To the editors:

Some regard other universes—domains of space and time that we cannot observe, perhaps even in principle—as being in the province of metaphysics rather than physics; it is claimed that we cannot observe them and can neither confirm nor refute their existence. I would argue that the question “Do other universes exist?” is a genuinely scientific one. I shall outline why it is an interesting question, why it may be answered within a few decades, and why I already suspect that the answer may be yes.

Does something that is part of science have to be observable? I think the answer is no, not least because there is actually a blurred transition between the readily observable and the absolutely unobservable, with a very broad grey area in between. To illustrate this, let us envisage a succession of horizons, each taking us further than the last from our direct experience. There is a limit to the parts of reality that our present-day instruments can probe. Obviously there is nothing fundamental about this horizon; it is constrained by current technology. Many more galaxies will undoubtedly be revealed by the bigger telescopes now being planned. And new techniques will allow us to probe the subatomic microworld, discovering new particles that we can now just theorize about. We do not demote objects from the realm of proper scientific discourse simply because they have not been seen yet. When we speculate about what might swim in the oceans of Jupiter’s frozen moon Europa, we are asking a scientific question.

But there are some things that we cannot observe, even in principle. We cannot see inside a black hole. We cannot see galaxies that are so far away that their light has not had time to reach us since the big bang. There are definite horizons to our observation. If you were voyaging on the ocean, it is conceivable the water actually ends immediately beyond your horizon—but it would be unlikely. Likewise, we would expect galaxies beyond our horizon. We are now receding from remote galaxies at an ever-increasing rate, so if their light has not yet reached us, it never will. Such galaxies are not merely unobservable in principle now—they will forever be beyond our horizon. But such galaxies surely still exist. Many arguments suggest that galaxies beyond the horizon outnumber those we see by a vast factor—perhaps so vast that all combinatorial options would occur repeatedly, and we would all, far beyond the horizon, have avatars.

These never-observable galaxies would have emerged from the same Big Bang as we did. But suppose that we imagine separate Big Bangs. Are space-times that are completely disjoint from ours any less real than regions forever unobservable which are the aftermath of our Big Bang? Surely not. These other universes too should count as part of physical reality.

I present the above step-by-step argument to those who are prejudiced against the multiverse as an exercise in phobia therapy. Following this technique, if you are terrified of spiders you are first reconciled to a small one a long way away, and then, stage by stage, to giant tarantulas crawling all over you. From a reluctance to deny that galaxies just beyond reach of present-day telescopes are proper objects of scientific inquiry, you are led toward taking seriously quite separate space-times, perhaps governed by quite different laws.

There are credible theories for an eternal cosmos where many universes sprout from separate Big Bangs into disjoint regions of space-time; these models are based on physical assumptions which are clearly spelled out, but which are untested. The best-known and most-studied version is eternal inflation, espoused especially by Andrei Linde. But there are other scenarios. Alan Guth, Edward Harrison, and Lee Smolin have, from different viewpoints, suggested that a new universe could sprout inside a black hole, expanding into a new domain of space and time inaccessible to us. Others have envisioned universes separated from us in an extra spatial dimension. Bugs crawling on a large sheet of paper—their two-dimensional universe—would be unaware of other bugs on a separate sheet of paper. We would be unaware of our counterparts on another brane separated in an extra dimension, even if that separation was only by a microscopic distance. 

Another important issue then arises. The physical laws seem to be universal within our observational domain; spectra of distant galaxies reveal that they are made of atoms and stars governed by the same physics that we study in the lab. But this need not imply that they are the same far beyond the horizon in our universe—still less in the aftermaths of other Big Bangs.

So there are two key questions for twenty-first-century physics:

1. Are there many big bangs rather than just one?
2. If there are many big bangs, are they all governed by the same physics?

The first question is crucial. The traditional idea has been that the physical laws are somehow unique; they are there in a Platonic sense independent of the universe. If so, we should have to regard the laws as a brute fact. Insofar as they are tuned to allow our existence—some say that we should not be surprised by this—we would not exist otherwise. This attitude has been countered by John Leslie, one of the few philosophers that has tackled this problem, and who offers a nice metaphor. Suppose you were before a firing squad. A dozen bullets are fired, but they all miss. Had that not happened, you would not be alive to ponder the matter. But your survival is still a surprise—one that it is natural to feel perplexed about. 

Many of us have been puzzled for a long time about why the laws of nature allow complexity to emerge, and how stars, galaxies, planets, planets and people could emerge from a simple Big Bang. That is an enigma. We can easily imagine laws not all that different from the ones that actually prevail, but that would have led to a rather boring universe: laws which led to a universe containing dark matter and no atoms, laws where you have hydrogen atoms but nothing more complicated, and therefore no chemistry and no nuclear energy to keep the stars shining, laws where there was no gravity, or a universe where gravity was so strong that it crushed everything. The lifetime of the universe might be so short that there was no time for evolution or its expansion too fast to allow gravity to pull stars and galaxies together.

Many things in our cosmic environment—the exact layout of the planets and asteroids in our Solar System—are accidents of history. If the answer to the second question is no, the recipe for an entire universe may be arbitrary. Or some aspects may be arbitrary and others not. As an analogy, which I owe to Paul Davies, consider the form of snowflakes. Their ubiquitous six-fold symmetry is a direct consequence of the shape of water molecules. Snowflakes display an immense variety of patterns because each is molded by its micro-environments; how each flake grows is sensitive to the fortuitous temperature and humidity changes during its growth as it falls towards the ground.

If physicists achieved a fundamental theory, it would tell us which aspects of nature were direct consequences of the bedrock theory—just as the symmetrical template of snowflakes is due to the basic structure of a water molecule—and which are—like the distinctive pattern of a particular snowflake—the outcome of accidents. If the answer to the second question is no, we would not expect to find ourselves in a typical universe. Rather, we would be in a typical member of the subset of universes in which an observer could evolve.

Here is an analogy. People used to wonder: why is the earth in this rather special orbit around this rather special star, which allows water to exist, has oceans and continents, and offers an environment for life to evolve? We now perceive nothing remarkable in this apparent fine tuning. That’s because we know that there are millions of stars with retinues of planets around them; among that huge number there are bound to be some that have the conditions for life. We should not be surprised that we live on one of that special subset of planets. So there is no mystery about the fine-tuned nature of the earth’s orbit; it is just that life evolved on one of millions of planets where things were right.

If we are perplexed by the fine tuning of the physical laws, it seems an attractive idea that our Big Bang is just one of many. Our earth is a planet that happens to have the right conditions for life, among the many, many planets that exist. Our universe, and our Big Bang, is the one out of many which happens to allow life to emerge, to allow complexity.

When the multiverse is mentioned, it is sometimes asserted that domains that are not observable do not count as a part of science. I think that is the wrong way to look at it. We cannot observe the interior of black holes, but we believe what Einstein says about what happens there because his theory has gained credibility by agreeing with data in many contexts that we can observe. If we had a theory describing physics at 10¹⁶ GeV that had been corroborated in other ways—by yielding insights into the masses of neutrinos, and the nature of the nuclear and electric forces—then such a theory would thereby acquire credibility. If this theory, applied to the very beginning of our universe, were to predict many big bangs, or separate branes, then we would have as much reason to believe in separate universes as we now have for believing inferences about quarks inside atoms, or about the unobservable interiors of black holes.

We may one day have a convincing theory that accounts for the very beginning of our universe, tells us whether a multiverse exists, and, if so, whether some so-called laws of nature are just parochial by-laws in our cosmic patch. If they are, then some seemingly fine-tuned features of our universe could then only be explained by anthropic arguments. Although this style of explanation raises hackles among some physicists, it is analogous to what any observer or experimenter does when they allow for selection effects in their measurements: if there are many universes, most of which are not habitable, we should not be surprised to find ourselves in a habitable one!

But while we are waiting for that theory—and it could be a long wait—we can check whether anthropic selection offers a tenable explanation for the apparent fine tuning.  Such a hypothesis could even be refuted. This would happen if our universe turned out to be even more specially tuned than our presence requires.

Let me give a simple example of this style of reasoning.

Even if we knew nothing about how stars and planets formed, we would not be surprised to find that our Earth’s orbit was fairly close to circular. Had it been highly eccentric, water would boil when the Earth was at perihelion, and freeze at aphelion—a harsh environment, unconducive to our emergence. A modest orbital eccentricity, up to 0.1 or so, is plainly not incompatible with life. Had it turned out that the Earth moved in a near-perfect circle with eccentricity 0.000001, then some quite different explanation would be needed; anthropic selection from orbits whose eccentricities had a Bayesian prior that was uniform in the range 0 to 1 could plausibly account for an eccentricity of 0.1, but not for one as tiny as this.

We can apply this style of reasoning to the important numbers of physics, such as the cosmological constant Λ, to test whether our universe is typical of the subset that could harbor complex life. The methodology requires us to decide what range of values is compatible with our emergence. It also requires a specific theory that gives the relative Bayesian priors for any particular value within that range. In the case of Λ, are all values equally probable? Or, conversely, does the physics favor values especially close to zero? Are there even just a finite number of discrete possible values? With this information, one can then ask if our actual universe is typical of the subset in which we could have emerged. If it is a grossly atypical member even of this subset, not merely of the entire multiverse, then we would need to abandon our hypothesis.

Most physicists would consider the natural value of Λ to be large, because it is a consequence of a very complicated microstructure of space. Perhaps there is only a rare subset of universes where Λ is below the threshold that allows galaxies and stars to form. In our universe, Λ obviously had to be below that threshold. But if our universe were drawn from an ensemble in which Λ was equally likely to take any value, we would not expect it to be too far below it.

Current evidence suggests that the dark energy has an actual value five to ten times below that threshold. That would put our universe between the tenth or twentieth percentile of universes in which galaxies could form. Our universe is not significantly more special, with respect to Λ, than our emergence demanded. But suppose that, contrary to current indications, observations showed that Λ made no discernible contribution to the expansion rate, and was thousands of times below the threshold, not just five to ten times. This overkill precision would raise doubts about the hypothesis that Λ was equally likely to have any value, and suggest that it was zero for some fundamental reason—or that it had a discrete set of possible values, and all the others were well about the threshold.

I have taken Λ just as an example. We could analyze other important numbers of physics in the same way, to test whether our universe is typical of the habitable subset that could harbor complex life. The methodology requires us to decide what values are compatible with our emergence. It also requires a specific theory that gives the probability of any particular value. For instance, in the case of Λ, is there a set of discrete vacuua or a continuum of values? In the latter case we need to know whether all values are equally probable, or whether the probability density is clustered at a low value. The fact that our universe contains an ugly mix of ingredients—five percent baryons, twenty-five percent dark matter, and seventy percent dark energy—need occasion no surprise if the ensemble rings the changes on these proportions, and these are within the anthropically allowed range. 

In Before the Beginning I argued that what we have traditionally called our universe is just one island in an infinite cosmic archipelago—the multiverse.¹ My arguments were motivated by the seemingly biophilic and fine-tuned character of our universe; they had been bolstered by the cosmic inflation theory of the 1980s. New lines of evidence elevate the multiverse concept to a new level of credibility.

I have, ever since, had a close-up view of this shift in opinion and the emergence of these, admittedly speculative, ideas.

In 2001, I helped to organize a conference on this theme. It took place in Cambridge, but not at the university. I hosted it at my home, a farmhouse on the edge of the city, which has a converted barn that provided a good location for our discussions. Some years later, we had a follow-up conference. This time the location was very different: a rather grand room in Trinity College, with a portrait of Newton behind the podium.

The great physicist Frank Wilczek attended both meetings. As the second, where he gave the summary talk, he contrasted the atmosphere at the two gatherings. At the first they were:

physicists who subsisted on the fringes, voices in the wilderness who had for many years promoted strange arguments about conspiracies among fundamental constants and alternative universes. Their concerns and approaches seemed totally alien to the consensus vanguard of theoretical physics, which was busy successfully constructing a unique and mathematically perfect Universe.²

But at the second meeting “the vanguard has marched off to join the prophets in the wilderness.”

At my conferences and their successors the agenda focused on two issues. First, how real is the fine-tuning? How sensitive is our existence to a small tweak in the governing laws? What are the requirements for life’s emergence? There is obviously a danger that our imaginations are too limited. We have evolved in symbiosis with our environment as have all animals. We should not be surprised that our legs are just long enough to reach the ground.

If there is a multiverse, it will take our Copernican demotion one stage further—our solar system is one of billions of planetary systems in our galaxy, which is one of billions of galaxies accessible to our telescopes—but this entire panorama may be a tiny part of the aftermath of our big bang—which itself may be one among billions. It may disappoint some physicists if some of the key numbers they are trying to explain turn out to be mere environmental contingencies—no more fundamental than the parameters of the earth’s orbit round the sun. But in compensation, we would realize space and time were richly textured—but on scales so vast that astronomers are not directly aware of it—any more than a plankton whose universe is a spoonful of water would be aware of the world’s topography and biosphere.

Appendix 1. Prerequisites for a Complex Cosmos

We can trace back to one second after the beginning. Indeed, we can probably be confident back to a nanosecond; that is when each particle had about 50 GeV of energy—as much as can be achieved in the LHC—and the entire visible universe was squeezed to the size of our solar system. But questions like “Where did the fluctuations come from?” and “Why did the early universe contain the actual mix we observe of protons, photons and dark matter?” take us back to the even briefer instants when our universe was still more hugely compressed—the size of a tennis ball, and perhaps even of microscopic size—where experiments offer no direct guide to the relevant physics.

There is much evidence to support the so-called inflationary universe model. It may be useful to summarize the essential requirements for the emergence of our complex and structured cosmos from simple amorphous beginnings.

The first prerequisite is, of course, the existence of the force of gravity—which, as explained earlier, enhances density contrasts as the universe expands, allowing bound structures to condense out from initially small-amplitude irregularities. It is a very weak force. On the atomic scale, it is about forty powers of ten weaker than the electric force between an electron and a proton. In any large object, positive and negative charges almost exactly cancel. By contrast, everything has the same sign of gravitational charge, so when sufficiently many atoms are packed together, gravity wins. Stars and planets are so big because gravity is weak. Were gravity stronger, objects as large as asteroids, or even sugar-lumps, would be crushed by it. Although gravity is crucial, it is also crucial that it should be very weak.

There must be an excess of matter over antimatter.

Another requirement for stars, planets, and biospheres is that chemistry should be non-trivial. If hydrogen were the only element, chemistry would be dull. A periodic table of stable elements requires a balance between the two most important forces in the microworld: the nuclear binding force—the strong interactions—and the electric repulsive force that drives protons apart.

There must be stars—enough ordinary atoms relative to dark matter. Indeed, there must be at least two generations of stars: one to generate the chemical elements, and a second able to be surrounded by planets.

The universe must expand at the right rate, not collapse too soon, nor expand so fast that gravity cannot pull together the structures.

There must be some fluctuations for gravity to feed on—sufficient in amplitude to permit the emergence of structures. Otherwise the universe would now be cold ultra-diffuse hydrogen—no stars, no heavy elements, no planets and no people. In our actual universe, the initial fluctuations in the cosmic curvature have an amplitude of 0.00001. According to inflationary models, this amplitude is determined by quantum fluctuations. Its actual value depends on the details of the model.

Appendix 2. Nomenclature

I have learnt that it is necessary, in discussions with philosophers, to inject an apology at the start for ambiguous nomenclature. They would define universe as everything there is—and if that is the definition, then there plainly cannot be more than one. If there are other domains—perhaps originating in other big bangs, and perhaps differing from our universe in size, content, or dimensionality—then we should really define the whole enlarged ensemble as the universe. We then need a new word, such as metagalaxy, to denote the domain to which cosmologists and astronomers have observational access. So long as this whole idea remains speculative, it is probably best to leave the term universe undisturbed with its traditional connotations intact, even though this then demands a new term, multiverse, for the whole (still hypothetical) ensemble.

Martin Rees

George Ellis replies:

I have debated the multiverse issue with Martin Rees on various occasions, and his response is a nice presentation of his views, but it is also full of what ifs. There is nothing wrong with such speculation, but that is what it is—speculation. As I have said elsewhere, of all the options he discusses, the one I prefer most is Smolin’s, but that is as yet only a partially completed theory. “So if we are perplexed,” writes Rees, “by the fine tuning of the physical laws, it seems an attractive idea that our Big Bang is just one of many.” Yes that is a nice philosophical argument; it is a good philosophically-based scientific proposal. But is it testable? No.

Do we know what is in the unobservable interior of black holes? Yes, insofar as Einstein’s theory of relativity is correct, and it is well tested.

We may one day have a convincing theory that accounts for the very beginning of our universe, tells us whether a multiverse exists, and (if so) whether some so-called laws of nature are just parochial by-laws in our cosmic patch.

Maybe so—but will that theory be testable? If not, which would probably be the case, we are back at square one.

Rees then goes on to give philosophical arguments for a multiverse based on the anthropic principle and the idea of fine tuning. But first deciding what is and what is not fine-tuned depends on having a well-defined measure on the family of universes in a multiverse. Because they are alleged to be infinite in number, most measures proposed diverge and so are not well defined. In any case, there is no way to experimentally or observationally test if a proposed measure is correct or not—it is back to a philosophically based decision. The whole underlying concept of naturalness can be questioned, as in the writing of Sabine Hossenfelder.³ It is one philosophy against another: John Baez agrees with Hossenfelder.⁴

I like Rees’s article, it is a good defense. In the end, it is a version of Dawid’s no alternatives argument—see the letter from Dawid and my response for more on this. There is, of course, an alternative: maybe things simply were indeed fine-tuned to be that way. That is yet another possible philosophical position. As stated by Paul Davies:

For a start, how is the existence of the other universes to be tested? To be sure, all cosmologists accept that there are some regions of the universe that lie beyond the reach of our telescopes, but somewhere on the slippery slope between that and the idea that there are an infinite number of universes, credibility reaches a limit. As one slips down that slope, more and more must be accepted on faith, and less and less is open to scientific verification. Extreme multiverse explanations are therefore reminiscent of theological discussions. Indeed, invoking an infinity of unseen universes to explain the unusual features of the one we do see is just as ad hoc as invoking an unseen Creator. The multiverse theory may be dressed up in scientific language, but in essence it requires the same leap of faith.⁵