Kant on the Ontological Argument

Kant claims the ontological argument is unsound because existence is not a predicate. Since existene is not "something' that can be part of our concept of a thing, it can not be part of our concept of God. Basically Kant thinks existence consists in the application of a concept to reality. He foreshadows the Russellian view of existence as a second order property.

My concern with Kant's critique lies in the relationship between two claims:

(1) Existence is not a real predicate

and


(2) Existence is not a first order property


I don't think (1) entails (2)

To be a "real predicate" is to be part of the concept of a thing. To be a property is to be part of a thing (or instantiated by the thing--the point is properties deal with reality, predicates with concepts)

Suppose there was a feature of reality that could not be part of the concept of a thing. Surely we cannot rule out a priori the possibility of such a property.

But how could such a property be known? if it is not part of the concept, how do we even thing about it?

My answer is to suggest that thinking by means of concepts is only one way to understand something. I can think of pain by means of a concept (i had a bad headache yesterday) OR I can think of pain by directly experiencing the pain. Only the former requires a concept.

I think that the cogito provides us with an example of a property being present phenomenologically, but which may, as Kant says, never be a part of our concept of something. When I know I exist, I know something about me intuitively, not by means of concepts. So I have reason to think that existence is a property even if Kant is right that existence is not a real predicate.

Another way of making the same point is to consider whether it is proper to think of consciousness as an object at all. Perhaps our concepts are limited to cognition of objects, but that which is aware of objects is of a radically different ontological kind.( compare Berkeley on notions and ideas.)

Could there be a non-existent consciousness?

My earlier argument pretty quickly assumed that If existence is a property of consciousness, it is a necessary property. I took for granted that there could not be non-existent consciousness, though there may be non-existent objects.


But why suppose this to be the case. Of course there are those who either deny existence is a property or hold that it is a property that everything has. I think there is at least a phenomenological distinction between:


(1) I am aware of a chair


and


(2) I am aware of my self (or my awareness)



Only (2) contains the guarantee of existence. Indeed, if (1) guaranteed existence, skeptical arguments would no longer keep us up at night.



But consider this case: In a dream I encounter Sherlock Holmes. Sherlock Holmes is presented as a person, not as an inanimate object. Therefore there is a sense in which if we allow non-existent objects, we need, on similiar grounds to allow non-existent persons. And if we allow that persons are conscious, it then follows that there is non-existent consciousness.


This conclusion can be avoided by denying non-existent objects altogether. Stubborn Meinongian that I am, I am not going to take that route.


Rather I propose we recognize that there are two different ways in which we can be aware of a consciousness, and it is only the first person way, my awareness of my awareness, that guarantees existence.


Sherlock Holmes is presented, in the dream, as being conscious--but his consciousness is not presented to me in an intuitive way. Rather it is presented after the manner that the back of a building is presented. In Husserl's terminology, it is presented as "merely meant." There is no intuitive grasp of consciousness when we are presented with the other

But this does not solve the problem. We are considering not whether or not we have an intuitive grasp of Sherlock Holmes' consciousness--if we did, we would know it to exist, and Sherlock Holmes would cease to be a nonexistent person. And here we see a disanalogy between consciousenss and nonexistent objects. While I may merely think of a unicorn, i can also have an intuitive awareness of a unicorn (in a dream or hallucination). But I can never have an intuitive grasp of a non-existent consciousness--if I did, I would then know the consciousness to exist.

Furthermore, if we suppose that "there are" nonexistent consciousness, then we would have to allow that Sherlock Holmes could perform the reductio. Just as the nonexistent holmes can have the property of walking, so to the nonexistent holmes can, it seems, have the property of thinking about his thinking and concluding that he exists.

But this, I take it, is absurd.

What we need to do, I think, is to go back and limit the class of nonexistent objects to those that can be presented in a intuitive way. An "object" which can only be "meant", which can never be present in person, must be understood in some other way. Similiarly, just because I can seem to think of an impossible object (the last prime number, for example) is not, by itslef, sufficient to suppose that there are impossible non-existent numbers. I have never been happy with the example of the round square--to me, all that I can think of in this case is "roundness," squareness" and the incompatibility of the two properties.

an ontological argument for immortality

Recall Descartes' famous dictum that "I exist" can be known with certainty prior to any other knowledge. Now if my existence is really what is first known, it cannot be an inference. If Descartes reasoned "Everything that thinks, exists"/"I think" "THerefore, I exist, my existence would not be the first thing known. it would depend on some general and controversial premises. For these reasons I think the cogito has to be understood as an immediate intuition--an insight into the necessary connection between "me" and "existence"

But the "I" or "me" may be too metaphysically robust. We may argue, contra Descartes, that what is known to exist is not the metaphysical subject, but consciousness--particular moments of awareness.

If that is so, then we should more conservatively claim: There is a necessary connection between consciousness and existence.

why a necessary connection? Well, if it was merely contingent, if there could be a consciousness that did not exist, then even this truncated cogito fails. But this more limited cogito is probably the least controversial of all philosophical theses.

Nor can one say that the intuition involved in the cogito only applies to consciousess--that the existence part is really superfluous. This is because it is not obvious that "everything" exists. Hallucinations and dreams seem to provide prima facie examples in which we are aware of nonexistent things. Of course, those of a Russellian bent will be inclined to explain away these appearances. But the theory of descriptions is hardly a premise we can use to buttress the soundness of the cogito. Again, the cogito stands on its own--its soundness relies on an immediate insight, not an argument that involves other premises.

But if existence is necesarily connected to consciousness, have we not then shown that consciousness necessarily exists? is not this parallel to the conclusion of the ontological argument with respect to God?

An argument against molinism

Molinists hold that God has knows what choices any possible free creature will make in any possible circumstance. To supporters of molinism, the view has the twin advantage of allowing for a very strong variety of divine providence while still insisting on a libertarian view of creaturesly freedom (it is crucial to the view that God knows what choices a creature would *freely* make in the circumstances specified). To others, like myself, the view makes no sense at all. The most common objection is the so-called "grounding problem." If a future choice is really free, then there seems nothing (now) that would make it true that a person really would make any particular choice. What God has, with respect to free choice, is at best probabilities. there is no fact, prior to creation, of how I will freely choose to spend my evening, and if there is no such fact, there is nothing that couldmake these "counterfactuals of freedom" true (though the non-existence of such facts might make these counterfactuals all false).

I think the grounding objection is a good one, but I want to explore the possibility of another argument against molinism. The following seems to be true about liberatarian views of free will.

(1) If a person has free will, then the person's choice is not determined.

Note that (1) asserts only a necessary condition for free will. All free actions are undetermined, but not all undetermined actions are necessarily free. One of the odd things about free will is that while it requires on the one hand a lack of determinism, it also requires, it seems to me, a kind of responsibility-If I freely choose, it is *I* that is doing the choosing. A purely random event is not sufficient to capture this sense of personal responsibility.

Be that as it may, lets now consider what God knows, according to the molinist scheme. One example of God's knowledge would be this:

(2) God knows that if Obama wins the election Joe will freely choose to drink a beer.


If free will requires indeterminism, then God must also know the following.


(3) Joe's choice of drinking a beer is not determined.

Here is the thing: I think (2) and (3) contradict one another. But its not obvious why this should be so. Consider:


(4) God knows that yesterday I freely chose to mow my yard

and

(5) God knows that Yesterday I was not determined to mow my yard.

There is clearly nothing contradictory about these two statements. From the fact that I did an act, it does not follow that the act was determined. Indeed, it has long been recognized that
the argument for the incompatibility of freedom and foreknowledge requires not

Necessarily, if God knows Joe will drink a beer, joe will drink a beer

But:

If God knows Joe will drink a beer, then, Necessarily, Joe will drink a beer.

But we have to be careful of the sense of necessity here. If we think of necessity in terms of possible worlds, then all sorts of unfree actions are not going to be necessary.

There is a possible world in which I go to the moon tommorow, but this is not something I can actually choose to do. Likewise, the fact that the current phase of the moon is determined does not preclude there being other possible worlds in which in which the moon is in a different phase or even in which there is no moon at all.

To capture the sense of necessity required to deny free will, we must consider not the entire set of possible worlds, but the set of possible worlds which are identical in all respects up to the point of the choice in question.

Once we have available this subset of possible worlds, we can see that an action will be determined if all worlds that are exactly like this one up to this point also contain the action in question and an action will be undetermined if there is at least one possible world with exactly the same history and in which the choice is different

Now it seems to me that already we have a refutation of molinism based on the grounding objection: Either there is something about this world which allows God to know what Joe will choose or there is not. If there is something, then all identical possible world segments will also contain this fact (maybe, as Plantinga says, its a brute counterfactual fact), and so in all those worlds Joe will choose to drink the beer-and Joe's choice will not be free.

But what about my claimed contradiction? To see it, we need to realize that God knows what Joe would choose to do in a way that does not involve temporal modalities. God is knowing all of this, on the Molinist view, prior to creation. So if joe's choice is free, God knows its not determined. Yet God knows also, "at the same time," that Joe has a finite probability of making this choice.
But if Joe does make the choice, the probability in question is 1--. But God cannot know both
(1) the probability of Joe making the choice is 1 and (2) the probability of Joe making the choice is less than one. These are inconsistent. So, once again, Molinism is false

After writing this, I think the earlier argument regarding possible world segments may be more convincing. But is this an example of another, independent argument?