Before the Big Bang
June 3, 2000
HOW DID THE UNIVERSE BEGIN? NOT WITH A BANG BUT WITH A WHIMPER, SAYS MARCUS CHOWN
WHAT HAPPENED before the big bang? Try asking a cosmologist this and they'll usually fob you off by saying it's a meaningless question. Stephen Hawking famously likened it to asking "What's north of the North Pole?" The big bang, the idea goes, was the ultimate beginning. Time and space came into being then. There was no "before".
But that may not be true after all. One daring physicist has come up with an answer to the question no one is supposed to ask. If he is right, the Universe began an unimaginably long time before the big bang. "Far from being the beginning of time, the big bang was merely an important turning point in the Universe's history," says Gabriele Veneziano of CERN, the European Laboratory for Particle Physics near Geneva.
Veneziano's quest for the era before the big bang started in the early 1990s, when he and his colleague Maurizio Gasperini, then of the University of Turin, set out to fix some of the major shortcomings of the standard big bang model. It has plenty of them. For instance, if you imagine the expansion of the Universe running backwards like a movie in reverse, the density and temperature increase remorselessly until they skyrocket to infinity.
This infinite point, known as a singularity, is a big problem for physicists. "The singularity tells us that our description of the Universe--Einstein's theory of gravity, or general relativity--is not applicable in the earliest moments of the Universe," says Veneziano. That's because there is a point known as the Planck time--within 1043 seconds of the big bang--when gravity is comparable in strength to the other forces of nature. Because of this, to deal with the physics of the singularity you have to come up with a quantum version of gravity--and no such theory exists.
But physicists do have some ideas about how they might apply quantum theory to gravity. Perhaps the most promising is string theory. According to this, the fundamental particles of nature are impossibly tiny "strings" vibrating in a space of nine dimensions, with all but three dimensions "rolled up" smaller than an atom. One of the fundamental vibration modes turns out to be a massless particle that looks just like the hypothetical carrier of the gravitational force, the "graviton". That's why Veneziano decided to see whether string theory could help with the singularity problem. "I was convinced that cosmology of the very early Universe was the right arena for applying and testing string theory," he says.
Beyond the wall
What Veneziano found was that string theory can remove the troublesome singularity at time zero. Because strings have a finite size, the Universe still shrinks as time runs backwards but it never reaches zero volume so the singularity never arises. In the standard model of the big bang, the singularity acts like a brick wall. In string cosmology, there is no such wall and it is possible to venture back into the era before the big bang.
Having an epoch before the big bang solves another problem with the standard big bang model. If you imagine the expansion of space running all the way back to the earliest time reliably described by general relativity, you discover something extremely odd about the Universe. Although the content of today's observable Universe was squeezed into a volume only about 1 millimetre across, the distance light could have travelled since t = 0 was 1031 times smaller.
This has profound implications for the early Universe. The only way one region of space can "know" about the conditions in another is if there has been time for some kind of influence to go back and forth between them--and the maximum speed of any influence is that of light. Hence, the early Universe consisted of 1093 regions which were entirely separate, or "causally disconnected".
The problem is that today's Universe is homogeneous. Not only is the density of matter essentially the same everywhere but so too is the temperature of the leftover radiation from the big bang. Contrary to all expectations, the 1093 disconnected regions must have found a way to learn about each other.
In the standard model of the big bang, there is not enough time for the regions to have learned about each other and equalised their density and temperature in the ultra-short interval between the Universe's birth and the Planck time. Cosmologists have no choice but to patch things up later by postulating a brief phase of superfast expansion that "inflates" the entire observable Universe from a single causally connected region. Inflation doesn't just explain why the Universe is so uniform. It also provides a mechanism for blowing up tiny quantum fluctuations into the giant conglomerations of galaxies we see in today's Universe. However, the theory has its own problems. For instance, the mechanism responsible for blowing up space--the "inflaton" field--is plucked from thin air by theorists, and for it to work you have to pick the field's initial state with extraordinary care.
All this means that life would be far easier for cosmologists if the big bang was not the beginning of the Universe, and a long prehistory preceded it. There would be plenty of time before the Planck time for the densities and temperatures in 1093 regions to be equalised, and it would not be necessary to patch things up with standard inflation. "A prehistory is the obvious way to homogenise the Universe," says Veneziano. And that's just what string theory can provide.
So what would the pre-big-bang period be like? You can see clues to the answer, Veneziano says, by looking at the Universe's symmetry. The standard cosmological solutions of Einstein's general relativity equations--known as Friedmann big bang models--possess a very simple symmetry. They are invariant under time-reversal: in other words, if you take a solution of the equation and replace time by minus time you get another solution. Rather than expanding from a big bang, this time-reversed Universe contracts down to a big crunch.
The cosmological solutions of string theory possess this same symmetry, and another one too: they are also symmetrical if you replace the scale factor of the Universe, which sets its size, by its reciprocal. Apply both of these changes--time reversal and the scale-factor change--and you get an unusual type of expanding universe in which the expansion is accelerating rather than decelerating. In other words, it is an inflationary universe. "For every big bang solution, there turns out to be a solution in which space inflates from t = f to t = 0--that is, towards the big bang," says Veneziano.
String theory therefore suggests that the Universe underwent a phase of accelerated expansion from t= f to until just before t = 0. Just after t = 0, this changed into the more leisurely decelerating expansion of the old Friedmann big bang models.
Now we see the big bang in a new light. It was the time of transition from inflation to the expansion we see around us today. "At the big bang, the Universe had maximum curvature, maximum expansion rate and maximum temperature," says Veneziano. "The big bang emerges not as the beginning but an important turning point in the history of the Universe."
Some others are in broad agreement. "I think it's very likely that the big bang is indeed a later stage of the Universe," says Gordon Kane, a particle physicist at the University of Michigan.
One of the huge advantages of the string picture is that the pre-big-bang era is automatically inflationary and there is more than enough time for inflation to equalise the density and temperature of the Universe. But a description of how space evolved in the pre-big-bang era is only half the picture. To get the rest of it we need to know what the Universe evolved from--its initial conditions. What was the Universe like at t = f?
In the standard model of the big bang, the Universe started out in an extraordinarily special state in which the temperature and density are fine-tuned to be precisely the same in 1093 separate regions. Faced with a special state, a physicist's instinct is always to look for a not-so- special state from which it evolved. In the case of a crystal, for instance, everyone recognises that the fantastically ordered arrangement of atoms evolved via a phase transition from a more amorphous, and far less fantastic, liquid state. "In the same way, the Universe must have evolved from the simplest, most ordinary state that can be imagined," says Veneziano.
The simplest possible universe, according to Veneziano is infinite in extent, empty, cold and "flat"--which means space has a very low curvature. Veneziano and his colleagues Alessandra Buonanno at Caltech and Thibault Damour at the Institute of Advanced Scientific Studies (IHES) near Paris, call it a principle of "asymptotic past triviality".
Not everyone is happy with this. "It's equivalent to assuming homogeneity in the first place," says cosmologist Paul Steinhardt of Princeton University. "Gabriele has not solved any initial conditions problems--he has 'trivially' set them by his initial conditions."
Actually, complete flatness does not fit the bill. Veneziano points out that the most general solution of the string equations is filled with a chaotic sea of gravitational waves--random fluctuations in the curvature of space. "If space were completely flat it would be trivial in a special way," he says.
Clearly, the Universe did not stay trivial or we wouldn't be here today. As time passed--and there's been an awful lot of that--there was an opportunity for classical fluctuations to create regions that had a greater energy density than average. As in today's Universe, these fluctuations could become pronounced enough to give birth to black holes.
It is here that we see the real birth of the Universe we recognise. According to Veneziano, the solution of string theory that applies to the inside of a black hole is exactly the same as the accelerated expansion that string theory predicts for the pre-big-bang era. "Our Universe is a patch of the inside of a black hole," he says.
Veneziano and Gasperini have shown over the past few years that as the Universe expands towards t = 0, space-time becomes increasingly curved, resulting in a dramatic increase in temperature and energy density. A split second after time zero, a millimetre-sized, three-dimensional region within this vast expanse could look just like the superdense, superhot patch in standard inflation theory. Production line
Of course, to be valid any scenario must supply the particles of matter we see around us today. According to Veneziano, particles such as electrons, positrons and photons are conjured into existence by fluctuations in the geometry of space. This is where quantum mechanics comes into play. Just as strong electric fields can create electron-positron pairs, strongly varying gravitational fields lead to quantum production of all sorts of particles. Furthermore, the particles are born with a large amount of kinetic energy, and as a result the Universe gradually warms up. "This can be contrasted with the standard picture in which the particles are produced and heated up after the end of inflation," says Veneziano.
For the pre-big-bang scenario to be anything more than science fiction it has to make predictions about the Universe. Fortunately it does make several testable predictions which are at odds with those of the standard model of inflation. If Veneziano is right, the Universe should be filled with a chaotic sea of gravitational waves left over from its trivial past. The waves will be weak, and the prospects of seeing them with the current generation of detectors--or even their immediate successors--are not good. However, Veneziano maintains that third-generation detectors should be able to pick up this gravitational-wave background.
Effects from before the big bang should also be visible in the cosmic microwave background, the "afterglow" of the big bang. Veneziano predicts different angular positions of the peaks in the power spectrum of the cosmic background, compared with standard theories. This could be tested by the American MAP satellite, which is due for launch in November, or the European Planck probe scheduled for 2007.
The pre-big-bang scenario still has problems. Many might accuse Veneziano of simply moving the beginning of the Universe to an earlier time, without doing any more to explain how it came into being. Veneziano pleads that this is too harsh. "In the pre-big-bang scenario, the beginning is still not nothing but it's far more trivial and natural than the big bang," he says. "I believe this is a big improvement on the standard big bang."
Another problem is what happens at time zero. Somehow, the accelerating Universe of the pre-big-bang era must change seamlessly into the more leisurely expanding Universe we see around us today. This "patching together" problem is exactly what string theory, with its abhorrence of singularities, should solve. But it is not yet clear exactly how the patching would work.
While Veneziano doesn't yet have all the answers, many cosmologists think he has done them a huge service. "There is a tendency to be very sceptical on the part of most people who think about this stuff," says Andreas Albrecht of the University of California at Davis. "But others are happy to hear that someone is trying to do some cosmology with stringy connections."
"Gabriele's ideas are extremely good," agrees Kane. "It's less clear they are right, but if not they will surely stimulate thinking about these issues. This is probably the most important point--studying serious, testable (at least in principle) models about before-the-big-bang physics is now good physics."
Others are not so positive. Steinhardt says: "While everyone really appreciates Gabriele for his creativity and charm, I think most of us think that he has been on the wrong track, or a very unlikely track, for a number of years."
The risk that he's wrong doesn't bother Veneziano. He is perfectly willing to admit this is a possibility. But, he says, "we have shattered a wall". Want to know what came before the big bang? Then ask away. It's no longer taboo.