Proofs for the Existence of God (#AMA with Dr. Edward Feser)
EDITOR'S NOTE: Dr. Edward Feser just released a new book, titled Five Proofs of the Existence of God (Ignatius Press, 2017). You probably know Dr. Feser from his sharply reasoned posts here at Strange Notions, or from his popular blog, which mainly focuses on the philosophy of religion.
Dr. Feser has written several other excellent books, including:
- Neo-Scholastic Essays (St. Augustine's Press, 2015)
- Scholastic Metaphysics: A Contemporary Introduction (Scholastic Editions, 2014)
- The Last Superstition: A Refutation of the New Atheism (St. Augustine's Press, 2010)
- Aquinas (A Beginner's Guide) (Oneworld Publications, 2009)
- Philosophy of Mind (A Beginner's Guide) (Oneworld Publications, 2006)
He is a Thomistic philosopher, meaning he specializes in the thought of St. Thomas Aquinas, and has written extensively on Aquinas' Five Ways (or five proofs) to God. But in his new book, he examines not just the Thomistic arguments for God, but several more. Here's a brief summary:
Five Proofs of the Existence of God provides a detailed, updated exposition and defense of five of the historically most important (but in recent years largely neglected) philosophical proofs of God's existence: the Aristotelian proof, the Neo-Platonic proof, the Augustinian proof, the Thomistic proof, and the Rationalist proof.
This book also offers a detailed treatment of each of the key divine attributes—unity, simplicity, eternity, omnipotence, omniscience, perfect goodness, and so forth—showing that they must be possessed by the God whose existence is demonstrated by the proofs. Finally, it answers at length all of the objections that have been leveled against these proofs.
This book offers as ambitious and complete a defense of traditional natural theology as is currently in print. Its aim is to vindicate the view of the greatest philosophers of the past—thinkers like Aristotle, Plotinus, Augustine, Aquinas, Leibniz, and many others—that the existence of God can be established with certainty by way of purely rational arguments. It thereby serves as a refutation both of atheism and of the fideism which gives aid and comfort to atheism.
We recently invited Dr. Feser to do an #AMA (Ask Me Anything) here at Strange Notions, and after he accepted, the questions poured in from all of our commenters, both believer and skeptic alike.
We chose several of the most popular questions to ask Dr. Feser below. Enjoy!
QUESTION (Jonathan): I read once on a blog post that the proofs for God were not intended as rhetorical or polemical proofs, in the sense of being intended to persuade unbelievers. They were more like edifying exercises for the faithful, but medieval theologians would not say that such philosophical arguments were sufficient to instill faith. Is this true?
DR. FESER: That is not true, and I suspect that the writers you read who said this misunderstand what “faith” means for a medieval theologian like Aquinas. The proofs were indeed meant to be completely rationally convincing even to someone who is initially coming to the question as an atheist. No faith is required at all.
The reason is that faith, as a thinker like Aquinas understands it, is a matter of believing something because it has been revealed by God. But before you can do that, you first have to establish that God really does exist in the first place and that he really has revealed something. And that requires evidence and argumentation.
Showing that God really does exist is where the proofs come in. So far, faith doesn’t enter the picture. Then we need to establish that God really has revealed something, and that involves showing that some purported divine revelation was associated with a miracle, because only a miracle – understood as a suspension of the natural order that only God could possibly bring about – could justify the claim that a revelation has really occurred. That requires a mixture of philosophical and historical argumentation. The traditional label in Catholic theology for these sorts of arguments for the authenticity of a revelation are “motives of credibility.”
In traditional Catholic apologetics, it is only after all this argumentation is set out that one can know that something really has been revealed, and so it is only then that the question of faith really arises. And when it does, what it means, again, is believing something because you have rationally come to know that God really did reveal it. It is not a matter of believing something just because you want to, or working yourself up into an emotional state, or taking an irrational leap beyond the evidence, or anything like that.
It is true that faith is said to be a gift of God, but in no way does that entail any sort of irrational “will to believe” or any of the other caricatures or distortions of the concept of faith. If someone says that his eyes are a gift from God, he isn’t saying that he is relying for his vision on a will to believe, or on an irrational leap beyond the visual evidence, or the like. Similarly, to say that faith is a gift of God in no way implies that it is contrary to reason or involves an irrational leap beyond the evidence.
QUESTION (Bradley Robert Schneider): To what extent are the arguments in your book just different versions of, or different ways of looking at, the same (cosmological) argument? That is, can you rationally reject one of the proofs but accept another? Also, what are some of the other arguments you consider persuasive but did not include among these five?
DR. FESER: Of the five, only four of them – the Aristotelian proof, the Neo-Platonic proof, the Thomistic proof, and the rationalist proof – might be considered variations on the cosmological argument. However, the expression “cosmological argument” might be a little misleading, because it makes it sound as if the arguments start from some claim about the cosmos or universe as a whole. And that is not the case. As I argue in the book, one could in each case start with something much less grand than that. For example, in the Thomistic proof, you could start with the fact that some particular stone exists here and now, and proceed from that to show that there must be a single eternal, immaterial, immutable, necessary, omnipotent, omniscient, perfectly good uncaused cause of its existing here and now. No fancy claim about the universe as a whole is needed as a premise.
The fifth argument, the Augustinian proof, is very different. It is not a causal argument from the existence of some thing in the world to God as its cause, but rather an argument from mathematical and other necessary truths to an infinite divine intellect.
One could accept one of the proofs while rejecting one or more of the others. For example, a reader might find the Aristotelian proof compelling but not be persuade by the Neo-Platonic proof, or might find the Augustinian proof powerful but not any of the causal proofs, or vice versa.
There are several other arguments I think are compelling but which I did not put into the book, because I judged that they required just too much in the way of controversial background metaphysical argumentation to be useful for the particular purposes of this book. For example, I think that all of Aquinas’s Five Ways are sound arguments, and I have defended them all in various other writings. But to defend the Fourth Way (for example) requires first defending so much in the way of background metaphysical theses well beyond what I already cover in the book that it just isn’t a suitable argument for the kind of audience I intend to address in the book. In the Further Reading section of the book I direct readers to sources that defend the other arguments that I think are persuasive.
QUESTION (Doug Shaver): Regardless of one's worldview, any proof must, by logical necessity, rest on one or more assumptions, which are premises that are stipulated to be unprovable. Can Dr. Feser list some of the assumptions on which at least one of his proofs of God's existence depends?
DR. FESER: Depending on what you mean by “unprovable,” I don’t necessarily agree with the premise behind your question. Take, for example, the principle of sufficient reason, which the rationalist proof appeals to as a premise. Is it “unprovable”? That depends on what you mean. If by “provable” you mean “derivable by deductive inference from premises that are more certain or fundamental,” then no, it is not provable. But if by “provable” you mean “defensible by arguments which any rational person ought to find compelling,” then I would say yes, it is provable.
The reason is this. The principle of sufficient reason, correctly understood – and as I argue in the book, lots of people don’t understand it correctly – is a “first principle.” What that means is that it is more clearly correct than anything that could be said either for it or against it. When we reach claims like this, we’ve hit bedrock. They are so basic to rationality that they are presupposed by other arguments, but don’t in turn presuppose anything deeper than they are themselves. Hence we can’t defend them by the same kind of reasoning by which we defend less fundamental claims.
But that does not by any means entail that we cannot say anything in defense of them, or that they rest on faith, or an act of will, or anything like that. We can rationally defend them in an indirect way. We can show, for example, that objections to them are mistaken in various ways. More importantly, we can defend them by the method of retorsion, which involves showing that one cannot deny them on pain of self-contradiction or incoherence.
This method is sometimes misunderstood. Some people think it merely involves showing that we can’t help thinking a certain way, but where this leaves it open that this way of thinking might nevertheless not correspond to reality. In other words, they think that retorsion arguments are essentially about human psychology. That is not at all the case. Rightly understood, such arguments are a species of reductio ad absurdum argument. They involve defending a claim by showing that the denial of the claim entails a contradiction, and thus cannot as a matter of objective fact (and not merely as a contingent matter of human psychology) be correct.
QUESTION (Surroundx): When you speak about proofs of God, what epistemic status do you ascribe to the conclusion of each? Are they epistemically infallible, ontically infallible, or something else?
DR. FESER: The word “proof” has, historically, been used in different senses. Naturally, I don’t mean that the arguments are proofs in exactly the same sense in which a mathematical proof is a “proof.” They are mostly not a priori arguments, for one thing. But I used the word deliberately, and I certainly claim a high degree of certainty for the claim that God exists. For example, I would claim that it is as certain that God exists as it is that the world external to our minds is real and not an illusion foisted upon us by a Cartesian demon or the Matrix.
How can I say that? Well, the point of the book to show this. The arguments are “proofs” in that, first of all, the conclusion is claimed to follow deductively from the premises. They are not mere probabilistic inferences, arguments to the best explanation, or “God of the gaps” arguments. (I hate “God of the gaps” arguments.) The claim is that the arguments show, not merely that God is the most likely explanation of the facts asserted in the premises of the arguments, but rather that God is the only possible explanation in principle of those facts.
Second, the premises are knowable with certainty. The premises include both empirical premises (for example, the premise that change occurs) and philosophical premises (for example, the premise that everything has an explanation or is intelligible). The premises in turn can be defended in various ways that show them to be beyond reasonable doubt. For example, some of them can be defended via retorsion arguments (which, again, are a species of reductio ad absurdum argument). That is to say, such arguments try to show that anyone who denies such-and-such a claim is implicitly contradicting himself.
So in arguments of the sort I am defending, the conclusion is claimed to follow necessarily from the premises, and the premises are claimed to be knowable beyond any reasonable doubt. That sort of argument fits one traditional use of the word “proof.”
Naturally, I am aware that some people will nevertheless challenge the arguments or remain doubtful about one or more of them. But that’s true of every single argument one could give for any conclusion, even mathematical proofs. A determined and clever enough skeptic will always be able to come up with some grounds for doubt, even if the grounds are bizarre or far-fetched. That doesn’t mean that the grounds are, all things considered, going to be reasonable ones.
Anyway, my calling something a “proof” doesn’t entail that I think every reader, even every fair-minded reader, is immediately going to be convinced. What it is meant to indicate is the nature of the connection between the facts described in the premises and the fact described in the conclusion. It is a metaphysical claim, not a sociological claim. Too many people mix these things up. They think that as long as a significant number of people are likely not to agree with some argument, you can’t call it a “proof.” That just misunderstands the way the term is being used.
QUESTION (Ryan Beren): Is God definable at all? If he is, is the definition one that can be known by humans, or only by God himself? If he is not, then how else can the truth (or falsehood) of the statement "God exists" be guaranteed?
DR. FESER: It depends on what you mean. For Thomists (i.e. thinkers in the school of thought originating with Thomas Aquinas) to define something, in the strict sense of the term “define,” is always to locate it in a species of thing. That in turn requires identifying the genus the species falls under and what differentiates it from other species in the same genus. (Here I am using the terms “species” and “genus” in the broad sense in which they are historically used in logic, not the narrower sense that they later came to have in biology.)
So, take human beings, for example. The traditional Aristotelian definition of human beings is that they are rational animals. “Human being” is the species of thing we are defining, “animal” is the genus or more general class to which this species belongs, and “rationality” is what distinguishes them from other species in the same genus. (Whether this definition is correct, and what exactly it is claiming, are irrelevant to the present point. It’s just an illustration. Pick a different example if you like.)
Now, to define something in this way entails attributing metaphysical parts to it. For example, in defining human beings as rational animals, we are attributing both animality and rationality to human beings. But anything with parts requires some cause or explanation outside itself to account for why it exists. With human beings, there needs to be some explanation of how rationality and animality get together so that you have human beings. These things are not of their nature necessarily co-occurring, after all.
Now, according to arguments of the sort I am defending, God is not like that. He cannot have any parts at all, but must be simple or non-composite – that is to say, not composed of anything. For if he were, then he would require a cause just like everything else does, in which case he would not be the primary or ultimate cause of things.
For that reason, there cannot be any distinction in God between some genus he belongs to and some differentiating feature that distinguishes him from other things in that genus. Again, if there were, then God would have parts and thus require a cause of his own. Hence, strictly speaking, God is not part of a species of things, he is not in a genus, and thus in the strict sense of the term he is not definable. That is why the human mind inevitably finds God difficult to grasp. Our normal mode of understanding things is to define them in terms of the genus they fall under and what differentiates them from other things in that genus, and this method cannot apply to God.
However, that does not mean that we cannot use the term “God” intelligibly, and it does not mean we cannot know anything about God. We can say, for example, that by “God” we mean the primary or fundamental cause of there being anything at all; that when we analyze what something would have to be like in order to play such a role, it would have to have such-and-such attributes; we can note that those are precisely the sorts of attributes traditionally attributed to God; and so forth.
In short, if you mean “Can we use language about God intelligibly, so that we can discuss the question whether God exists, what he would be like if he exists, etc.?” the answer is Yes. If you mean “Can we have the kind of penetrating knowledge of God’s nature that we have when we are able to identify the genus a thing falls under and what differentiates it from other species in that genus?” then the answer is No. Again, it depends on how you define “define.”
QUESTION (Alexander): Is it conceivable that God does not exist?
DR. FESER: It depends on what you mean. If you mean “Can we coherently form the thought that there is no God?” then yes, we can do that. To be sure, if we had a complete and penetrating grasp of God’s nature, we would understand that, given that nature, it is metaphysically impossible that he not exist. In that case we would be contradicting ourselves if we said he did not exist. We would see that it is inconceivable that he not exist. However, we mere human beings do not in fact have such a complete and penetrating grasp of the divine nature. Hence we are unable directly and immediately to see the inconceivability of God’s non-existence and thus cannot deduce God’s existence via an ontological argument. We have to arrive at knowledge of God in another way.
Now, when we reason to the existence of an uncaused cause of things as we do in arguments like the kind I defend in the book (the Aristotelian proof, the Thomistic proof, and so forth), and then we analyze what something would have to be like in order to play that role, we find that it must be something that of its nature exists in an absolutely necessary way. We can then deduce that it must be the kind of thing which, if we had a complete grasp of its nature, we would see directly that it could not possibly not exist. But we had to get there by reasoning from the existence of things in the world to God as their cause. We can’t skip this procedure and just cut to the chase via an ontological argument, as Anselm tries to do.
So, if you mean “Is God the sort of thing that exists of absolute necessity, so that it is metaphysically impossible that he not exist?” then the answer is Yes. But if you mean “Can we reason to God’s existence just by carefully unpacking the content of the concept of God, as in Anselm’s ontological argument?” then the answer is No.
QUESTION (Steven Dillon): Monotheism asserts the proposition that "Only one God exists." In quantifying the amount of Gods that exist, this proposition treats of a plurality of "Gods." In denying existence of all but one in this plurality, monotheism separates Gods from "existence", and thus treats of a plurality of abstractions, or "essences" as Thomists may say. It would seem, therefore, that monotheism is committed to a view on which a God's essence is separable from his "existence." But, for Aquinas, the essence of God just is his existence. Was Aquinas thus not a monotheist? If not, what was he?
DR. FESER: Aquinas is a monotheist, and he argues – correctly in my view, as I argue in the book – that there could not even in principle be more than one God. One of the reasons for this is indicated in the answer I gave above to the questioner who asked about whether we can define God. As I noted there, given that God is absolutely simple or non-composite, he cannot be defined in terms of a genus and some differentiating feature that sets him apart from other species in the genus. Now, whenever there is more than one instance of a kind of thing, there is some genus to which it belongs, and something that differentiates it from other things in that genus. Since these notions don’t apply to God, it follows that there is no way for him to be merely one instance of a kind of thing. There is no genus or general class to which he belongs. He is of his nature unique. We get the same result when we analyze the implications of something’s being purely actual, or its having an essence that is identical to its existence, as I show in the book.
Your question seems to suppose that because we can stick an “s” at the end of the word “God,” that suffices to show that there is a general class of things we call “Gods,” and then we can ask how many things are in that class. But that is a fallacy. Essentially, it confuses grammar with metaphysics. To borrow an example from Chomsky, I can form the sentence “Colorless green ideas sleep furiously.” But though the sentence is perfectly well-formed, it is still nonsense. Ideas aren’t green or any other color, if they were then they wouldn’t be colorless, they don’t sleep, and it makes no sense to speak of sleep as something one could do furiously. Mere grammatical possibilities don’t by themselves entail anything about reality.
Similarly, we can stick an “s” on the end of the word “God” and then go on to ask questions like “How many Gods are there?” But that doesn’t entail that it really makes sense to think of “Gods” as a class of things that might in theory have more than one member. Again, in fact that makes no sense when we unpack the implications of what it is to be absolutely simple or non-composite, to be purely actual, and to have an essence identical to one’s existence. It seems otherwise only if we confuse grammar with reality.
QUESTION (Paul Brandon Rimmer): There has been a new movement in Christian apologetics arguing for idealism, the idea that all of reality is made of minds, and that ultimately there is no such thing as matter. Is this consistent with Thomism? Is this consistent with Catholic doctrine? In each case, if not, what would have to be changed about Thomism/Catholic doctrine, in order to make room for idealism?
DR. FESER: It is not consistent with Thomism. One reason is Thomism’s Aristotelian conception of matter as what limits form to a particular individual time and place and thus individuates instances of a species of thing. If you have two or more stones, for example, then you need, in addition to what they have in common – the form of being stone – something to differentiate them, and that’s what matter does. Different bits of matter instantiate the same form. For the Thomist, then, it makes no sense to say “There are two stones, and neither one is material.” Berkeley can say that (given his different conception of matter), but Aquinas cannot.
Furthermore, to get rid of matter you’d really be reducing all of reality to a collection of angelic minds – that is to say, minds which are of their very essence divorced from matter. That would mean identifying human minds with angelic minds of a sort. You’d be saying that we are essentially really minds without matter, and what seems to be the material world is really just a collection of our perceptions. But for the Thomist, that cannot be right, because our minds are simply not like angelic intellects, as Aquinas understands them. For example, angelic intellects don’t have a stream of sensory experiences, as we do. Sensation is, for the Thomist, essentially bodily, so that what lacks a body lacks sensation. For that reason, angels don’t acquire knowledge the way we do, by learning things from a series of sensory perceptions. Their knowledge is “built in.”
Neither is idealism compatible with Catholicism, for reasons that might be evident from what has been said already, because Catholic doctrine is deeply committed to the reality of the material world. That is, for example, what the doctrine of the Incarnation is all about. The Second Person of the Trinity became flesh in Jesus of Nazareth, and if you say that the flesh was really just a collection of perceptions, it is hard to see how that avoids collapsing into Docetism.
There is also the fact that modern idealism stems from general metaphysical and epistemological premises that Thomists regard as deeply mistaken. Berkeley’s idealism is a byproduct of the modern empiricist reduction of concepts to mental images. Leibniz’s idealism is a byproduct of his working within the Cartesian dichotomy of res cogitans and res extensa. But for the Thomist, these are just bad starting points, and in particular they get badly wrong both the nature of substance and the nature of our knowledge.
This is a large topic, and much more could be said. Suffice it to say that the divergence between the views is so deep that there is no way to reconcile them, and for the Thomist there is no good reason to want to reconcile them, since idealism is (the Thomist would argue) riddled with philosophical and theological errors.
QUESTION (Camainc): How do you square divine simplicity with the personal God of the Hebrew and Christian scriptures? Are the scriptural references to God changing his mind, getting angry, etc., just anthropomorphic?
DR. FESER: It depends on what you mean. Take anger, for example. If by anger you have in mind the way that we human beings go from a state of tranquility to a state of emotional agitation, then no, there is nothing like that in God. The reason is that, for one thing, emotional states are bodily and God is immaterial. For another thing, emotional fluctuations entail going from potential to actual, and there is no potentiality in God. However, by anger you could mean the will to inflict a punishment on an evildoer. And there is something like that in God.
Here it is crucial to understand the differences between the univocal, metaphorical, and analogical uses of language, which I discuss in the book. When we use terms univocally, we are using them in the same sense. For Thomists and many other classical theists, theological language is not to be understood univocally. Hence, when we say that Bob is angry and God is angry, we are not to be understood as attributing to God the exact same thing that Bob has.
Are we speaking metaphorically or non-literally, then? That depends. If when saying “God is angry,” you mean that God feels highly agitated, then this cannot literally be true, and so at best could be a metaphor.
But if instead you mean “God intends to punish evildoers,” then this is not metaphorical, but literally true. However, not all literal language is univocal. Some of it is analogical. For example, when I say that the cheeseburger I am eating is good and that the book I am reading is good, I am not using “good” in exactly the same sense – the goodness of a book and the goodness of food are very different – but I am not speaking metaphorically or non-literally either. Rather, I am speaking analogically. There is something in the goodness of a book that is analogous to the goodness of food, even if it is not the same thing.
Now, that, for the Thomist, is how to understand a claim like “God intends to punish evildoers.” There is something in God that is analogous to what we call an intention to punish evildoers, even though in God it doesn’t involve the kind of thing that goes on when we have this intention. (For example, we have fluctuating emotional states, we weigh various considerations when deciding whether to punish – which involves going from one thought to another, and thus actualizing potentialities – and so on, and none of that exists in God.)
So, some of the anthropomorphic attributes that scripture attributes to God are to be understood metaphorically, but by no means all of them are. Some of them are to be understood in an analogical sense, which is a kind of literal sense. The rule of thumb would be: If scripture attributes some bodily characteristic or emotional state to God, that is to be understood merely metaphorically or non-literally. But if scripture attributes to God something having to do with intellect or will, then that is to be understood literally, though in an analogical sense (which is one kind of literal sense) rather than in a univocal sense.
QUESTION (Brian Seets): I am an atheist. I choose to do what I see as good. That is, I do nothing that could have a negative impact on others. I do this because I want to live in a world where that is the standard behavior and so that (I hope) others will choose to not negatively impact my life. What makes my choice less worthy than a Christian's? Doesn't living well without hope of reward or fear of punishment make God irrelevant?
DR. FESER: Unlike some other theists, I don’t myself think it is quite correct to say that morality could have no foundation on an atheistic conception of reality. It’s a little more complicated than that. What I would say is that the possibility of morality presupposes the reality of what Aristotelians call formal and final causes. We have to be able to say that there is, as a matter of objective fact, such a thing as the nature or essence of a human being, and also that there is, as a matter of objective fact, such a thing as a set of final causes or ends or goals inherent in human nature, the realization of which defines what is good for us.
Now, in theory someone could accept this much while at the same time denying the existence of God. For example, Thomas Nagel at least flirts with something like this metaphysical position in his book Mind and Cosmos. For the Aristotelian and the Thomist, morality is grounded in human nature, and human nature would still be what it is, and still be knowable, even if per impossibile there were no God. Hence, just as you can do chemistry and thereby discover the facts about the causal properties of sulfur, phosphorus, etc. whether or not you affirm the existence of a divine first cause, so too can you, at least to a large extent, know what is good or bad for human beings just by studying human nature. For morality is not about arbitrary divine commands, but is, again, grounded in human nature.
However, it is also true that this is an unstable position. Just as, for the Thomist, causality is only ultimately intelligible if there is a divine uncaused cause (for reasons I set out in Five Proofs), so too, final causality ultimately makes sense only if there is a divine intellect which directs things toward their natural ends or goals (for reasons set out in Aquinas’s Fifth Way, which I have defended at length in a couple of places, though it’s not an argument that is covered in Five Proofs). Still, the issues can be distinguished. The question “Is there final causality in nature?” is one thing, and the question “Does all final causality presuppose the existence of God?” is another. If you could defend a Yes answer to the first question while at the same time defending a No answer to the second – and some people have, historically, tried to do this (even if, in my view, such a project at the end of the day won’t work) – then you could have a foundation for morality while avoiding theism.
But very few atheists are willing to do this. Most of them reject the whole idea of formal and final causality, along with theism. That is part of the general package of modern philosophical naturalism, even if in theory one could adopt some alternative atheist metaphysics. And if you reject the whole idea of final causality or teleology, then I think you will not be able to give any rational foundation for morality. This isn’t a topic I get into in Five Proofs, though I have addressed it elsewhere. (See my book Neo-Scholastic Essays for treatment of some of the issues I’ve been referring to, such as the foundations of morality and Aquinas’s Fifth Way.)
Now, does that mean that someone who is both an atheist and rejects the whole idea of final causes or purposes in human nature will in fact be without any moral virtue? Of course not. Many atheists have many admirable character traits. But that is not the point. The question is not whether atheists will in fact sometimes do the right thing – of course they will – but rather whether they can give a rational philosophical justification of morality in the context of a naturalistic metaphysics. And that, I would argue, is not possible.
Note: Our goal is to cultivate serious and respectful dialogue. While it's OK to disagree—even encouraged!—any snarky, offensive, or off-topic comments will be deleted. Before commenting please read the Commenting Rules and Tips. If you're having trouble commenting, read the Commenting Instructions.