The Arguments for God's Existence Fail

Undoubtably, the most persuasive arguments for God's existence are the Kalam cosmological argument (which suggests that God caused the universe) and the fine-tuning theological argument (which suggests that God designed it). Now, in truth, even if our universe needs a cause or designer, it need not necessarily be the perfect God of traditional theism. Still, establishing the existence of such a cause or designer would go a long way towards vindicating theism. As we shall now see, however, both arguments fail to do so.

The Kalam Cosmological Argument

The Kalam cosmological argument was originally formulated to make up for certain deficiencies in other cosmological arguments.Footnote ⁷ It gets its name from the ʿIlm al-Kalām, a Muslim school of thought, or ‘science of discourse’, used to defend Islam. One of its practitioners, the eleventh century's Al-Ghazālī, found fault with the theistic arguments of a competing school of thought: the falāsifa, which was inspired by Greek philosophy (or falsafa). The falāsifa argument suggested that God must exist because all material things depended on the existence of a necessary entity (an entity that must exist). Even if that's true, Al-Ghazālī realized, the argument provided no reason that necessary entity couldn't just be the universe itself, rather than God (or Allah).

Similar worries arise about the later arguments of Thomas Aquinas, who argued (for example) that there must be a God because there must be an uncaused causer that lies at the ‘beginning’ of the causal chains that exist in the world.Footnote ⁸ The problem is, even if his arguments work (which itself is debatable), why must the thing they suggest exists be God? Why couldn't the universe itself be Aquinas’s uncaused causer? After all, wouldn't that be the simpler explanation?

Modern versions of the Kalam argument, obviously inspired by the arguments of Al-Ghazālī, essentially set out to answer that question. Why can't the universe be the uncaused causer, but God can? Because the universe began to exist; God did not. This, it seems, is a relevant difference between the universe and God that would imply that the former must have a causal explanation, but the latter need not. Things that begin to exist need causes. Things that don't, do not.

Modern versions of the Kalam argument go like this:

1. (1) Whatever begins to exist has a cause of its existence.

2. (2) The universe began to exist.

3. (3) Thus, the universe must have a cause of its existence (from 1 and 2).

4. (4) God (if he exists) did not have a beginning.

5. (5) Thus, God (if he exists) does not need a cause (from 1 and 4).

6. (6) If the universe must have a cause, but God need not, then God could be the uncaused causer but the universe cannot be.

7. (7) Thus, God could be the uncaused causer, but the universe cannot be (from 3, 5, and 6).Footnote ⁹

This argument is also deficient,Footnote ¹⁰ but it's a little harder to explain why. To begin, consider what it means for something to begin to exist. As examples, those who present this argument often point to ordinary everyday objects, like chairs; they even use them as evidence for the first premise. But what does it mean for something like a chair to come into existence? It means for already existing matter (e.g. wood and screws) to become arranged in such-and-such a way. No new matter comes into existence. Yet matter coming into existence is exactly what happened when the universe began to exist; and that's what premise 2 is about. So the argument equivocates on the word ‘begin/began’. In premise 1, ‘begin’ refers to ‘matter being arranged’. In premise 2, ‘began’ refers to matter coming into existence. And arguments that equivocate are invalid.

The mistake becomes obvious if we disambiguate these premises.

• (1*) When existing matter gets arranged to form a new object, there is a causal explanation for how the object's matter came to be arranged as it is.

• (2′)  The matter of the universe came into existence.

• (3)   Thus the universe must have a cause.

Premises 1* and 2′ have nothing to do with each other; one is about ‘matter arrangement’, the other is about ‘matter coming into existence’. Premise 3, therefore, does not follow.

Now, of course, one can change the first premise so that it is also about matter coming into existence; this will allow us to make a valid derivation.

• (1′) When matter comes into existence it must have a cause of its existence.

• (2′)  The matter of the universe came into existence.

• (3′)  Thus when the matter of universe came into existence, it must have had a cause.

But now the derivation begs the question; it assumes the truth of what it is trying to prove. Why? Because the one and only time that matter came into existence is when the universe came into existence, so the only way one can accept the first premise (that when matter comes into existence it has a cause) is if one already accepts the conclusion: that the matter of the universe has a cause.

To make the mistake clearer, consider this argument.

1. (1) Anytime Hillary loses an election to Donald, she will win the popular vote.

2. (2) Hillary lost to Donald in 2016.

3. (3) Thus, when Hillary lost to Donald in 2016, she won the popular vote.

The only way that you would accept premise (1) as true is if you already knew that (3) was true. (Indeed, you would have to know that (3) was true, and that 2016 was the only time Hillary ran against Donald.) So the argument begs the question. In the same way, the only way that you could believe that ‘whenever matter comes into existence it needs a cause’ is if you already accepted that the matter of the universe must have a cause. So the Kalam argument begs the question.

Now the scientifically adept theist might rightly point out that, technically, the beginning of the universe was not the only time that matter came into existence. On the quantum level, it's common for electron, positron, and photon trios to randomly come into existence, from nothing, and then annihilate one another. It's called a ‘vacuum fluctuation’. But this will hurt the theist's cause, not help it, because it shows that the Kalam cosmological argument's first premise is false. Since these events are truly random, matter can come into existence without a cause. Worse still, Ed Tryon has convincingly argued that our universe has all the properties it would if it were merely the result of a large-scale vacuum fluctuation (Tryon Reference Tryon1973: 396–7). So, scientifically speaking, the universe actually could be ‘one of those things which [randomly] happen [uncaused] from time to time’ (Tryon, Reference Tryon1973: 397). The Kalam cosmological argument's attempt to show that this is not the case is a failure.

The Fine-Tuning Argument

The fine-tuning argument invokes God, not as an explanation for the universe's existence, but for its design. The basic idea is this: the laws and constants that make life possible in this universe seem to be balanced on a razor's edge for life; if any one of them were slightly different from what it is, life could not exist. And this is much more likely on theism than atheism. So, God probably exists.

The list of supposedly fine-tuned laws and constants is extensive, but a few are worth mentioning:

• • [I]f the electric charge of the electron had been only slightly different, stars either would have been unable to burn hydrogen and helium, or else they would not have exploded. (Hawking Reference Hawking1988: 125)

• • If the mass of neutrinos were 5 × 10⁻³⁴ kg instead of 5 × 10⁻³⁵… the additional gravitational mass would result in a contracting rather than expanding universe. (Davis Reference Davis1987: 140–1)

• • If the ratio of electrons to protons differed by one part in 10³⁷, then life would be impossible. If it were larger, electromagnetism would dominate over gravity and galaxies would not form. If smaller, gravity would dominate and chemical bonding would not occur. (Stenger Reference Stenger2009: 92)

No stars? No life. Contracting universe? No life. No galaxies? No life. No chemical bonding? No life. The laws and constants of the universe have exactly the only values they must have for the universe to produce life. The universe, it seems, is an engine, fine-tuned for life's creation – or so the argument goes.

Despite how convincing this argument seems at first, it is monumentally flawed. There are essentially five problems, each of which derails the argument completely.

First, if the universe is fine-tuned for anything, it is not for the formation of life. The vast majority of the universe is completely hostile to life, and where it does exist, it only exists briefly. Less than 5 per cent of the universe is normal (atomic) matter (the rest is dark energy and dark matter).Footnote ¹¹ Of that 5 per cent, most of it is free hydrogen and helium; and the rest is mostly stars. In the end, only 0.03 per cent of the universe comprises heavy elements (of which planets are composed), and only a tiny fraction of planets (perhaps only one!) house life.Footnote ¹² If the universe is an engine fine-tuned for the creation of life, it is practically the most inefficient such engine possible. So, if the universe has a designer, that designer either cares nothing for life or is monumentally incompetent. Either way, it cannot be the God any theist believes in.

Second, the argument is fallacious, committing what I call the ‘only one way to make a square’ fallacy. Consider a square with sides each 1 foot in length. It's true that, if any one side of that square differed in length at all, it would no longer be a square. But from this fact no one would ever conclude that all squares have sides 1 foot in length – that the sides of the square in question are the one and only length they must be to form a square. Obviously, there are a great many ways to make squares. If you started varying the size of the square's sides randomly, quite often you would get another square – just one of a different size. To think otherwise commits an obvious logical mistake.

Yet this is exactly the logic that the fine-tuning argument utilizes. ‘If any one [law or constant] were slightly different than it is, life could not exist.’ Even if that's true, it does not follow that ‘The laws and constants of the universe have exactly the only values they must have for the universe to produce life.’ Changing another value might balance things back out, so that life could still form. And, indeed, when Victor Stenger allowed all the values of the laws and constants to be varied randomly in simulations, more than 50 per cent of the time he got universes that were conducive to life.Footnote ¹³ If there is a radio station on every other frequency, it is no coincidence when you spin the dial randomly and land on a station.

Third, the fine-tuning argument presents no reason for thinking that the values of the laws and constants it mentions could be different from what they are. Yes, we can assign them a number and then think of a different number, but perhaps they are what they are necessarily. Indeed, many of them couldn't be different. For example, the aforementioned ratio of electrons to protons is a result of the conservation of energy, a fundamental property of space time (which simply states that energy cannot be created or destroyed). The same is true for the mass density of the universe and the ratio of neutrinos to their mass.Footnote ¹⁴ Indeed, if neutrinos were heavier than 5 × 10⁻³⁵, there would a directly proportional smaller number of them. Their cumulative mass would therefore be the same, and the expansion rate of the universe would be unaffected.

But fourth, even if the mass of neutrinos (or similar small particles) could be different, their mass would still not be fine-tuned. The fine-tuning argument expresses their mass in kilograms, but the kilogram is a unit we use to measure the weight of full-grown humans. Neutrinos are so small that they pass through normal matter, so, of course, any change in their mass expressed as a fraction of a kilogram is going to have a monumental effect. Indeed, changing it from 5 × 10⁻³⁵kg to 5 × 10⁻³⁴ kg, as the argument suggests, would increase their mass exponentially, by tenfold! But, in reality, their mass could vary wildly from what it actually is, and there wouldn't be any dramatic effect.

To illustrate: suppose I said that Michael Jordan was the one and only height he had to be to play basketball because he's 2 × 10^−16th of a light year tall; if he were 2 × 10^−17th he'd be too short, and if he were 2 × 10^−15th he'd be too tall. These figures are all true – but at 2 × 10^−17th he'd be 18 cm tall, and at 2 × 10^−15th he'd be 59 feet! In reality, of course, Jordan's height could differ significantly from what it is, and he would probably still be the greatest player to ever play. But by using a disproportionate unit of measure, I've ruled out ‘moderate’ fluctuations in his height, and made Jordan's height seem much more ‘fine-tuned’ than it is.Footnote ¹⁵

A similar point can be made when proponents of the fine-tuning argument say that if some factor (like the mass density of the universe) had differed by 1 part in 10⁵⁹, life could not have evolved (Stenger Reference Stenger2009: 91). The exponent's value depends solely on how finely you divide out, or measure, what you are calling ‘parts’. If the parts are bigger, that number is smaller and less impressive. And there is no objective criterion by which to divide up or measure such ‘parts’.

And that brings me to the fifth problem with the fine-tuning argument: the probabilities it expresses are either meaningless or worthless. Why? The fine-tuning argument is only impressive if the probability of the laws and constants having the ‘right’ value is low; but the assumptions one has to make to meaningfully determine such probabilities are completely arbitrary. How so?

Assume that the value of some particular constant of our universe is X. What are the odds that it would have that particular value? One is tempted to say that the odds are 1/∞ (‘one in infinity’) because there's an infinite amount of numbers, and presumably any one of those numbers could have represented the constant's value. But 1/∞ is a meaningless probability. When put in terms of fractions, the denominator of a probabilistic statement must be a whole number greater than 0 (e.g. 1/2, 99/100, etc.).Footnote ¹⁶ So, in order for a such a statement to be meaningful, we must restrict, or ‘normalize’, the values which that constant could have to some finite range. But how narrowly the range is restricted will determine everything about how ‘impressively low’ the probability in question is. Since any decision about how to restrict the range can only be made arbitrarily, any statement about how likely it is that some constant had the ‘right’ value will be worthless.Footnote ¹⁷

What's more, probabilities are usually not evenly distributed, or fair – as they are with dice, where each outcome is equally probable. Usually, natural factors create a bias that makes certain outcomes more likely than others; a bell curve describes how often certain outcomes would result in the long run, and thus how probable those outcomes are. But unless we know how flat/curved the bell is – and where on the curve the actual value lies – we can't say how likely the actual value is. And, again, any such declaration would be completely arbitrary. So any declaration like ‘the probability that such-and-such constant would have the right value is low’ is either meaningless or worthless.

So both of the most ‘persuasive’ arguments for God's existence fail. The Kalam cosmological argument either equivocates or begs the question. The fine-tuning argument is logically fallacious and misstates (in multiple ways) how unlikely ‘fine-tuning’ is. But as we shall now see, not only do the arguments for the existence of God fail, but the main argument against God's existence is successful.

The Problem of Evil Succeeds

Perhaps the most common reason philosophers give for not believing in God is the presence of evil in the world. Rape, hunger, poverty, animal suffering, human misery, the holocaust … the list goes on. ‘If God exists, such evils would not exist’, the argument goes. ‘The fact that such evils do exist seems to be direct evidence that God does not.’

Again, usually the argument is directed at the traditional tri-omni God of traditional theism. Truth be told however – regardless of whether he is perfect in every way – the god that any theist believes in would desire, be able, and know how to prevent the evil that exists in this world. Since that evil clearly exists, it seems that no version of God that any theist believes in exists. Formally, we could express the argument as such:

1. (1) If God existed, he would desire, be able, and know how to prevent the evil of the world.

2. (2) If God desired, was able, and knew how to prevent the evil of the world, the evil of the world would not exist.

3. (3) But the evil of the world does exist.

4. (4) Thus, God does not exist.

This argument is valid; if the premises were true, the conclusion would follow. But many philosophers, including some atheists, argue that premise (2) is false. Why? Because there might be something else that God desires even more than the elimination of evil, but that he can't have unless evil exists.

What might that be? The most common suggestion is free will – or, more precisely, the moral good that can only exist if free will does. A person can get moral credit for doing the right thing only if they do so of their own accord, by exercising their free will. But for an agent to truly be free to choose to do some action, it must be possible for that agent to not do that action – to do evil. And if it is possible for agents to do evil, inevitably some (many!) will. And since the existence of moral good is more important than the elimination of evil, the existence of evil is something that God is willing to tolerate. Or so the argument goes. But even if we set aside the obvious problem that human free will can't be used to excuse away evil if humans don't have it (raised by my previous piece ‘Does Free Will Exist’), there are still two objections to this solution that completely derail it.

First, it entails that God is nothing like what the vast majority of theists believe God is like. On this view, the reason that God didn't, for example, prevent the Holocaust is because doing so would have required him to interfere with Hitler's free will. But if preserving free will is that important, it must be the case that God never interferes in free will. If free will is so sacred that God won't violate it to save 6 million Jews, he definitely won't violate it to, for example, make your boss give you a pay rise. In order for this solution to work, God must maintain a complete non-interference policy when it comes to free will. Yet, obviously, theists pray for things that would require God to interfere with free will all the time – from parking spaces to election results. Many theists even believe that God controls all of the world's affairs (and there is no way to do that without controlling people's choices). Justifying evil in the name of God preserving free will simply doesn't align with how most theists view God and his relation to the world.

But even if one does think that God maintains a non-interference policy, there is a second problem: not all evil is a result of human freely willed decisions. A grandiose amount of human suffering is caused by natural disasters: earthquakes, hurricanes, tornadoes, diseases, famines, genetic defects, and so forth. The suffering such things cause is called ‘natural evil’ for there is nothing that humans do to bring it about. And there is no good explanation for why God would allow natural evil.

Indeed, it's not even a question of why God would allow natural evil, but why God would be the author of it. According to theists, God is both the creator and designer of the universe – and that includes (as we saw last section) the laws that govern the universe. Yet it is those laws that cause – indeed, necessitate – natural disasters. So natural evil is a direct consequence of laws that God himself wrote. This is not only incompatible with God being all-good; it's incompatible with God even being minimally decent.Footnote ¹⁸ If I built puppy-killing machines into the walls of my house, I could hardly be said to be a loving master of my puppies. Yet that is essentially the kind of world that God (if he exists) has forced us to live in; one with human-killing machines (e.g. natural disasters) embedded into its very laws.

Some theists, like Alvin Plantinga, have tried to avoid this problem by blaming Satan and his demons; it's they, not God, who are responsible for natural disasters. But this, of course, is preposterous. And rational objections to such a view do not merely consist of ‘amused contempt or instinctive revulsion’ (Plantinga Reference Plantinga and van Inwagen2004: 16). Indeed, the rational and empirical refutation of such a view is one of the most noteworthy accomplishments of the natural sciences. We used to think that natural disasters were the work of supernatural beings; we now know they are not. Earthquakes, for example, are caused by the release of pressure built up between tectonic plates – diseases by viruses, bacteria, and genetic defects. And the existence and activity of such things are necessitated and governed by the laws of nature – not the devil and his demons.Footnote ¹⁹

In reply, the theist might insist that they can still embrace ‘sceptical theism’, the position that no evil can ever count against God's existence because, if God exists, he could have reasons for allowing evil that we simply can't understand. But (a) sceptical theism is mathematically unsound (even if God could have such reasons, evil still lowers the probability of God's existence)Footnote ²⁰ and (b) most philosophers reject sceptical theism because it renders theism unfalsifiable and thus irrational.Footnote ²¹ The problem of natural evil thus remains unanswered, and the existence of (especially natural) evil in the world provides direct evidence against theism.

You Gotta Have Faith

So, as we've seen, the arguments for a supernatural creator fail and the evil in the world seems to be direct evidence that one doesn't exist. This still doesn't explain, however, why most philosophers are atheists. Belief in God, after all, is often said to be a matter of faith. So why don't philosophers just have faith that God exists? The answer, in short, is because faith is belief without sufficient evidence and most philosophers agree with William Clifford: ‘It is wrong always, everywhere, and for anyone to believe anything on insufficient evidence.’Footnote ²² But a bit more elaboration is in order.

First of all, there are two kinds of faith – what we might call ‘simple faith’ (belief without sufficient positive evidence) and ‘blind faith’ (belief despite sufficient evidence to the contrary). Blind faith, by all accounts, is irrational – epistemically unvirtuous. If it was acceptable to just continue believing whatever you want despite all evidence and arguments to the contrary, the entire purpose and very existence of education (and philosophy itself) would be folly. Yet this is what the philosopher must do if they continue to believe in God without an answer to the problem of evil. Since no acceptable solution is forthcoming, it is no wonder that atheism is so common among philosophers.

Second of all, when it comes to existential matters (matters regarding whether something exists), the burden of proof is on the believer. If you want to believe that bigfoot exists, it's your job to provide convincing evidence of his existence. Until you do, you are not justified in believing, and I am justified in doubting. Indeed, disbelief should be embraced. And the same, of course, is true for belief in a deity. So even if we ignore the problem of evil, the failure of the arguments for God's existence, by itself, is enough to demand atheism. Since the burden of proof has not been met, belief in God by simple faith is not rational either.

In short, contrary to common opinion, faith is not a virtue. Philosophers generally recognize this and thus avoid it. If, as we've seen, belief in a deity requires faith, it should be no surprise that most philosophers avoid it. In response, some academic theists might insist that I am taking belief in God too literally. Karen Armstrong, for example, thinks religious language is largely non-literal, and sees religion as a set of practices, rather than a set of doctrines.Footnote ²³ This is a third kind of faith I call ‘mythical faith’, and I have relegated my full comments on this issue to elsewhere.Footnote ²⁴ Here I will simply say that, for the vast majority of religious believers (including pastors, priests, and popes), to admit that one does not believe in God's literal existence (to suggest that ‘God exists’ is only mythically true) is tantamount to admitting that one is an atheist. So, what I have laid out here not only explains why most philosophers are atheists, but also why academics, who are aware of what this article has laid out but who also feel they can't reject theism, embrace notions like mythical faith.