The deeper we look, the more densely populated the universe appears. In the reverse movie of the universe, the matter gets progressively denser as if some giant piston were squeezing it ever tighter. That piston is, of course, gravity. Moreover, it is a property of matter that as it is compressed, it grows hotter as well as denser. Today, the average temperature of the universe is only about 3 degrees above absolute zero, or -270 degrees centigrade. But as we follow the universe into the past, the temperature rises, first to room temperature, then to the boiling point, and eventually to the temperature on the surface of the sun.
The sun is so hot that the atoms that it is composed of have been torn apart by their violent thermal motion. The nuclei are intact, but the more loosely attached electrons have been torn free and can roam throughout the sun’s hot gases, which are now electrically conducting plasma.8
Electrical conductors are generally the least transparent of materials. The freely moving electrons easily absorb and scatter light. This scattering of light makes the sun opaque. But as we move outward to the sun’s surface, the temperature and density decrease to the point where it becomes transparent. That is where we see the sun’s surface.
Now let us travel backward in time and outward in space to the last visible shell, where the conditions are similar to the sun’s surface. Again the light comes to us from a surface like the sun’s—a giant shell of hot plasma surrounding us on every side. Astronomers call it the surface of last scattering. Sadly, looking through the conducting plasma to an even earlier and more distant shell is no more possible than looking through the sun.
Immediately after the Big Bang, the light from the surface of last scattering was every bit as bright as the sun’s surface. That raises an interesting question: why, when we look at the sky around us, don’t we see the bright glare of ionized hot primordial plasma? To ask it another way, why isn’t the sky uniformly illuminated with the same brightness that we would see if we were to look straight into the sun? Fortunately, the Doppler shift rescues us from that awful prospect. Because of the Hubble expansion, the plasma that originally emitted the primordial light is receding away from us with a large velocity. In fact, using the Hubble Law, we can calculate the velocity of this recession, and the result is only slightly less than the speed of light. This means that the emitted radiation was Doppler redshifted way past the visible and infrared, all the way to the microwave spectrum. Here, one of the earliest discoveries of quantum mechanics plays an important role: the energy of a photon depends on wavelength in such a way that a microwave photon has about one thousand times less energy than a photon of visible light. For this reason, the photons that eventually reach us from the surface of last scattering are not very potent. They have no more effect on our retinas than the radio waves that continually surround us.
There is another way to understand the diminished potency of the cosmic radiation by the time it reaches us. The photons from the surface of last scattering were very hot, about as hot as the sun’s surface. They filled space, forming a kind of photon gas, and like all gases, when they expand, they cool. The expansion of the universe, since the time of the Big Bang, cooled the photon gas to the point where it lost most of its energy. Today, the CMB (cosmic microwave background) radiation is very cold: fewer than 3 degrees above absolute zero. The two explanations of the CMB’s loss of power are mathematically completely equivalent.
George Gamow was the first to have the idea of a Big Bang. Soon after, two of his younger colleagues, Ralph Alpher and Robert Herman, got the idea for CMB as a kind of leftover afterglow. They even estimated the temperature of the radiation today, and got 5 degrees: within two degrees of the right answer. But physicists at that time believed that such weak radiation could never be detected. They were wrong, but it took until 1964 for the CMB to be accidentally discovered.
At that time, the Princeton cosmologist Robert Dicke wanted to test the idea of CMB by measuring the radiation left over from the hot Big Bang. While he was in the process of building a detector, two young Bell Laboratory scientists were doing precisely the kind of experiment that Dicke was aiming for. Arno Penzias and Robert Wilson were scanning the sky for microwave signals, not for the purpose of discovering the birth of the universe, but for communications technology. They couldn’t identify a strange background static that was getting in the way of their real goal. Legend has it that they thought it was bird droppings on the detector.
Princeton University and Bell Labs are close neighbors in central New Jersey. As fate would have it, Dicke found out about the Penzias-Wilson “noise” and realized that it was the CMB from the Big Bang! Dicke got in touch with the Bell Labs scientists and told them what he thought was going on. Subsequently, Penzias and Wilson got the Nobel Prize for the discovery. It is one of those twists of fate that had Princeton and Bell Labs been farther apart, Dicke might have finished his experiment and been the first to make the discovery.
The Penzias-Wilson detector was a crude affair mounted on the roof of Bell Labs. By contrast, the most modern CMB detectors are extremely sophisticated and are mounted in space, high above the atmosphere. The detectors can be pointed in different directions to measure the CMB from each point in the sky. The results are presented as a kind of map of the sky.
One of the most striking features of the CMB is how dull these maps are. To a very high degree of precision, the microwave-sky is a featureless, homogeneous expanse. It seems that, in early times, the universe was almost perfectly homogeneous and isotropic. The microwave radiation coming from the surface of last scattering is almost identical in all directions of the sky. This extraordinary degree of homogeneity is somewhat puzzling and needs an explanation.
As smooth as the universe was at that early time, it could not have been perfectly smooth. Some small, primordial lumpiness had to be there to seed the formation of galaxies. If the seeds were too weak, galaxies would not have formed; if too strong, the lumps would have grown too rapidly and collapsed to black holes. Cosmologists strongly suspected that under this boring homogeneous background, the seeds of future galaxies were there to see. Even better, theoretical cosmologists had a pretty good idea how strong the density contrasts had to be in order to create the galaxies as we see them now. The difference between the microwave intensity in different directions would have to be about 100,000 times smaller than the average intensity.
How on earth is it possible to detect such incredibly small density contrasts? The answer is that you don’t do it on earth. You’ve got to get high above the polluting environment of the planet. The first experiments to see small variations in the microwave radiation were done by detectors suspended from balloons flying above the South Pole. The South Pole is good for a number of reasons, not the least of which is that a balloon doesn’t wander very far from its launch site. The prevailing winds would carry a balloon clear around the world, but around the world is not very far when you’re at the South Pole. The experiment was named Boomerang!
High above the South Pole, microwave detectors compared the intensity at pairs of locations and automatically determined the difference between them. Theorists had their expectation, but no one knew for sure that anything of interest would be seen. Perhaps the sky would just continue to be a featureless, gray background. Then they would have to go back to the drawing board and redesign the theories of galaxy formation. Everyone who had any interest in cosmology waited for the jury’s verdict with nervous anticipation. The verdict that came back was everything a defense attorney could hope for. The theorists told the truth. Lumps in the cosmic oatmeal were there and at exactly the right strength—10–5—one part in 100,000.
Outer space is an even better place from which to measure the cosmic microwaves. The data from the orbiting Wilkinson Microwave Anisotropy Probe, known by the nickname WMAP (pronounced “double-u map”), are so incredibly precise that they not only measured that 10–5 lumpiness but also detected the roiling, oscillating motions of huge blobs of hot plasma that radiated th
e CMB.
The large blobs of coherently moving plasma were not at all unexpected. Theoretical cosmologists had predicted that the expansion of the universe would start the lumps in the plasma oscillating like ringing bells. At first the smaller lumps should begin contracting and expanding. Later, with a lower frequency, larger blobs would join in: a perfectly predictable symphony. The detailed calculations indicated that at any time the largest visible oscillating blobs would have a certain definite size. Thus, when WMAP saw such oscillating blobs, cosmologists already knew a great deal about the size of the largest ones.
Knowing the size of the largest oscillating blobs had an incredible serendipitous payoff: it now became possible to survey cosmic triangles and measure the curvature of space. Here’s how it was done.
Suppose you know the size of an object and also just how far away it is. This will enable you to predict how big it will look in the sky. Consider the moon. The moon is about 2,000 miles in diameter and about 240,000 miles away. Just from that information I can predict that it will occupy an angle of half a degree in the sky. By pure coincidence the sun is four hundred times bigger than the moon but also four hundred times farther away. The result is that the sun and the moon look the same size in the sky, namely half a degree. If we were on the moon looking at the 8,000-mile-diameter earth, it would look four times bigger than the moon as seen from the earth, i.e., two degrees.
Actually, in making these claims I made a tacit assumption, namely, that space is flat. Think of the diameter of the moon as the third side of a triangle. The two other sides are straight lines from our point on the earth to two diametrically opposite points on the moon.
If space is flat between the moon and the earth, my claims are correct. But if space is appreciably curved, the situation is different. For example, if space is positively curved, the moon will look bigger than a half degree. The opposite is true if the curvature is negative.
Now suppose we had independent confirmation that the diameter of the moon is 2,000 miles and that it was 240,000 miles away. We can use the apparent size of the moon to tell us the curvature of space. To a very high degree of accuracy, space is flat between us and the moon.
Let’s get back to surveying the cosmos. Here is what we know: the largest oscillating blobs that were active at the time the CMB was emitted were about 200,000 light-years in diameter. Bigger blobs than that had not yet begun to ring.
Today, the source of the CMB is about ten billion light-years away, but at the time the CMB started its journey, our distance from the surface of last scattering was a thousand times smaller, i.e., ten million light-years away. That’s enough to compute how big the largest CMB blobs should look to WMAP if space is flat, namely, about two degrees: as big as the earth seen from the moon. If space is not flat, the apparent size of the blobs would tell us how curved it is.
What did WMAP find? It found that Euclid was right! Space is flat.
Let me qualify that a bit. By measuring triangles on the surface of the earth, it is possible to tell that the earth is a curved sphere. But in practice, unless we can measure very big triangles, we would find that they behave as if the earth were flat. Obviously, Columbus could not convince the king of Spain that the earth was round by drawing a few triangles near the king’s palace. He would have had to measure triangles at least a few hundred miles on a side, and even then he would have had to do it with great accuracy. All Columbus could say by surveying small triangles was that the earth is very big.
The same is true for cosmic surveying: all we can really conclude is that the universe is flat on scales of ten or twenty billion light-years. If the universe is finite, it is a lot bigger than the portion we can see.
So here is what we know with good confidence. First, the ordinary mass in the universe, stars, gas clouds, and dust, is not sufficient to make the universe flat. By past standards it’s not so far off, only by a factor of fifty. But cosmology is no longer a qualitative science. By today’s standards it’s not close at all. Without other hidden sources of matter, the universe would be ruled open and negatively curved. But there is more matter in the universe, about ten times more, that we know about only by its gravitational effects. It may be made up of new elementary particles that hardly interact with the usual kind. These dark-matter particles, if that’s what they are, would fill the galaxy, passing right through the sun, the earth, and even us. But they are still not enough to make the universe flat or closed. If the universe is flat, another kind of mass or energy must be pervading space.
Second, the age of the universe appears to be too young unless the history of its expansion is different than expected. The only conventional explanation is that there is a cosmological constant which accelerates the expansion. Although completely unexpected, it is confirmed by Type I supernova data that provide a kind of reverse film clip of the evolution. The best explanation of the age problem is that a cosmological constant exists at just about the level predicted by Weinberg’s anthropic argument.
Third, cosmic microwave data directly show that the universe was extremely homogeneous in early times. Moreover, it is also very large, large enough to appear flat to cosmic surveyors. The bottom line is that the universe is many times bigger than the portion that we can see, and its expansion is accelerating under the influence of a small cosmological constant.
Inflation
It used to be a joke in the United States how Soviet Communist ideologues would claim that everything had been invented first in Russia. This included the radio, television, lightbulb, airplane, abstract painting, and baseball. In my own field of physics, there was sometimes truth to the joke. Soviet physicists were so badly isolated that a number of extremely important discoveries went unnoticed in the West. One of them was a remarkable guess about how the universe began. More than a quarter century ago, the young cosmologist Alexy Starobinsky had the idea that the universe started out with a brief period of prodigious exponential expansion. I’m not sure exactly what his motivation was, but in any case, only a few other isolated Russians appreciated Starobinsky’s thought until some time later, when a young physicist in my own university rediscovered the idea. Alan Guth was a young postdoc working on high-energy theoretical physics at the Stanford Linear Accelerator Center (SLAC).
When I first met him, in 1980, I assumed that he was working on the ordinary problems of particle physics. At that time very few elementary-particle physicists knew much about cosmology. I was an exception because two years earlier Savas Dimopoulos and I had worked on the problem of why nature made so many more particles than antiparticles. My friend Bob Wagoner, one of the early pioneers of cosmology, had asked me if particle physics provided any explanation for the overwhelming preponderance of matter over antimatter. Dimopoulos and I had the right idea, but we were so ignorant of basic cosmology that we had confused the horizon size with the scale factor. That’s like an auto mechanic not knowing his steering wheel from a hole in his muffler. But under Bob’s wing we learned fast and eventually wrote the first paper outside the USSR about a subject to be called baryosynthesis. Ironically, barosynthesis was another subject that had first been invented in the USSR, this time by the great Andrei Sakharov, twelve years earlier.
Anyway, despite the fact that I was interested in the subject, I don’t think I knew that Guth was interested in cosmology, that is, I didn’t know until he gave a seminar on something that he called inflationary cosmology. I imagine I was one of the two or three people in the room who knew enough to be impressed.
Alan was after big game: the biggest. Why was the universe so big, flat, and so extremely homogeneous? To see why this is such a puzzle, let’s go back to the CMB and focus on two separated points on the sky. At the time when the CMB was produced by the hot plasma, those two points were a certain distance apart. In fact, if they were more than a few degrees apart, the distance at that time would have been large enough that no light or other signal could possibly have gotten from one point to the other. The universe was on
ly about half a million years old, so if the points were separated by more than half a million light-years, they could never have been in contact. If they had never been in contact, what made those two places so similar? In other words, how did the universe become so homogeneous that the CMB looked exactly the same in every direction?
To make this point clearer, let’s return to the balloon theory of the universe. Imagine that the balloon began in a deflated state, badly shriveled up, with lots of wrinkles like a dried prune. As the balloon expanded, the wrinkles would have begun to smooth out. At first, small wrinkles; then later, the bigger wrinkles would be ironed out. There is a rule about how wrinkles smooth out: a wrinkle of a given size can get smoothed only if there is enough time for a wave to propagate across the wrinkle. In the case of the universe, that means enough time for a light wave to cover the distance.
If there were insufficient time for large wrinkles to smooth out when the CMB originated, we should see them imprinted on the sky map. But we don’t see such wrinkles. Why was the universe so smooth? Might there have been a long prehistory, hidden from view by the opaque early plasma, during which the wrinkles were stretched out? That’s what Inflation theory is all about—a prehistory during which wrinkles were removed.
Alan immediately seized on the possibility that Starobinsky’s exponential expansion might well be the key to this puzzle. The universe, according to Guth, had inflated like a balloon, but an extraspecial balloon. A real balloon will inflate only so far, and then it will burst. Alan’s universe grew exponentially, and in a short time it became enormous. You can think of the Inflation as taking place before usual cosmology began. By the time the conventional Big Bang started, the universe had already grown to immense proportions. And in growing, all the wrinkles and inhomogeneities got stretched out so that the universe became exceedingly smooth.