Cosmological Argument: Objections

| | Comments (11) | TrackBacks (1)

This is the the nineteenth post in my Theories of Knowledge and Reality series. Follow the link for more on the series and for links to other entries as they appear. In the last post, I presented the cosmological argument for the existence of God. In this post, I'll address what I consider the two best objections to the argument before offering some concluding thoughts.

First, we might think that the universe itself is self-existent. Then the conclusion of the argument is true, but it doesn't give us anything like the traditional theistic God. Suppose that is right. This commits us to a certain view about the universe, namely that it is the sort of thing that couldn't fail to exist. It means it is false to say that there might not have been a universe. This is certainly not a conclusive argument, but many philosophers want to avoid this conclusion.

Suppose you are comfortable with that conclusion. Do we really have an explanation for why there are any dependent things at all? Being self-existent simply because your parts are all explained still doesn't give an explanation of why there are any such parts. The traditional conception of God explains it more fully. It's God's nature to exist. God is the sort of thing that has to exist, but God is also viewed as a creator. Would we see the universe as a creator in the same way? It's hard to see how, which might leave us thinking that the universe as a whole doesn't serve as the kind of explanation that God does. In short, theism as a view explains why God would be self-existent, but I know of no explanation of why the universe would be self-existent. I don't think of this response as a disproof of the objection, but I do think of it as a good reason to prefer the theistic account.

The second objection I have in mind is William Rowe's (see the reference in the previous post in the series). His strategy is to deny PSR altogether. He says there could be a third kind of answer to explanation questions. Something's nature could explain something about it. Something else could explain something about it. But if you deny PSR, you can also simply have facts without any explanation. Philosophers call these brute facts. If PSR is true, there are no brute facts. Every fact is explained. But Rowe wonders why there couldn't simply be one brute fact -- the existence of dependent beings. Then there's no reason why any dependent things exist. Some will think the question is meaningless (like the question of where the universe is or when the timeline is). I get the impression that Rowe doesn't think it's meaningless, but he just thinks there's no answer to it. Either way, this response takes PSR to be right about individual things but not about the kind of explanation this argument calls for.

My basic response to this is that we assume PSR all the time. It seems intuitively true. Rowe insists that that doesn't mean it's right. We're often wrong about principles that seem right. But don't claims that seem intuitively true deserve the benefit of the doubt unless we can come up with good reason to think they're false? Rowe doesn't think it is intuitively true, however. His main piece of evidence for this is that many philosophers doubt it. But I don't think we have any good reason to doubt it. They just claim that they don't have any evidence for it. But lack of evidence doesn't mean we have no reason to believe it. I'll refer you to earlier posts in this series for my thoughts on evidence and what counts as a good belief, but one conclusion I've argued for is that it's perfectly fine to believe things that we have no evidence for, under certain kinds of conditions, and this may well be one of them. In particular, if we are constructed in such a way that inclines us to ask for explanations, and the reason we're constructed that way is because things do have explanations, then isn't PSR not just a good belief to have for the sake of understanding the world scientifically but also a completely justified belief, even knowledge? If what I've argued about knowledge in the first part of this series of posts is correct, I have to answer positively.

Rowe complains that a theist offering the cosmological argument is begging the question by assuming a premise that guarantees what they're trying to prove. If he's right that there's no reason to believe PSR, then the argument does beg the question. Don't assume anything that will guarantee what you want to prove if the only basis of your assumption is that you want the conclusion to be true. But that's not what the theist is doing here. Since it seems intuitive, and we rely on it all the time, why not assume that it's true, even apart from my argument that it might be genuine knowledge? Even if it's assumed for purely pragmatic purposes in our search for scientific explanations, it's still not being assumed to prove the existence of God, and thus it doesn't beg the question. Scientific pursuit need not involve any attempt to prove God's existence.

In fact, to dismiss a principle that seems intuitively true (and we also happen to rely on, and may even be knowledge if certain other things are true) without argument just because you want to resist the conclusion that God exists seems to be begging the question the other way. If you deny PSR because you don't believe in God, isn't that just as bad? So Rowe may well be doing exactly what he accuses the theist of doing if he denies the principle to avoid theism. He doesn't seem to me to give any reason for denying it other than that he doesn't think it's been proved without a doubt, but hardly anything has, and it's not as if the considerations for believing it are all that weak. I conclude that we should believe PSR absent a real objection to it. (I will note that some have offered objections to certain versions of PSR, but the version I've been using here isn't subject to those objections. If someone wants to take this up in the comments, that's fine. Otherwise I'll just leave it to the side.)

So what do we conclude if the argument is successful, as I believe it is? Well, it's not all that much. There's something that's self-existent. It doesn't seem to me that the universe is a good candidate, at least not as good a candidate as God is. God seems to me to be the best candidate, in fact, though this leaves open a number of conceptions about what God might be like. As an argument for theism, this seems to me to be one of the stronger in terms of establishing its conclusion but one of the weaker in terms of what the conclusion tells us about God. Many theists place this argument together with others to establish a fuller picture, and together with a larger apologetic for a particular perspective on what God is like I think this argument really can play a role, but it has to be just one small part of a much bigger picture, and many parts of that picture will be established inductively and without 100% guarantee. Even this argument has some of that due to PSR's not being proved 100%. But is that nothing? Many people dismiss this argument as ineffective simply because it doesn't do all that one might hope an argument for God could do, and that just seems to me to think too much in all-or-nothing terms. Philosophy simply isn't like that most of the time.

The next post will consider the design argument for the existence of God.

1 TrackBacks

Listed below are links to blogs that reference this entry: Cosmological Argument: Objections.

TrackBack URL for this entry:

I'm doing a series at my own blog based on the content of an intro to philosophy course that I teach from time to time called Theories of Knowledge and Reality. This is designed to reflect that course, so it's... Read More


Great posts, Jeremy.

We do have a strong intuition that the PSR is true. I wanted to see if you have comments on this thought:

As you said in the post and in the comment dialogue with wink, your version of the cosmo. argument depends on using "self-existent" and "dependent" in place of the commonly used "necessary" and "contingent". But while dependent doesn't imply contingent, you agree that self-existent does imply necessary and that it follows that the dependent things are also necessary.
But the intuition that the things in our world are contingent is also very strong. I'm not sure I would give this up to save the cosmo. argument.

Actually, I'm not completely sure if self-existence implies necessity. It seems to me that it does. But what I was conceding is that the existence of the self-existent thing is necessary. I don't think I conceded that everything the self-existent thing does is necessary. That makes the difference in terms of whether dependent things are necessary. If the explanation for what the self-existent thing does is a final cause, then it doesn't necessitate in the way it would if it were efficiently caused by the self-existent thing's nature. You need not get Spinoza, in other words. You might just have Augustine, Aquinas, or Leibniz, all of whom think of God's providence in terms of final causation in cases where God doesn't efficiently cause events.

If the universe is self-existent then it must have existed eternally since there is nothing in nothing which can produce anything, let alone a universe.

If universe has existed eternally then according to the second law of thermodynamics (entropy is always increasing) it should now be in a state of heat death even if it has Banged and Crunched - assuming such events are true and possible - eternally. (That's the trouble with the infinite past. You can always mentally extend it further backwards.)

But the universe is not now in a state of heat death. Therefore it cannot have existed eternally. Therefore it cannot be self-existent. Something greater than the universe and outside the universe must have caused the universe to come into existence.


As you said, though, there's the possibility (if an infinite past is possible) of the eternal crunch. There's no strong argument for that in physics, a far as I've seen. The standard view is the Big Bang with a first moment (or at least a moment before which is nothing). But I wouldn't rule it out on philosophical grounds, except if we can rule it out by ruling out an infinte past.

The other danger of your argument is that it assumes current physics is entirely accurate even about a time we know a lot less about. We may have the laws very accurately now, at least for things we can now observe, but they've been finessed a good deal since Newton, because we've observed a lot more since Newton that showed he had the laws wrong. If that happens again, the whole thermodynamic picture we have may turn out to be an approximation, with the more precise laws not leading to the consequence of heat death. So we do have to be a little careful in applying our best physics in a philosophical argument. At best we can say that the current understanding of laws seems to make this impossible.

My objection to PSR is this: What is the explanation for the existence of PSR?

First of all, PSR is not merely an observation; it does not merely say "Everything we've seen so far has an explanation of some sort." It hold the force of Law. It mandates that everything has an explanation. Therefore it is a thing in and of itself.

Since it is a thing, not merely an observation or abstration, then it is subject to PSR, and must have an explanation. However, PSR governs the existence of all things. (If it didn't, then it wouldn't be useful in the Cosmological argument, because anything which preceeded PSR could then be without explanation.) Since this is the case, PSR must be self-existent. But there is nothing in the nature of PSR that indicates that it is its own explanation. Furthermore, not only must PSR be self-existent, it must also be the first of all self-existent things. This is certainly not going to be acceptable to any Theists for whom the first of all self-existent things must be God.

Ultimately, it seems easier for me to abandon the Cosmological argument and instead take the stance that God is a brute fact--He needs no explanation; He simply IS.

If you mean what makes it true, it's just that it's metaphysically necessary, just as the laws of mathematics and logic are. That doesn't make it a thing in the sense that we are things. It just makes it a true principle. The truth that 1+1=2 isn't a thing, even if Platonism about numbers is true and 1 and 2 are things.

Theists do generally ground all necessary truths in the existence of God. I can understand thinking of necessary truths as existing in some hierarchy. For instance, if it's necessary that God create, then it seems that the necessity of creation has got to be metaphysically dependent on the necessity of God. But I get lost quickly. Some necessary truths are true simply because of logical relations. Should we think that logic is grounded in God or that God's existence is grounded in logic (such that the denial of God is contradictory, even if we can't figure out how)? Or could both be true? This is where I lose touch with any intuitions that might guide me. I don't know how to think about logical relations among logically necessary truths, because they all logically entail each other, and logical relations among metaphysical truths will all metaphysically entail each other. It makes thinking of things as logically or metaphysically prior really difficult.

Also, you don't need to think of God as a brute fact to deny this argument. You could just say that the argument doesn't establish the conclusion. You might still say that God is necessary. You can deny PSR even if you think that. I believe that's Peter van Inwagen's view, but I'm not sure. Some contemporary theists do think of God as brute, though. Richard Swinburne is probably the most influential.

There is a different objection to the cosmological argument which you don't adresss. PSR may be valid without restriction but it may be a category mistake to apply it to the universe as a whole.

What I mean is that every particular thing, process, event, or feature of the world may have an explanation, and that explanation a further one, and so on, so that "for all x, x has an explanation" is true. But there may be no real totality called "the universe" or "All" to which apply a global demand for explanation. To ask for an explanation for "the whole universe" after each particular thing has been given an explanation may be the same sort of mistake, to paraphrase Russell, as saying "yes, each human being has a mother, but who is the mother of the whole human species?"

Sorry for posting without reading first the full thread in your previous post on the subject; I see now that mine is more or less the same as Wink's objection. Your strongest answer to it is:

"Wink: I think one thing that bugs me about an infinite regress of causes is as follows. Suppose A causes B, which causes C. If A wholly causes B, and B wholly causes C, then A is really the cause of C. B only serves as an intermediate cause. So B doesn't really serve as the explanation for C, because it in turn is dependent on A. But with an infinite regress of causes, nothing is a real explanation. Nothing is a real cause of anything, in this sense. This is one of Aquinas' arguments, and I find it a lot more convincing than most philosophers today seem to."

Well, I can only say that I don´t share your intuitions about explanation. I think that B can truly and fully be an explanation of C even if it has to be explained by as well. This is clearer if we remember that normally we don´t have a linear chain of explanations as in your example. A thing Z may be explained by the conjunction of X and Y; X by the conjunction of U, V and W, and so on. So we can´t ask directly for "the explanation of the explanation of Z", but only for explanations for each of the particular elements of the explanation. (For example, if Z is an event, X a natural law and Y an initial condition, to ask for the explanation of "X and Y" as a whole would mean little; the real explanations are separate and of completely diferent kinds for X and for Y, and the true explanation of Z is just "X and Y", not the conjunction of wildly unrelated explanations for X and Y).

Wildly unrelated? If A necessitates B, and B necessitates C, then A necessitates C. I can't see how that's wildly unrelated.

I did address the illegitimate totality question in the first post, I believe. It's illegitimate to ask where the universe is if the only meaning for 'where' is locations within the universe. It's illegitimate to ask when time began if the only times are within time. This objection extends that, saying it's illegitimate to ask why the universe is if the universe is all there is and all there could be. But the way the terms have been defined this is simply not an issue at all. You are using the term 'universe' to describe what you presumably are taking to be a set of dependent things. That begs the question already. If the theist is right, then the totality of things includes God. That totality would be the totality that needs no explanation, if your charge is correct. That doesn't mean the set of dependent things needs no explanation. So I think the charge begs the question against theism. They only way it could work is if theism is false. That doesn't make the objection false. It just means it shouldn't move the theist.

If PSR is an intuitively plausible principle, as I think it is, then we ought to seek explanations for the existence of things and for any positive facts (things being a certain way). I'm not sure how your move is anything more than Rowe's, which simply says that one positive fact is unexplained. The charge you have to avoid is not appearing arbitrary in singling out one fact not to explain. I'm not sure you've done that. You want to do it in terms of seeing the universe as a whole that can't be explained by anything apart from itself, but I was never saying that the explanation had to be apart from itself. It the universe is everything there is, and theism is true, then God is part of the universe. You can't then equivocate and see the universe just as the part God created and say that that needs no explanation.

Can PSR be applied to God?

Yes, it would have to be. That's why this couldn't be an argument for any old being. It has to be an argument for a self-existent being (i.e. a being whose explanation is within its own nature). If God's nature explains God's existence, then we have a fully satisfactory answer according to PSR. If God's nature does not, then we need to figure but what explains God, and we haven't gotten a real answer.

Since traditional theism takes God to be a self-existent being, God is a plausible candidate for the self-existent being that the conclusion of this argument requires.

Leave a comment


    The Parablemen are: , , and .



Books I'm Reading

Fiction I've Finished Recently

Non-Fiction I've Finished Recently

Books I've Been Referring To

I've Been Listening To

Games I've Been Playing

Other Stuff


    thinking blogger
    thinking blogger

    Dr. Seuss Pro

    Search or read the Bible

    Example: John 1 or love one another (ESV)

  • Link Policy
Powered by Movable Type 5.04