# Worlds with no beginning and no end

In today’s blog we talk to Dr. Paul J. Steinhardt, PhD from Harvard University, and the Albert Einstein Professor in Science at Princeton University.

**David**: The apparent insensitivity of the laws of physics to the direction of time flies in the face of everyday experience. In your view, is this a deep discovery about nature?

**Paul**: First of all, this is not quite true. Physicists observe that the laws are not sensitive to simultaneous time reversal, parity and charge (CPT) but, at the same time, violate CP alone (charge and parity flip). Hence they violate time-reversal symmetry; that is, they are sensitive to the direction of time. Experiments are underway to search for similar sensitivity when strong nuclear forces are involved, but so far none has been achieved.

But you are probably referring to the bigger effect of the 2nd law of thermodynamics, the notion that entropy increases on average with time. That means the entropy in the past must be smaller than the entropy today, and the entropy today must be smaller than in the future. This is a deep discovery about nature that we have known for some time.

Some have argued that this means the ultimate future of the universe is one in which the universe reaches maximal entropy – which means the combination of all the contents of the universe reach thermal equilibrium. This is called heat death. Some point out that the 2nd law creates a cosmological puzzle – to explain the time flow we observe, it requires the entropy produced in the big bang must have been very small. How did that happen? Curiously, this is the opposite of what quantum physics would predict – large quantum fluctuations should have excited gravitational and matter and radiation degrees of freedom bringing them all to high temperature and high entropy. But if that were the case, heat death would have occurred long ago.

But in thinking about bouncing and cyclic models, especially with my collaborator, Anna Ijjas, who has pioneered these ideas, other possibilities emerge. First of all, there is no reason why there had to be a big bang. (There is not a shred of experimental evidence for it – the idea of the big bang is based on a simplistic extrapolation.)

An obvious alternative is that, instead of a bang, the universe underwent a bounce – a smooth transition from a contracting universe to an expanding one. Once you allow that, you avoid the mathematical singularity and the infinite temperatures of the bang. What about the entropy? Well, now we have to be more careful about what we mean when we say that entropy always increases. If we have a system that is closed – where we can observe everything inside it and account for all the entropy – then the 2nd law says that the total entropy must increase. But in cosmology, an observer only sees a tiny patch of the universe called the Hubble volume. That entropy is finite within that volume. And now it can be that the entropy in the universe overall is growing, but Hubble volume is growing (during expansion) and shrinking (during contraction) such that the entropy within the Hubble volume after a bounce is small, becomes bigger as that volume grows, and then most of that entropy leaves the Hubble volume as it contracts so that the entropy within is small again. That’s roughly how it works in a cyclic model.

**David**: Is there tension between the laws of physics and notions of beginning?

**Paul**: We don’t know, but one must admit that a “beginning” is certainly a strange idea. Nothing we know in physics experimentally has an “after” without also having a “before.” The big bang is a hypothetical idea which only has an after. We have no direct evidence for it. All our tests of cosmology have to do with periods long after any hypothetical big bang. Nor can we say it is impossible.

But we can say today that having a big bang creates problems aside from having to explain the bang itself. With a bang, it is hard to understand how we could have a bang and end up with a distribution of energy that is smooth as we observe existed in the early universe. And also hard to understand how the entropy of the part of the universe we observe (the Hubble volume) should have been so low in the first place. Attempts to patch these problems with a period of inflation have encountered numerous problems. In sum, we cannot rule out the idea of a beginning, but it is a strange idea compared to our physics experience in other realms; it creates numerous problems that are unresolved thus far; and the problems can potentially be resolved if there were no beginning/bang but bounces and cycles instead.

**David**: Can one square the time asymmetry of the laws of physics with the explanation of lower entropy density causing the low entropy of the next cycle?

**Paul**: Yes. As explained in answering the first question: we only observe a small patch of the universe (the Hubble volume). In the cyclic models we consider, the total entropy in the universe increases over the course of a cycle, but so does the volume so that entropy density remains low. The next cycle emerges from a small patch of the space that was observable the cycle before, so it contains only a small fraction of its entropy. So the apparent entropy is low at the beginning of the next cycle because most of it is beyond the horizon that we can observe -- spread over a much larger volume that we cannot observe – perhaps even an infinite one.

**David**: The value of the cosmological constant cries out for explanation. How natural in your estimation is its resolution using the contracting phase of the cyclic model? Are there features of that explanation that remain unresolved?

**Paul**: The cosmological constant is a puzzle if you think of it as corresponding to the energy density of the vacuum today and assume that the vacuum today is the stable, lowest energy state of the universe. Then the magnitude of the cosmological constant (whether it be positive or negative), seems extraordinarily small in magnitude (or absolute value) compared to what one would naturally expect given all the large contributions plus and minus to the vacuum energy that we know must come from different interactions and physical effects. Large positive or very negative values would be plausible, but very close to zero by 100 decimal points compared to what we would estimate?? That’s a big problem. Some say the biggest numerical flaw in scientific thinking.

But in the cyclic picture, the vacuum today is __not__ the lowest energy state. The lowest energy state is predicted to have a large negative vacuum density or cosmological constant consistent with estimates. A corollary is that the current state of the universe is unstable – it cannot be expanding as it does forever. It is that instability that ultimately leads to a period of contraction and bounce back to a similar unstable state that repeats periodically.

**David**: What has been the feedback on your ideas on cyclic cosmology? How much of it is rooted in critical thinking and how much in other instincts such as the pressure to fit in?

**Paul**: For those familiar with earlier types of cyclic models, there are various famous problems that they had that they assume apply in our case: (a) violation of the 2nd law of thermodynamics; (b) violent collisions between black holes as the universe contracts; (c) so-called chaotic mixmaster behavior that disorders the universe; (d) the singularity at the big crunch. Consequently, when first hearing about anyone pursuing bouncing and cyclic models, many assume that these problems apply and are highly skeptical right from the start. Some stop there and do not consider our ideas further.

Those who take more time to study the novel kind of cyclic models Anna Ijjas and I have been exploring learn that all four problems can be overcome. Those people become very interested, especially since these models make distinctive predictions that are testable (unlike, say, the multiverse which is described as allowing any possible outcome).

**David**: Thank you Professor!