Refuting Richard Dawkins

The Flawed Spectrum: Why Richard Dawkins’ Scale of Theistic Probability Collapses Under Scrutiny

Richard Dawkins’ The God Delusion, first published in 2006, stands as one of the defining texts of the New Atheist movement. Written with characteristic clarity, wit, and polemical energy, the book sold millions of copies and helped shift public conversation about religion in many Western societies. At the heart of its case against theism lies a deceptively simple conceptual tool: the “spectrum of theistic probability.” Dawkins invites readers to locate their own beliefs on a continuous scale from 1 to 7. Position 1 represents the strong theist who is 100 percent certain that God exists; position 7 represents the strong atheist who is 100 percent certain that God does not exist. Intermediate positions capture varying degrees of confidence or doubt. Dawkins himself has consistently placed himself at 6, later refining the figure to 6.9: “I cannot know for certain but I think God is very improbable, and I live my life on the assumption that he is not there.”

The spectrum is presented as an exercise in intellectual honesty and scientific rigor. It appears to replace crude binary thinking with a nuanced probabilistic assessment. In reality, the device rests on a series of deep conceptual errors. It treats a metaphysical question as if it were an ordinary empirical hypothesis amenable to the same kinds of probability assignments used in the natural sciences. It depends on a philosophically thin and historically uninformed understanding of divine complexity. It confuses the limits of current scientific explanation with the ontological improbability of a transcendent reality. And it systematically under-engages the classical theistic tradition it claims to dismantle. What follows is a sustained examination of these flaws, drawing on work by analytic philosophers, theologians, historians of ideas, and scientists who have subjected Dawkins’ central arguments to careful scrutiny. The aim is not to demonstrate the existence of God, but to show that the spectrum, and the broader atheistic case built upon it, fails on its own stated terms.

 The Spectrum Itself: A Misleading Continuum

Dawkins articulates the spectrum with seven “milestones.” At one extreme stands the strong theist who declares, in the spirit of Carl Jung, “I do not believe, I know.” At the other extreme stands the strong atheist who claims absolute certainty of non-existence. Between these poles lie positions of high but incomplete confidence in theism, leaning toward theism, pure agnosticism (exactly 50 percent probability), leaning toward atheism, and de facto atheism (very low probability short of zero). Dawkins observes that many religious believers occupy position 1 because their traditions discourage doubt, while few atheists claim position 7 because a properly scientific mind remains open to new evidence.

The Spectrum Itself: A Misleading ContinuumDawkins’ milestones are:Strong theist. 100% probability of God.  

• De facto theist. Very high probability but short of 100%.  
• Leaning toward theism.  
• Completely impartial agnostic. Exactly 50%.  
• Leaning toward atheism.  
• De facto atheist. Very low probability but short of zero.  
• Strong atheist. 100% certainty of non-existence.

This framing already contains a revealing asymmetry. Certainty on the theistic side is portrayed as closed-minded dogmatism; residual uncertainty on the atheistic side is celebrated as epistemic humility. Yet the spectrum itself presupposes that the proposition “God exists” is the sort of claim that can be assigned a meaningful numerical probability in the first place. That presupposition is far from obvious.

In classical theism—the tradition stretching from Augustine and Anselm through Aquinas, the Reformers, and much of contemporary analytic philosophy of religion—God is not conceived as one more contingent entity inside the spatiotemporal universe. God is understood as a necessary being, the uncaused cause, the ground of all contingent existence rather than an item within it. If the concept of a necessary being is coherent, then the probability of such a being’s existence is either 1 or 0; intermediate values make little sense. Assigning a probability of, say, 0.01 to a necessary being is not an exercise in careful calibration; it is a category error. The spectrum quietly imports a naturalistic ontology in which every existent must be treated as an empirical hypothesis competing for probability mass. That ontology is precisely what is at issue.

Even if one grants that some form of probability assignment is appropriate, the middle position of exact equiprobability (position 4) is almost never rationally warranted for complex, high-stakes propositions. Bayesian epistemology requires priors informed by background knowledge, theoretical virtues, and the total evidence. Treating “God exists” and “God does not exist” as initially equally likely is no more justified than treating “the number of stars in the observable universe is odd” and “the number is even” as 50-50 once cosmological data are taken into account. The appearance of numerical precision in Dawkins’ scale is therefore largely rhetorical. It gives the impression of scientific exactitude while resting on arbitrary or question-begging assignments of prior probability.

Further difficulties arise when one considers the nature of the evidence that would move a person along the spectrum. Dawkins repeatedly insists that the existence of God is a scientific hypothesis like any other. Yet many of the considerations that theists actually advance—arguments from the contingency of the universe, from the applicability of mathematics to nature, from the existence of moral obligations that appear to transcend evolutionary utility, or from the fine-tuning of physical constants—are not straightforwardly empirical in the laboratory sense. They are philosophical arguments that draw on scientific data but are not themselves settled by further experimentation. The spectrum therefore forces an ill-fitting template onto a question that straddles metaphysics, epistemology, and the philosophy of science.

 The Core Argument: The Ultimate Boeing 747 and Divine Complexity

The engine that drives Dawkins’ low probability assignment is the argument he calls the “Ultimate Boeing 747 gambit.” The reasoning proceeds roughly as follows. The universe, and especially living organisms, exhibit organized complexity that creates a powerful appearance of design. Design, if real, requires a designer. Any designer capable of producing such dazzling complexity must itself be at least as complex—and therefore even more statistically improbable—than the phenomena it is invoked to explain. Positing God therefore multiplies rather than reduces the improbability. By contrast, Darwinian natural selection explains the appearance of design through gradual, cumulative processes that begin with simplicity and build complexity step by step. Consequently, the God hypothesis is “very close to being ruled out by the laws of probability.”

This argument has been subjected to sustained philosophical criticism, and the objections cut deep.

The first and most fundamental problem concerns the notion of complexity itself. Dawkins treats complexity largely in engineering or biological terms: the number of parts, their interrelations, the improbability of their random assembly. A Boeing 747 or a living cell is complex because it contains a vast number of finely coordinated components. He then projects this same metric onto God. Yet the classical theistic tradition has long insisted on the doctrine of divine simplicity. On this view, God is not a composite entity assembled from more basic parts or properties. God’s knowledge, power, and will are not modular add-ons that increase complexity in the way additional circuitry increases the complexity of a computer. In the language of Thomas Aquinas, God is pure actuality (actus purus); in the language of much contemporary analytic philosophy, God is a necessary mind whose essence is identical with its existence. Alvin Plantinga has argued that Dawkins’ claim is circular: it assumes that any mind capable of designing a universe must be complex in the biological or artifactual sense, then concludes that such a mind is highly improbable. But if God is not the kind of entity to which that metric applies, the analogy collapses.

Even if one rejects the traditional doctrine of simplicity, a second problem remains. Probability estimates of the sort Dawkins invokes typically depend upon background laws of nature and observed frequencies. The probability that a hurricane will assemble a functioning aircraft in a scrapyard is vanishingly small relative to the laws of physics and the distribution of materials. But if one is asking about the existence of a being hypothesized to be the source or ground of those very laws, the frequency approach loses its footing. Philosopher Mark F. Sharlow has formalized this difficulty: methods of estimating probability that presuppose a fixed set of natural laws cannot be straightforwardly applied to an entity whose existence is proposed as the explanation of those laws. The argument therefore either begs the question against theism or changes the subject to a different, more anthropomorphic deity that classical theists do not affirm.

A third difficulty is logical. Even if one grants every premise about complexity, the conclusion that God is extremely improbable does not follow. At best the argument shows that a certain conception of God would not function as a good scientific explanation of biological complexity in the same way that natural selection does. It does not show that God does not exist. William Lane Craig and other philosophers have noted that Dawkins confuses the explanatory adequacy of theism for one particular domain (the appearance of design in biology) with the overall epistemic status of theism. One can fully accept evolutionary biology while still regarding cosmological, ontological, or axiological arguments as providing positive support for theism. Dawkins largely sidesteps the contemporary philosophical literature on these arguments, offering instead brief, often dismissive treatments that many professional philosophers regard as inadequate.

Alister McGrath has observed that Dawkins’ language of improbability remains vague when compared with careful probabilistic treatments of theism, such as those developed by Richard Swinburne. Swinburne argues that theism can be simpler than naturalism in important respects: it postulates one ultimate personal explanation rather than a host of brute physical facts, laws, and initial conditions. Given the existence of a law-governed, finely tuned, conscious universe, the posterior probability of theism is, on Swinburne’s accounting, not vanishingly small. One need not accept Swinburne’s specific numerical estimates to recognize that Dawkins’ alternative remains under-argued. The claim that God is “almost certainly” nonexistent is presented as the outcome of rigorous probabilistic reasoning; in practice it functions more as a statement of metaphysical preference.

 Scientism and the Boundaries of Science

Beneath the spectrum and the Boeing 747 argument lies a broader methodological stance: the existence of God is a scientific hypothesis like any other and must be evaluated exclusively by the methods and standards of the natural sciences. This is a form of scientism—the elevation of scientific method to the status of the sole reliable path to knowledge about reality.

The difficulty is that the thesis “only scientific methods yield reliable knowledge” is not itself a scientific hypothesis. It is a philosophical claim about the scope and limits of knowledge. More importantly, many of the concepts on which Dawkins relies—logical consistency, mathematical probability, the uniformity of nature, and even the moral evaluation of religious practices—cannot be derived from empirical science alone. They function as presuppositions of scientific inquiry rather than as its products. Critics have pointed out that Dawkins repeatedly violates his own naturalistic commitments when he appeals to objective standards of rationality or morality while simultaneously denying any transcendent ground for those standards.

The physicist Marcelo Gleiser has argued that categorical atheism sits uneasily with the scientific method. Science proceeds by formulating testable hypotheses and remaining open to revision in light of evidence. Absolute declarations of non-existence about entities that lie outside the current empirical reach of science are not scientific conclusions; they are metaphysical ones. The absence of laboratory evidence for God is not automatically evidence of absence once the discussion moves into questions of ultimate explanation, the origin of the laws of nature, or the existence of necessary being. Treating theism as “almost certainly false” solely on the basis of the success of evolutionary biology therefore constitutes an overreach. Evolution may undercut certain popular design arguments, but it leaves untouched deeper questions: why there exists a law-governed universe capable of evolution at all; why the fundamental constants of physics appear finely tuned for the emergence of complexity; and why conscious, rational, moral agents exist within that universe.

The spectrum’s insistence on treating God as a scientific hypothesis also flattens important distinctions between different kinds of explanation. Scientific explanations typically cite laws, initial conditions, and mechanisms operating within the natural order. Metaphysical explanations may appeal to the conditions that make such an order possible. Conflating the two leads to the familiar but ultimately unhelpful demand that theists produce a “mechanism” by which God creates, as if God were one more causal agent inside the system rather than the ground of the system itself.

 Misrepresentation of the Theistic Tradition

A persistent criticism of The God Delusion is that it attacks a straw man. Dawkins focuses heavily on a highly anthropomorphic, interventionist deity who functions essentially as a super-powerful agent operating inside the universe. He then demonstrates that such a being would be improbable. Yet the mainstream of classical theism has long denied that God is a being among beings. The God of Augustine, Aquinas, Maimonides, and much of the subsequent tradition is not a magnified version of a human designer. Arguments for God’s existence in this tradition—the Five Ways, the ontological argument in its various forms, cosmological arguments from contingency, and contemporary developments in analytic philosophy—operate with a different conceptual framework.

Dawkins’ engagement with these arguments is widely regarded as thin. The ontological argument is dismissed as “infantile” and “dialectical prestidigitation.” Cosmological arguments are reduced to the child’s question “Who made God?” Fine-tuning is met primarily with speculation about a multiverse, without sustained attention to the philosophical and scientific difficulties that multiverse hypotheses themselves face. Terry Eagleton’s well-known review captured the problem with characteristic force: “card-carrying rationalists like Dawkins… are in one sense the least well-equipped to understand what they castigate, since they don’t believe there is anything there… worth understanding. The more they detest religion, the more ill-informed their criticisms of it tend to be.”

The result is a book that is highly effective at confirming the views of those already inclined toward unbelief, while leaving many philosophically informed theists unpersuaded. The spectrum inherits this limitation. By framing the question exclusively in terms of the probability of an empirical hypothesis, it sidelines the very conceptual resources that sophisticated theists use to articulate their position.

 Broader Problems with the New Atheist Framing

The spectrum does not stand alone. It forms part of a larger New Atheist package that treats religion primarily as a failed scientific theory and as a uniquely potent source of social harm. Both claims are overstated.

On the scientific side, evolutionary biologists such as David Sloan Wilson have criticized Dawkins for inadequate engagement with the empirical literature on religion. Wilson and others have explored the possibility that religious beliefs and practices can function as group-level adaptations that enhance cooperation and cohesion. Whether or not one accepts particular adaptive hypotheses, the point is that Dawkins’ treatment of religion as a simple “delusion” or “virus of the mind” does not fairly represent the range of scientific work in the field.

On the historical and moral side, the claim that religion is uniquely responsible for violence and oppression ignores the complex interplay of ideology, politics, economics, and ethnicity in human conflict. It also underplays the secular body counts of the twentieth century and the capacity of non-religious ideologies to generate their own forms of dogmatism and cruelty. Philosophical claims that morality can be fully grounded in evolutionary psychology confront the persistent is-ought gap and the difficulty of accounting for the binding, categorical force of moral norms. These are not minor side issues; they form part of the broader intellectual context in which the spectrum is offered as a decisive tool.

None of these observations rehabilitates every religious claim, institution, or practice. Many religious beliefs are false, many religious institutions have caused harm, and many believers hold their convictions for poor reasons. The point is narrower: the confident assignment of near-zero probability to theism, presented as the outcome of careful probabilistic and scientific reasoning, is not in fact supported by the arguments Dawkins advances. It reflects a particular set of background metaphysical commitments—methodological naturalism elevated to a comprehensive naturalism—combined with selective engagement with the opposing literature.

 Alternative Approaches and Intellectual Humility

A more careful approach would distinguish several different questions. One can ask whether particular empirical claims made by religious traditions are true. One can ask whether the success of evolutionary biology undercuts certain design arguments. One can ask whether the concept of a necessary being is coherent, and whether any of the traditional arguments for such a being succeed. One can ask about the moral and social consequences of religious belief in specific historical contexts. These are distinct inquiries. Collapsing them into a single seven-point scale of “theistic probability” obscures more than it illuminates.

Philosophers who remain unpersuaded by theism often do so for reasons that have little to do with Dawkins’ spectrum: the problem of evil, the apparent hiddenness of God, the success of naturalistic explanations in specific domains, or a preference for ontological parsimony understood in a certain way. These are serious considerations that deserve serious response. They are not, however, the same as the claim that God is “almost certainly” nonexistent because a designer would have to be more complex than the designed. The latter claim, which powers the spectrum, remains one of the weaker links in the New Atheist chain.

Intellectual humility cuts in more than one direction. Theists who claim absolute certainty on the basis of religious authority alone open themselves to legitimate criticism. Atheists who claim that the non-existence of God has been established to a degree approaching scientific certainty, on the basis of the arguments examined here, likewise overreach. The spectrum’s greatest weakness is that it presents a contested and under-argued metaphysical conclusion in the language of measured, scientific probability.

A Revealing Live Encounter: Dawkins’ Uneasy Performance Against Cardinal Pell

One of the more telling public tests of Dawkins’ approach came not in the controlled environment of a book or a prepared lecture, but in a live televised exchange. On 9 April 2012, the Australian Broadcasting Corporation’s Q&A program staged a special Easter episode pitting Dawkins against Cardinal George Pell, then Archbishop of Sydney and Australia’s most senior Catholic cleric. Moderated by Tony Jones, the hour-long discussion ranged across the existence of God, evolution, the afterlife, the nature of “nothing,” suffering, and the moral record of religion and atheism. What unfolded was less a polished intellectual contest than a revealing display of Dawkins’ limitations when forced to respond in real time to a theistic interlocutor before a live studio audience.

Dawkins arrived jet-lagged from international travel and appeared visibly tired, at times slumped in his chair. From early in the program he grew irritated by audience reactions. On multiple occasions the studio audience laughed at statements he regarded as straightforward scientific or philosophical points. Dawkins repeatedly snapped back with visible frustration: “What? Why is that funny?” The laughter was not always directed at obvious gaffes; sometimes it greeted his insistence that questions about the purpose of the universe are “silly,” or his attempts to clarify what physicists mean by “nothing” in cosmological discussions. When Pell pressed him on the coherence of claiming that something can come from nothing, and the audience responded with amusement at the exchange, Dawkins’ irritation only deepened. The moment underscored a recurring pattern: the gap between the confident, quasi-scientific rhetoric of The God Delusion and the more contested, philosophically slippery terrain of live debate.

Pell, for his part, was no polished philosopher of religion. He made well-publicized factual errors, including a confused claim about human descent from Neanderthals, and at times wandered into less rigorous theological territory. Yet the dynamic of the evening did not unfold as a straightforward triumph for the New Atheist side. Audience laughter and applause were distributed unevenly, and several of Dawkins’ sharper dismissals—particularly his reduction of ultimate “why” questions to scientific irrelevance—landed poorly with portions of the room. One especially awkward sequence occurred when Pell began an anecdote about preparing “some young English boys” for first communion; the audience erupted in laughter (a reaction clearly informed by the broader cultural awareness of clerical abuse scandals), forcing Pell to press on while the moderator interjected. Dawkins’ own moments of discomfort with the audience’s sense of humor, however, were more frequent and more personal. He treated the laughter as an affront to rational discourse rather than a signal that his framing was failing to persuade.

Contemporary reports described the encounter as often descending into a “feral snarl-off.” Dawkins threatened, in effect, to discipline an audience that dared titter at his assertions. The spectacle illustrated a deeper problem with the spectrum-of-probability approach and the broader New Atheist style. In the book, Dawkins can assign theism a vanishingly low probability, declare purpose questions meaningless, and move on. In a live setting, confronted with a representative of the tradition he dismisses and an audience not uniformly predisposed to his worldview, the same moves produced visible friction. The laughter was not evidence that theism had been vindicated; it was evidence that Dawkins’ preferred mode of argument—treating metaphysical questions as settled scientific ones and reacting with impatience when others disagreed—did not travel as smoothly as the sales figures of The God Delusion might suggest.

The Q&A exchange remains a useful case study precisely because it was unscripted. It showed a leading public advocate of the view that God is “almost certainly” nonexistent struggling to maintain composure when ordinary audience reactions challenged the self-evidence of that claim. The episode did not settle any philosophical question. It did, however, expose the practical limits of a style that relies heavily on rhetorical certainty and probabilistic shorthand while remaining relatively thin on engagement with the actual conceptual architecture of classical theism. In that sense, the laughter that so clearly irritated Dawkins was less a personal slight than a symptom of the larger intellectual mismatch the rest of this essay has examined.

Watch it here: Debate: Richard Dawkins vs Cardinal George Pell on Q&A | DevoutNone - YouTube

 Conclusion: Probability, Metaphysics, and the Limits of a Scale

Dawkins’ spectrum of theistic probability is rhetorically powerful and pedagogically convenient. It is not philosophically robust. It applies a framework suited to contingent empirical hypotheses to a claim that, if true, is not of that kind. It rests on a contested and often question-begging account of divine complexity. It treats residual uncertainty as a mark of intellectual virtue for atheists while treating analogous caution among theists as special pleading. And it proceeds with insufficient attention to the conceptual resources of the classical theistic tradition.

One may reject theism for many reasons—evidential, moral, existential, or aesthetic. What one cannot responsibly claim is that the spectrum, or the Ultimate Boeing 747 argument that underwrites it, has shown the existence of God to be “almost certainly” false in anything approaching the sense in which a well-confirmed scientific theory is almost certainly true. The device confuses the current reach of scientific method with the boundaries of reality itself. That confusion remains a central and enduring weakness of the New Atheist project. A more careful conversation about theism and atheism will require greater precision about the kinds of questions being asked, greater familiarity with the history of ideas, and greater willingness to acknowledge the limits of any single methodological approach. The seven-point scale, for all its popular appeal, does not meet that standard.

 References

- Aquinas, Thomas. Summa Theologica. Translated by the Fathers of the English Dominican Province. Benziger Bros., 1947.  

- Craig, William Lane. Critiques of the Ultimate 747 gambit and related arguments in debates, articles, and Reasonable Faith resources.  

- Dawkins, Richard. The God Delusion. Bantam Press, 2006.  

- Eagleton, Terry. “Lunging, Flailing, Mispunching.” London Review of Books, Vol. 28, No. 20, 19 October 2006.  

- Gleiser, Marcelo. “Atheism Is Inconsistent with the Scientific Method, Prizewinning Physicist Says.” Scientific American, 19 March 2019.  

- McGrath, Alister. Dawkins’ God: Genes, Memes, and the Meaning of Life. Blackwell, 2005.  

- Plantinga, Alvin. Review of The God Delusion and essays on divine simplicity, warrant, and theism.  

- Sharlow, Mark F. “Playing Fast and Loose with Complexity: A Critique of Dawkins’ Atheistic Argument from Improbability.” PhilSci-Archive, University of Pittsburgh, 2009.  

- Swinburne, Richard. The Existence of God. 2nd edition. Oxford University Press, 2004.  

- Additional sources: Corlett, J. Angelo. “Dawkins’ Godless Delusion.” International Journal for Philosophy of Religion 65 (2009): 125–138; Internet Encyclopedia of Philosophy entry on “New Atheists”; Wilson, David Sloan. Critiques of Dawkins’ treatment of the evolutionary study of religion; various papers in Synthese and related journals on fine-tuning, Bayesian confirmation, and the philosophy of cosmology.