Ontological Arguments for the existence of God are an odd grouping. Not much holds them together except for their most common rebuttal: that they attempt to define a God into existence. This is not my favourite rebuttal to the arguments, although it is true: ontological arguments for God try to sneak the idea of existing into the definition of a God. They do this either by shifting between the idea of a God and an actual God, or employing a many worlds model of possibility and then condensing those worlds into this world. That is their basic mechanism.
For our purposes here, we’ll focus mainly on the original version of the Ontological Argument by Anselm (but that doesn’t mean we will neglect talking about other versions). This version asks us to imagine the greatest being we can. It has extra steps later down the line, but I want to start with this, because I already have two issues.
What can you imagine? Is your imagination actually limited to reality or practical considerations? I want to walk through a series of issues here, before steel-manning the premise to make it stronger.
A naive version of the Greatest Imaginable being might be Omnipotent (infinitely powerful). And, if you didn’t know any better, you might assume you are imagining such a thing ― oblivious to the fact that there is a problem here. People who enjoy conservations like these may be aware of the omnipotence paradox: can God create a boulder so heavy that not even God can lift it? The answer cannot be no without violating God’s omnipotence, but neither can it be yes, for that would create a boulder God cannot lift.
There is also a paradox around omniscience (being all knowing) and omnipotence. An omniscient God knows how the future will unfold. An omnipotent God can change how the future unfolds. If God, today, knows that the sun will rise in the morning, It cannot then act to stop it ― for if God does stop the sun from rising then It did not know it would rise.
These are not my objections. There are ways around the omni-paradoxes which we will come to. The objection is the limited nature of the imagination to conceive of some relevant being here. And the issues with human imagination go further.
Spiderman is an imaginable but impossible being. The webbing that Spiderman produces comes from literally nowhere: Peter Parker (or Miles Morales) simply doesn’t eat enough to produce that. Similarly, Superman flies without propulsion or even (seemingly) the expenditure of energy. Imagined. Not possible.
What you have to do, then, to Steel-Man this argument (make it as strong as you can) is limit human imagination only to what is possible and assume a detailed analysis of that imagination such that it makes coherent sense at all levels. You need not the Greatest Being imaginable at all, but the Great Being computable. There needs to be some calculus that maximises levels of omnipotence and omniscience without entering into a paradox (that works with all the other omni-properties as well ― which opens up a conversation about which properties are relevant: is mass relevant? Why?).
The various omni-properties now need to be measurable in a comparative way. It needs to make sense that if you take 5 points from power, leading to a 6 point gain in knowledge, this is a real gain and Greater Being (see figure 1). But, what happens if adding 1 point to power takes one from knowledge, leading to the same total score (see figure 2)? Now there are two maximal beings. And, until someone actually does this we have no idea how impotent or ignorant a Being we have computed. It might be blasphemy at this point.
It is also worth pointing out that we are still talking, at this point, about a rendering on a computer (maybe even a supercomputer). But, it’s not projected into reality by this. It’s just a constructed idea in a computer.
Some versions of the Ontological Argument do away with this issue of what can be imagined by instead talking of what is possible. And that also opens up an interesting set of issues for anyone trying to use that argument. What is possible? Or, more importantly, what can we rationally claim is possible?
‘Possible’ does not mean that which has not been shown to be impossible. If anyone were to try that argument, we could simply refuse to accept it and instead offer the definition of ‘impossible’ as that which has not yet been shown to be possible. These two definitions are inadequate, but do hint at the proper next step: assigning a burden of proof (is there an Atheist Cliche Bingo doing the rounds?). From your point of view, as the reader, I could be any gender, age or ethnicity. But, it is also possible that I don’t exist and this entire blog is the result of a laptop being left on in a house with a wifi connection and a lot of cats running over the keyboard.
To a classical determinist, ‘possibility’ is only a statement about uncertainty, because the universe as it actually is and as it actually unfolds is all that is possible. Everything else, no matter how compliant with imaginations or physics, is impossible because the details (not the laws) of the universe won’t unfold that way. A tossed coin comes up heads not because it was 50:50 and had to come up one way, but because it was always going to be heads and could not be tails.
There’s another idea of possibility within Model Dependent Realism. When you take the model that best describes the system you’re interested in and run the numbers, you may end with probability distributions as your answer. Probability Theory, for example, states that the 50:50 chance of getting heads is not a statement about our knowledge of the universe, but an actually true statement about the system. Some weather forecasts work like this: they take the weather system as it currently is and look at historical events to see how systems like that evolve. It then counts up all the end results and groups them to give a probability. (Under that assessment, it is difficult to see how a claim like 70% chance of rain isn’t actually a statement about the modelling methods and the equipment, instead of a claim that a weather front can will itself against its influences, but that’s a different discussion.)
Suffice it to say, then, at this point, that possibility is also a slippery topic to place in the argument without a robust meaning. What’s more, having seen that premise delivered before, it can look like a premise that an interlocutor simply has to accept because the alternative looks like saying God is impossible. As discussed, this is not the case ― it is the claim that possibility has not been established. But, slipping a rhetorical trick passed an audience is often just as much a part of the game as honesty and integrity.
I did say I wanted to focus primarily on Anselm’s version (‘Greatest Conceivable’), and so we’ll return our focus to that and the bait and switch that happens. It’s transparently obvious once the language is pointed out. If I ask you to imagine a God, you are not actually creating a God in your mind. You are creating an idea of a God. To say that there you have a God in your mind sounds natural because it is the normal shorthand for when you are thinking about a thing. But, to be precise, when you are thinking about a thing, what you have in your mind is an idea of that thing.
We are not talking about narrative versions of reality, nor the idea that we all hallucinate normal objects from what are actually precisely arranged quantum particles. It is simply the recognition that reality exists out there, and our best model of that reality exists in our heads (or computers). Perhaps more powerfully, there is a one-way causal relationship between objects in reality and the ideas we have about them: the object can influence how we think about it, how we think about an object doesn’t change what it is. Imagine a coin and discovering both faces were heads, that changes how you think about the coin; in reverse, take out a coin and imagine both faces are tails. Did it change anything?
A thing that exists in the mind simply is not the same object as exists in reality, even if one appears to represent the other. God does not exist both in your mind and outside it. An idea in your head is, at best, a simulacrum and, at worst, a fantasy.
The idea of a being doesn’t even have to accurately track the being. As I am not a vet, I will have ideas about dogs that are wrong and dogs themselves are more complex than my ideas of their biology, behaviour and psychology.
Therefore, the step in Anselm’s Ontological Argument that tries to create a shared set of attributes between the Being and the idea of the Being is the wrong step to take.
In the introduction to this, I said that forcing existence into the definition of a Being isn’t my favourite rebuttal to the argument. But, there is an argument from absurdity from The Massianic Manic that is worth sharing. He creates an idea: the realicorn. A realicorn is just like a unicorn, except it exists. And so, if you really understand what a realicorn is, you must believe in it. If you think you are imagining a proper understanding of it, but don’t believe in it, then you are only imagining a unicorn, not a realicorn. Apologists have tried to get around objections like this by saying that a realicorn (or largest island, or any superlative object) isn’t necessary or is contingent, unlike God. If that’s the line of argument, they can abandon the Ontological Argument altogether and rest on that bald semi-coherent assertion of being a Necessary Being ― if that were a credible claim, the Ontological Argument for God would be redundant.
The other version of the Ontological Argument worth looking at is Platinga’s Modal Ontological Argument for the existence of God. This modal argument takes a ‘many-worlds’ definition of what is possible. That is, to be possible means to exist in some imaginable world. Platinga takes advantage of that conceptualisation by distinguishing between two different types of ‘maximal’ beings: maximally excellent (to be the Greatest Computable Being in a world) and maximally Great (to be the Greatest Computable Being in all worlds). Assert that a Maximally Great Being is possible; point out that being possible means existing in some world; point back to the definition that says ‘all worlds’; remind everyone that our world is a part of all worlds. Therefore God.
We can throw the realicorn at this argument. But, why be done there?
I have used the word ‘computable’ in my version of the argument to make it stronger. (If you want to use ‘conceivable’, revisit the start of this post for all the issues therein.) So, we’re looking for a being that maximises its scores across the relevant fields of Greatness (power, knowledge etc.) But, would that calculation behave the same in all possible worlds? If it is possible that a Being could ― for example ― be both more powerful and knowledgeable in a given world, because it is a simpler world, then the definition of a Maximally Excellent being isn’t consistent across all worlds. And so, a Maximally Great Being would not be possible, because the Maximally Excellent Beings are necessarily different in at least one possible world.
We can also flip the argument. If we instead plug in the premise that it is possible that a Maximally Great being does not exist, then that would wholly block a Maximally Great Being in all worlds because Platinga is playing an all-or-nothing game.