Skip Navigation

Mysterious cosmos

August 6, 2004 By Philip Ball This article courtesy of Nature News.

Why is our Universe so exquisitely tuned to host life? Using the anthropic principle to explain the world might be a tempting alternative to invoking God, but it’s not science, says Philip Ball.

Please log in to rate this page.

View Comments

We are lucky to be alive. Extraordinarily lucky. So lucky, in fact, that some people can only see God's hand in our good fortune.

Creationists are fond of pointing out that if you mess with the physical laws of the Universe just a little, we wouldn't be here. For example, if the neutron were just 1% heavier, or the proton 1% lighter, or the electron were to have 20% more electrical charge, then atoms could not exist. There would be no stars, and no life.

But although creationists rejoice in the divine providence that has made the Universe exquisitely contrived to support life, science has long argued for an alternative explanation: the anthropic principle.

The theory has been supported by several leading physicists and astronomers, from Fred Hoyle to Steven Weinberg, who claim it reduces the mystery of our existence to a logical necessity.

Yet the whole idea is roundly trashed by Lee Smolin, a renowned quantum-gravity theorist at the Perimeter Institute in Waterloo, Canada. Smolin asserts, in a preprinted paper on Arxiv1, that the anthropic principle is not a scientific theory at all, because it lacks the basic requirement of falsifiability. It is impossible to prove the anthropic principle wrong - hence it is outside the remit of science.

Circular argument

In truth, the idea always had an air of circularity about it. Crudely put, the anthropic principle states that our universe must look the way it does (that is, primed for life), because if it didn't, we wouldn't be here to argue about it.

But there is a little more sophistication to the idea, which has enabled some researchers to claim that they have put the principle to the test.

The argument usually goes something like this. Let us assume that our Universe is in fact just one of many. This collection of universes is called the multiverse, and its members cannot affect each other in any way. Assume also that physics, by which we mean the fundamental constants of nature, such as Planck's constant and the speed of light, differs more or less at random in each universe.

Most of these universes would be unable to support life. But if there were enough universes, a tiny fraction of them should have just the right parameters to give rise to a cosmos like ours.

Then, the anthropic principle asserts, the intelligent beings in those lucky universes would marvel at how their universe seemed fine-tuned for life, unaware of the countless other universes that remain barren forever. There is no need to invoke God to explain our precarious existence; chance alone suffices.

How one arrives at the multiverse is another matter, but there are possible mechanisms for that. For example, an extension of inflationary theory called eternal inflation suggests that new universes could continually blossom from tiny regions of a precursor universe2.

Alternatively, new universes could be spawned by bouncing black holes. General relativity predicts that sufficiently large stars can collapse without limit under their own gravity to produce a point of space-time that is infinitely small and infinitely dense: the 'singularity' of a black hole. But the quantum-mechanical effects that must take hold at very small scales could in theory cause this collapse to reverse, so the black hole could rebound to form an entirely new universe, which would be a region of space isolated from the universe in which the black hole originally formed.

Put to the test

So in principle one can make new universes, and the anthropic principle could be true. But the real question is, does it actually predict or explain anything? This is what Smolin disputes. He thinks that not only has the idea failed to produce any testable predictions, it cannot do so even in principle.

That is quite a controversial view. Fred Hoyle, for example, used the anthropic principle to successfully predict the existence of a certain energetically excited state of the carbon nucleus. He argued that there would be no life as we know it without carbon, which can only be produced by stars. And he calculated that carbon atoms could only be made in stars in significant quantities if carbon possessed this particular state, which had never at that stage been observed.

Armed with his prediction, astronomers duly looked for the excited state, and found it. It sounds impressive. But Smolin points out that there is nothing anthropic about this reasoning. We know that carbon exists in the Universe irrespective of the existence of life, so life plays no essential role in the logic.

Look at it another way: suppose the predicted state of carbon had not been seen. Would we conclude that the anthropic principle was wrong? Clearly not, because carbon does, in fact, exist, and so do we. We would have to conclude instead that there must be some way of generating carbon other than the one Hoyle proposed, or that his calculations were wrong.

Much the same applies to a prediction of the value of the cosmological constant, made by Steven Weinberg in 1987. The cosmological constant was originally proposed by Einstein as a way of stopping his equations of relativity from predicting an expanding Universe (it was at that time assumed to be static). The idea of such a constant is now back in favour, because it might, this time, be used to rationalize the recently observed acceleration of the Universe's expansion.

Weinberg invoked the anthropic principle to argue that the constant could not be very large, or the Universe would expand too fast for galaxies, stars (or us) to form. A subsequent refinement of his argument predicted that the probability that the cosmological constant has the value suggested by current astrophysical observations is only around 10 %.

In other words, the principle seems to make a prediction that, if not stunningly accurate, is certainly plausible. But again, Smolin says, "What is actually doing the work in the arguments is never the existence of life or intelligent observers, but only true observed facts about the universe, such as that carbon and galaxies are plentiful." Our own existence doesn't 'predict' anything.

"The anthropic principle is never going to give falsifiable predictions for the parameters of physics and cosmology," Smolin asserts.

No need for God

So are we forced back to the hypothesis of God? Not at all. Smolin adduces an alternative that, he claims, is scientifically falsifiable. He calls it 'cosmological natural selection'.

The most obvious scientific theory that accounts for a situation that, at first sight, seemed highly improbable is natural selection, points out Smolin. Darwin's idea abolished the need for the Reverend William Paley's heavenly 'watchmaker' to fashion the beautifully 'designed' products of biology.

Smolin believes that a similar principle could save us from making the same mistake about the Universe. This does not involve indulging in any "mysticism about living universes", he reassures us. Rather, he suggests that if there exists some process by which parent universes spawn new universes with small, random changes in their physical parameters, and if the characteristics of a universe determine how many progeny it produces, then fine-tuned universes like ours can arise by cosmological natural selection.

In particular, if new universes are produced by black-hole bounces, then universes in which stars (and thus black holes) can form are 'fitter' than others. After a period of time, you would expect the universes produced by this process to have a set of cosmological parameters that maximizes the number of black holes that can form.

That conclusion helps to explain why we are here, since universes in which complicated structures such as stars and black holes can form are also likely to be hospitable to life, Smolin argues. It also gives us a way to test the idea.

Smolin points to astrophysical measurements that we are able to make now, that could refute cosmological natural selection. For example, the existence of neutron stars with a mass greater than 1.6 times that of the Sun would scupper the idea. So, it could be wrong...but at least it's science.


  1. Smolin S., et al. (2004).
  2. Linde A., et al. (2003).
  3. Weinberg S., et al. Phys. Rev. Lett., 59. 2607 (1987).


User Tools [+] Expand

User Tools [-] Collapse

Pinterest button


Please log in to add this page to your favorites list.

Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.