January 28, 2004

The initial conditions of the Universe

The standard cosmology does a remarkable job at explaining the gross features of the universe from a few simple principles. One starts off with Einstein’s general theory of relativity and the physics of gases at different pressure and temperature. One then makes a crucial assumption, that the universe is completely homogeneous and isotropic. A homogenous universe looks the same at every single place. For example the temperature, or the density of electrons, is the same wherever we look. The temperature may decrease with time, or the electrons may become more dilute, but this happens the same way, everywhere. An isotropic universe looks the same in all directions. In other words if we swivel around at any observation point, we always see the same thing. With these assumptions it is possible to construct a cosmology, the Big Bang which can predict a number of features: the fact that distant galaxies are moving away from us, the ambient bath of fossil radiation left over from when the universe was 300,000 years old and even the abundances of the lightest elements of the periodic table.
The assumption of homogeneity can only be an approximation. The real universe is far more complex, with structures spanning the whole sky. We see a wealth of stars, stacked into galaxies which themselves swarm around in clusters. All we have to do is look at our own solar system to realize that the universe is clumpy and not smooth as the Big Bang cosmology would have has assume. One of the outstanding questions in contemporary science is what is the origin of this large scale structure?
If you just think about it naively, you come up with two possibilities. The first possibility is that the universe started off in a completely chaotic mess, very much like water in a thunderstorm. Slowly but surely things would have evened themselves out on the largest scales to look smooth while the restless beginning would have been left imprinted on the smallest scales. Order would have emerged from chaos. We know however that this is not possible. On the largest scales, the dominant force is gravity. Gravity does not smooth things out. On the contrary it tends to make things more and more clumpy. To see this, think of how it works. If you have two objects of a given mass they tend to attract each other and clump together. The more massive an object is the more efficient it is at attracting other stuff to it. So if you start off with some stuff thrown around, with some regions having more mass than others, gravity will tend to push it around so the that more massive regions get even fuller while the empty region become even emptier. If we start off with a chaotic set of initial conditions, gravity will ensure that the universe becomes even rougher. It is not possible to make it smooth on large scales with gravity alone.
From the above description, it should be possible to see what the viable alternative is. We can start off with an almost perfectly smooth universe and lay down very small initial irregularities. As if they were small ripples on what is otherwise a very placid pond. Gravity will do its job and, slowly over time these irregularities will build up to form big peaks and valleys. The peaks are where the galaxies will accumulate while the troughs will become large expanses of emptiness, voids in the distribution of matter. As long as we have the right initial seeds, gravity will do the job for us.
Until the early 80s, the initial seeds were put in by hand. There were some half hearted attempts at proposing a mechanism for generating them, but nothing really worked. In the beginning of the 80s, a number of people (Guth, Albrecht, Steinhardt, Linde to name but a few) proposed that the universe underwent an incredibly short period of very fast expansion. During the period of expansion, a small patch of the universe was blown up to gigantic scales. Such a small patch of the universe could have undergone many physical processes before it grew and it would possible be smooth. The only roughness left in it would be due to the inevitable bubbling that comes from quantum mechanics, the quantum noise of the fabric of space and time. This period of expansion, known as inflation, would stretch and amplify these quantum ripples to cosmic scales. After the end of inflation, these ripples would be left to trigger the gravitational growth which leads to the galaxies and clusters of galaxies we see today.
Inflation as the origin of cosmic structure is highly speculative but it is, by far, the best worked out theory we have. A cause for concern is that it is the only theory on the market at the moment with any level of rigour. This is not to say there aren’t competitors. For example, a possibility which has recently been proposed is that our universe is but a slice, or membrane of a higher dimensional universe. Our universe can slap into another membrane and in the process, ripples can be set up. Just as in inflation, these ripples then seed structure which grow through gravitational collapse.
But we would like to have more alternatives. At the very least, to be sure that inflation is the most plausible.

Posted by Pedro Ferreira at January 28, 2004 11:49 PM