and still believes an ontological argument is possible, though more successful if reframed. His version of the argument, called the “modal ontological argument” goes as follows, according to the Stanford Encyclopedia of Philosophy:
(M.i) Possibly, God exists.
(M.ii) Necessarily, if God exists, then it is necessary that God exists. (God is, by definition, a necessary being)
(M.iii) Therefore, it is necessary that God exists .(M.i-ii)[6]
Put simply: if ThatThan can exist, then it must exist, because it is definitionally necessary, not contingent.
Attempting to defeat this argument is beyond the scope of this paper, though I suspect it may share a similar frailty as discussed in the previous section. If Kant is right, then ThatThan’s existence is secretly entailed in (M.i-ii), if it is true that one is not saying anything extra when asserting that a thing exists, if one is already discussing qualities of the thing that may or may not exist. But again, this paper is aimed at defeating Anselm’s argument, and Plantinga is simply out of my league. My agreement with Kant is more for the purpose of defeating Anselm’s argument, and not necessarily to disprove all ontological arguments at large—though if Kant is right, then this follows.
OPPY’S RESPONSE TO ANSELM
Atheist philosopher Graham Oppy takes a similar approach to disproving Anselm. He says:
However, no fool worth his salt will allow that it is possible that there is a being than which no greater can be conceived, if nothing could properly be labelled ‘that than which no greater can be conceived’ that fails to be a necessary existent. That is, any fool worth his salt will now either deny that the expression ‘that than which no greater can be conceived’ is understood, or he will deny that ‘that than which no greater can be conceived’ refers to something that is necessarily existent, necessarily omnipotent, necessarily omniscient, necessarily perfectly good, and so forth.[7]
Essentially, Oppy claims Anselm is making the mistake of assuming that ThatThan is a coherent concept that can be understood, and that when one thinks they understand it, they aren’t just aware of the meaning of several words strung together. He asks:
Does one, for example, count as understanding the expression ‘the round square’ because one understands the expressions ‘round’ and ‘square’ and [recognizes] that the definite description ‘the round square’ is syntactically well-formed?[8]
But even if one grants that it is a coherent concept, it also has to be granted that the fool agrees that ThatThan would have all these necessary qualities.
However, in more alignment with my argument, Oppy correctly points out that Anselm is committed to at least two domains of existence. Hidden in his argument are these claims: (1) Some objects exist in reality, (2) some objects exist in the understanding, (3) some objects exist both in reality and in the understanding, (4) some objects exist only in the understanding, and not reality. Oppy seems to agree that these two domains of existence are unlikely, and wonders aloud why something existing in the understanding doesn’t also count as existing in reality. However, he argues that even if they are granted, the argument is still unstable. Unlike me, he does not explain exactly how their denial would break Anselm’s argument, but he does not need to, since they are crucial premises in the First Argument; their denial renders it unsound.
CONCLUSION
Anselm’s ontological arguments for God’s existence are some of the most controversial and heavily debated pieces of logical work in history, and this paper is surely barely scratching the surface into the conversation. However, I hope it will at least be granted that if it is even remotely possible that I am correct about the non-predicability of existence, and the denial of different of levels of existence, then Anselm’s argument is reduced to a probabilistic one, where the likelihood of it being true is, at most, the probability of the truth of his preconception that existence can be predicated, and has different levels.
A very special thank-you to Mr. David Gilbert: the most joyful, kind, and dependable teacher and mentor I’ve ever had, without whom none of my progress in this class, or the program at large, would have been possible.
[1] Anselm, Proslogion: With the Replies of Gaunilo and Anselm, trans. Thomas Williams (Indianapolis: Hackett Publishing Company, 1995), 6.
[2] Alvin Plantinga, God, Freedom, and Evil (Grand Rapids: William B. Eerdmans Publishing Co., 1977), 85.
[3] "Ontological Arguments," in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta and Uri Nodelman, Winter 2023 ed. (Stanford University,2023), https://plato.stanford.edu/entries/ontological-arguments/.
[4] Immanuel Kant, Critique of Pure Reason, abridged, trans. Werner S. Pluhar (Indianapolis: Hackett Publishing Company, 1999), 202.
[5] Critique of Pure Reason, 197-8.
[6] “Ontological Arguments”, SEP.
[7] Graham Oppy, “Anselm’s First Argument”, 4.
[8] Ibid.