Saturday, February 4, 2017

A Changing Arrow of Time?

This is one post I am not looking forward to writing.  Some of my readers have asked me to comment about alternative theories to the Big Bang which remove the necessity of our universe having a beginning.  I have been thinking for some time about how to write on this subject in a non-technical manner, which is the tone I strive for on this blog.  Because most of these ideas are quite theoretical, requiring complex mathematics and intricate nuances, it is quite a challenge for me to give an accurate and adequate description of most of these proposals, yet still be comprehensible.   Nevertheless, in this post I want to try to discuss the paper by Anthony Aguirre and Steven Gratton (AG) that describes a scenario which they claim requires no beginning.1   I also want to give some thoughts on how their idea fits into the whole discussion of evidence for or against a deity, particularly the Christian God.  My attempt may be an epic fail.

In the model proposed by Aguirre and Gratton, they claim to avoid a beginning by proposing a thermodynamic arrow of time that points in different directions depending on whether the universe is expanding or collapsing.  To understand what this means I need to first take a diversion to discuss what the thermodynamic arrow of time means.  Actually, no one really knows for sure why we experience time moving forward but we do know that a quantity called entropy must increase in any non-reversible process.  Entropy strictly has to do with the number of microstates available to a system. The concept of a microstate can be illustrated by considering two six-sided dice. There is only one microstate available for the dice to roll 2: both must show a one. However, there are six possible microstates available for the dice to roll a 7. The combinations are 1 and 6, 2 and 5, 3 and 4, 4 and 3, 5 and 2, or 6 and 1. Because there are more available microstates, the dice will more often roll a 7. A macroscopic system with more available microstates has a greater entropy than one with fewer microstates. The second law of thermodynamics states that an isolated system will evolve spontaneously to the state with maximum entropy.   This is a statistical idea.  In general, all processes move toward those that are more statistically probable.  That gives us the arrow of time.   Time moves in the direction where entropy increases.

The entropy of the universe is an important topic when discussing its beginning.  Some early theories in the 1970's that tried to avoid a real beginning of the universe by proposing an oscillating universe, one that expanded, then contracted, then expanded and repeated that pattern over and over, ran into problems with entropy.  Thermodynamic considerations showed that such universes could only have a few oscillations because entropy always increases.

(As an aside, since entropy always increases our universe had to begin in a very low entropy state, one with very low probability.  This low probability beginning, in and of itself, is a "problem" that is hard to explain.  Of course, it is not difficult to explain if one hypothesizes that an outside intelligent designer began the universe because intelligent designers create systems with low entropy all the time.)

Anyway, Aguirre and Gratton propose a universe with a twist, in that entropy will decrease during any collapsing phase, so that time basically runs backward.  In some sense if you were standing at the moment of the Big Bang and looked forward into our universe you would see time moving forward away from you and also if you looked backward at what existed "previously" you would also see time moving forward away from you, rather than looking backward in time.  This seems to go totally against any reasonable ideas of time but we can't totally dismiss it since we don't really have a definitive idea of what time is anyway.  Their idea seems to resolve some of the previous problems with oscillating universes and entropy but introduces complete havoc with our intuition about time.  I think you could make a good argument that most physicists would say this idea of time is not physical.  In addition, it is unclear whether or not their proposal even removes the need for a beginning.  In a paper discussing the AG scenario and other such ideas, Alexander Vilenkin says, "Even though the spacetime has no boundary in the AG model, it does include a hyper-surface on which the low-entropy (vacuum) boundary condition must be enforced by some mechanism. This surface of minimum entropy plays the role of the beginning of the universe in this scenario." In other words it isn't even clear that the AG model removes the need for a practical beginning of our universe.

Although the AG model doesn't directly correlate with an oscillating universe, it does share the common feature that it allows a period of contraction and a period of expansion.  So if you're still reading and not totally lost yet, I'd like to make a few observations about the AG scenario and its ramifications for ideas about a transcendent beginning, the Christian God, and the biblical record.  Let me try to address two important questions regarding the ramifications of the theory.  The first question is, "Would a universe with a pre-history of contraction diminish the correlation between what science has discovered and the words of Genesis 1:1?"  If I read the opening verse of the Bible carefully, I see that it claims our universe had a beginning of time and matter.  It doesn't say anything about whether or not there were other universes before ours.  In just about any theory of a universe that contracted before it expanded, our universe still had a beginning.  So the writer of Genesis got it right regardless of whether or not there were other universes before ours.  Throughout written history there have been many opinions about whether or not our universe had a beginning.  The Bible claims it did and any modern theories of multiple universes still require a Big Bang with an expanding hot, dense, universe that had a practical beginning.  That is in agreement with the biblical claim.  I'll talk more about multiple universes (multi-verse) in future post but I'll maybe intrigue you by saying I do believe there have been, and will be, more than one universe.   The second question is, "Does an oscillating universe, or a universe that derived its existence from a previous history, remove the need for a transcendent cause?"  If by "transcendent" you mean outside of this universe, then of course we still need a transcendent cause since the space, time, matter, and energy of this universe had a beginning.  If the question is more theistically worded as, "Does an oscillating universe remove the need for a supernatural transcendent cause?" that is a harder question.  Philosophically I think the answer is still that God is the best hypothesis for the existence of something rather than nothing since something must have existed forever, either the laws of nature, or the stuff of nature, or some self-existent being.  I'll explore those options more in a future post when I discuss the ideas of Lawrence Krauss.  In any case, the actual scientific observations currently give no evidence of any oscillating universe scenarios or any universe before ours.  All the observations and laws of physics that have been confirmed point to a single universe that had a beginning.

As I have said and will continue to repeat, we don't really know what happened in the first 10-35 seconds or our universe, nor do we have any observational evidence for any pre-history.  But everything we do know, and even speculative ideas like the AG model, still agree with the first verse of the Bible, that for our universe, "In the beginning God created the heavens and the earth," and that a transcendent being remains a cause that is supported by all the evidence.

Appendix:  A couple notes about this post:

  • In my attempt to keep this post as simple as possible I have skipped some important details about the AG scenario including the condition that their solution closely resembles a de Sitter spacetime near the boundary conditions, and other such details.  I apologize to those who are mathematically advanced and contemplate such ideas even before their first cup of coffee.
  • For a reasonable discussion of some of these ideas from a theistic point of view, you may want to read New Proofs for the Existence of God: Contributions of Contemporary Physics and Philosophy, by Robert J Spitzer, William B. Eerdmans Publishing Company, Grand Rapids, MI/ Cambridge, U.K. (2010)

1Aguirre, Gratton, "Inflation without a beginning, a null boundary proposal," Phys. Rev. D 67 (2003) 083515.
2Vilenkin, "Arrows of time and the beginning of the universe," Phys. Rev. D 88 (2013) 043516.

The figure above is taken from Vilenkin's paper describing Aguirre and Gratton's model.  The caption on the figure says, "Arrows of time in the Aguirre-Gratton scenario with a low-entropy boundary condition [left] at the de Sitter bounce and [right] on the null hypersurface t = -∞."


  1. Dr. Strauss, the two figures in the image appear to represent two different models. Which one is the AG model?

    1. Great question. Actually they are both possible scenarios. I think I probably left out too many details in this post and didn't make some things clear. In the AG scenario, the universe does seem to appear from a vacuum state and has a "bounce surface" at which point the arrow of time points in opposite directions. Depending on the exact boundary conditions of that surface you can get either one of the pictures. Vilenkin says this about the two options. "The bounce surface in de Sitter space is defined up to boost transformations. The surfaces obtained from one another by such transformations are completely equivalent, but in the limit of infinite boost the surface becomes null." [That limit is the picture on the right.]

    2. What is "de Sitter space"? Google has an explanation, but its still a bit technical. Can you break it down? Thanks!

  2. Thanks Mike, In any ocillating process or evolution leading to our expanding universes how time sensitive would the alternate phases be in order to get the universe we observe. Does such consideration lead to the same sort of semi-infinite improbability as when considering the spectrum of precise physical constants. Thanks Keith