The God Delusion (Richard Dawkins)

Richard Dawkins is one of the supreme expositors of modern evolutionary biology. "The Selfish Gene" was a book that set the standard of how scientific ideas should be presented. It completely entranced me in my high school years.

Reading "The God Delusion" provides a very different experience. There are points of light, but the book has a tone of bluster and arrogance, the arguments are presented sloppily and often incompletely, and there are too many self-indulgent anecdotes. In the preface Dawkins says he intends that "religious readers who open it will be atheists when they put it down", but the rest of the book is so scornful of religious believers that I cannot imagine them feeling anything but a visceral resistance to his arguments.

Setting aside issues of tone and editing, what are the central flaws? For all his proven talent at explaining evolutionary biology, Dawkins does not lay out philosophical arguments in a clear and persuasive way. He mixes up different arguments in a single exposition, makes circular arguments, and does not maintain a clear distinction between arguments against the existence of God, and refutations of arguments for the existence of God. Moreover, he systematically over-relies on scientific theories in confronting philosophical arguments. Given the problems that science educators already face from religious pressure groups in the USA, this is dangerous as well as confused, because it gives the impression that science is the source of arguments against the existence of God, when in fact these arguments have a much more generic basis.

(1) How not to argue for atheism

Let's review what Dawkins ought to be saying. The basic atheist position is two-pronged, depending on the theist's definition of God. If God is defined so as to have no observable manifestations then the atheist points out that it makes no sense to say anything about him. If God has observable manifestations then God's existence is an empirical matter, and the atheist argues that God is an extraordinary hypothesis that requires extraordinarily strong empirical evidence to support it. The crucial goals for the atheist are to establish that empirical evidence is the basis for any specific claims about God, and to make it clear that the burden of proof lies with the theist.

Dawkins may appreciate these points but he does not highlight them clearly enough, and continually tries to make stronger points for which he does not have solid arguments. This has made it easy for intelligent and relatively open-minded theists (his supposed target audience) to sidestep the powerful basic argument outlined above, and dismiss the whole case for atheism. For example, "The Dawkins Delusion", a very reasonable critique by theologian Alister McGrath and religious psychologist Joanna McGrath, finishes with the words: "the fact that Dawkins relies so excessively on rhetoric rather than the evidence... clearly indicates that something is wrong with his case. Ironically, the ultimate achievement of The God Delusion for modern atheism may be to suggest that this emperor has no clothes to wear. Might atheism be a delusion about God?"

(1.1) Use a muddled definition of atheism

The main challenge in mounting a critique of religion is to avoid misrepresenting atheism or naturalism as "just another faith". Dawkins leaps directly into that quicksand. His definition of atheism is (p14)

An atheist in this sense of philosophical naturalist is someone who believes that there is nothing beyond the natural, physical world, no supernatural creative intelligence lurking behind the observable universe, no soul that outlasts the body and no miracles---except in the sense of natural phenomena that we don't yet understand.

What is so disappointing about this pivotal passage, and similar ones throughout the book, is the way Dawkins contaminates the fundamental point about atheism, which is to believe only in things for which adequate evidence has been presented, with much more specific and debatable claims such as the exclusivity of nature and the non-existence of the soul. His definition raises tangential questions like what do you mean by "nature", exactly? What is a soul and what would count as evidence for or against it? Dawkins doesn't stay focussed on the real point, which is to do with how we gain knowledge, not the specifics of what we have found up to now. The exclusivity that he dimly senses lies not in the nature of "nature" but in the insistence that there is only one kind of truth, only one way to acquire reliable knowledge, and that is via evidence.

(1.2) Frame a simple question as a technical one

The sections on agnosticism give a fuller picture of Dawkins's over-inflated brand of atheism. He notes that there is a spectrum between agnosticism and atheism, and is rightly dissatisfied with those who adopt "agnosticism in principle" as a permanent position that silences debate on the topic of God. He senses that what matters is the strength of the evidence for and against the existence of God. The proper next step would be to tackle the question of what evidence would be needed to establish a conclusion. Which claims are ordinary (plausible in the absence of detailed evidence) and which are extraordinary (requiring strong evidence to support them)? Dawkins comes close to these essential points in his discussion of "Russell's teapot", a classic example of a highly implausible claim (that there is a teapot floating out in space between Earth and Mars) which therefore would require clear evidence to justify believing in it, and of which it is reasonable to be skeptical in the absence of relevant evidence.

Unfortunately, Dawkins lurches off in less productive directions. Instead of following though with a demonstration that belief in God is an extraordinary empirical claim that would require extraordinary evidence to give it any plausibility, he characterizes the existence of God as a scientific claim (p48), to be answered in terms of probability. This over-specificity raises another set of tangential questions. What is the difference between a scientific question and an ordinary empirical question? How exactly are numerical probabilities going to be derived from the evidence at hand?

What Dawkins has overlooked is the true point of Russell's teapot. It is not science that makes us disbelieve in the existence of the teapot. It is the inherent implausibility (extraordinariness) of the hypothesis combined with a lack of evidence (scientific or otherwise).

(1.3) Prematurely claim a knockout: "Ultimate Boeing 747"

With a loud fanfare, Dawkins announces his knockout argument that there is "almost certainly no God". This is the title of Chapter 4. Unfortunately, what he actually presents is a set of refutations of the theist Argument from Design. These refutations do indeed show that the appearance of design does not imply the existence of a designer. However on their own they certainly don't imply the absence of a designer.

He starts out with the point that

However statistically improbable the entity you seek to explain by invoking a designer, the designer himself has got to be at least as improbable. (p113-114)

He spends the rest of the chapter on the evidence that natural selection is the explanation for complex life forms on earth, and invokes the anthropic principle to explain why life arose at all and why the laws of physics are conducive to life. At the end of the chapter he gives a summary, which goes roughly like this:

(1) The designer hypothesis immediately raises the larger problem of who designed the designer.
(2) There is strong evidence that life on earth was "designed" by natural selection and evolution.
(3) Therefore the only way a complicated thing like God could have arisen is via a similar mechanism.

Dawkins and others have indeed shown that natural selection is a wonderful explanation of the appearance of design. But where is the argument that it is the only possible explanation?

At the minimum, perhaps one could set aside Dawkins's bluster, and read this chapter simply as an extension of the previous chapter, where he presented refutations of various arguments for the existence of God. Does Dawkins at least do a good job of persuading theists that they don't have any knock-down arguments of their own?

(2) How not to disprove arguments for the existence of God

Dawkins somehow fails to deliver convincing death blows to even some of the weaker arguments for the existence of God. Perhaps their fallacy is so obvious to him that he can't really be bothered.

After this discussion of some of the less plausible arguments for the existence of God, Dawkins goes on to illustrate how not to deal with the more well-regarded ones: the arguments based on design or other apparently special features of nature.

(2.1) Present natural selection as the only basis for refuting the Argument from Design

At the start of Chapter 3, Dawkins gives a nice economical refutation of the argument from first cause. He makes the standard point, one does not have to resort to a multifaceted explanation like God (a being who cares about us, hears prayer, sets moral standards, etc) in order to address questions like "what was the first cause?". Instead of a first cause that is also an all-knowing and all-powerful God, how about an equally mysterious but much simpler "first event"? If God can be his own cause, why can't the big bang be its own cause?

However, Dawkins does not seem to notice that an equally economical approach is sufficient to deal with the argument from design. Instead of a super-intelligent designer that crafted the universe by unknown means, how about a mechanical "complexity force" that (by equally mysterious means) favors the development of more complicated objects? It is true that natural selection gives a satisfyingly fleshed-out understanding of many details of living organisms, but it is worth noting that the argument from design would be flawed even if natural selection were to be disproved.

(2.2) Take the "origin of life" bait

Like "design" and "first cause", the occurrence of life on earth does not provide any argument for believing in God. The atheist's best approach is to point out that the various Gods of the world's religions have many extra properties beyond what would be needed to explain the occurrence of life on earth, so life does not provide evidence of their existence. Just because there is a mystery doesn't make God a plausible hypothesis: any modest non-divine mechanism will suffice. In other words, as always, keep the burden of proof on the theist and keep the focus on his/her extraordinary claims.

Unfortunately, Dawkins does not stick to this minimalistic strategy. He claims that "there are two hypothesis to explain what happened, the design hypothesis and the scientific or 'anthropic' hypothesis" and proceeds to defend some vague scenario based on the anthropic principle. His discussion is not wrong, but his formulation of the issue gives the false impression that in order to enter the discussion of God one must have a viable hypothesis about the origin of life on earth. Moreover, unlike the origin of species via natural selection, scientists have only vague speculations about the origin of life on earth, and it is a mistake to present these as if they were a persuasive theory, since their weaknesses will then seem to be weaknesses in the atheist worldview.

(2.3) Take the "finely-tuned laws of nature" bait

Dawkins then takes up a recently popular variant of the argument from design, the "Goldilocks" argument that the laws of nature themselves are improbably well-suited to the existence of life in the universe, and hence there "must" be a God. Again, the best response is the minimal one, pointing out firstly that we have no idea where the laws of nature come from, so there is no way to say if they are improbably "finely tuned" or not, and also there is no reason to adopt such an over-engineered solution as a personal God, with all the detailed properties that world religions have attributed to him.

Again Dawkins bites off more than he can chew. In his over-eagerness to show that science has everything under control, he expounds a very recent and controversial piece of speculative cosmology, namely the "multiverse". This is the idea that somehow there is a vast number of different universes with different laws of nature, so naturally we find ourselves in one of the few that have life-friendly laws of physics. This sort of argument may be plausible when one is dealing with planets, where we have reasonable evidence that a vast number of them might exist, but when applied to unobserved parallel universes it becomes a science-accented echo of the theists's use of this argument as a proof of the existence of God. In what sense is a gigantic multiplicity of unobservable universes so much more plausible than one universe with some specific set of laws of nature? Like the deity himself, the multiverse is an overwrought solution that is greatly under-motivated. Bear in mind that there is no independent evidence for the validity of the theories that predict multiverses! Yet again, by presenting a dubious and controversial argument for a case that he doesn't need to make, Dawkins hands the theists an opportunity to claim that atheism has come up short.

(3) What Dawkins does well

After all this criticism, it is only fair to mention the places where I thought Dawkins made valuable or interesting points.

  1. He points out that religious beliefs are favored with special treatment, benefiting from a free pass through the normal filters of legality and reasonableness. He gives various examples, such as the exemptions from drug laws for religions that "require" the drugs for their ceremonies.
  2. He emphasizes that most religious people or organizations do not in practice subscribe to Stephen Jay Gould's "Non-overlapping magisteria" idea, that science has no bearing on religious beliefs. He discusses the experiments that have shown that prayer has no effect on recovery from surgery, and notes, very plausibly, that if these experiments had shown a positive result, believers would have loudly trumpeted this as vindication of their beliefs.
  3. He does a good job of showing that the Jewish/Christian bible does not provide absolute moral guidance. Atheism's lack of a divinely-mandated moral code is one basis for theistic criticisms, but Dawkins shows clearly that the behavior and values that are presented as exemplary in the bible would in many cases be reprehensible by current moral standards, and are not in practice followed by decent people.
  4. He provides an interesting survey of highly speculative evolutionary scenarios for how moral sense, altruism, and religiosity might arise. The scenarios are fine as examples of how evolutionary theory might account for these phenomena, i.e. as refutations of the claim that one cannot account for them without a God. However they have little empirical support, and one must resist the temptation to use them as arguments that we have a scientific explanation of morality or religiosity. In fact, if taken too seriously, they lead to significant problems. For example, if natural selection can favor delusional beliefs about Gods, why shouldn't it have favored other delusions; perhaps the cognitive processes on which science is based could themselves be flawed in some way. It is vital to remember that science is a development and application of more basic knowledge-gathering procedures: it is founded on them, and does not provide a separate foundation in its own right.

Copyright © Mark Alford (2008)

alford(at)physics.wustl.edu

Mark Alford's home page

Valid XHTML 1.0!