John Loftus on the Ontological Argument

For a summary of the ontological argument see here.

Loftus echoes David Hume’s criticism of the ontological argument: his distinction between “matters of fact” and “relations of ideas.” “The former has to do with abstract ideas that are deductively known. The latter has to do with that which actually exists, and so they can only be known inductively through experience. Hume argued that the question of the existence of God is about a matter of fact, which can only be known inductively through experience. Therefore the ontological argument incorrectly uses the relations of ideas to show the existence of a matter of fact, and the argument fails. Basically, no mathematical proof can ever show that something really exists.”[i]

The trouble with this distinction is that it rules out all deductive knowledge as being something other than a “matter of fact.” This means that we cannot know that the conclusions of mathematical theorems are facts, which draws into doubt a great deal of modern physics. Loftus doesn’t present any argument for this arbitrary bifurcation of knowledge. If abstract objects and universals were not really facts, we couldn’t truly know anything. Abstract objects and universals structure the world of experience and make it meaningful. The barrage of sense data would make no sense to us without universals. For example, without the universal “chair”, the wooden thing in front of me would make no sense. The distinction between abstract objects and matters of experience is a false dichotomy for this reason.  This distinction also seems to be self-defeating, because it is a distinction that is not found in the empirical world and so cannot be a matter of fact according to its own principle. Also, as I show in the article on the meaningfulness of religions language, there are many things we cannot know through experience, including the external world, other minds, objective morality, and truths of logic (like the scientific method). Like elsewhere in his book, Loftus just makes an assertion without backing it up. Like other atheists, Loftus just seems to assume empiricism while not realizing that empiricism cannot account for many of the self-evident beliefs almost every person has, and without confronting the objection that empiricism is self-defeating. The idea that there can be no such thing as deductive knowledge or deductive facts is irrational. It is a deductive fact that Socrates is mortal if it is a fact that he is a man. You may say that this appeals to experiential phenomena for a deductive conclusion. Perhaps Loftus means that the ontological argument is purely deductive in a way that the above argument is not. But, like any other argument, the ontological argument has empirical features. What we would consider to be the greatest being is informed empirically, even if it is not entirely an empirical concept. For example, our concepts of power and knowledge are informed by what we see around us. In fact, it is arguable that there is no such thing as knowledge or propositions that are completely devoid of empirical content, and the ontological argument is not devoid of empirical content. Loftus also makes the assumption that numbers and abstract objects are not things that really exist, which is a claim that a great many philosophers would dispute.

After a brief nod to the Kantian objection to the ontological argument ( which I address here), Loftus launches into an attack on the coherence of theism. If theism is incoherent then this would hurt the first premise of the ontological argument ( it is possible that a greatest conceivable being exists). To do this, Loftus makes an appeal to the Christian theologian Paul Helm who says that it seems unintelligible that any being should be spaceless and timeless. Loftus doesn’t quote Helm further or explain why Helm regards these properties as unintelligible. He doesn’t make an argument so there’s nothing to respond to. Loftus seems to claim the debate over open theism as some sort of proof of the incoherence of theism. He concludes, “I think this whole debate indicates God’s existence is unintelligible, regardless of what we might conceive him to be, and hence ontological arguments can’t even get off the ground.”[ii] Here as elsewhere, Loftus’s treatment is superficial and he doesn’t make a real argument for his claim. Disagreement between Christian theologians is not evidence of the incoherence of theism anymore than debate between scientists show that the scientific method is a sham.  Loftus’s claim that he thinks that the “whole debate indicates God’s existence is unintelligible” is just a claim without support. Also, it is difficult to respond to Loftus claiming that something that is spaceless and timeless is unintelligible, because he does not tell us what he means by it. Does he mean it is self-contradictory? Does he mean that it is simply difficult to understand? It isn’t clear. Spacelessness and timelessness are not self-contradictory concepts, and while it may be difficult to grasp all of the implications of such an existence, they are not inconceivable. Many philosophers, for example, believe in the existence of abstract concepts and universals, which are spaceless and timeless.

Loftus finishes off his treatment of the ontological argument with an appeal to religious diversity and differences in human intelligence and morality.[iii] Who is to say that the god that I can conceive of is really the greatest conceivable being? Someone who is better or more intelligent can probably think up a greater god. And who are we to say that Eastern conceptions of God are not better than Western conceptions of God? Just because someone who isn’t intelligent cannot come up with the correct concept doesn’t mean there is no correct concept. In the same way, just because someone cannot come up with the conclusion of a mathematical theorem or comes up with the wrong answer doesn’t mean that there isn’t a correct answer. This response seems to assume without argument that the properties of God in the ontological argument are not objective great-making properties. A being that is omnipotent is greater than a being that is less powerful. A being that is morally perfect is better than a being who is morally flawed. A being that is personal is better than a thing that is impersonal ( what is better: a rock or a human being? Personal or impersonal?). A greatest conceivable being must be the greatest with regard to every property. In terms of power, that being must be the most powerful. In terms of knowledge, that being must the most knowledgeable. In terms of morality, that being must be most moral. If that being is the greatest in terms of power, it must be all-powerful, if the greatest in terms of knowledge, omniscient, if the greatest in terms of morality, morally perfect. Perhaps you may respond: “who is to say that a being that is omnipotent is better than a being that is less powerful?” Power is simply the ability to do things. More power is the ability to do more things than someone with less power. It seems obvious that power is a good thing , but not when it is pursued at the exclusion of everything else. Power is not a moral good – one is not morally good or bad based on how much power one has, but it is still a valuable property. It is still a good. To deny that is to deny that it would be preferable to be able to prevent bad things and encourage good things for anyone. But even if a being that is most powerful is not the greatest, you would have made an argument for this conclusion. This means it would still not be subjective  or non-rational what we regard as the greatest conceivable being.  In other words, there would still be an objective fact of the matter with regard to what sort of being is the greatest conceivable being.

An impersonal force like Brahman ( the Eastern concept of the supreme being) cannot be the greatest conceivable being, because there would be other beings that surpass or that are greater than it in terms of specific properties. Given that it does not have a mind, it cannot know anything, which means that almost any being is greater than it in terms of knowledge. Also, human beings would be greater than this being in terms of personhood, because it is not a person.  But you might say that other things would surpass the Western God with respect, for example, to being impersonal. But impersonality is not something you can excel at. You cannot be “very” impersonal, because impersonality is just a lack of personhood. It is a negation. Impersonality is also not valuable. It is not a good to be impersonal. To deny that is to deny that there is any value in being a human being over, say, a rock.

Loftus alleges, quoting John Hick, that the logic of the ontological argument can also be used to say that a maximally evil being exists. He doesn’t say how. The ontological argument contends that a greatest being must exist because the greatest conceivable being is better if it existed in reality than if it did not, so it cannot be a greatest conceivable being if it does not exist. But why should a most evil being exist? Existence doesn’t make a being more evil.  Existence would mean that it can give full consequence to its evil, but it wouldn’t make the being itself more evil. One can more easily argue that death or non-existence is evil, which means that a most evil conceivable being must be nonexistent rather than existent. There is no reason why a most evil conceivable being should exist rather than not. Perhaps you would phrase it as worst conceivable being rather than the most evil conceivable being. But again, a being is worse if it does not exist in reality than if it does. Therefore, the existence of a worst possible being is impossible.


[i]John Loftus, Why I Became An Atheist, Kindle Edition (Amherst: Prometheus Books)

p. 82

[ii] P.83

[iii] p. 83 – 84

2 thoughts

  1. I think you misconstrue the differentiation made between “relations of ideas” and “matters of fact.” I believe the point is the former is about “relations of (abstract) ideas” and the latter about “matters of 9concrete existential) fact. Yes 2+2=4 is a “matter of fact.” The argument does not imply that it is not a fact. But it is a matter of fact about abstractions, not a matter of fact about concrete existences. So, the problem is that the ontological argument tries to use an argument concerning the relations of ideas to prove a matter of concrete existential fact, and that is reasonable to consider that problematic.


  2. Hi Tom. Thank you for commenting. You don’t really explain the construal you prefer apart from saying “I believe the point is the former is about “relations of (abstract) ideas” and the latter about “matters of 9concrete existential) fact.” This is a circular definition, because you have used the terms you mean to define in an effort to define them ( adding “abstract”, “concrete”, and “existential” does not really help.) I’m getting some sense of what you’re getting at, but it’s not clear. If you explain what you mean here more fully, I’ll be able to engage with it. Then you said that 2+2=4 is a matter of fact. However, the way Loftus defines “relations of ideas” and “matters of fact” twice rules out that mathematical truths could be “matters of fact”. It is ruled out once because he assigns things which are “deductively known” to the category “relations of ideas”. Since mathematical truths are deductively known, that would make them, along with certain truths of logic, relations of ideas and not matters of fact. Secondly, he says that “matters of fact” are things which “actually exist” and are “inductively known.” This again rules out mathematical truths as being matters of fact, because mathematical truths are deductively, not inductively, known.

    You say that 2+2 = 4 is a matter of fact about abstractions, not a matter of fact about concrete existence. This does not fit Loftus’s explanation of the distinction. This also seems to contradict the definition you’ve given, because you contrast “matters of fact” with relations of “abstract” ideas. But let’s run with it. Why can’t we say then that the premises of the ontological argument are “matters of fact” about abstractions rather than matters of fact about concrete existence? Also, we regularly use relations of ideas to establish matters of fact about the physical world, both through everyday rationality, but also through the role that mathematics plays in physics. So why can something similar not be the case with the ontological argument.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s