# Ontological arguments

**URL:** <https://boards.straightdope.com/t/ontological-arguments/231958>\
**Category:** Great Debates\
**Created:** [March 1, 2004, 6:59pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958 "2004-03-01T18:59:06Z")\
**Posts on this page:** 20\
**Page:** 2

<div class="post-metadata">

**Author:** ![erislover](https://avatars.discourse-cdn.com/v4/letter/e/71e660/32.png) [@erislover](https://boards.straightdope.com/u/erislover)\
**Post date:** [March 3, 2004, 3:27pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/21 "2004-03-03T15:27:37Z")

</div>

But isn’t that what I just indicated by suggesting that the logic itself obtains in all possible worlds?

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 3, 2004, 3:54pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/22 "2004-03-03T15:54:54Z")

</div>

Yes. I think that’s a fair way of putting it, Eris.

For the OP’s information, not all ontological proofs are necessary valid. Some are even circulus in probando. (Those are valid but trivial.) But some are quite sophisticated, and that includes the modern modal arguments. They have been developed by some of the greatest thinkers of the 20th century from Godel to Plantinga.

I agree with those who say that the argument, even if valid and nontrivial, is not _convincing_. But I also agree with Plantinga who said that what a proper valid argument ought to convince _anyone_ of is that those who hold it are reasonable people. Sentient Meat, for example, respects my views and I respect his. It is not reasonable that people who believe in God be treated as intellectual savages who hold to nothing but superstition.

---

<div class="post-metadata">

**Author:** ![Ludovic](https://avatars.discourse-cdn.com/v4/letter/l/7ab992/32.png) [@Ludovic](https://boards.straightdope.com/u/Ludovic)\
**Post date:** [March 3, 2004, 3:57pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/23 "2004-03-03T15:57:00Z")

</div>

And if I mean exactly what I said? What then?

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 3, 2004, 4:17pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/24 "2004-03-03T16:17:00Z")

</div>

Well, then I guess you’ll have to explain what you mean. There’s no formal meaning for the term “communication” that I know of in any system of K.

---

<div class="post-metadata">

**Author:** ![Ludovic](https://avatars.discourse-cdn.com/v4/letter/l/7ab992/32.png) [@Ludovic](https://boards.straightdope.com/u/Ludovic)\
**Post date:** [March 4, 2004, 3:03pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/25 "2004-03-04T15:03:17Z")

</div>

> [@Libertarian](#):
>
> Well, then I guess you’ll have to explain what you mean. There’s no formal meaning for the term “communication” that I know of in any system of K.

While accessibility relationships are another problem I have with the proof (namely,that they are not defined, the closest I can fathom would be if all universes are mutually accessible if they follow the rules of logic, in which case I have yet _another_ issue with it…) my beef with “communication” is in the assumption of “greatness”.

Now, most proofs along these veins assign qualities to greatness other than mere size. But in the proof you use, “greatness” is determined by how many universes an entity appears in.

I, however, would not assign an entity this “greatness” even if it existed in all possible worlds. That is because the entities, while possessing the exact same properties, are not part of the same entity. For instance if g exists, and there are X worlds, there are X g’s, all possessing the same properties, but are not the same entity. The only way they _could_ be the same entity is if information could be exchanged between worlds, thus my reference to communication.

To use another example, **a** human is not greater, even in breadth, than **an** orangutan simply because there are more humans.

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 4, 2004, 4:00pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/26 "2004-03-04T16:00:34Z")

</div>

> [@](#):
>
> While accessibility relationships are another problem I have with the proof (namely,that they are not defined…

Since it uses the S5 axioim, the accessibility relation is Euclidean: (wRv&wRu) -\> vRu. That is, if v and u are modes onto w, then v is accessible to u and u is accessible to v.

> [@](#):
>
> But in the proof you use, “greatness” is determined by how many universes an entity appears in.

Not exactly. A modality is a view, or frame of reference, onto a world. Greatness is not determined by cardinality but by perfection — that is to say, there are no frames of reference not accessible to the entity.

---

<div class="post-metadata">

**Author:** ![Ludovic](https://avatars.discourse-cdn.com/v4/letter/l/7ab992/32.png) [@Ludovic](https://boards.straightdope.com/u/Ludovic)\
**Post date:** [March 4, 2004, 4:26pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/27 "2004-03-04T16:26:02Z")

</div>

> [@Libertarian](#):
>
> Since it uses the S5 axioim, the accessibility relation is Euclidean: (wRv&wRu) -\> vRu. That is, if v and u are modes onto w, then v is accessible to u and u is accessible to v.

Perhaps you misunderstand. Yes, S5 uses that sort of accessibility relation, perhaps I used misleading terminology. I meant that the subconditions under which accessibility exists in the first place are not well defined.

I see no reason to limit the accessibility function we are using to the definition of “g”. There may be other worlds which I believe are accessible to ours which do not include “g”. Of course, if I assume that they are accessible, this means that “g” cannot exist per the assumptions of this MOPoG.

> [@](#):
>
> Not exactly. A modality is a view, or frame of reference, onto a world. Greatness is not determined by cardinality but by perfection — that is to say, there are no frames of reference not accessible to the entity.

In what way is this perfect, considering that the set of frames of reference is arbitrary?

---

<div class="post-metadata">

**Author:** ![erislover](https://avatars.discourse-cdn.com/v4/letter/e/71e660/32.png) [@erislover](https://boards.straightdope.com/u/erislover)\
**Post date:** [March 4, 2004, 4:36pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/28 "2004-03-04T16:36:28Z")

</div>

> [@Ludovic](#):
>
> For instance if g exists, and there are X worlds, there are X g’s, all possessing the same properties, but are not the same entity. The only way they _could_ be the same entity is if information could be exchanged between worlds, thus my reference to communication.

Are the axioms of modal and predicate logic the same in all worlds, or individual but the same? Is there communication between worlds if the fundamental theorems of the calculus are derived in them?

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 4, 2004, 5:34pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/29 "2004-03-04T17:34:52Z")

</div>

> [@](#):
>
> Perhaps you misunderstand. Yes, S5 uses that sort of accessibility relation, perhaps I used misleading terminology. I meant that the subconditions under which accessibility exists in the first place are not well defined.

I’m not sure what you mean by “subcondition”, but I’ll take a stab at answering. The Euclidean relation posits that \<\>A -\> \<\>A. In other words, if it is possible for A to be true in the actual world, then in every world it must be possible for A to be true. That is a postulate that most materialists would accept readily since it forms the basis of their worldview.

> [@](#):
>
> I see no reason to limit the accessibility function we are using to the definition of “g”. There may be other worlds which I believe are accessible to ours which do not include “g”. Of course, if I assume that they are accessible, this means that “g” cannot exist per the assumptions of this MOPoG.

Well, that’s the point of the proof: to determine in what world(s), if any, “g” is accessible. It is not known until the penultimate inference that “g” is accessible in all worlds. And it is not known until the final inference that “g” is accessible in the actual world. We begin only with what we know — that “g” possibly is accessible in some world.

> [@](#):
>
> In what way is this perfect, considering that the set of frames of reference is arbitrary?

Perfect in the sense of complete. Noncontingent. That is, all frames — no matter how many and no matter any other consideration — are accessible.

---

<div class="post-metadata">

**Author:** ![hansel](https://avatars.discourse-cdn.com/v4/letter/h/6bbea6/32.png) [@hansel](https://boards.straightdope.com/u/hansel)\
**Post date:** [March 5, 2004, 4:59am UTC](https://boards.straightdope.com/t/ontological-arguments/231958/30 "2004-03-05T04:59:36Z")

</div>

> [@ultrafilter](#):
>
> The law of the excluded middle is very well-established. It’s the tautology _A_ v ~_A_, where _A_ is any wff of propositional logic.

Slight nitpick: the law of the excluded middle is well established in _bivalent_ logic calculi, which are dominant but not the only logical calculi around (much like Euclidian geometry isn’t the only plausible geometry). Intuitionistic logic comes to mind as a logical calculus that specifically denies the law of the excluded middle.

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 5, 2004, 1:01pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/31 "2004-03-05T13:01:57Z")

</div>

In paraconsistent logic, not only does excluded middle fail, but so does noncontradiction.

---

<div class="post-metadata">

**Author:** ![Faldage](https://avatars.discourse-cdn.com/v4/letter/f/f4b2a3/32.png) [@Faldage](https://boards.straightdope.com/u/Faldage)\
**Post date:** [March 5, 2004, 3:20pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/32 "2004-03-05T15:20:58Z")

</div>

Where all these arguments fall apart is in the unstated premise that the Universe is somehow compelled or constrained by our ability or lack of ability to comprehend. This is, of course, utter nonsense. But y’all keep it up. Y’all’re funnier than a yard full of headless chickens.

---

<div class="post-metadata">

**Author:** ![Ludovic](https://avatars.discourse-cdn.com/v4/letter/l/7ab992/32.png) [@Ludovic](https://boards.straightdope.com/u/Ludovic)\
**Post date:** [March 5, 2004, 3:22pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/33 "2004-03-05T15:22:57Z")

</div>

> [@](#):
>
> Perfect in the sense of complete. Noncontingent. That is, all frames — no matter how many and no matter any other consideration — are accessible

except for _im_possible worlds. Impossible worlds are those without NE (the only accessibility parameter I can find in the proof.)

So NE is contingent upon being within the S5 Euclidean “circle”, as it were, of worlds with NE. There are plenty of imaginable worlds that are not “possible” according to the parameters of the proof. For instance, as someone has said, what if the Prime World does not follow the rules of bivalent logic?

---

<div class="post-metadata">

**Author:** ![SentientMeat](https://avatars.discourse-cdn.com/v4/letter/s/43a26b/32.png) [@SentientMeat](https://boards.straightdope.com/u/SentientMeat)\
**Post date:** [March 5, 2004, 3:25pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/34 "2004-03-05T15:25:47Z")

</div>

> [@](#):
>
> Where all these arguments fall apart is in the unstated premise that the Universe is somehow compelled or constrained by our ability or lack of ability to comprehend.

We are not _talking about_ the universe. The universe is known epistemically. It is ontologically irrelevant.

Please, do pay attention if you wish to join in.

---

<div class="post-metadata">

**Author:** ![Faldage](https://avatars.discourse-cdn.com/v4/letter/f/f4b2a3/32.png) [@Faldage](https://boards.straightdope.com/u/Faldage)\
**Post date:** [March 5, 2004, 3:40pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/35 "2004-03-05T15:40:08Z")

</div>

> [@SentientMeat](#):
>
> We are not _talking about_ the universe. The universe is known epistemically. It is ontologically irrelevant.
> 
> Please, do pay attention if you wish to join in.

OK. For _Universe_ substitute _God_. I didn’t want to use that word because it carries so much extraneous garbage with it, which is another probelm with the whole argument, it doesn’t really prove anything, it merely attaches some sort of sophistic pseudorelevance to a poorly defined word. If the ontological argument does prove anything it is not any God of the book religions, it is the Buddhist concept of Godhead.

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 5, 2004, 7:41pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/36 "2004-03-05T19:41:29Z")

</div>

> [@Ludovic](#):
>
> except for _im_possible worlds. Impossible worlds are those without NE (the only accessibility parameter I can find in the proof.)
> 
> So NE is contingent upon being within the S5 Euclidean “circle”, as it were, of worlds with NE. There are plenty of imaginable worlds that are not “possible” according to the parameters of the proof. For instance, as someone has said, what if the Prime World does not follow the rules of bivalent logic?

Impossible worlds don’t exist. And you have the contingency backwards. S5 is contingent on there being NE. Saying that NE is contingent on S5 is like saying that gravity didn’t work until Newton wrote his equation. Finally, the bivalent system of S5 is the appropriate system for determining the modal status of “g” for the reasons already stated. If a triangle is proved to be a right triangle by Euclidean geometry, it doesn’t matter that it cannot be proved to be a right triangle by Peano arithmetic.

---

<div class="post-metadata">

**Author:** ![Anduril](https://avatars.discourse-cdn.com/v4/letter/a/c0e974/32.png) [@Anduril](https://boards.straightdope.com/u/Anduril)\
**Post date:** [March 5, 2004, 9:13pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/37 "2004-03-05T21:13:53Z")

</div>

> [@Pochacco](#):
>
> Lib’s argument is flawed. As I pointed out to him in a previous thread, one can use a virtually identical proof to demonstrate that God _doesn’t_ exist. Here it is, for the curious:
> 
> **1. G -\> []G** – “If God exists he exists in all possible worlds.”
> 
> This is Libertarian’s definition of God.
> 
> **2. ~[]G** – “It’s possible that God doesn’t exist.”
> 
> A reasonable premise, since we have no evidence for God’s existence. And it’s easy to imagine a possible universe that contains nothing, not even God.
> 
> (This premise is slightly different from Libertarian’s. He chooses “It’s possible that God does exist” as his second step.)
> 
> The rest of the proof mirrors Libertarian’s proof exactly:
> 
> **3. []~G -\> ~G  
> 4. []G v ~[]G  
> 5. ~[]G -\> []~[]G  
> 6. []G v []~[]G  
> 7. []~[]G -\> []~G  
> 8. []G v []~G  
> 9. []~G  
> 10. ~G**
> 
> Look, I’ve proven that God doesn’t exist!
> 
> Unless you can demonstrate why Libertarian’s proof is valid and mine is not, you’re forced to conclude that modal logic is an inadequate tool for the task at hand … .

This counter-example is quite a convincing proof that there is a flaw in the method (even for one with no background in modal logic). You only have to determine whether or not the second premise is reasonable and you are on your way to debunking this proof. So far, no one has shown an acceptable reason to reject the second premise short of begging the question.

---

<div class="post-metadata">

**Author:** ![Liberal](https://avatars.discourse-cdn.com/v4/letter/l/848f3c/32.png) [@Liberal](https://boards.straightdope.com/u/Liberal)\
**Post date:** [March 5, 2004, 9:52pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/38 "2004-03-05T21:52:30Z")

</div>

It fails because of the coherent definition of g as the greatest possible being. The assertion “\<\>~g” reads “it is possible that the greatest possible being does not exist”. If it is the greatest possible being, then it cannot be impossible. Therefore, \<\>~g (or ~g) is a contradiction. On the other hand, the assertion “\<\>g” reads “it is possible that the greatest possible being exists”, and that assertion is intuitively tenable. Therefore, whatever g might represent in the statement \<\>~g, it cannot represent God as defined.

---

<div class="post-metadata">

**Author:** ![erislover](https://avatars.discourse-cdn.com/v4/letter/e/71e660/32.png) [@erislover](https://boards.straightdope.com/u/erislover)\
**Post date:** [March 5, 2004, 9:59pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/39 "2004-03-05T21:59:03Z")

</div>

I’m suddenly seeing this proof in a whole new light. Maybe it is the vicodin.

As a theist, it would be in my interest to prove god’s existence with this. To do so, one of the assumptions is that God is not impossible, where impossibility is symbolically ~G, that is, “necessary not”. So we’re saying god is not impossible, or it is not necessary that there is no god. ~~G, or \<\>G. Definitionally equivalent. What this actually means, though, is that there is at least one logical world where G is the case.

Suppose a strong atheist wants to use this proof. The counter proposition runs, ~G, or “God is not necessary.” What this says, in fact, is that there is at least one world where G is not the case. And so we come to the point where if g fails to obtain in one world, it fails to obtain in all of them.

The weak atheist, or agnostic, would have to counter the theist as such: either the other assumption (G-\>G) is not a logical quality god would have, or would have to hold onto strong atheism which, as ~G implies.

Interesting. This almost forces weak atheists to become agnostics or strong atheists, if I’m understanding my head properly (again: drugs).

**Lib** , I still strongly dislike the idea that G is _defined_ into existence like that. The proof, however strong it is, seems just as strong without it, since the hypothesis \<\>G assumes existence anyway.

[sub]This is not meant to usurp my previous comments wrt the meaning we may actually assign these symbols, which still might only add up to “Necessary existence exists.” Also, it is a prescription from dental work, nothing illegal here folks.[/sub]

---

<div class="post-metadata">

**Author:** ![Anduril](https://avatars.discourse-cdn.com/v4/letter/a/c0e974/32.png) [@Anduril](https://boards.straightdope.com/u/Anduril)\
**Post date:** [March 5, 2004, 11:02pm UTC](https://boards.straightdope.com/t/ontological-arguments/231958/40 "2004-03-05T23:02:37Z")

</div>

> [@Libertarian](#):
>
> It fails because of the coherent definition of g as the greatest possible being. The assertion “\<\>~g” reads “it is possible that the greatest possible being does not exist”. If it is the greatest possible being, then it cannot be impossible. Therefore, \<\>~g (or ~g) is a contradiction. On the other hand, the assertion “\<\>g” reads “it is possible that the greatest possible being exists”, and that assertion is intuitively tenable. Therefore, whatever g might represent in the statement \<\>~g, it cannot represent God as defined.

But, if using this method, you can’t even posit the possibility of g’s non-existence, how “credible” is the method? Talk about un-falsifiable.

(Feel free to educate me / expose me as an idiot. I really don’t get this part.)

[Previous page](https://boards.straightdope.com/t/ontological-arguments/231958.md?page=1)

[Next page](https://boards.straightdope.com/t/ontological-arguments/231958.md?page=3)
