What is the Janus Universe and its implications for the arrow of time?

[windy miller]

One idea I have been hearing about recently is the Janus universe
https://physics.aps.org/featured-article-pdf/10.1103/PhysRevLett.113.181101
Im hoping someone might have a go at giving a layperson explanation of this paper. As I understand they are saying a system will eventually come to a point of low entropy naturally due to gravity and then bounce back at out so there is a reversal of the arrow of time at the bounce point. Is that right?
How is this different to Poincare recurrence?

 

[Greg Bernhardt]

paging @Janus

 

[spacejunkie]

One idea I have been hearing about recently is the Janus universe
https://physics.aps.org/featured-article-pdf/10.1103/PhysRevLett.113.181101
Im hoping someone might have a go at giving a layperson explanation of this paper. As I understand they are saying a system will eventually come to a point of low entropy naturally due to gravity and then bounce back at out so there is a reversal of the arrow of time at the bounce point. Is that right?
How is this different to Poincare recurrence?

The difference is that Poincare recurrences arise repeatedly in confined systems as long as enough time passes while a Janus point is an unique point in an unconfined system which goes to infinity.

The idea is not so much that the arrow of time changes direction at the Janus point but that the emergence of the arrow of time always leads to the Janus point being in the past.

This paper by the same authors is a little more readable than the one you posted:-
Janus Points and Arrows of Time
https://arxiv.org/abs/1604.03956

Barbour also has a laymans level book just published about the subject.

 

[PeterDonis]

As I understand they are saying a system will eventually come to a point of low entropy naturally due to gravity and then bounce back at out so there is a reversal of the arrow of time at the bounce point. Is that right?

Not quite. You are implicitly assuming an arrow of time that is not the one in the model, when you say it "bounces" and the arrow of time "reverses". But the whole point of the model is that there is no "time" external to the model. Instead, what you have is a spacetime model that is divided into two regions by the Janus point; each of the two regions has its own arrow of time, pointing away from the Janus point. Observers living in each of the two regions will perceive a "flow of time" in the direction of the arrow of time in their region, i.e., away from the Janus point. That means no observer will perceive a "bounce".

 

[windy miller]

Not quite. You are implicitly assuming an arrow of time that is not the one in the model, when you say it "bounces" and the arrow of time "reverses". But the whole point of the model is that there is no "time" external to the model. Instead, what you have is a spacetime model that is divided into two regions by the Janus point; each of the two regions has its own arrow of time, pointing away from the Janus point. Observers living in each of the two regions will perceive a "flow of time" in the direction of the arrow of time in their region, i.e., away from the Janus point. That means no observer will perceive a "bounce".

Ah ok thanks for that.