Is the Modal Ontological Argument Compelling?

In a comment, Patrick Toner writes,

. . . there is no substantive philosophical position for which there is *better* philosophical support than theism. I'm open to the possibility that at least one other philosophical position--namely, dualism--is at least as well supported by philosophical argument as theism. But nothing's got better support.

[. . .]

That said, I find St. Thomas's second way indubitable. I also find the modal ontological argument compelling. The kalam cosmological argument seems pretty much irrefutable.

In another comment in the same thread, Toner writes,

But we still do (or can) know God and the soul with certainty through the use of natural human reason. (emphasis added)

What interests me in this entry is Toner's explicit claim that the modal ontological argument is (rationally) compelling, and his implicit claim that this argument delivers (objectively) certain knowledge of the existence of God.  While I consider the argument in question to be a good argument, I don't find it to be compelling.  Nor do I think that it renders its conclusion certain. My view is that no argument for or against theism is rationally compelling.  No such argument resolves the issue.  I think it would be wonderful if there were a compelling argument for the existence of God.  The metaphysical knowledge generated by such an argument would be the most precious knowledge that one could possess.  So I would be much beholden to Toner if he could show me the error of my ways.  

Perhaps there is a theistic argument that is rationally compelling. If there is I should like to know what it is.  I am quite sure, however, that the following argument does not fill the bill.

A Modal Ontological Argument

'GCB' will abbreviate 'greatest conceivable being,' which is a rendering of Anselm of Canterbury's "that than which no greater can be conceived."  'World' abbreviates 'broadly logically possible world.' 'OA' abbreviate 'ontological argument.'

1. Either the concept of the GCB is instantiated in every  world or it is instantiated in no world.

2. The concept of the GCB is instantiated in some world.  Therefore:

3. The concept of the GCB is instantiated in every world.  (1, 2 by Disjunctive Syllogism) 

4. The actual world is one of the worlds. Therefore:

5. The concept of the GCB is instantiated in the actual world. (3, 4 ) Therefore:

6. The GCB exists. (5)

This is a valid argument: it is correct in point of logical form.  Nor does it commit any informal fallacy such as petitio principii, as I argue in Religious Studies 29 (1993), pp. 97-110.  Note also that this version of the OA does not require the controversial assumption that existence is a first-level property, an assumption that Frege famously rejects and that many read back (with some justification) into Kant.  (Frege held that the OA falls with that assumption, cf. Die Grundlagen der Arithmetik, sec. 53; he was wrong: the above version is immune to the Kant-Frege objection.)

(1) expresses what I call Anselm's Insight.  He appreciated, presumably for the first time in the history of thought, that a divine being, one worthy of worship, must be noncontingent, i.e., either necessary or impossible.  I consider (1) nonnegotiable.  If your god is contingent, then your god is not God. There is no god but God.  God is an absolute, and no absolute worth its salt is contingent.  End of discussion.  (If, however, (1) is reasonably disputable, then this only strengthens my case against compellingness.)

It is premise (2) -- the key premise -- that ought to raise eyebrows.  What it says -- translating out of the patois of possible worlds -- is that it it possible that the GCB exists.

Whereas conceptual analysis of 'greatest conceivable being' suffices in support of (1), how do we support (2)?  Why should we accept it?  How do we know that (2) is true?  Some will say that the conceivability of the GCB entails its possibility.  But I deny that conceivability entails possibility.  

Conceivability Does not Entail Possibility

The question is whether conceivability by finite minds like ours entails real possibility.  A real possibility is one that has a mind-independent status.  Real possibilities are not parasitic upon ignorance or on our (measly) powers of conception.  Thus they contrast with epistemic/doxastic possibilities.  Since what is epistemically possible for a person might be really impossible (whether broadly-logically or nomologically), we should note that 'epistemic' in 'epistemically possible' is an alienans adjective: it functions like 'decoy' in 'decoy duck.'  Ducks don't come in two kinds, real and decoy.  Similarly, there are not two kinds of possibility, epistemic and real.  To say that a state of affairs is epistemically/doxastically possible for a subject S is to say that the obtaining of the state of affairs is logically compatible with what S knows/believes.  For example, is it possible that my State Farm insurance agent Tim be working his office during normal business hours today ?  Yes, epistemically: it is not ruled out by anything I know.  But if Tim unbeknownst to me 'bought the farm' last night, then it is not really possible that Tim be working in his office today.

By 'conceivability' I mean thinkability by us without apparent logical contradiction.  

First Argument

Why should the fact that a human being can conceive something without apparent logical contradiction show that the thing in question can exist in reality? Consider the FBI: the floating bar of iron. If my thought about the FBI is sufficiently abstract and indeterminate, then it will seem that there is no 'bar' to its possibility in reality. (Pun intended.) If I think the FBI as an object that has the phenomenal properties of iron but also floats, then those properties are combinable in my thought without contradiction. But if I know more about iron, including its specific gravity, and I import this information into my concept of iron, then the concept of the FBI will harbor a contradiction. The specific gravity of iron is 7850 kg/cu.m, which implies that it is 7.85 times more dense than water, which in turn means that it will sink in water.

The upshot is that conceivability without contradiction is no sure guide to (real) possibility. Conceivability does not entail possibility.

Second Argument

Both the existence and the nonexistence of God are conceivable, i.e., thinkable by us without apparent logical contradiction.  So if conceivability entails possibility, then both the existence and the nonexistence of God are possible.  If so, God is a contingent being.  But this contradicts the Anselmian Insight according to which God is noncontingent.  So if the Anselmian Insight is true, then conceivability-entails-possibility is false and cannot be used to support premise (2) of the modal OA.  The argument can be put in the form of a reductio:

a. Conceivability entails possibility.  (assumption for reductio)

b. It is conceivable that God not exist. (factual premise)

c. It is conceivable that God exist.  (factual premise)

d. God is a noncontingent being. (true by Anselmian definition)

Ergo

e. It is possible that God not exist and it is possible that God exist.  (a, b, c)

Ergo

f. God is a contingent being. (e, by definition of 'contingent being')

Ergo

g. God is a noncontingent being and God is a contingent being. (d, f, contradiction)

Ergo

~a. It is not the case that conceivability entails possibility. (a-g, by reductio ad absurdum) 

Or, if you insist that conceivability entails possibility, then you must give up the Anselmian Insight.  But the modal OA stands and falls with Anselmian insight.  

 

Is Conceivability Nondemonstrative Evidence of Possibility?

 

We don't need to discuss this in any depth.  Suppose it is.  This won't help Toner's case.  For if it is not certain, but only probable that (2) is true, then this lack of certainty will be transmitted to the conclusion, which will be, at most, probable but not certain. In that case, the argument will not be compelling.  I take it that an argument is compelling if and only if it renders its conclusion objectively certain.

Are There Other Ways to Support the Possibility Premise?

I can think of one other way.  It has been suggested that the possibility premise can be supported deontically:

A. A maximally perfect being ought to exist.
B. Whatever ought to exist, is possible.
Therefore
C. A maximally perfect being is possible.

I discuss this intriguing suggestion in a separate post  wherein I come to the conclusion that the deontically supercharged modal OA is also not compelling.

What is it for an Argument to be Compelling? 

My claim on the present occasion is that the modal OA provides no demonstrative knowledge of the truth of theism. Demonstrative knowledge is knowledge produced by a demonstration.  A demonstration in this context is an argument that satisfies all of the following conditions:

1. It is deductive
2. It is valid in point of logical form
3. It is free of such informal fallacies as petitio principii
4. It is such that all its premises are true
5. It is such that all its premises are known to be true
6. It is such that its conclusion is relevant to its premises.

To illustrate (6).  The following argument satisfies all of the conditions except the last and is therefore probatively worthless:

Snow is white
ergo
Either Obama is president or he is not.

On my use of terms, a demonstrative argument = a probative argument = a proof = a rationally compelling argument.  Now clearly there are good arguments (of different sorts) that are not demonstrative, probative, rationally compelling.  One type is the strong inductive argument. By definition, no such argument satisfies (1) or (2).  A second type is the argument that satisfies all the conditions except (5). 

And that is the problem with the modal OA. Condition (5) remains unsatisfied.  While the possibility premise may be true for all we know, we do not know it to be true.  So while the modal OA is a good argument in that it helps render theism rational, it is not a compelling argument. 

Related articles

Soloveitchik on Proving the Existence of God

Modality, Possible Worlds, and the Accidental-Essential Distinction

Does the Atheist Deny What the Theist Affirms?

Comments

[Elliott]

I’m inclined to agree that human conceivability does not entail logical possibility. The fact that a premise does not appear logically contradictory to the human mind does not entail that the premise is free from contradiction (especially regarding complicated premises), although I'm inclined to believe that conceivability provides a good reason in favor of logical possibility.

But I’m wondering if a problem is lurking. Condition five in the conditions for a rationally compelling argument doesn't seem rationally compelling.

[Elliott]

I’ll try to articulate the problem (if it is one -- I'm not sure) in the form of a dialogue:

Smith: So, you hold that a rationally compelling argument must meet these six conditions?

Jones: Right.

S: Conditions one through four are obvious, as is condition six. But what do you mean by “known to be true” in condition five? Do you mean that the premises must be known with certainty? Does knowledge require objective certainty?

(Suppose Jones says “no.” Then condition five might be paraphrased as “It is such that all the premises are recognized as more plausibly true than false.” But suppose Jones says …)

J: I hold that knowledge requires objective certainty.

S: Some competent thinkers would disagree. They’d say that knowledge doesn’t require objective certainty.

J: There’s a reasonable disagreement here.

S: Why believe that knowledge requires objective certainty? Is there a rationally compelling argument for that claim?

At this point, either there is or there is not a rationally compelling argument. If there is, then the argument must meet condition five. But the truth of five hasn’t been demonstrated in a compelling manner.

If there is not, then the discussion remains open concerning the nature of a rational demonstration, and the question of whether or not the modal OA is compelling remains open.

[BV]

Very intelligent response, Elliot. Whether or not knowledge requires objective certainty, I may re-write condition (5) as

5* It is such that all its premises are objectively certain.

In this way I side-step controversy over whether knowledge requires certainty. I don't need to claim that knowledge requires certainty; all I need to claim is that an argument that issues in a certain conclusion must have certain premises. In fact, I don't need to use 'knowledge' at all.

Toner claims that the existence of God can be certainly known on the basis of the modal OA. Well, then, if there is the least bit of uncertainty with respect to any of the premises, than that uncertainty will be transmitted to the conclusion. Therefore, it is in the spirit of what I am proposing that (5) be re-formulated as above.

[BV]

If I am told that I 'set the bar too high,' I will reply that I set it in exactly the right place. A rationally compelling argument is one whose conclusion is such that, if you reject it, then you are irrational.

[Elliott]

>> all I need to claim is that an argument that issues in a certain conclusion must have certain premises. <<

That is a good point, and 5* seems an apt re-formulation of (5).