Richard Dawkins concludes chapter 3 of his book The God Delusion with the following claim:

There is a much more powerful argument, which does not depend upon subjective judgement, and it is the argument from improbability. It really does transport us dramatically away from 50 per cent agnosticism, far towards the extreme of theism in the view of many theists, far towards the extreme of atheism in my view. I have alluded to it several times already. The whole argument turns on the familiar question ‘Who made God?’, which most thinking people discover for themselves. A designer God cannot be used to explain organized complexity because any God capable of designing anything would have to be complex enough to demand the same kind of explanation in his own right. God presents an infinite regress from which he cannot help us to escape. This argument, as I shall show in the next chapter, demonstrates that God, though not technically disprovable, is very very improbable indeed.

Throughout chapter 4, which is entitled, “Why There Almost Certainly is No God,” he struggles with the view of the theologians that God is simple, as opposed to his own idea that God, if he exists, must be extremely complicated. He begins the chapter:

The argument from improbability is the big one. In the traditional guise of the argument from design, it is easily today’s most popular argument offered in favour of the existence of God and it is seen, by an amazingly large number of theists, as completely and utterly convincing. It is indeed a very strong and, I suspect, unanswerable argument— but in precisely the opposite direction from the theist’s intention. The argument from improbability, properly deployed, comes close to proving that God does not exist. My name for the statistical demonstration that God almost certainly does not exist is the Ultimate Boeing 747 gambit.

The name comes from Fred Hoyle’s amusing image of the Boeing 747 and the scrapyard. I am not sure whether Hoyle ever wrote it down himself, but it was attributed to him by his close colleague Chandra Wickramasinghe and is presumably authentic. Hoyle said that the probability of life originating on Earth is no greater than the chance that a hurricane, sweeping through a scrapyard, would have the luck to assemble a Boeing 747. Others have borrowed the metaphor to refer to the later evolution of complex living bodies, where it has a spurious plausibility. The odds against assembling a fully functioning horse, beetle or ostrich by randomly shuffling its parts are up there in 747 territory. This, in a nutshell, is the creationist’s favourite argument— an argument that could be made only by somebody who doesn’t understand the first thing about natural selection: somebody who thinks natural selection is a theory of chance whereas— in the relevant sense of chance— it is the opposite.

There follows a discussion of evolution, creation, and intelligent design. He concludes the section by stating,

A deep understanding of Darwinism teaches us to be wary of the easy assumption that design is the only alternative to chance, and teaches us to seek out graded ramps of slowly increasing complexity. Before Darwin, philosophers such as Hume understood that the improbability of life did not mean it had to be designed, but they couldn’t imagine the alternative. After Darwin, we all should feel, deep in our bones, suspicious of the very idea of design. The illusion of design is a trap that has caught us before, and Darwin should have immunized us by raising our consciousness. Would that he had succeeded with all of us.

The argument here is basically that evolutionary theory has been fairly successful in explaining living things as having resulted from a slow and detailed process in which they became increasingly complex through natural causes. Consequently Dawkins is optimistic that this manner of explanation can in principle be applied to everything else. In fact, according to him, no one has ever offered any other plausible explanation of things:

Turning Watchtower’s page, we find the wonderful plant known as Dutchman’s Pipe (Aristolochia trilobata), all of whose parts seem elegantly designed to trap insects, cover them with pollen and send them on their way to another Dutchman’s Pipe. The intricate elegance of the flower moves Watchtower to ask: ‘Did all of this happen by chance? Or did it happen by intelligent design?’ Once again, no of course it didn’t happen by chance. Once again, intelligent design is not the proper alternative to chance. Natural selection is not only a parsimonious, plausible and elegant solution; it is the only workable alternative to chance that has ever been suggested. Intelligent design suffers from exactly the same objection as chance. It is simply not a plausible solution to the riddle of statistical improbability. And the higher the improbability, the more implausible intelligent design becomes. Seen clearly, intelligent design will turn out to be a redoubling of the problem. Once again, this is because the designer himself (/ herself/ itself) immediately raises the bigger problem of his own origin. Any entity capable of intelligently designing something as improbable as a Dutchman’s Pipe (or a universe) would have to be even more improbable than a Dutchman’s Pipe. Far from terminating the vicious regress, God aggravates it with a vengeance.

He says something similar while discussing multiverse hypotheses:

It is tempting to think (and many have succumbed) that to postulate a plethora of universes is a profligate luxury which should not be allowed. If we are going to permit the extravagance of a multiverse, so the argument runs, we might as well be hung for a sheep as a lamb and allow a God. Aren’t they both equally unparsimonious ad hoc hypotheses, and equally unsatisfactory? People who think that have not had their consciousness raised by natural selection. The key difference between the genuinely extravagant God hypothesis and the apparently extravagant multiverse hypothesis is one of statistical improbability. The multiverse, for all that it is extravagant, is simple. God, or any intelligent, decision-taking, calculating agent, would have to be highly improbable in the very same statistical sense as the entities he is supposed to explain. The multiverse may seem extravagant in sheer number of universes. But if each one of those universes is simple in its fundamental laws, we are still not postulating anything highly improbable. The very opposite has to be said of any kind of intelligence.

Beginning to address the response of theologians, he says:

But what attempts have theists made to reply? How do they cope with the argument that any God capable of designing a universe, carefully and foresightfully tuned to lead to our evolution, must be a supremely complex and improbable entity who needs an even bigger explanation than the one he is supposed to provide? The theologian Richard Swinburne, as we have learned to expect, thinks he has an answer to this problem, and he expounds it in his book Is There a God?. He begins by showing that his heart is in the right place by convincingly demonstrating why we should always prefer the simplest hypothesis that fits the facts. Science explains complex things in terms of the interactions of simpler things, ultimately the interactions of fundamental particles. I (and I dare say you) think it a beautifully simple idea that all things are made of fundamental particles which, although exceedingly numerous, are drawn from a small, finite set of types of particle. If we are sceptical, it is likely to be because we think the idea too simple. But for Swinburne it is not simple at all, quite the reverse. Given that the number of particles of any one type, say electrons, is large, Swinburne thinks it too much of a coincidence that so many should have the same properties. One electron, he could stomach. But billions and billions of electrons, all with the same properties, that is what really excites his incredulity. For him it would be simpler, more natural, less demanding of explanation, if all electrons were different from each other. Worse, no one electron should naturally retain its properties for more than an instant at a time; each should change capriciously, haphazardly and fleetingly from moment to moment. That is Swinburne’s view of the simple, native state of affairs. Anything more uniform (what you or I would call more simple) requires a special explanation. ‘It is only because electrons and bits of copper and all other material objects have the same powers in the twentieth century as they did in the nineteenth century that things are as they are now.’ Enter God. God comes to the rescue by deliberately and continuously sustaining the properties of all those billions of electrons and bits of copper, and neutralizing their otherwise ingrained inclination to wild and erratic fluctuation. That is why when you’ve seen one electron you’ve seen them all; that is why bits of copper all behave like bits of copper, and that is why each electron and each bit of copper stays the same as itself from microsecond to microsecond and from century to century. It is because God constantly keeps a finger on each and every particle, curbing its reckless excesses and whipping it into line with its colleagues to keep them all the same. But how can Swinburne possibly maintain that this hypothesis of God simultaneously keeping a gazillion fingers on wayward electrons is a simple hypothesis? It is, of course, precisely the opposite of simple. Swinburne pulls off the trick to his own satisfaction by a breathtaking piece of intellectual chutzpah. He asserts, without justification, that God is only a single substance. What brilliant economy of explanatory causes, compared with all those gigazillions of independent electrons all just happening to be the same!

Note that Richard Swinburne is not the only one who thinks it too much of a coincidence that electrons are not all different and randomly changing their properties from moment to moment. David Hume, praised by Dawkins, believes the same thing. In any case, in terms of the argument here, Swinburne is exactly right. There is only one first cause, and it does indeed explain why all electrons behave in the same way. Some such thing would have to be the case in any event, but the only way the activity of electrons (or of anything else) can be understood is in relation to a final cause, the formal aspect of an efficient cause.

Dawkins however objects that such an explanation is not simple at all, but supremely complex:

Swinburne generously concedes that God cannot accomplish feats that are logically impossible, and one feels grateful for this forbearance. Having said that, there is no limit to the explanatory purposes to which God’s infinite power is put. Is science having a little difficulty explaining X? No problem. Don’t give X another glance. God’s infinite power is effortlessly wheeled in to explain X (along with everything else), and it is always a supremely simple explanation because, after all, there is only one God. What could be simpler than that?

Well, actually, almost everything. A God capable of continuously monitoring and controlling the individual status of every particle in the universe cannot be simple. His existence is going to need a mammoth explanation in its own right. Worse (from the point of view of simplicity), other corners of God’s giant consciousness are simultaneously preoccupied with the doings and emotions and prayers of every single human being— and whatever intelligent aliens there might be on other planets in this and 100 billion other galaxies. He even, according to Swinburne, has to decide continuously not to intervene miraculously to save us when we get cancer. That would never do, for, ‘If God answered most prayers for a relative to recover from cancer, then cancer would no longer be a problem for humans to solve.’ And then what would we find to do with our time?

Outraged by this idea of simplicity, Dawkins considers another example of this position:

Not all theologians go as far as Swinburne. Nevertheless, the remarkable suggestion that the God Hypothesis is simple can be found in other modern theological writings. Keith Ward, then Regius Professor of Divinity at Oxford, was very clear on the matter in his 1996 book God, Chance and Necessity: “As a matter of fact, the theist would claim that God is a very elegant, economical and fruitful explanation for the existence of the universe. It is economical because it attributes the existence and nature of absolutely everything in the universe to just one being, an ultimate cause which assigns a reason for the existence of everything, including itself. It is elegant because from one key idea— the idea of the most perfect possible being— the whole nature of God and the existence of the universe can be intelligibly explicated.”

Like Swinburne, Ward mistakes what it means to explain something, and he also seems not to understand what it means to say of something that it is simple. I am not clear whether Ward really thinks God is simple, or whether the above passage represented a temporary ‘for the sake of argument’ exercise. Sir John Polkinghorne, in Science and Christian Belief, quotes Ward’s earlier criticism of the thought of Thomas Aquinas: ‘Its basic error is in supposing that God is logically simple— simple not just in the sense that his being is indivisible, but in the much stronger sense that what is true of any part of God is true of the whole. It is quite coherent, however, to suppose that God, while indivisible, is internally complex.’ Ward gets it right here.

Important things here are the statement that “Ward mistakes what it means to explain something,” and that “he also seems not to understand what it means to say of something that it is simple.” And lastly there is Dawkins’s attempt at doing theology when he says that “Ward gets it right here.” I will return to this shortly. In any case, Dawkins continues by recounting his experiences at a conference at Cambridge:

At a recent Cambridge conference on science and religion, where I put forward the argument I am here calling the Ultimate 747 argument, I encountered what, to say the least, was a cordial failure to achieve a meeting of minds on the question of God’s simplicity. The experience was a revealing one, and I’d like to share it.

After some discussion of the background of the conference, Dawkins explains his experience with his argument against the existence of God:

For better or worse, I attended two days at the Cambridge conference, giving a talk of my own and taking part in the discussion of several other talks. I challenged the theologians to answer the point that a God capable of designing a universe, or anything else, would have to be complex and statistically improbable. The strongest response I heard was that I was brutally foisting a scientific epistemology upon an unwilling theology. Theologians had always defined God as simple. Who was I, a scientist, to dictate to theologians that their God had to be complex? Scientific arguments, such as those I was accustomed to deploying in my own field, were inappropriate since theologians had always maintained that God lay outside science. I did not gain the impression that the theologians who mounted this evasive defense were being willfully dishonest. I think they were sincere. Nevertheless, I was irresistibly reminded of Peter Medawar’s comment on Father Teilhard de Chardin’s The Phenomenon of Man, in the course of what is possibly the greatest negative book review of all time: ‘its author can be excused of dishonesty only on the grounds that before deceiving others he has taken great pains to deceive himself’. The theologians of my Cambridge encounter were defining themselves into an epistemological Safe Zone where rational argument could not reach them because they had declared by fiat that it could not. Who was I to say that rational argument was the only admissible kind of argument? There are other ways of knowing besides the scientific, and it is one of these other ways of knowing that must be deployed to know God.

There are basically three possibilities here. Either Dawkins did not understand the theologians, the theologians did not understand Dawkins, or the theologians did not understand their theology. The third possibility is very plausible given the criticism of St. Thomas by Keith Ward and Sir John Polkinghorne mentioned by Dawkins earlier. Most likely all three are the case.

Dawkins continues to what perhaps is the heart of the issue between himself and the theologians:

Time and again, my theologian friends returned to the point that there had to be a reason why there is something rather than nothing. There must have been a first cause of everything, and we might as well give it the name God. Yes, I said, but it must have been simple and therefore, whatever else we call it, God is not an appropriate name (unless we very explicitly divest it of all the baggage that the word ‘God’ carries in the minds of most religious believers). The first cause that we seek must have been the simple basis for a self-bootstrapping crane which eventually raised the world as we know it into its present complex existence. To suggest that the original prime mover was complicated enough to indulge in intelligent design, to say nothing of mindreading millions of humans simultaneously, is tantamount to dealing yourself a perfect hand at bridge. Look around at the world of life, at the Amazon rainforest with its rich interlacement of lianas, bromeliads, roots and flying buttresses; its army ants and its jaguars, its tapirs and peccaries, treefrogs and parrots. What you are looking at is the statistical equivalent of a perfect hand of cards (think of all the other ways you could permute the parts, none of which would work)— except that we know how it came about: by the gradualistic crane of natural selection. It is not just scientists who revolt at mute acceptance of such improbability arising spontaneously; common sense baulks too. To suggest that the first cause, the great unknown which is responsible for something existing rather than nothing, is a being capable of designing the universe and of talking to a million people simultaneously, is a total abdication of the responsibility to find an explanation. It is a dreadful exhibition of self-indulgent, thought-denying skyhookery.

I am not advocating some sort of narrowly scientistic way of thinking. But the very least that any honest quest for truth must have in setting out to explain such monstrosities of improbability as a rainforest, a coral reef, or a universe is a crane and not a skyhook. The crane doesn’t have to be natural selection. Admittedly, nobody has ever thought of a better one. But there could be others yet to be discovered. Maybe the ‘inflation’ that physicists postulate as occupying some fraction of the first yoctosecond of the universe’s existence will turn out, when it is better understood, to be a cosmological crane to stand alongside Darwin’s biological one. Or maybe the elusive crane that cosmologists seek will be a version of Darwin’s idea itself: either Smolin’s model or something similar. Or maybe it will be the multiverse plus anthropic principle espoused by Martin Rees and others. It may even be a superhuman designer— but, if so, it will most certainly not be a designer who just popped into existence, or who always existed. If (which I don’t believe for a moment) our universe was designed, and a fortiori if the designer reads our thoughts and hands out omniscient advice, forgiveness and redemption, the designer himself must be the end product of some kind of cumulative escalator or crane, perhaps a version of Darwinism in another universe.

We can see here what Dawkins means when he says that Ward mistakes what it means to explain something. “The very least that any honest quest for truth must have in setting out to explain such monstrosities of improbability as a rainforest, a coral reef, or a universe is a crane and not a skyhook.” Otherwise, according to Dawkins, you haven’t explained anything. And what does he mean by a crane rather than a skyhook? A skyhook, identified with what he considers a complex God, would be something that already has such complexity within itself. A crane is something simple, and simple in the sense intended by Dawkins. Explanation, therefore, according to Dawkins, requires an original simplicity, this being understood as he understands it.

In reality, attempting to explain things is to look for their causes. And correspondingly, there are different kinds of explanation and different kinds of causes. But Dawkins is identifying certain types of causality and explanation in particular, namely those that are found in Darwinian evolution. It is likely that he is doing this because he feels satisfied by such explanations, and therefore tends to think that other accounts are not real explanations, since they leave him dissatisfied. In reality, however, there are various types of explanation and thus various types of cause.

What did Dawkins mean when he said that Ward “seems not to understand what it means to say of something that it is simple”? And why does he say that “Ward gets it right here” when Ward opposes St. Thomas on the understanding of the simplicity of God?

St. Thomas asserts that God is simple in the sense that he is not composed of parts. Given his supposed activities, Dawkins considers this absurd, and thus he says that Ward gets it right when he admits that God is “internally complex.” In other words, despite believing that God does not exist, Dawkins is making the theological claim that God cannot be simple in the sense asserted by St. Thomas, but must be composed of parts.

Why does he say this? Why doesn’t he think that since he doesn’t believe in God, this is none of his concern and he should just leave it to the theologians as they apparently told him?

Dawkins is reasoning from the supposed activities of God to his nature. God is supposed to be “a being capable of designing the universe and of talking to a million people simultaneously.” Designing the universe seems to involve planning, which involves a plan, which has various parts. Talking to people seems to involve words and sentences, which are distinct from one another, and also thoughts, which seem to be distinct insofar as they are thoughts about diverse things. In other words, it is obvious that when we design and plan things, and when we speak with people, we are capable of doing so because we consist of parts. Consequently if God can do these things, he must have parts as well.

In fact, in terms of the argument for a first cause, Dawkins nearly admits that he cannot refute the argument:

Yes, I said, but it must have been simple and therefore, whatever else we call it, God is not an appropriate name (unless we very explicitly divest it of all the baggage that the word ‘God’ carries in the minds of most religious believers). The first cause that we seek must have been the simple basis for a self-bootstrapping crane which eventually raised the world as we know it into its present complex existence.

His problem is not the argument for a first cause, therefore, but the things that are typically said of that cause, and he objects to these things because they seem to him to imply that the first cause is not simple.

We already saw that Dawkins objects to the idea that God has no parts. But is this his real objection? Simply that he thinks that the first cause must be partless, and therefore that it cannot do things like designing, planning, and talking that seem to involve parts?

This is not his real objection, whether or not he understands this fact himself. For the correct response to this objection, from a theological point of view, is exactly that God is simple in the sense defined by St. Thomas. And he does not perform the activities mentioned by Dawkins in the way that he supposes. God does not pass from one thought to another. He does not think of one part of a plan, and then another. If he speaks, he does not go from word to word in his mind. To the extent that parts are implied by such things, they are to be denied of God, and the theologian only believes that they exist in God by analogy.

But Dawkins will still have a problem with this response, if it implies that God still performs those activities, even in an analogous way. If for example God ever directly produces a voice in my mind telling me to do something, Dawkins will have a problem with this, even if I say that God does not have parts. Only the voice has parts. Dawkins will still insist that this explanation is “not simple.”

And why not? Because it is not the kind of explanation that is pleasing to him, where complexity comes from simplicity, not just in the sense that a partless being causes beings with parts, but in the sense that mathematical complexity is caused by mathematical simplicity. This is ultimately what he means when he talks about a crane rather than a skyhook. If we give a mathematical explanation of the voice in my head, it will be a mathematically complex one, and if the only cause is God, it may not be clearly possible to reduce that mathematical complexity to something mathematically simple. Evolutionary explanations, on the other hand, allow something mathematically complex to be explained in terms of laws which are mathematically simple. And this is the only kind of explanation that Dawkins considers reasonable, satisfying, or true.

We can divide all of this discussion into various questions:

1. Is there a first cause at all? We have established that there is, and Dawkins does not deny it.
2. Does the first cause have parts? We have established that it does not, and in principle Dawkins does not assert that it does. To some extent he could be taken to be conceding that it does not, since his objection is that if God exists, he has many parts and is extremely complicated, and therefore cannot be the first cause.
3. Does the first cause produce mathematically complex things from mathematically simple ones? It is certain that it does in general. Our discussions of mathematical laws in nature and of the order of the world are both relevant, as well as the issue of simplicity and probability. Dawkins agrees with this, and in fact his position is that this is the only way that mathematical complexity is ever produced.
4. Does the first cause ever produce mathematical complexity without doing this through mathematically simple things? Nothing in our discussions establishes that such a thing is impossible, nor that it is actual. Dawkins denies that this is possible or at least that it is reasonable, but he does not seem to have a particular argument for this other than the fact that such a claim leaves him feeling dissatisfied, feeling that something has been left unexplained which should be explained. But as we have seen, this is not a question about the nature of explanation in general, but the kind of explanations which are pleasing to him.
5. Is a first cause which does not directly produce such mathematical complexity worthy of being called God? This is mainly a question about the meaning of words, although there also could be questions about what that being would be like. Dawkins denies that this is a reasonable way to use the word “God”, because, according to him, God is always understood to intervene directly in the world, causing things which are meaningful on a human level and consequently which are already mathematically complex.
6. Do God’s activities imply that he has parts? Dawkins assumes that they do, and apparently the theologians at the conference that he attended were unable to explain otherwise.

It is problematic to discuss the question of “whether God exists” with someone like Richard Dawkins because these separate questions end up being mixed together. Dawkins gives a negative response to question 5, but if this is in fact a reasonable way to use the name “God,” then Dawkins should not deny that God exists, even if the rest of his position is correct. Likewise, Dawkins assumes an affirmative answer to question 6, and therefore concludes that if the answer to question 2 is negative, God cannot be the first cause, and therefore that if he exists he must be caused. Discussing these questions with him separately would possibly be much more productive.