Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Cosmology: The untestable multiverse

George Ellis reminds us that Brian Greene's beguiling book on parallel worlds is more theory than fact.

The Hidden Reality: Parallel Universes and the Deep Laws of the Cosmos

  • Brian Greene
Knopf/Allen Lane: 2011. 384 pp. $29.95/£25 9780307265630 | ISBN: 978-0-3072-6563-0

Cosmology must seem odd to scientists in other fields. More and more accurate data about the distant Universe are being generated by high-tech observational techniques, giving rise to an era of 'precision cosmology' — but to explain these impressive data, cosmologists are increasingly turning to untestable theories.

In The Hidden Reality, theoretical physicist Brian Greene explores one of the strangest proposals: that we live in a multiverse. This fashionable concept supposes that large, perhaps infinite, numbers of separate universes exist in parallel to ours. Using well-constructed analogies, Greene explains nine different multiverse proposals, ranging from simple extrapolations of cosmological models to those based on quantum field theory, string theory and pure logic.

Greene carefully sets out the reasoning behind each proposal and the grounds for why it might be true. His arguments are absorbing and convincing on their own terms. But they inexorably lead us farther from the gold standard of testability that is the scientific method. The beguiling text moves beyond established science into philosophical speculation.

Izabella Godlewska de Aranda's painting Cosmic Joy! (2009) hints at the idea of many universes. Credit: THE BRIDGEMAN ART LIBRARY

Greene's nine types of multiverse are as follows. First, if space extends forever, an infinite number of domains similar to ours might lie beyond the part of the Universe that we can see. Second, some versions of inflationary theory — the idea that the newborn Universe had a fleeting period of super-fast accelerating expansion — predict the existence of innumerable other universes, with different characteristics from our own. Third, string theory, the pre-eminent theory of quantum gravity, suggests that our Universe might be one of many four-dimensional 'braneworlds' floating in a higher-dimensional space-time.

This option is developed further in the fourth and fifth proposals, which involve cyclic universes, or variations on physical parameters that are possible in the string-theory landscape. The sixth is a quantum mechanics idea that many worlds simultaneously exist as branches of the wave function of the Universe. The seventh suggests that the Universe is a holographic projection. The eighth states that we live in one of a set of artificial universes created as simulations on a super-advanced computer. The ninth argues that it is a philosophical necessity that every possible universe must be realized somewhere, in “the grandest of all multiverses”.

By presenting this plethora of theories, Greene gives the impression that the multiverse is on a sound scientific footing, but these nine arguments are mutually exclusive. We do not know how to test which is right, if any, because we cannot make direct observations of domains beyond the observational horizon — the greatest distance that light can have travelled towards us since the Universe became transparent to radiation 300,000 years after the Big Bang. Given this lack of evidence, there is a viable tenth option: that there is no multiverse at all.

Greene cites indirect evidence to support the multiverse idea. The values of the physical parameters seem to be fine-tuned to allow life. For example, if the strength of the cosmological constant — currently causing the accelerated expansion of the Universe — was much different, then galaxies would not exist and we would not be here to make measurements. Similarly, the strength of the strong nuclear force permits atoms, and hence humans, to exist. Such anthropic reasoning invokes multiverses where there is some likelihood that physical constants take on different values in each.

But probabilistic arguments only make sense if these parallel universes actually exist. And logic cannot prove their existence. For instance, a multiverse model may predict a likely value of the cosmological constant, but the reverse is not true. A particular measurement of the cosmological constant does not require a multiverse. Nor can the multiverse concept be disproved by any specific observationally determined value of the cosmological constant, for multiverses can accommodate any value. These arguments can only provide probabilistic consistency tests for some kinds of multiverse.

So one can motivate multiverse hypotheses as plausible, but they are not observationally or experimentally testable — and never will be. It is easy to support your favourite model over others because no one can prove you wrong — you can simply adjust its parameters to fit the latest information. If the Universe is a simulation (option eight), then anything is possible. However, the existence of a computer allowing such a simulation is not remotely feasible. Scientists are beginning to confuse science with science fiction.

Greene, to his credit, devotes a chapter to the question of whether the multiverse idea is a scientific theory or not. He believes it is, and even supports the extravagant claim that infinities exist — infinite numbers of universes hosting countless galaxies. This leads to well-known paradoxes, such as the infinite repetition of everything because of the finiteness of possibilities. But again, there is no way to test it, because infinity is always beyond reach — and so will not plausibly exist in physical reality, as mathematician David Hilbert argued.

The gap in current theories that warrants pursuing such untestable theories is our inability to predict firmly why physical constants have the values they do. If a fundamental theory were to be proposed that explained them, the drive for a multiverse explanation would fall away. But the puzzle of why these values allow life would remain.

The multiverse argument is a well-founded philosophical proposal but, as it cannot be tested, it does not belong fully in the scientific fold. Read The Hidden Reality with enjoyment, but beware its misleading title. Greene is not presenting aspects of a known reality; he is telling of unproven theoretical possibilities.

Author information

Authors and Affiliations


Rights and permissions

Reprints and Permissions

About this article

Cite this article

Ellis, G. Cosmology: The untestable multiverse. Nature 469, 294–295 (2011).

Download citation

  • Published:

  • Issue Date:

  • DOI:


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing