Why Even Infinite Universes Do Not Save Atheism

One of the most common responses to modern cosmology, especially to the fine tuning of the universe, is an appeal to the multiverse. When confronted with the astonishing fact that the laws of physics appear delicately balanced for life, many skeptics reply that no explanation beyond nature is required. If there are enough universes, they argue, then one of them was bound to look like ours.

This response has become so common that it is often treated as decisive. Yet when examined carefully, using the words of leading physicists and philosophers themselves, the multiverse turns out not to be a solution but a postponement. It does not remove fine tuning, it does not eliminate a beginning, and it does not explain why anything exists at all.

I do not believe the multiverse exists. But even if it did, it would not do the philosophical work atheism often demands of it. This article aims to show precisely why.

What the Multiverse Is and What It Is Not

The multiverse is not a single theory but a family of speculative proposals. Max Tegmark’s well known classification describes multiple levels of multiverse, ranging from regions beyond our observable horizon to radically abstract mathematical realities in which every consistent structure exists.¹

What is often overlooked is that none of these proposals are observational discoveries. They arise as theoretical extensions of inflationary cosmology, quantum mechanics, or string theory, frameworks that themselves remain incomplete.

George F. R. Ellis, an eminent cosmologist and longtime collaborator of Stephen Hawking, has repeatedly emphasized that multiverse proposals do not arise from direct observation but from philosophical extrapolation beyond what science can test.² Universes beyond our causal horizon are, by definition, unobservable. No signal can reach us from them. No experiment can confirm their existence.

Jeffrey Zweerink, an astrophysicist with Reasons to Believe, makes the same point from within the scientific community. In Who’s Afraid of the Multiverse?, Zweerink notes that multiverse models depend on mechanisms that occur outside our observable universe and therefore remain speculative. They are not discoveries forced upon us by data but inferences layered onto existing theories.¹³

This does not make the multiverse incoherent. But it does mean it belongs to metaphysical interpretation rather than empirical science.

Objection: Science Often Posits Unobservable Entities

It is often objected that science regularly posits unobservable entities such as quarks or fields, so unobservability alone cannot disqualify the multiverse.

The distinction is crucial. Unobservable entities in physics are accepted because they produce testable and falsifiable predictions within our observable domain. Multiverse proposals do not. Other universes cannot causally affect ours in any measurable way.

Physicist Sabine Hossenfelder, an atheist, is unusually direct on this point. She argues that universes forever beyond observation are not part of science precisely because they cannot be falsified.³ This is not a theological objection but a methodological one. Science advances by constraint. The multiverse removes constraint.

Fine Tuning Remains and Is Merely Displaced

There is widespread agreement among physicists that the universe appears finely tuned for life. The fundamental constants and initial conditions fall within extraordinarily narrow ranges that allow for stable matter, chemistry, stars, and conscious observers.

Paul Davies, who does not hold to Christian theism, states plainly that fine tuning is a real feature of modern physics.⁴ The multiverse is often introduced as a way to blunt the force of this observation. If enough universes exist, then life permitting ones become statistically inevitable. But this move does not eliminate fine tuning. It pushes it back one level.

Zweerink emphasizes that every multiverse model presupposes a universe generating mechanism governed by specific laws. Inflationary multiverses require a particular inflaton field and potential. String theory landscapes require stable higher dimensional mathematics. Quantum many worlds interpretations require a rigorously defined wavefunction and evolution rule.¹⁴

Davies himself acknowledges this limitation. Even if a multiverse exists, the mechanism that generates it must be exquisitely ordered to produce any universes at all.⁵ The question of why those laws exist in a life permitting form remains unanswered.

Objection: The Multiverse Explains Fine Tuning Without God

This objection only works if one assumes that laws, mechanisms, and mathematical structure require no explanation. But that assumption is precisely what is under dispute. Appealing to a multiverse does not eliminate the need for explanation; it simply relocates it. Instead of asking why this universe has life permitting laws, we are asked to accept without explanation a universe generating mechanism capable of producing vast numbers of universes governed by highly specific rules.

Tim Maudlin, an atheist philosopher of physics, has pressed this point forcefully. He argues that multiverse proposals do not explain fine tuning so much as multiply the number of unexplained entities. The explanatory burden increases rather than decreases, because one must now account not only for the existence of a life permitting universe but for the existence of a law governed system capable of generating universes at all.⁶

This is a crucial distinction. Explanations in science typically reduce mystery by showing how complex outcomes arise from simpler, more fundamental principles. The multiverse does the opposite. It replaces one finely tuned universe with a finely tuned universe generator. The laws governing inflation, quantum fields, or string landscapes must themselves fall within narrow ranges for any universes to exist, let alone life permitting ones.

Paul Davies acknowledges this limitation. He notes that a multiverse does not explain why the laws of physics have the form they do. It merely shifts the question to a deeper level. A universe generating mechanism must still be governed by mathematically elegant, stable laws, and there is no reason to think such laws are inevitable or self explanatory.⁵

The problem is not merely philosophical. Without a principled reason why the multiverse generating mechanism exists in the form it does, the appeal to the multiverse becomes an exercise in postponement rather than explanation. Saying that “everything happens somewhere” does not explain why anything happens at all.

Fine tuning, therefore, is not dissolved by the multiverse. It is deferred. The explanatory weight is pushed back one step, where it reappears with even greater force. The question remains unchanged: why does reality possess the kind of order, mathematical structure, and law like regularity that makes life possible in the first place?

A Bayesian perspective sharpens the problem even further. Bayesian reasoning asks which hypothesis makes the observed data more probable. The relevant data here is not merely that a life permitting universe exists, but that the universe is governed by simple, elegant, mathematically describable laws that fall within an extraordinarily narrow life permitting range. The multiverse hypothesis does not increase the probability of such laws existing. It presupposes them. A multiverse generating mechanism must already operate according to stable meta laws capable of producing a wide range of universes, including some that are life permitting. But the probability of such a mechanism existing is itself unexplained. As a result, the multiverse does not raise the likelihood of fine tuning on atheism. It merely shifts the improbability from one level to another, leaving the overall explanatory balance unchanged or even worsened.

By contrast, a theistic explanation predicts law like order from the outset. If reality is grounded in a rational mind, then mathematical structure, stable laws, and intelligibility are not surprising. On Bayesian grounds, fine tuning is not an unexpected coincidence under theism but a natural consequence of a universe intended to support conscious, embodied agents. Even critics of theism have acknowledged that fine tuning poses a serious challenge for purely naturalistic worldviews when evaluated probabilistically.

At this point, some atheists retreat to brute fact explanations. The laws of physics, the multiverse, and even the universe generating mechanism are said to exist “just because.” No deeper explanation is available or required. But this move comes at a significant cost. Brute facts are not explanations. They mark the end of inquiry, not its success. To insist that the most fundamental features of reality simply exist without reason is not a scientific conclusion but a philosophical stopping point.

Moreover, brute fact atheism undermines the very impulse that drives scientific investigation. Science proceeds on the assumption that reality is intelligible and that deeper explanations are worth seeking. Declaring the foundational structure of reality to be inexplicable abandons that assumption precisely where it matters most. If anything can exist without explanation at the deepest level, then there is no principled reason to deny that a necessary, self existent mind could exist as the ultimate explanation.

In the end, the multiverse does not rescue atheism from fine tuning. It either postpones explanation indefinitely or terminates it arbitrarily. Neither option provides the kind of explanatory depth that fine tuning demands. The question remains unavoidable, and no number of hypothetical universes makes it disappear.

Probability Arguments and the Inverse Gambler’s Fallacy

Another major pillar of multiverse reasoning is probability. We are told that observing a life permitting universe is not surprising if enough universes exist. Given a sufficiently large number of trials, any improbable outcome will eventually occur. Therefore, the existence of a finely tuned universe allegedly requires no explanation beyond sheer numbers.

At first glance, this sounds plausible. On closer inspection, however, it commits a well known logical error called the inverse gambler’s fallacy. The fallacy occurs when one infers the existence of many unobserved trials solely from the observation of a rare outcome. In ordinary probability reasoning, observing an unlikely event does not justify concluding that countless prior attempts must have occurred unless there is independent evidence that they did.

A simple illustration makes the problem clear. Suppose someone walks into a casino and sees a roulette wheel land on double zero five times in a row. It would be irrational to conclude, on that basis alone, that the wheel must have been spun thousands of times earlier that evening. The observation of a rare outcome does not license inferences about unobserved trials. Without independent evidence for those trials, the inference is invalid.

This is precisely the mistake made by multiverse probability arguments. From the fact that our universe is finely tuned, it does not follow that there must be a vast ensemble of universes in which different constants were tried. That conclusion only follows if there is prior evidence that such an ensemble exists.

Philosopher Philip Goff, who is not a Christian apologist and is critical of traditional theism, explicitly applies this critique to multiverse reasoning. He argues that inferring many universes from fine tuning alone commits the inverse gambler’s fallacy and therefore fails as an explanation. Without independent empirical evidence for the multiverse, probability based arguments collapse under their own logic.⁷

Jeffrey Zweerink offers a parallel illustration that captures the intuitive force of the problem. Imagine standing before a firing squad of expert marksmen who all fire simultaneously, yet you remain alive. The rational response is not to conclude that countless other firing squads must have failed elsewhere in order to make your survival unsurprising. The rational response is to ask why this particular firing squad failed. Multiplying imagined attempts does not reduce the need for explanation. It intensifies it.

The same applies to cosmic fine tuning. The existence of a life permitting universe remains remarkable regardless of how many hypothetical universes one imagines. Without independent evidence that such universes exist, appealing to probability does not explain fine tuning. It merely redescribes it.

Once again, fine tuning is not explained away by the multiverse. It is deferred. The improbability does not disappear. It is simply pushed out of sight, where it continues to demand an explanation.¹⁵

Framed in Bayesian terms, without independent evidence that a multiverse exists, the multiverse hypothesis does not increase the probability of observing fine tuning and therefore provides no confirmatory advantage over rival explanations, making the appeal to sheer numbers epistemically empty rather than explanatory.

Objection: Anthropic Selection Explains Everything

It is often claimed that anthropic selection explains everything. We observe a life permitting universe, we are told, simply because only in such a universe could observers exist. On this view, no further explanation is required. Once the observer effect is acknowledged, fine tuning supposedly dissolves.

This response trades on a subtle but important confusion. Anthropic reasoning explains why we observe certain conditions if they obtain. It does not explain why those conditions exist in the first place. It functions as a selection filter, not as a causal account. It tells us that observers necessarily find themselves in environments compatible with their existence. It does not tell us why the universe possesses life permitting conditions rather than lifeless ones.

A simple analogy makes the distinction clear. Suppose a condemned prisoner awakens in a hospital after a firing squad execution and is told, “Of course you survived. If you had not survived, you would not be here to observe it.” That statement is trivially true, but it does nothing to explain why the firing squad missed. The observer’s existence explains the observation, not the outcome.

This limitation has been acknowledged even by those most closely associated with anthropic reasoning. Steven Weinberg, an atheist Nobel laureate who helped popularize anthropic arguments in cosmology, conceded that anthropic explanations do not provide causal accounts of why the laws of physics have the form they do. He recognized that such explanations explain at most why observers find themselves in certain kinds of universes, not why those universes exist at all.⁸

The problem deepens when anthropic reasoning is used in conjunction with the multiverse. In that context, anthropic selection becomes dependent on an already existing ensemble of universes with varying laws. But the existence of such an ensemble is precisely what is in question. Without independent evidence for a multiverse, anthropic reasoning lacks an explanatory foundation. It cannot generate universes. It can only sort among them after the fact.

Anthropic reasoning therefore does not remove the need for explanation. It presupposes it. The life permitting structure of the universe remains an unexplained given, merely accompanied by a reminder that observers can only notice universes that permit observation. That reminder is true, but it is not an explanation.

In the end, anthropic selection does not dissolve fine tuning. It leaves it untouched. The question remains why the universe possesses the kind of order, mathematical structure, and law like regularity that makes observers possible at all.

As an explanation, anthropic selection lacks explanatory adequacy because it does not identify a cause, increase predictive power, or reduce explanatory arbitrariness, and therefore cannot function as an inference to the best explanation for the fine tuning of the universe.

The Multiverse and the Problem of a Beginning

A frequent claim in multiverse discussions is that the multiverse avoids the problem of cosmic origins. Even if our universe had a beginning, it is argued, the larger multiverse might be eternal. On this view, the multiverse functions as a kind of metaphysical backdrop against which universes bubble into existence without any ultimate beginning.

This claim does not survive careful scrutiny. Even the most speculative cosmological models remain subject to fundamental constraints on spacetime structure. The Borde–Guth–Vilenkin theorem demonstrates that any spacetime which, on average, has been expanding throughout its history cannot be extended infinitely into the past. This result applies not merely to our observable universe but to inflationary frameworks that generate multiple universes as well.⁹

Alexander Vilenkin, one of the theorem’s authors and no friend of theism, has been unambiguous about the implication. Inflationary models, including those invoked to support a multiverse, require a past boundary. They cannot be eternal in the past. A beginning is unavoidable.¹⁰ Attempts to avoid this conclusion by postulating earlier phases or higher-level multiverses do not remove the boundary. They merely relocate it.

The philosophical problem underlying this result is straightforward. An actual infinite sequence of physical events cannot be formed by successive addition. If the past consisted of an infinite number of events, the present moment could never be reached. No matter how far one goes back, there would always be an infinite distance remaining to traverse. The fact that the present exists is evidence that the sequence of past events is finite.

This difficulty is not avoided by appealing to cyclic or oscillating multiverse models. Cycles still involve successive events. Each cycle adds to the total number of past states. If the sequence were infinite, the present would remain unattainable. Physics does not permit the traversal of an actual infinite.

Objection: The Multiverse Exists in a Timeless State

At this point, some proponents retreat to the claim that the multiverse exists in a timeless or quasi-timeless state, from which universes emerge with time. This move, however, abandons physical explanation altogether. A timeless physical system that gives rise to temporal effects is no longer a physical mechanism in any meaningful sense. It is a metaphysical postulate.

Moreover, a timeless state capable of producing time and space already possesses the attributes traditionally associated with a metaphysical foundation of reality. It exists without temporal succession, is not governed by change, and serves as the ultimate source of all contingent spacetime realities. At that point, the multiverse has ceased to function as a naturalistic explanation and has become a placeholder for something beyond physics.

Even if such a timeless multiverse existed, the central question would remain unchanged. Why does this timeless structure exist at all, and why does it have the capacity to generate temporal universes governed by intelligible laws? The appeal to timelessness does not answer the question of origin. It reframes it.

The upshot is clear. The multiverse does not eliminate the problem of a beginning. It either inherits the need for a beginning within physical time or relocates the origin question into a realm that physics cannot address. In neither case does it remove the need for an ultimate explanation.

In this respect, the multiverse offers no advantage over a single-universe cosmology. Both confront the same fundamental question. Why does reality exist rather than nothing, and why does it exist in an orderly, intelligible form capable of giving rise to universes, laws, and conscious observers?

Viewed through a Bayesian lens, the problem becomes even sharper. The relevant data is not merely that a universe exists, but that reality appears to have a finite past and a temporal beginning governed by intelligible laws. Under theism, this data is not unexpected. A transcendent, necessary mind is not temporally bound and can freely initiate a finite temporal order. A beginning is therefore probabilistically coherent within a theistic framework. By contrast, a brute fact multiverse hypothesis assigns no principled reason to expect either a beginning or an eternal past. The presence of a past boundary becomes an unexplained contingency. As a result, the probability of observing a universe with a beginning is not increased by the multiverse hypothesis, whereas it is naturally accommodated by theism, giving theism greater explanatory coherence with respect to cosmic origins.

Why the Multiverse Fails as an Ultimate Explanation

Even if the multiverse were granted for the sake of argument, it would still fail as an ultimate explanation of reality. At most, it would function as a descriptive framework for how many universes might exist. It would not explain why anything exists in the first place.

The deepest question is not why this universe has certain properties rather than others, but why there is a reality governed by laws at all. The multiverse does not answer that question. It presupposes it. A multiverse, no less than a single universe, requires the existence of something rather than nothing. It requires laws rather than chaos. It requires mathematical structure rather than incoherence.

Appeals to the multiverse therefore leave untouched the most fundamental explanatory issues. Why do laws exist? Why do those laws have mathematical form? Why is reality intelligible to rational minds? Why is there a universe generating mechanism rather than nothing whatsoever? None of these questions are answered by multiplying universes. They remain as pressing as ever.

Martin Rees, former Astronomer Royal and not a Christian apologist, is explicit on this point. He acknowledges that even if a multiverse exists, it does not explain why the laws of physics are what they are.¹¹ The multiverse, at best, shifts the question one level deeper. Instead of asking why this universe has its laws, we are asked to accept without explanation why the multiverse itself is governed by laws capable of producing universes at all.

This shift is often mistaken for progress, but it is not. An explanation that merely postpones the deepest question does not solve it. It delays it. The multiverse replaces one contingent reality with a larger contingent reality and leaves the explanatory gap intact.

Philosophically, the problem is unavoidable. An ultimate explanation cannot consist of an infinite regress of contingent states. If every feature of reality depends on something else for its existence, then nothing is ever truly explained. Explanation must eventually terminate in something that exists necessarily, not hypothetically, and not merely as a brute fact.

The multiverse does not provide such a terminus. Whether finite or infinite, whether temporal or timeless, it remains contingent. It could have failed to exist. It could have had different laws. It could have lacked the capacity to generate universes altogether. For that reason, it cannot function as the final explanatory ground of reality.

Some respond by embracing brute fact explanations. The multiverse simply exists, they say, and no further explanation is available or required. But this move abandons explanation rather than completing it. To say that the most fundamental features of reality exist without reason is not a scientific conclusion but a philosophical stopping point. It signals the end of inquiry, not its success.

Moreover, brute fact multiverse explanations sit uneasily with the very enterprise of science. Science is driven by the assumption that reality is intelligible and that deeper explanations are worth seeking. Declaring the foundational structure of reality to be inexplicable undercuts that assumption at precisely the point where it matters most.

By contrast, a theistic explanation does not merely stop inquiry. It grounds it. If reality ultimately depends on a necessary, self existent mind, then laws, mathematical order, and intelligibility are not accidental features of the universe. They are expected. The success of science itself becomes intelligible rather than surprising.

For these reasons, the multiverse fails as an ultimate explanation. It does not answer the question of existence. It does not ground the laws of nature. It does not terminate explanation in anything necessary. It simply postpones the deepest questions and leaves them unresolved.

The multiverse may be an interesting speculative model. But speculation is not explanation. And when it comes to the ultimate nature of reality, the multiverse does not do the explanatory work that atheism often asks of it.

Taken together, the failure of the multiverse as an ultimate explanation is not isolated to any single weakness. Fine tuning is not removed but displaced. Probability arguments collapse under the inverse gambler’s fallacy. Anthropic reasoning explains observation without explaining existence. Attempts to avoid a beginning either fail physically or retreat into metaphysics. And at the deepest level, the multiverse remains contingent, law governed, and unexplained. Each move designed to rescue atheism merely shifts the explanatory burden without discharging it. The cumulative result is that the multiverse does not rival theism as an explanation of reality. It depends on precisely the kinds of order, intelligibility, and necessity that theism predicts from the outset.

The Irony: The Multiverse Fits Better Within Theism

There is a deep irony at the heart of the multiverse proposal. A reality capable of generating countless universes governed by elegant, mathematically precise laws looks less like brute chaos and far more like rational order. The very features invoked to avoid design instead point toward it.

Christian theism has never claimed that God must create only one universe. What it claims is more fundamental. All contingent reality, whether singular or plural, depends on a necessary, self existent mind. Whether God creates one universe or many is a secondary question. What matters is that creation itself is grounded in something that does not depend on anything else.

George Ellis has made this point explicitly. A universe, or even a multiverse, that produces life purely by chance is not a simpler explanation than one produced by design. It is a more complex one, requiring additional assumptions and unexplained structure.¹² Complexity is not reduced by multiplying worlds. It is increased.

Jeffrey Zweerink reaches the same conclusion from within astrophysics. Even if a multiverse were shown to exist, it would fit comfortably within a theistic framework that already affirms a rational, law giving Creator. Atheism gains no explanatory advantage by appealing to countless universes. It merely enlarges the system that still requires explanation.¹⁶

For this reason, the multiverse does not weaken the case for God but clarifies it, revealing that even the most expansive naturalistic proposals ultimately depend on the kind of rational, necessary foundation that theism has always affirmed.

Conclusion

Christians should not fear the multiverse. Atheists should not treat it as a conversation ending trump card.

There is no observational evidence for it.
It does not eliminate fine tuning.
It does not avoid a beginning.
It does not explain why anything exists.

From a Bayesian perspective, the situation becomes even clearer. Bayesian reasoning does not ask whether a hypothesis is possible, but whether it makes the observed data more probable than its rivals. The relevant data is not merely that a universe exists, but that reality exhibits elegant mathematical laws, fine tuning for life, and a finite temporal beginning.

The multiverse hypothesis does not increase the probability of these features. It presupposes them. A multiverse requires a law governed mechanism capable of generating universes, stable mathematical structure, and boundary conditions that permit complexity. The probability that such a universe generating system exists is left unexplained. As a result, the multiverse offers no independent reason to expect fine tuning, intelligibility, or a beginning. It simply relocates those features to a deeper level.

By contrast, theism predicts these features from the outset. If reality is grounded in a rational, necessary mind, then law like order is expected rather than surprising. Mathematical elegance is not accidental. A finite temporal order is coherent. On Bayesian grounds, the probability of observing a finely tuned universe with a beginning is therefore higher on theism than on a brute fact multiverse hypothesis.

This highlights a crucial asymmetry. The probability of a multiverse is not established by evidence, and its explanatory payoff is minimal. The probability of a single universe governed by intelligible laws is not only observed but fits naturally within a theistic framework. When the data is weighed comparatively, the multiverse does not outperform theism. It underperforms it.

At best, the multiverse remains a speculative proposal that postpones the deepest questions without answering them. It explains neither existence nor order. It does not terminate explanation in anything necessary. It multiplies worlds while leaving the foundation untouched.

The multiverse is not an escape hatch. It is a detour that leads back to the same question humanity has always faced, a question no number of hypothetical universes can dissolve.

Why is there something rather than nothing?

“On Bayesian grounds, the probability of observing a finely tuned universe with a beginning is therefore higher on theism than on a brute fact multiverse hypothesis.” – Richard Swinburne¹⁷

---

Footnotes

1. Max Tegmark, “Parallel Universes,” Scientific American 288, no. 5 (2003): 40–51.
2. George F. R. Ellis, “Does the Multiverse Really Exist?” Scientific American, August 2011.
3. Sabine Hossenfelder, “Why the Multiverse Is Not Science,” Backreaction, July 2019.
4. Paul Davies, The Goldilocks Enigma (Boston: Houghton Mifflin, 2006), 3.
5. Paul Davies, Cosmic Jackpot (Boston: Houghton Mifflin Harcourt, 2007), 198.
6. Tim Maudlin, The Metaphysics Within Physics (Oxford: Oxford University Press, 2007), 131–134.
7. Philip Goff, “Fine Tuning and the Multiverse,” Synthese 198 (2021): 5907–5925.
8. Steven Weinberg, Dreams of a Final Theory (New York: Vintage, 1994), 251.
9. Arvind Borde, Alan Guth, and Alexander Vilenkin, “Inflationary Spacetimes Are Not Past Complete,” Physical Review Letters 90, no. 15 (2003).
10. Alexander Vilenkin, Many Worlds in One (New York: Hill and Wang, 2006), 176.
11. Martin Rees, Just Six Numbers (New York: Basic Books, 1999), 121.
12. George F. R. Ellis, “Why the Multiverse Hypothesis Is Unscientific,” in Universe or Multiverse?, edited by Bernard Carr (Cambridge: Cambridge University Press, 2007), 124.
13. Jeffrey Zweerink, Who’s Afraid of the Multiverse? (Covina, CA: Reasons to Believe, 2011), 6–9.
14. Jeffrey Zweerink, Who’s Afraid of the Multiverse?, 12–18.
15. Jeffrey Zweerink, Who’s Afraid of the Multiverse?, 22–24.
16. Jeffrey Zweerink, Who’s Afraid of the Multiverse?, 28–30.
17. Richard Swinburne, The Existence of God, 2nd ed. (Oxford: Oxford University Press, 2004), 97–132; Robin Collins, “The Teleological Argument,” in The Blackwell Companion to Natural Theology, ed. William Lane Craig and J. P. Moreland (Oxford: Wiley-Blackwell, 2009), 202–281; Alexander R. Pruss, The Principle of Sufficient Reason (Cambridge: Cambridge University Press, 2006), 165–187.