Well, Isn't That Special?

Stars as well as human beings are born, grow old, and die. In the 19th century scientists proposed the dismaying notion of the "heat death” of the universe, according to which every hot thing becomes tepid while all cool things become warm, so that in the end all matter exists at the same middling temperature and the future is an eternal unchanging tedium. Physicists have a word for this general tendency toward decay and dissipation: entropy. And entropy, as Sean Carroll, a physicist and cosmologist at the California Institute of Technology, ably explains, is all about the directionality of time. The onward march of time fundamentally derives from something peculiar about the way the universe was born, and that’s the puzzle Carroll attempts to resolve.

If you could watch a movie of two atoms bashing into each other and then bouncing apart, you could not tell which way time was running. A collision run backward in time obeys the laws of mechanics exactly as well as the same collision run forward. But think of a lot of atoms crashing about - milk being stirred into black coffee, for example - and a clear direction of time emerges. Stir that coffee as long as you like, and you will never see the milk collect itself in one spot to form a white island in a black sea.

Entropy, you may have heard, explains this. Entropy is a measure of disorder, and the second law of thermodynamics says that it can only increase. Highly ordered arrangements of atoms (the milk all in one place, surrounded by coffee) inevitably evolve, through the general commotion of atoms, into disorderly arrangements (the milk mixed up throughout the coffee). The fundamental reason is simple: There are far more disorderly arrangements than orderly ones, so, as a matter of straightforward probability, changing systems are more likely to end up in disordered states.

Having set out these basic ideas with great lucidity, Carroll delivers the kicker. It’s all very well to say that systems tend to go from lowprobability states to high-probability states, and that that progression is fundamental to our sense of time marching forward, but who said that we have to start from a place of low probability? If the initial point of any system, whether we’re talking about milk in coffee, life on Earth, or the big bang itself, were drawn randomly from the list of all possible states such a system might be in, it most likely would be a high-probability state with, entropically speaking, nowhere to go. Why, that is, wasn’t the universe born into a state of heat death - by far the most likely state for any universe?

For a century or more, physicists have had to simply assume that the universe began in a special state endowed with low entropy, so as to allow thermodynamic room for all the interesting later stuff (galaxies, planets, people) to happen. But that’s an uncomfortable assertion, because saying that the universe must have started in some special way implies a deep reason for that specialness that we haven’t been able to figure out yet. Carroll claims that at least the glimmer of an explanation is finally in sight. To support that conclusion he travels through deep and mysterious realms of physics, ranging from the existence of time machines to the fate of quantum information swallowed by black holes. Carroll is an affable and enthusiastic guide, but I suspect many readers will need to take frequent time-outs to let their minds unboggle.

In his final chapter, Carroll draws on the increasingly commonplace (to cosmologists, anyway) idea that our universe is just one of many. In the picture he sketches, each universe grows old and boring in line with the classical idea of heat death, but in its dotage it can spawn baby universes that start with low entropy and live through an exciting phase of growth and activity before becoming boring themselves, so continuing the cycle. We know that we live in just such an exciting phase because our universe still has a lot going on. As for how elderly universes give birth to new ones, and why those young’uns start life with low entropy, you’ll have to read Carroll’s book, and you’ll have to pay attention.

Carroll is admirably honest in acknowledging the degree of speculation here. It’s legitimate to worry, I think, that his low-entropy baby universes only come about because he has put together a number of unproven physical assumptions in just the right way. The specialness that he and other cosmologists are so eager to explain away may be there still, having been flushed out of one dark corner only to scurry into another.