The ontological argument redux

Since this thread disappeared into the archive before it ran its course, some readers may have missed it.

What we see here is an all-too-typical exchange between a believer and an unbeliever. The unbeliever raises an intellectual objection to some aspect of Christian theism or Christian apologetics.

The Christian rises to the challenge and patiently answers the objection.

At this point our militant atheist suddenly plays the anti-intellectual card, expressing his boredom at the technicality of the reply.

**********************************

Monday, August 07, 2006


Ontological arguments

According to Exbrainer, “Ontological arguments suck…Theists assert that a god does exist in the actual world. It is their responsibility, then, to demonstrate this.”

To which a friend sent me the following:

Here's the version from Alvin Plantinga’s God, Freedom, and Evil:

i) Key assumption: "There is a possible world in which maximal
greatness is instantiated"

(1) That is, possibly, maximal greatness is exemplified.
(2) Maximal excellence: depends on the properties it has in a given
world. Entails omniscience, omnipotence, and moral perfection.
(3) Maximal greatness: depends on what the being is like in other worlds.
(4) Maximal greatness entails maximum excellence in every world.
(5) Thus, "a being has the maximal degree of greatness in a given
world W only if it has maximal excellence in every possible world".


ii) The argument:

(1) Possibly, maximal greatness is exemplified (say, in W').
(2) Thus, "had W' been actual, there would have been a being with
maximal greatness."
(3) Thus, "if W' had been actual, there would have existed a being who was omniscient and omnipotent and morally perfect and who would have had these properties in every possible world."
(4) Thus, "if W' had been actual, it would have been impossible that there be no omnipotent, omniscient, and morally perfect being."
(5) "But… while contingent truths vary from world to world, what is logically impossible does not."
(6) "Therefore, in every possible world W it is impossible that there be no such being."
(7) Therefore, "it is impossible in the actual world (which is one of
the possible worlds) that there be no omniscient, omnipotent, and
morally perfect being."
(8) Thus, "there really does exist a being who is omniscient,
omnipotent, and morally perfect and who exists and has these
properties in every possible world."

posted by steve at 10:07 AM


14 comment(s):
Daniel Morgan said:

Daniel Morgan's variation of Plantinga --

i) Key assumption: "There is a possible world in which purple winged unicorns are instantiated"

(1) That is, possibly, purple winged unicorns are exemplified.
(2) Purple winged unicorn: depends on the properties it has in a given
world. Entail: is horse-like, is the color purple, has wings, and a horn in the center of its forehead.
(3) Purple winged unicorns: depends on what the being is like in other worlds.
(4) Purple winged unicorns entail the qualities: horse-like, is the color purple, has wings, and a horn in the center of its forehead in every world.
(5) Thus, "a being has the qualities: horse-like, is the color purple, has wings, and a horn in the center of its forehead in a given
world W only if it has purple winged unicorns in every possible world".


ii) The argument:

(1) Possibly, purple winged unicorns are exemplified (say, in W').
(2) Thus, "had W' been actual, there would have been a being who is a purple winged unicorn" (190).
(3) Thus, "if W' had been actual, there would have existed a being who is horse-like, is the color purple, has wings, and a horn in the center of its forehead and who would have had these properties in every possible world."
(4) Thus, "if W' had been actual, it would have been impossible that there be no beings which are horse-like, the color purple, have wings, and a horn in the center of the forehead."
(5) "But… while contingent truths vary from world to world, what is logically impossible does not."
(6) "Therefore, in every possible world W it is impossible that there be no such being."
(7) Therefore, "it is impossible in the actual world (which is one of the possible worlds) that there be no being who is horse-like, is the color purple, has wings, and a horn in the center of its forehead."
(8) Thus, "there really does exist a being who is horse-like, is the color purple, has wings, and a horn in the center of its forehead, and who exists and has these properties in every possible world."

Do you believe...? I do!
8/07/2006 10:46 AM
aquascum said:

Hi Daniel,

A couple of problems with your reply here.

First, you failed to notice the distinction between maximal excellence and maximal greatness, in the original presentation. In Plantinga's version of the argument, the first gets defined in i.(2) and the second gets defined in i.(3)-(4). Unfortunately, in your subsequent parody, you use the same term ("purple-winged unicorns") to define what were two different concepts in Plantinga's presentation.

But no matter. There's a way to patch up your parody so that this infelicity of terminology gets removed.

Second, and more importantly, you're failing to register the difference between Plantinga's definition of "maximal excellence," and your definition of "purple-winged unicorn". Your ii.(3) speaks of a purple-winged unicorn that exists "in every possible world." But that seems incoherent. As Craig and Moreland put it, in discussing the "necessarily existent lion" objection which is formally similar to yours: "For as a *necessary being* such a beast would have to exist in every possible world we can conceive. But any animal that could exist in a possible world in which the universe is comprised wholly of a singularity of infinite density just is not a lion [or a unicorn]. In contrast, a maximally excellent being could transcend such physical limitations and so be conceived as necessarily existent" (p. 497 of J. P. Moreland and William Lane Craig, _Philosophical Foundations for a Christian Worldview_ (Downers Grove, IL: InterVarsity Press, 2003)).

So, how about it? Could your purple-winged unicorn exist in a world without oxygen? Without matter? I think that the more you Chisholm your original concept of a purple-winged unicorn, in order to deal with the Craig/Moreland response, the more your beast will come to resemble what Plantinga calls a maximally great being. And so, in the end, your objection just provides us more roundabout way to confirm Plantinga's conclusion.
8/07/2006 11:22 AM
aquascum said:

Err, that should have been, "provides us *a* more roundabout way..."
8/07/2006 11:24 AM
John W. Loftus said:

Did you know that John Hick is known as the most influential philosopher of religion of the last half of the last century, and did you know that he did what Daniel Morgan just did to refute Plantinga's ontological argument too? Hick used equivalent terms that Plantinga used and concluded that "therefore there is an omniscient, omnipotent, and absolutely depraved being." But since two such opposite beings cannot exist, and since both arguments lead to contradictory conclusions, that therefore Plantinga's argument is wrong.

See Hick's An Interpretation of Religion, p. 78
8/07/2006 11:43 AM
steve said:

In other words, Daniel Morgan is reproducing the same flawed argument as John Hick. Thanks for pointing that out to us, John.

So Aquascum just killed two Dodo birds with one stone.
8/07/2006 12:45 PM
John W. Loftus said:

Steve, are you serious here? Duh. Do you even understand what's being said? Duh.
8/07/2006 1:48 PM
aquascum said:

Hi John,

One very general objection to all ontological arguments is that the concept of God employed -- "maximal greatness," "greatest conceivable being," etc. -- is incoherent. That would be the quickest way to block the central assumption that maximal greatness is possibly exemplified. Any defender of an ontological argument must recognize that the argument rises and falls on this coherence question.

But, of course, the same stricture would apply to any parodies of the ontological argument, such as the one Hick gives. On Hick's view, "maximal malignness" involves having "omniscience, omnipotence and absolute moral depravity" (78). And being "maximally evil" involves having "maximal malignness in every world" (78)

The problem here is that it's quite implausible to say that maximal malignness is possibly exemplified. Upon examination the idea looks incoherent, though for a reason which is different from why a necessarily-existing purple-winged unicorn looks incoherent. Could there really be a being who is omniscient, omnipotent, and absolutely morally depraved? If we're objectivists about morality, then morally good actions just are those actions we have the most reason to do. And if a being is omniscient, then he knows which actions he has the most reason to do. And if he's omnipotent, then -- being perfectly free, and subject to no non-rational causal factors influencing his choices -- he will always do those actions he has overriding reason to do. (In particular, he'll do a morally best action [if there is one] and no morally bad action.) And so, as Swinburne argues at length in ch. 11 of _The Coherence of Theism_, "it is logically necessary that an omniscient and perfectly free being be perfectly good" (188, rev. ed.). The first two attributes imply the third. So, contra Hick, maximal malignness is not possibly instantiated, although maximal greatness is.

I grant that this response assumes moral objectivism. It also assumes that omniscience and omnipotence entail perfect goodness. Disputing those assumptions might prove interesting. But the arguments for both are given in the chapter just referenced. If you think the arguments there fail, that would be the next logical step, but the arguments are there to be engaged.
8/07/2006 2:22 PM
Daniel Morgan said:

I thought no omnipotent being could be viewed by a standard of parallel or equivalent morality, since any standard would be contingent upon something the being has created [a lower standard]? The Being can either be evil or good from the perspective of a lesser being, but as the presups like to say, that's "just an opinion" (or something).

I thought this was why God escapes [supposedly] the Euthyphro Dilemma? I thought the standard defense is, from the Being's perspective: How is there a "good" or "bad" action for an omniscient, omnipotent creature? It just acts. Its creations could view its actions as "bad" for them and towards them (which is obvious, since I think the problem of evil is a valid argument). In that sense, the malignancy of Hick's creature must be relative to its actions as they affect its creations. How can a god-figure hurt itself, or act as to lessen itself? It cannot.

I'd never heard of John Hick, nor his counterargument. I had, however, heard of the ontological argument, and as soon as I heard it, the same sort of response came to my head as Hick says in p.77 of An Interpretation of Religion: Human Responses to the Transcendent, Second Edition -- that it is an attempt to prove God by definitional fiat:

"As in the case of other formulations of the ontological argument, the reasoning looks suspiciously like an attempt to prove divine existence by definitional fiat...This is perhaps fortunate; for Plantinga's argument for a maximally excellent being, if valid, would also work for a maximally evil being."

He goes on to basically replace "excellent" with "evil" in all of Plantinga's argument, as I did "purple unicorns". Will someone please explain why what Hick did was invalid? [given that the above attempt at nullifying the counterargument fails, since omnipotent beings may do what is good or bad for their creations?] According to Hick, Plantinga admits that this is indeed a valid refutation of this form of the argument (Plantinga, 1977, 110). I'm supposing the reference is to: "Actualism and Possible Worlds," Theoria (1977).

Perhaps more interesting than this specific case of the argument is Hick's assessment of the general issue of divining (no pun intended) the ontological necessity from necessary existence. Perhaps someone will fill me in more on how the following doesn't invalidate any attempt to "create" something by showing that it follows from definitional properties (that if it exists, it exists in X way)?

On p.76 he says,
"In assessing this argument [the ontological argument for God's existence] a distinction has to be drawn between logical and factual or ontological necessity. Logical necessity is the property that some propositions have of being true in virtue of the meanings of the terms composing them. But existential propositions, declaring that x exists, cannot have the kind of necessary or analytic truth because, as we noted above, existence does not name a defining property but is a term used to assert that a certain concept is instantiated. Thus whilst it may be necessarily true, not only that 'triangles have three sides,' but also that 'God is good', it cannot be necessarily true that there exist any objects with the properties of a triangle or any entity with the characteristics that would constitute it God. For logical necessity has no purchase on matters of fact and existence. There cannot be a logically necessary existent being. Nor indeed has classical theism generally supposed that there should." (emphasis mine)

How do we go from there to "proof" of existence?
8/07/2006 3:54 PM
John W. Loftus said:

aquascum, I think I could reason with you. I like how you argue, and yes, you allow for me to deny certain assumptions. I look forward to more of your comments.
8/08/2006 6:31 AM
John W. Loftus said:

Daniel can do research "on the spot." I like it!
8/08/2006 6:32 AM
Anonymous said:

i) Key assumption: "There is a possible world in which maximal greatness is NOT instantiated."

ii) The argument:

(1) Possibly, maximal greatness is NOT exemplified (say, in Y').
(2) Thus, had Y' been actual, there would NOT have been a being with maximal greatness.
(3) Thus, if Y' had been actual, there would NOT have existed a being who was omniscient and omnipotent and morally perfect and who would have had these properties in every possible world (e.g. this being would not exist in possible world Y').
(4) Thus, if Y'had been actual, it would have been POSSIBLE that there be no omnipotent, omniscient, and morally perfect being.
(5) Therefore, in ANY possible world Y it is POSSIBLE that there be no such being.
(6) Therefore, "it is POSSIBLE in the actual world (which is one of the possible worlds) that there be no omniscient, omnipotent, and
morally perfect being."
(7) Thus, IT IS POSSIBLE THAT a being who is omniscient, omnipotent, and morally perfect DOES NOT EXIST and DOES NOT HAVE these properties in every possible World (e.g. in World Y').

Let's continue:

(i) Key assumption: "a being has the maximal degree of greatness in a given world W only if it has maximal excellence in every possible world".

(1) A godless world (say, Y') is a world without a being with the maximal degree of greatness.
(2) If a godless world is possible, then a being who has a maximal degree of greatness does not exist (i.e. because a being with a maximal degree of greatness has a maximal degree of excellence in EVERY possible world).
(3) A godless world is possible.
(4) Therefore a being who has a maximal degree of greatness does not exist.
8/08/2006 9:33 AM
aquascum said:

Hi Daniel,

Re: your comment above, I don't quite follow the reasoning in your first two paragraphs. You seem to be relying on some assumptions about God and morality that I don't endorse. But it could be that I've misunderstood your point. Could you rephrase? For instance, I don't know what is meant by "a standard of parallel or equivalent morality". As far as I can tell, what you're saying is that objective morality can't apply to an omnipotent being. But that's just to dispute moral objectivism, a possible move I already noted in my final paragraph. Notice, though, that if you say that objective morality can't apply to an omnipotent being, then we've been given even *more* reason to think that Hick's parody-premise is implausible, for it claims that an omnipotent and absolutely morally depraved being is possible. This clearly involves the claim that it's coherent to apply moral categories to an omnipotent being. So I don't think you can defend Hick's parody-argument by disputing its premise :-)

The Euthyphro Dilemma is not solved by making morality relative, but by making God's nature the truthmaker for objective moral truths. That solution is quite compatible, I think, with the defense of the ontological argument I offered above.

You refer to Plantinga's "Actualism and Possible Worlds," but (i) that was published in 1976, not 1977, and (ii) it contains no mention of the ontological argument, apart from a brief mention in section II.2 that "If the ontological argument is correct, the property of knowing that God does not exist is necessarily coextensive with that of being a square circle; but surely these are not the same property, even if that argument is correct." No concession to Hick there! So, what is Hick referring to? Is he citing a Plantinga paper which he doesn't bother to name?

Indeed, as of 1998, Plantinga continues to endorse his modal ontological argument. Cf. his entry in the _Routledge Encyclopedia of Philosophy_, under "God, arguments for the existence of". In section 4 he outlines the argument and defends its central premise. Here's his closing paragraph, FWIW:

"So stated, the ontological argument breaches no laws of logic, commits no confusions and is entirely immune to Kant's criticism. The only remaining question of interest is whether its premise, that maximal greatness is possibly exemplified, is indeed true. That certainly seems to be a rational claim; but it is not one that cannot rationally be denied. A remaining problem with the argument, perhaps, is that it might be thought that the epistemic distance between premise and conclusion is insufficiently great. Once you see how the argument works, you may think that asserting or believing the premise is tantamount to asserting or believing the conclusion; the canny atheist will say that he does not believe it is possible that there be a maximally great being. But would not a similar criticism hold of any valid argument? Take any valid argument: once you see how it works, you may think that asserting or believing the premise is tantamount to asserting or believing the conclusion. The ontological argument remains as intriguing as ever."

Finally, it's quite true that -- in the material you cite, and elsewhere -- Hick distinguishes between logical and factual necessity. His writing on this point is widely anthologized. But his argument that no existential proposition can have logical necessity is flawed.

First, as with most analytic philosophers writing in a pre-Kripkean context, Hick conflates logical necessity with analyticity. Notice how he uses the terms "necessary" and "analytic" interchangeably. He says that "Logical necessity is the property that some propositions have of being true in virtue of the meanings of the terms composing them." But defining (as he does here) logical necessity in terms of analyticity isn't what *Plantinga* means by the term (existing in all possible worlds). Indeed, Plantinga *critiques* that notion of necessity-as-analyticity in ch. 1 of _The Nature of Necessity_. Surely it's clear that, in the version of the modal ontological argument posted above from _God, Freedom and Evil_, when Plantinga says that the being with "maximal greatness" necessarily exists, he means that it "exists in all possible worlds". That has nothing to do with analyticity.

In the post-Kripkean era of analytic philosophy, philosophers regularly distinguish the semantic notion of analyticity from the metaphysical notion of logical necessity, and distinguish both from the epistemological notion of a prioricity. In his writings on this subject, Hick regularly conflates the first two categories.

Now, it may be that all analytic statements are logically necessary. That is, being analytic is *sufficient* for logical necessity. But it doesn't follow that *only* analytic statements are logically necessary (that is, that being analytic is *necessary* for logical necessity). The necessary is a larger category than the analytic, such that there are other ways to be a logically necessary statement besides "being true in virtue of the meanings of the terms".

For instance, "There is a prime number between fifty and fifty-five" is not analytic (to give an example from Plantinga), because being "between fifty and fifty-five" is not part of the *meaning* of "prime number". But its necessarily true all the same. Ditto for "Everything that has a shape has a size," "Everything green has some spatial property," and (I say) "It is wrong to torture babies for fun." None of these statements are analytic, but they are all necessarily true if true at all. So Hick is wrong to identify the logically necessary with the analytic.

Second, even if we accepted Hick's conflation here, it doesn't follow that that's a reason to reject the ontological argument. For perhaps "God exists" *is* analytic, and therefore (ex hypothesi) logically necessary. According to Brian Leftow (who has written a book defending divine simplicity), if divine simplicity is essential to deity (perhaps by way of perfect being theology), then God is identical with his own existence. But that means that God = God's existence, and so the subject and predicate of "God exists" have the same reference. So it's analytic, and therefore (even on Hick's assumption about logical necessity as analyticity) would be logically necessary.

Third, Hick is mistaken to think that no "existential proposition" can be logically necessary. As Norman Malcolm puts it, "Is the Euclidean theorem in number theory, 'There exists an infinite number of prime numbers,' an 'existential proposition'? Do we not want to say that in some sense it asserts the existence of something?" Likewise, it's an analytic truth that "No bachelors are married." But that also tells me something about existence in the actual world. I can deduce from it that there are no married bachelors. Given this, it would be fairly arbitrary to say that logically necessary statements can only have existential implications for nonexistence, but not for *existence*.

So when Hick says that "logical necessity has no purchase on matters of fact and existence. There cannot be a logically necessary existent being," there's little reason to think this. If by "logical necessity" he just means "analyticity," then Plantinga could justly charge him with rebutting a definition of necessity which he isn't endorsing in the first place. But if by "logical necessity" he means "existence in all possible worlds," then *a fortiori* that concept *does* have "purchase on matters of fact and existence." That is its central characteristic: existence in worlds.

In short, Hick's objection looks like a non-starter to me.

Appendix on Anonymous's comment above. He's on the right track here, in a distant sort of way. But he needs to look at the Plantinga argument originally posted, and then compare it with his parallel. He's not in fact reducing Plantinga's argument to absurdity, because he's not in fact reproducing Plantinga's reasoning. In particular, his inference from (3) to (4) doesn't match Plantinga's reasoning, and so he doesn't get Plantinga's consequences. What Plantinga gets in his premise (4) is an "impossibility" which, in virtue of the fact that impossibilities do not change from world to world, gets you an impossibility in the actual world. (Cf. the top of this post.) And that gets you the existence of God. But in Anonymous's parallel, what he gets in his premise (4) is a "possibility". All that gets him is a possibility in the actual world, that is, it's possible that a being with maximal greatness doesn't exist. But not only is that his original assumption (and so the argument goes nowhere), it doesn't establish the nonexistence of God in the actual world. At best, all you get with "possibly, maximal greatness is not exemplified" is that God doesn't have necessary existence. But you certainly don't get the nonexistence of God.

So I think he needs to Chisholm this a bit to make it work.
8/08/2006 11:34 AM
Daniel Morgan said:

Aquascum,

The problem here is that it's quite implausible to say that maximal malignness is possibly exemplified. Upon examination the idea looks incoherent, though for a reason which is different from why a necessarily-existing purple-winged unicorn looks incoherent. Could there really be a being who is omniscient, omnipotent, and absolutely morally depraved? If we're objectivists about morality, then morally good actions just are those actions we have the most reason to do.

To which I replied,
I thought no omnipotent being could be viewed by a standard of parallel or equivalent morality, since any standard would be contingent upon something the being has created [a lower standard]? The Being can either be evil or good from the perspective of a lesser being, but as the presups like to say, that's "just an opinion" (or something).

Now, I was specifically focused upon, If we're objectivists about morality, then morally good actions just are those actions we have the most reason to do.

I phrased myself carelessly. What I mean to ask, and honestly ask (rather than argue) is: How can we establish what "the most reason to do" is if we do not establish some context (make it relative)? Don't we have to say, "the most reason to do, if the Being wants to make its creations happier and healthier," or "the most reason to do, if the Being wants to bring its own nature about in its creations," or something similar?

In this sense, I quoted the Euthyphro dilemma only because I thought the typical attribution of morality to God as an objective construct failed in this regard. You further reinforce this notion in your reply that,
The Euthyphro Dilemma is not solved by making morality relative, but by making God's nature the truthmaker for objective moral truths. That solution is quite compatible, I think, with the defense of the ontological argument I offered above.

But if we simply use definitional fiat to make God's nature "good" or "malignant", why is one valid but not the other?

As to Hick's reference, I don't know which paper he cites, but it is in the Bibliography under Plantinga, 1977, 110.

At best, all you get with "possibly, maximal greatness is not exemplified" is that God doesn't have necessary existence. But you certainly don't get the nonexistence of God.
And I suppose this was all I was pointing to by quoting p.76 of Hick at length. Whether anonymous or Hick said it properly, it still appears invalid to "create" something by definitional fiat, rather than provide for its possibility.

I was too busy picking my nose and having my eyes glaze over to follow much of the obscurantism in analytic philosophy. Sorry. I'm not saying you're wrong (or right), it just doesn't terribly interest me (or bother me).

Thanks for the fruitful dialogue, though. Interesting thoughts.
8/09/2006 6:26 AM
aquascum said:

Hi Daniel,

Four main points here.

[1]
Re: morally good actions being "those actions we have the most reason to do," all I can do at this point is again refer you to ch. 11 of Swinburne's _The Coherence of Theism_, where he defends this conception at length. The argument is also given in Part I of his _Responsibility and Atonement_ (pp. 9-18). In particular, he rebuts the notion that objective moral principles are context-relative (good relative to humans, or relative to purposes, etc.).

You ask: "How can we establish what 'the most reason to do' is if we do not establish some context (make it relative)?" Well, defending moral objectivism is to defend a meta-ethical theory about morality. It does not involve, in addition, identifying *which* particular putative actions are moral and which are immoral. That would be doing moral epistemology, not meta-ethics. Even if we couldn't settle the issue of moral epistemology, much less judgment of specific cases as right or wrong, that wouldn't have any bearing on the general defense of an objectivist meta-ethic. An analogy: someone could defend a theory of justification as internalist, without having to pass judgment as to whether a person P in situation S was in fact justified in believing p on that occasion. Or someone could hold that natural science is always the best route to truth, without knowing whether or not -- in some particular case -- good science or bad science was being performed. There could be agreement about the former, but disagreement about particular cases re: the latter.

More importantly, defining morality as what you have the *most* reason to do (because it is of overriding importance that you should do it) will settle the kinds of contextual questions you've raised. Sure, we can ask: given that person P wants to make others "happier and healthier," what does he have most reason to do? And wouldn't that be different from what he would have reason to do if he had *another* purpose in mind? Sure. But we can also ask: what reason do we have to make others happier and healthier in the first place? In the end, you will reach a bedrock principle that is context-independent. It's morally right to make others happier and healthier, and there's an end on it. The judgments about what we have most reason to do will involve, in any context, the invoking of principles that are not context-dependent.

(This isn't a full defense of the view, you understand. I'm just sketching out what the theory is in fact claiming about morality, and why these contextual questions don't undermine the theory, but rather provide opportunity for illustrating it.)

[2]
I had proposed that "The Euthyphro Dilemma is not solved by making morality relative, but by making God's nature the truthmaker for objective moral truths." You reply: "But if we simply use definitional fiat to make God's nature 'good' or 'malignant', why is one valid but not the other?" Again, I think you're continuing to confuse (perhaps?) meta-ethical and epistemological issues. *If in fact* God's nature is (the truth-maker for) the standard for the good, then the Euthyphro Dilemma has been avoided, for on this view morality is neither arbitrary nor more ultimate than God. Now, *why* is it the case that God's nature has the content that it has? That's an interesting question. *How do we know* that God's nature is in fact good, rather than evil? Another good question. But neither of these are the Euthyphro Dilemma.

In the end, the whole reason for making the divine nature the truth-maker for objective moral truths is to *clearly eschew* the view that morality is a matter of "definitional fiat". On the theory in question, morality *isn't* defined by way of definitional fiat. It's not defined by any act of choosing whatsoever, much less by an act of choosing how we are to use certain words. Rather, morality is *constituted* by some aspect of God's nature ('definitions' be damned). So either you've radically misunderstood the Euthyphro-response I've proposed, or I've radically misunderstood your question. From my vantage point, my answer is about as far from "definitional fiat" as can be imagined.

The Euthyphro Dilemma has consequences more far-reaching than may at first appear. I agree with Paul Helm: "In the opinion of many philosophers, when in the Euthyphro Plato asked the question 'whether the pious or holy is beloved by the gods' he was in effect raising one crucial difficulty for any Divine Command Theory of ethics. (If he was, he was also raising a difficulty for the relation of morality to any authority whatsoever.)" ("Introduction" to Paul Helm (ed.), _Divine Commands and Morality_ (New York: Oxford University Press, 1981), p. 2). In other words, if the Dilemma works, it works at underming moral objectivism *simpliciter* (theistic or not). Even an atheistic advocate of objective ethics faces the Euthyphro-style question: "Given the objective standard for morality you propose, are actions moral because the standard says they are moral, or does the standard say they're moral because they're moral (independent of the standard)?" If the former, then morality looks arbitrary; if the latter, then the 'standard' looks superfluous and irrelevant. That's why I made clear, in my initial reply to John above, that the real issue here is moral objectivism, not theistic moral objectivism in particular.

[3]
You say that "it still appears invalid to 'create' something by definitional fiat, rather than provide for its possibility." But this isn't what's going on in the ontological argument (whether Anselm's original version or Plantinga's modal version). If either author were 'creating' God by definitional fiat, the presentation would most likely go something like this: "I define 'God' as an existing being. Therefore, God exists." Clearly, this *isn't* what's going on in the ontological argument. For starters, although the argument employs definitions, it doesn't employ them like *that*! Rather, an *argument* is being given. One starts with a definition of "maximal greatness" (which definition does *not* involve the claim that God actually exists), and then one *argues*, by way of the relevant modal principles of S5 logic, that if we grant that maximal greatness is possibily exemplified, we get the actual exemplification of maximal greatness; i.e., the existence of God. This isn't "creating something by definitional fiat." It's arguing to a conclusion on the basis of premises. One can dispute the premises, or dispute the inferences, but one can't deny that the premises and inferences are there to be disputed. That's because what we have here is an actual argument, and not a "definitional fiat".

Again, you say that "it still appears invalid to 'create' something by definitional fiat, rather than provide for its possibility." I don't understand the force of the "rather than". "Providing for the possibility" of God's existence is precisely what's going on in the argument, nothing more. If you think Anselm or Plantinga start with something more than this, I'd like to see the evidence.

In the end, what you're proposing here looks like what Stephen T. Davis calls the "Boy Scout objection" to the ontological argument. (Cf. p. 26 of his _God, Reason, and Theistic Proofs_.) Even as an initiate to the Boy Scouts, on his first campout, might express disbelief that you can rub two sticks together and get something like fire (isn't fire radically different from wood?), so the objector to Anselm's/Plantinga's argument might express disbelief that you can rub concepts together and get something like the actual existence of God (isn't God radically different from mere concepts)?

If that's the objection, then the reply is twofold. First, the claim is straightforwardly false. As Davis puts it, surely "we *can* use purely a priori procedures to show some things about reality, for example, that certain things do not exist in it" (26). This is how we know there are no married bachelors or square circles; not by looking around us! We simply reflect upon the associated concepts, and we see that they couldn't be instantiated in reality. So we come to conclusions about reality on the basis of concepts alone. And, as I stated last time, it's arbitrary to say we can go one way with a priori procedures (show nonexistence) but not the other way (show existence).

Second, the claim looks to be question-begging. You can't just say, at the outset, that no one *can* prove anything this way. (A paraphrase of Hick: "It's invalid to start with mere concepts and reason your way to substantive claims about existence!") Rather, you'll have to examine the offered proof and see if it works! Maybe this is the first time such a proof *does* succeed. Who knows? Rather than rule out the cogency of such a proof at the outset, on the basis of an *a priori* philosophical prejudice that such proofs can't work in general, what's needed is a response to the proof itself.

[4]
Finally, you seem to imply that my interaction with the Hick citation you provided was just so much "obscurantism in analytic philosophy," and that "it just doesn't terribly interest" you. Well, OK, but I posted the material above for a reason. If you recall, you were the one who said that "Hick's assessment of the general issue" was "more interesting," and who said that "Perhaps someone will fill me in more on how the following [paragraph from Hick] doesn't invalidate any attempt to 'create' something by showing that it follows from definitional properties (that if it exists, it exists in X way)?"

That clearly looked like a request to interact with Hick's criticism. As far as I can tell, that's what I did. I'm sorry if your eyes glazed over :-) but I can't just respond to Hick by dismissing him. Thoughtful challenges require thoughtful replies. The stuff about necessity and analyticity was, well, necessary, because that's how *Hick* chose to frame the issue. He's the one who offered a rebuttal by using "analytic" and "necessary" interchangeably. And as I argued, that's the Achilles' Heel of his response (along with several other observations I made).
8/09/2006 3:18 PM