An Interview with Martin Rees, Astronomer Royal


Martin Rees, Baron Rees of Ludlow, has been the Astronomer Royal since 1995, and is also Emeritus Professor of Cosmology and Astronomy at the University of Cambridge, and Honorary Fellow of King’s College, Cambridge. On behalf of the King’s Review, Professor Michael Proctor, Provost of King’s College, sat down with Professor Rees to talk about the past and future of cosmology, the big bang, and the possibility of alien life.

Michael Proctor: Within the twentieth and, even more so, the twenty-first century, we’ve seen an explosion in our understanding of the universe. How do you see this in the context of astronomy since Newton in the last few centuries?

Martin Rees: Since the earliest times, people everywhere have looked up at the stars and wondered about them. The night sky is the most universal feature of the human environment. Newton was the first person to actually quantify what was happening in the solar system. Of course, right back to the Babylonians, records of the sky had been taken. They didn’t understand what was happening, but they discerned regularities and predicted eclipses. After Newton, the structure of the solar system became better understood. Newton was lucky to hit on one of the few phenomena in nature which was both understandable and predictable. The orbits of the planets were understandable and predictable, and that was really the first exact science.

The next step forward was in the nineteenth-century when people realized that the stars were made of the same stuff as the Earth. Of course, the traditional view had been that the vault of heaven was made from some kind of ‘fifth essence’. From the mid nineteenth-century, astronomers were able to take spectra of the brightest stars and of the sun and realize that they are made of the same stuff. That allowed astronomers to do not just dynamics, but to do physics. That’s the process whereby we came to understand so much about the universe, by applying the physics that we know on Earth to the stars.

What has benefited us more through the last century is the burgeoning of technology. We’re certainly not as wise as Newton was; arm-chair theory alone wouldn’t have carried us far. Progress has stemmed from more powerful telescopes, being able to observe the sky from space, and so forth. But I think it’s amazing we’ve got as far as we have and that is because astronomy is based on physics and physics is easier than biology.

Another point I should have mentioned is that even where we can’t do experiments on what’s out there, we’ve benefited hugely from computer simulations – experiments in a virtual world. Computers are now powerful enough to do realistic simulations of what happens when stars explode or galaxies crash together – this is a really crucial development in the last ten or twenty years. Now we can use the kind of programs that aero-engineers use test aerofoils, as a substitute for wind tunnel tests. If they have enough confidence in them in that context, we should have reasonable confidence in the outcome of simulations of stars and galaxies.

Cosmology used to be described as a science with only two numbers in it: that was said by those who liked the subject. Those who didn’t like the subject said it wasn’t a science at all – because it dealt with a unique entity, and you couldn’t really say much about the Universe as a whole. But cosmology has now been transformed. Observations made by all kinds of techniques, both on the ground and in space, have revealed a great deal about the range of objects that exist in the Universe – from planets, up to galaxies. Also, we have, at least in outline, a ‘timechart’ of how our Universe has evolved, from some mysterious beginning about 13.8 billion years ago to its present structure. One challenge now is to fill in the details of how the first atoms, stars, galaxies and planets formed – to understand our ‘cosmic environment’, as it were. A second challenge it to probe back very close to the beginning, where there are still mysteries. We’d like to understand why the Universe contains the ‘mix’ of matter and radiation that we observe, why the Universe is expanding.

MP: There are a number of new observational devices available to cosmology – the Planck instrument, for example. Is it likely that further advances in observations might yield useful information concerning the challenges that you describe?

MR: We now, thanks to huge amounts of excellent data, have a very good understanding of galaxies. We know that they contain stars and gas, but they also contain dark matter: this is made up of a swarm of particles that behave as though they are affected by gravity but no other forces. But we don’t know exactly what these particles are: we only know their aggregate properties. Another big challenge is to understand how galaxies and clusters have evolved from the early Universe, which was hot, dense, and almost smooth. Here we’ve actually had wonderful success, especially by careful measurements of the microwave radiation which is an ‘afterglow’ of the big bang – and how its temperature varies between different directions.

But these advances bring into sharper focus all the things we don’t understand: one of course is what this dark matter is. We know it behaves as though it’s a swarm of electrically-neutral particles, but there’s a variety of possible candidates. In the early Universe there was not only radiation, and a lot of atoms, but there was this extra stuff, and we’d like to understand it. There are three lines of attack on that problem. The first is to hope that physicists will discover new particles that have properties consistent with the dark matter. That hasn’t happened yet. The experimenters at the Large Hadron Collider haven’t found any particles, apart from the Higgs particle. But they are still looking for so-called super-symmetric particles.

The second method is to actually detect the particles directly. This can be done because the dark matter pervades the entire galaxy. Some of these particles would be moving through this room at about 200 km per second. They mainlygo straight through matter. But occasionally one of them would collide with one of the atoms in the room; delicate experiments could detect the ‘recoil’ when this happens. Searches of this kind have to be done deep underground to get rid of all the backgrounds, and to seek the rare events when one of these dark matter particles interacts with an atom in a sensitive detector. Several such experiments have already been done. So far there are no compelling detections, but as instruments improve (and involve a larger mass of detector material) they are reaching the sensitivity level where they might expect to see something.

The third line of attack is to look up into space, because some ideas suggest that these particles would collide with each other and annihilate themselves, converting their energy into high energy photons, particles of light, or into positrons. There are detectors in space which are sensitive to energetic photons (gamma rays), and to fast-moving particles – but here again there are still no more than tantalising clues. It’s hard work actually pinning down what this dark matter is. All we can say now is that there’s about five times as much mass in it, in terms of gravitational effects, as there is in all the stars and gas we see. That’s why it’s so important.

MP: It’s just detected in the large by its gross gravitational effect?

MR: In the large, yes. And we can infer something about its gross properties. We know it’s not like a fluid or a gas by observing what happens when galaxies or clusters have collided. If it were a fluid you’d get shocks and compression waves. Whereas if it’s a swarm of independently moving particles, two swarms go through each other and ‘phase mix’. The latter option fits the data best.

MP: That’s quite exciting. But the path from observation to theory is quite indirect in this case. You have this great effect of dark matter and then you have to find the particles, before you can begin to understand them. So we are a long way from reaching such an understanding.

MR: Fortunately, one can make progress in understanding the effects of the dark matter even without understanding exactly what it is. Indeed, we cannot get a consistent picture for the formation of galaxies without it. Clusters of galaxies are the biggest self-gravitating structures in the Universe. One can measure the total mass of these systems in a number of ways. You can calculate the total mass needed to bind the system and stop it flying apart. And the temperature of gas in the cluster can be measured by detecting the energy of the X-rays that very hot gas emits, and that tells you the depth of the gravitational potential well it’s confined in. The third thing you can do is to look for the effect of gravitational lensing, light-bending, by a cluster of galaxies. Distant galaxies being viewed though a foreground cluster seem magnified, and distorted into streaky arcs, by the ‘lensing’ due to the cluster’s gravity. The cluster behaves like a poorly figured converging lens, and from the strength of the focusing you can infer its mass.

MP: And its distribution as well.

MR: Yes. And these three very different lines of evidence all give consistent estimates of how much dark matter exists in clusters of galaxies. I stress this because there have been conjectures that dark matter does not exist and that what we’ve got wrong is the theory of gravity. It’s indeed true that if the inverse square law of gravity broke down at large distances, the inferences would change. But you’d then be jettisoning not just Einstein but Newton too, and most of us are unwilling to do that except as a last resort. I personally don’t think such ideas should be taken very seriously unless and until we exhaust all the options for dark matter. Moreover, any alternative theory has to mimic the effects of dark matter consistently in very different ways – in light-bending, as well as its effect on galaxies and gas.

MP: Do you regard String Theory as a science? Are there any experiments that can validate it at this stage?

MR: The theory hasn’t yet been related to any phenomena we can observe or measure. Most of its proponents accept this: it won’t be taken seriously as science until there are such tests or until it yields formulae for basic numbers which we can’t otherwise explain. We are not going to have any direct measurements at the huge energies where its effects dominate because that’s billions of times higher than the energies we can achieve in the accelerator – indeed the best hope is cosmological observations which probe, at least indirectly, the initial instants of the big bang.

But there’s one important point here. One will never have the data to test every consequence of any theory. But that would be expecting too much. You have to be able to test a theory robustly enough to gain confidence in it; having done that, you then take seriously the predictions that you can’t directly verify or refute. For example, we believe what Einstein’s theory of gravity says about the inside of black holes, even though we can’t observe there, because we have been able to test and corroborate the theory in many other ways, outside black holes.

MP: You talked about the variations in the cosmic microwave background as being the starting point for numerical calculations, but is there any explanation for the anomalies that you actually observe?

MR: It’s very important in presenting cosmology, especially to non-specialists, to keep clear water between the parts that are well-established and the parts that are still speculative. I would criticize some popular writers for either over-egging the more speculative bits, or blurring that distinction. I would claim that we have great confidence in extrapolating back to when the Universe was one second old. That’s the era when nuclear reactions turn hydrogen to helium and deuterium, and the predictions match well with what we find. Indeed that extrapolation can be made with as much confidence as most inferences that geologists offer about the early history of our Earth. Cosmologists have ‘fossils’, and indeed data that are easier to interpret and more quantitative. So back to one second after the big bang, cosmology is now no more ‘speculative’ than geophysics.

Indeed, we have reasonable confidence extrapolating back a bit further – to a nanosecond. That’s the time at which all the particles are moving around with an energy similar to what is achieved in big accelerators like the LHC. If we try to extract it back still further, then of course the conditions are beyond the direct reach of experiments. When the universe was a nanosecond old, everything we now see out the limits of our horizon (including billions of galaxies) would be squeezed to the size of our solar system. But to address fundamental questions about why the universe is expanding the way it is we have to go back far further still – to an era when the Universe would be squeezed down to the size of a tennis ball. That is a huge backward extrapolation. There is a theory called inflation which would apply at that very early stage and which has survived some tests. The fluctuations predicted by this theory are consistent with what we actually observe. So it has something going for it. Inflation theories have been studied for 30 years, and there are a huge number of variants with regard to the detailed physics. We would like to narrow the range down by further observations. (And some results announced just last month, from an experiment called Bicep 2 which measured a new kind of fluctuation, are an important further step). In particular we would like to know whether our Big Bang the only one or were there others as well.

MP: There are physical quantities, like the fine structure constant and other physical constants, which obviously have an effect on the universe. Could these take different values from the observed ones, and still lead to a sensible universe? Are there any constraints which apply?

MR: The first thing to say is that we would not make any progress in cosmology were it not that these constants do seem to be ‘universal’, at least over the cosmic domain that we observe. One could envisage, in principle an ‘anarchic’ universe where different galaxies, or even different stars, were governed by different physics. It does seem that the strength of gravity, the mass of the electron, and other basic numbers of physics are the same everywhere we can measure them. And we can make that statement with pretty high precision. We can tell that gravity has not changed very much in the past because it would make stars evolve differently. If gravity had been stronger in the past, then the centres of stars would have been squeezed more and nuclear fusion would have burnt them up more quickly. And the early universe would have expanded faster. So we infer that gravity has not changed very much. We can also analyse the light from distant galaxies. In their spectra you see various features which correspond to different chemical elements. These spectra indicate that the atoms in distant galaxies are the same as those in the lab to precision of one part in a million, and that the mass of the electron has not changed more than that.

We have fairly good evidence that a basic uniformity prevails throughout the part of our Universe that we can see. But this opens up a very important question. Is the observable domain a large part of overall physical reality, or could it be a tiny (and perhaps unrepresentative) fragment of some still vaster whole? Even the most conservative cosmologists would accept that there is a lot of material beyond what we can see. Our ‘horizon’ is delineated by the distance light can travel since the big bang. There’s no more reason to believe that this is a limit to the overall universe, any more than you believe the ocean ends just beyond your horizon. There are indeed fairly strong reasons for thinking that the Universe extends thousands of times beyond the domain we can see. If you look as far as you can in one direction, and then in the opposite direction, then conditions look just the same. If there is an overall gradient, it is very slight – less than one part in a hundred thousand across the distance we can observe, That suggests that if we’re in a finite ‘island universe’ that has an edge, that edge is far further away than our horizon

MP: But that’s a philosophical question. If light can only have travelled as far as we can see since Big Bang, or does it mean to say there is something beyond that? Does that mean there has to be another Big Bang in another place?

MR: No, in fact one has to be wary of over-interpreting the speed of light because the famous ‘speed limit’ only applies in flat space, and when the speed is measured by a local and stationary clock. For instance, if we take the present favoured cosmology seriously, and the Universe is accelerating, what happens if you watch a particular galaxy? You will only see a finite part of its future. It gets more and more red shifted and its clock appears to us to go slower and slower, so however long we watch it for, we only see a finite part of its future. It’s rather like what happens if you watch something (or someone) falling into a black hole: the theory says you only see a finite part of its future, you don’t see the final spaghettification in the centre of the black hole. So most people would accept unobservable galaxies beyond the horizon as part of our physical reality. When we get to the idea of other Big Bangs then it becomes a bit more speculative.

These may well be unobservable, but I still think they are part of (albeit speculative) physics and not metaphysics. I like to present this argument as an exercise in aversion therapy. If you are scared of spiders, you start off with a little spider a long way away and end up with tarantulas crawling all over. You start off by noting there are galaxies which we can’t observe because they are too far away. In a decelerating universe they’d eventually come into view, but because of the acceleration they can never be observed even in principle. But that doesn’t make them less ‘real’. Most of us are happy with that, so why should we be less happy with the epistemological status of unobservable galaxies which are the aftermath of a different Big Bang?

If there are other Big Bangs, then would they be governed by the same physics as ours? This is a very interesting question because many theorists, like those who work with String Theory, suspect that there is nothing unique about our physics, and in particular they think that there could be spaces where the repulsive force that we call lambda could be much stronger. And the micro physics could be different too: for instance, electrons could have different masses. Some of these universes would be sterile or stillborn – the laws may not allow complex chemistry, emergence of stars and planets, enough time for evolution, etc. We would then find ourselves not in a typical universe, but a typical member of the subset that allows complexity to develop. This is an instance of what’s called ‘anthropic reasoning’ – something that makes some physicists foam at the mouth, but which I think may be unavoidable.

I often give talks about this theme and tell the audience that if you don’t like the idea of multiverses, then just regard this as an exercise in counter-factual history – rather as you might conjecture, say, what would have happened if the dinosaurs had not been killed off, or if the Brits had fought better in 1776. So in the same spirit, even those who don’t like the idea of multiple universes can develop their intuition by asking what the universe would be like if key parameters were a bit different. One definite requirement for a complex universe is a force of gravity – but it’s best if it’s very weak. Gravity is about 40 powers of 10 weaker than electric forces on the atomic scale. It ‘wins’ on big scales because, whereas positive and negative electrical charges are almost in balance in any large object, everything has the same sign of ‘gravitational charge’. And it is only because gravity is so weak for that reason that our cosmos can be so big, and entities like us can exist without being crushed by gravity. On the microscopic scale, of atoms and molecules, gravity is negligible.

Imagine a set of objects of increasing mass – sugar lumps, asteroids, planets and stars. For the sugar lumps, gravity is ineffective, for the asteroids, ditto, though for objects bigger than asteroids gravity is powerful enough to make it round – like a planet. Planets heavier than Jupiter would actually be crushed, and you then get into the realm of the stars. Because gravity is so weak, there are many powers of ten between the microworld and the cosmos. And that is essential for our existence, because structures like human beings are very large compared to atoms, with have layer upon layer of complex structure, but we can stand up rather than being crushed by planetary gravity.

So there has to be one large number in cosmology, a large number that reflects the weakness of gravity compared to the micro physical forces. Because gravity is weak, stars are big and live a long time. But there are other requirements for a complex cosmos. The existence of a periodic table of elements requires a rather delicate balance between the two most important forces in the microworld: the electrical force, which repels protons from each other, and the nuclear force which binds them together in a nucleus. It’s this same delicate balance which allows nuclear fusion to release the energy that powers the stars. Otherwise there would be just hydrogen – no fuel for the stars, no complex chemistry and no ‘us’.

MP: That is a very good explanation of why we are as we are. Of course, you could have a different large number and you get something similar. There must be a range.

MR: Yes, gravity isn’t fine-tuned. It just has to be weak. If it were still weaker that might be even better, insofar as there would be even more time, and an even larger object could assemble before gravity crushes it. In contrast, some of the other numbers have to be tuned more carefully to get the periodic table. The balance between the nuclear and electric forces has to be fairly close. The other number which is important is the number lambda which measures the cosmic repulsion. If that were much bigger then galaxies would never have formed because the Universe would have started to accelerate before they had had a chance to form and gravity would be overwhelmed.

MP: You’ve talked about very small intervals after the Big Bang. Is the Big Bang a singularity in time or only in space? If it’s not a singularity in time what happened beforehand?

MR: It’s in a sense a singularity in both, but we would suspect a singularity would be smoothed over if we had the right physics. The singularity is a signal that the physics is broken – that known physics has been applied beyond the range where it’s valid. As you go beyond the scales of the everyday world, either up or down, then you’ve got to jettison more and more common-sense notions. We’re used to the idea that the microworld confronts us with counterintuitive quantum effects. In the still more extreme conditions of the ultra-early universe then the idea of three spatial dimensions and time ticking away, may be oversimplified. Space and time may get screwed up and intermingled; extra dimensions may come into play and the idea of before and after might not be clear-cut. The idea of what’s before the Big Bang is a mystery at the moment. But I think what is remarkable is that we can talk with a straight face back to a nanosecond and say something about it and that’s progress in the last 40 years. When I was a research student back in the 1960s, even the idea of the Big Bang was controversial.

MP: Obviously, you think that the Big Bang is the correct explanation for the origins of the universe, as opposed to, say, the steady state theory. Do you feel there is a place for any sort of external force or creator?

MR: We don’t know about the very beginning. The Big Bang may be embedded in some grander structure where there are many, many Big Bangs. I think all bets are off with regards to how it started and whether there even was a beginning. I think the aim of studying the cosmos is to push back the causal chain. Newton explained why the planets move in ellipses, but he wrote in a couple of places that he found it mysterious that the planets were moving on more or less the same plane, what we call the ecliptic, whereas the comets come from more random directions. He thought that must be providence. We understand now why planetary orbits are more or less coplanar – it’s a consequence of their origin in a dusty protostellar disc. And we’ve pushed back the causal chain to understand the formation of atoms, stars and galaxies, but there’s always a further step as well.

MP: Is there ‘life out there’?

MR: That’s one of the most fascinating questions of all. But it’s a biological question. And biology is a more difficult subject than astronomy in that it deals with more complicated phenomena. We don’t even understand how life began on Earth. We understand how Darwinian evolution led from simple life to our complex biosphere. But people don’t understand the transition from complex chemistry to the first metabolising and reproducing systems. It’s gratifying that some really serious biochemists are now addressing this question When we understand that, it will tell us two things: whether it’s likely or unlikely that extraterrestrial life is widespread; and whether there is something particularly special about the chemistry on which terrestrial life is based. In other words, would we expect any other life to have the same DNA? So we don’t know that. Even the most firmly Earth-bound biologists would be fascinated by this issue. Another exciting prospect is that observations will tell us whether some of these planets have a biosphere. That will be do-able within ten or twenty years by the next generation of giant telescopes, powerful enough to provide sharp images, and collect enough light to identify the spectrum of a planet even if it’s millions of times fainter than its parent star.

MP: It’s all very exciting. Like the Universe, the discoveries are accelerating.

MR: Astronomy is a fundamental science, but it’s also an environmental science. If you look up at the stars, they’re the most universal part of the human environment. And what we’re discovering makes this environment seem more fascinating. I’m sometimes asked whether being an astronomer has any impact on my attitudes to everyday life. The one thing I certainly know, having lived among astronomers, is it doesn’t make them any more serene and relaxed about everyday matters (and you must know that as well). But there is one special perspective which they probably do have, and that’s an awareness of the far future. Most people are aware that we’re the outcome of about 4 billion years of evolution. But I’d guess that they tend to feel that we humans are the culmination – the endpoint of the process. No astronomer could think that way. We know the future is at least as long as the past: the sun is less than halfway through its life – the universe may even have an infinite future. So to us, it seems natural to suppose that humans are just a step on the way, maybe not even a halfway stage in the emergence of ever greater complexity. And this post-human evolution could happen here on Earth or somewhere beyond the Earth; it may be silicon-based rather than organic. This realisation gives us a different perspective on humans, but also perhaps an extra motive for concern about the future, because we realise that a catastrophe could foreclose an immense post-human era.

MP: It seems most unlikely considering the short time we’ve existed.

MR: Any creatures who will be alive to witness the death of the sun won’t be human – they could be as different from us as we are from protozoa, because the time between now and then is longer than the Earth’s resent age. Indeed future evolution is going to take place not on the Darwinian time scale, of natural selection, but on the technology time scale, because we’re obtaining the capacity to modify the genome. We might try to constrain such developments on Earth, but if there are communities, a few centuries from now, living on other planets or on asteroids we’d surely wish them good luck in deploying all known science to adapt their progeny to an alien and hostile environment

MP: Do you think that’s very likely?

MR: I don’t think manned space flight should be prioritized now, but I think it’s going to happen. One day there will be groups of pioneers living on Mars or on asteroids. I personally think the initial stages will involve crazy adventurers, bankrolled via independent, private funds. But I think it’s very important not to kid ourselves that we can solve Earth’s problems by mass emigration into space. There’s nowhere in our solar system even as clement as the top of Everest or the south pole, so it’s only going to be a place for pioneers – on cut-price private ventures and accepting higher risks than a western state could impose on civilians.

MP: you have had an enormously productive career as an astronomer, during a very exciting period. But do you think that if you’d been 30 years younger, you might have been a biologist instead? Biology has become much more dynamic in recent decades.

MR: I don’t know what I would’ve done. I think I am the kind of person who is better at writing the first paper on a subject than the last definitive one. I prefer to work on topics that are just opening up – trying to get the general picture, rather than tidying things up. We know that there are people of both those types in every science. And I prefer a more synoptic style of thinking – trying to make sense of fragmentary data, rather than pursuing elaborate mathematical modelling or deductions.

I’ve been quite lucky in that most of the work I’ve done has been on newly discovered phenomena and one can make some advances by applying general principles in a simple way and so I would’ve wanted to choose a subject of that kind. The advice I give to all students is to pick a subject where new developments are happening – either new observations or more powerful techniques. Otherwise you’re stuck trying to do the problems that the old guys got stuck on – and unless you’re cleverer than them, you won’t succeed either. So you’ve got to try to tackle a problem that they didn’t have a chance to do.

Looking back over my research, I was lucky because when I started, the subject was opening up: the first evidence of black holes, the first quasars, the first evidence of the Big Bang, the first pulsars, all came in the late 1960s. And so it was good to be a young astrophysicist then, because the experience of the old guys was at a discount.

Martin Rees is the Astronomer Royal. Michael Proctor is Professor of Applied Mathematics at the University of Cambridge, and the provost of King's College. Interview transcribed by Pat Wilson and Katie Fitzpatrick for the King's Review.