Philip Osborne

Interesting thread. I'll only address the ontological argument, since that is what I am familiar with.

First, it is perfectly possible to deny the existence of eunicorns. If we define eunicorns as existing, all we are saying is "For all x, if x is a unicorn, then x exists," which is trivially true. There is no formal contradiction in asserting both this and the proposition "No eunicorns exist."

I do not think a sentence with the form, "Possible, necessarily, P" is intelligible. In our case, P would denote "a being with all the properties of God except that may or may not exist necessarily." I just don't think this makes any sense, and it certainly does not work with a possible-worlds conception of modality.

You have defined P in a way that is inconsistent with the ontological argument. P denotes "a being which is God-like and which may exist necessarily." This is clearly different from P', which is "a God-like being which may or may not exist necessarily." P', I agree, is unintelligible. However, there is a difference between possibility and contingency. Possibility means truth in at least one possible worlds; contingency means truth in some worlds, falsehood in others. In other words, both necessary and contingent propositions are subsumed under the possible.

If we are to have an intellgible concept of modality, I think we will have to accept nested modal operators. I don't think anyone would deny that "Necessarily p implies p." In other words, if p is necessary, then "Necessarily p" is true in the actual world. Nor do I think it can be plausibly denied that "actually p implies possibly p." (Note: this can be directly inferred from the first principle I mentioned) But from these two propositions, can we not infer "possibly, necessarily p?"

The modal ontological argument does not use existence as a property; it uses Anselm's principle, which says that the properties of God cannot be of a kind that could be contingent. Whle existence is not a property, what is described in Anselm's principle clearly is a property, not a quantifier. From this conceptual truth, and the assumption that the existence of God is possible, it is valid to infer that God exists necessarily. So I think that Kant's criticisms are outdated against modal versions of the argument.

The only decisive objection against the modal argument, I think, is that the assumption that God is possible is unwarranted within the context of Anselm's principle. Every atheist would say that the non-existence of God is possible, and if we incorporate that into the argument, we get the conclusion that the existence of God is impossible. So we arrive at a conceptual stalemate. I think that the principle of the argument ("Possibly, necessarily p implies actually p") is very useful in analytic theistic arguments, but I doubt that the argument itself will be persuasive.

Sincerely,

Philip

Thomas Metcalf

Originally posted by Vanderzyden:

"Do you, as an atheist, really want to take this path? If the constants cannot be varied, then there is no lottery."

I do not need to believe that nothing could have varied. I'm simply denying that we can infer any sort of probability judgment merely from a conditional.

"In your version, '(neighbor) winning the lottery' is equivalent to 'permit life'. However, there is no equivalent for '(neighbor) cheating' in my argument. You have no equivalent for my 'God exists', since 'God exists' is not to ' permit life' as 'winning the lottery' is to 'cheating in the lottery' This, again, is why I don't quite understand the analogy."

The equivalent to "neighbor cheats" is "someone chooses the constants," in the original finetuning argument. "God exists" is to "someone cheated in the lottery" as "The universe permits life" is to "My neighbor won the lottery."

You conclude from massive improbability, "not chance." I suggest we do the same if my neighbor wins the Googolplex Lottery. Why not?

"I notice again that you avoid a particularly important question."

I do not know the probability that my monitor will spontaneously rise off the desk. I suspect it is very low. You seem to be arguing that fine-tuning provides some evidence for God. I do not dispute this. I dispute, however, that it provides enough evidence to counteract our intuitive background improbability. The lottery analogy is again relevant; yes, that my neighbor wins the Googolplex Lottery is some evidence that someone cheated, but it's not enough to make us think it's likely, is it?

rdalin

I don't understand why you say this. How do you know that science cannot close the gap you cite? Your position seems to be philosophical or theistic rather than scientific.

[ October 18, 2002: Message edited by: rdalin ]</p>

Thomas Metcalf

Originally posted by Philip Osborne:

"First, it is perfectly possible to deny the existence of eunicorns. If we define eunicorns as existing, all we are saying is 'For all x, if x is a unicorn, then x exists,' which is trivially true. There is no formal contradiction in asserting both this and the proposition 'No eunicorns exist.'"

Well, right. This is designed to show that some formulations of the ontological argument really stick together two steps, and the "eunicorns" point tries to get them unstuck. If we define God as a necessary being, all we're saying is "For all x and y, if x is God, then for all y, if y is a possible world, x exists in y."

"You have defined P in a way that is inconsistent with the ontological argument. P denotes 'a being which is God-like and which may exist necessarily.'"

Then within "God-like" are all of God's attributes except His alethic modal status? If this is so, then any time we assert it's possible that God exists, we're asserting that it's possible that it's necessary that P exists.

"I don't think anyone would deny that 'Necessarily p implies p.' In other words, if p is necessary, then 'Necessarily p' is true in the actual world. Nor do I think it can be plausibly denied that 'actually p implies possibly p.' (Note: this can be directly inferred from the first principle I mentioned) But from these two propositions, can we not infer 'possibly, necessarily p?'"

I agree that []p --&gt; p, and that p --&gt; &lt;&gt;p, but only if we restrict p to include only statements without modal operators. This was my original claim. In my experience, to assert it's possible that it's necessary that p is illictly appealing because the "possible" can be taken to be an epistemic rather than alethic modal operator. To assert &lt;&gt;[]p is not to assert that as far as we know, it might be the case that []p, as you no doubt agree, but rather, to assert that there is a possible world in which p is true in all possible worlds. My only intuition is that this simplifies to []p, which is how I believe we should take nested modal operators, to simplify to the most immediate operator.

"The modal ontological argument does not use existence as a property; it uses Anselm's principle, which says that the properties of God cannot be of a kind that could be contingent. Whle existence is not a property, what is described in Anselm's principle clearly is a property, not a quantifier."

Why is it not the case that to assert something is contingent is to assert that it exists in some, but not all, possible worlds? An assertion of alethic modal status is an assertion, I think, of some form of existence. I think it is rather doubtful that alethic modal status is a property, for the reasons I have suggested above and for this reason. If this is so, then Kant's criticism might be taken to be a more general principle, from which we can derive that alethic modal status is not a property.

"Every atheist would say that the non-existence of God is possible, and if we incorporate that into the argument, we get the conclusion that the existence of God is impossible."

Indeed. It is certainly conceptually possible to imagine that if we searched throughout all the possible worlds, we would not find a being Who happens also to exist in every other possible world.

"I think that the principle of the argument ('Possibly, necessarily p implies actually p') is very useful in analytic theistic arguments, but I doubt that the argument itself will be persuasive."

Do you mean in a Plantingan kind of way, that these analytic theistic arguments will help to support theism to people who are antecedently theists?

Philip Osborne

Then within "God-like" are all of God's attributes except His alethic modal status? If this is so, then any time we assert it's possible that God exists, we're asserting that it's possible that it's necessary that P exists.

On the Anselmian conception of God, the property of being God-like itself is enough to entail the impossibility of that entity being contingent.

I agree that []p --&gt; p, and that p --&gt; &lt;&gt;p, but only if we restrict p to include only statements without modal operators. This was my original claim. In my experience, to assert it's possible that it's necessary that p is illictly appealing because the "possible" can be taken to be an epistemic.....

Simplifying to the most immediate operator is exactly what the S4 and S5 modal logic systems are intended to do. In S5, &lt;&gt;&lt;&gt;&lt;&gt;&lt;&gt;p translates into &lt;&gt;p, and []&lt;&gt;[]p becomes []p.

Why is it not the case that to assert something is contingent is to assert that it exists in some, but not all, possible worlds? An assertion of alethic modal status is an assertion....

You may be right, if we conceive of modal operators as quantifiers over possible worlds. However, your argument makes a crucial error. To say that God cannot exist contingently is only to say ~&lt;&gt;(&lt;&gt;p & &lt;&gt;~p), or p =&gt; []p, both of which are not direct assertions of the thing's modal status. p =&gt; []p can be true whether the thing is impossible or necessary. p =&gt; []p does not quantify over possible worlds and/or tell us which worlds p exists or fails to exist in, and so does not seem vulnerable to your criticism.

Do you mean in a Plantingan kind of way, that these analytic theistic arguments will help to support theism to people who are antecedently theists?

Actually, I meant that the principle can be used in variations of traditional theistic arguments. Richard M. Gale's approach in "Why Traditional Cosmological Arguments Don't Work (And a Sketch of a New One That Does)" would be an example.

Sincerely,

Philip

[ October 18, 2002: Message edited by: Philip Osborne ]</p>

Thomas Metcalf

Originally posted by Philip Osborne:

"To say that God cannot exist contingently is only to say ~&lt;&gt;(&lt;&gt;p & &lt;&gt;~p), or p =&gt; []p, both of which are not direct assertions of the thing's modal status. p =&gt; []p can be true whether the thing is impossible or necessary. p =&gt; []p does not quantify over possible worlds and/or tell us which worlds p exists or fails to exist in, and so does not seem vulnerable to your criticism."

Let "p" be "a modally-neutral (may or may not be possible, necessary, contingent, or impossible) God-like entity." Is this acceptable, given the two sentences above? I don't see how anyone could derive either of them simply from p. If God's alethic modal status is defined simply as "not contingent," we've already made statements about existence in possible worlds. And if we're simply asked to accept the principle "If God exists, God is necessary, and if God does not exist, God is impossible," without special (and modally-specifying) definitions of God, well, I don't see how anyone could argue for that principle.

Vanderzyden

Thomas,

Even with your corrections, there are still problems with your analogous argument.

First, a question: What is "intuitive background probability"?

I would not say that winning a "Googleplex" lottery is any evidence of cheating. Proof of cheating would require a specialized investigation. High odds do not, as you say, indicate cheating. The trouble here is that cheating is not analogous to design. Teleology is the study of evidences for design in nature. In arguing the worthiness of such a study, one element of investigation is the probability that the ordered universe could come about from purely naturalistic means. Other elements are observing information, beauty, elegance, specificity, and variety.

Let me offer this definition:

Effective impossibility: an event is effectively impossible if the improbability of the event is greater than the number of atoms in the universe, which is approximately 10 to the power of 70.

Would you agree with this definition?

Vanderzyden

[ October 20, 2002: Message edited by: Vanderzyden ]</p>

HRG

How would you know ..

- which universes are "possible" ?
- what probability measure should be used on the "set of all possible universes" ?

Note that "equal probability for all universes" is meaningless, as soon as there are some parameters which may be varied continously (as the fine-tuning argument assumes).

- that unordered universes (whatever that means) are even possible, i.e. that "order" had to "come about" ? What else could be understood by "order" than "what the universe does?"
Sure. They exist inside the minds of intelligent organisms.

And what about ugliness, clumsiness etc. ?

No. First, it should be "particles".
Second, the threshold would be the number of possible interactions of those particles: 2^(10^80).

HRG.

Jack the Bodiless

Vanderzyden:
Blatant anthropomorphism. You're saying that sediments won't be graded by size as they settle out from muddy water without the action of the Mud Fairies?

For evolution, the selector is DEATH. Only a theist imagines that the Grim Reaper must be intelligent. Your example here is just one of the extreme examples of how far creationists take a notion and stretch it beyond its practical limits in order to just their wild hypothesi.
It's OK for purposes of illustration, for getting a handle on what "ridiculously improbable" means. I have myself used the estimated number of atoms in the Universe as an analogy against creationism: the perfection of the fossil record as proof that Genesis is false (the impossibility of "Flood sorting" and suchlike: it is easy to show that the odds against Genesis are much greater than the estimated number of atoms in the Universe).

However, we can perform such calculations only because we can estimate a large sample size (in this case, the number of permutations of a given number of fossils). We CANNOT do this with universal constants, because we do not know how they are interrelated and over what ranges they can vary. Such attempts are doomed to failure. We just can't get ANY numbers to play with.

Jack the Bodiless

I share rdalin's bafflement at your use of this assertion. Especially as the problems with this line of argument have already been pointed out to you on several occasions now.

1. Why assume science cannot close this gap?

2. Why ignore the SAME gap in theology? If there is a God, then WHY is there a God, rather than no God?