You are viewing the historical archive of The Philosophy Forum.
For current discussions, visit the live forum.
Go to live forum

Poincaré Reoccurrence Theorem And Time

Devans99 April 25, 2019 at 05:19 10125 views 20 comments
If time is infinite, the universe should go through all possible states eventually. A similar idea to this is Poincaré recurrence theorem:

https://en.wikipedia.org/wiki/Poincaré_recurrence_theorem

But surely some states are equilibrium states? One an equilibrium is reached, no further change in the universe is possible. Isolated systems naturally tend towards equilibrium with time. An example equilibrium state would be all matter in the universe in one gigantic black hole. Or all matter converted to energy (maximum entropy).

So because we are not in equilibrium, time must be finite?

A possible exception is if there is some mechanism that is preventing the universe from entering an equilibrium state. The only example I can think of is an endless cycle of Big Bang / Big Crunch, with entropy reset to zero at each Crunch. But that model is not compatible with current observations of the increase in the rate at the is universe expanding. Also, unless each crunch is ‘perfect’, entropy would increase on each cycle (which is not what is observed - the early universe was very low entropy - see the CMB radiation).

Comments (20)

Deleted User April 28, 2019 at 21:52 #283170
This user has been deleted and all their posts removed.
andrewk April 28, 2019 at 22:11 #283176
Quoting Devans99
If time is infinite, the universe should go through all possible states eventually.

For the theorem to apply, the system must have constant volume. That rules out infinite universes and finite universes with changing volumes, which are the only types predicted by current cosmological theories.

Also, note that equilibrium states are only asymptotically approached, never attained. So it is not possible that an equilibrium state will be fully attained and the universe will remain in that state forever. It is however expected that the universe will asymptotically approach 'heat death' which can be thought of as an equilibrium state to any pre-defined degree of approximation.
Metaphysician Undercover April 28, 2019 at 23:32 #283194
Quoting Devans99
If time is infinite, the universe should go through all possible states eventually


Wouldn't this allow that the universe would eventually come back to the exact same state again, making time cyclical rather than infinite?
Terrapin Station April 28, 2019 at 23:36 #283198
Quoting Devans99
If time is infinite, the universe should go through all possible states eventually. A similar idea to this is Poincaré recurrence theorem:


It could vacillate between just two states forever, or any other arbitrary number of states.

Devans99 April 29, 2019 at 05:33 #283306
Reply to tim wood Reply to Metaphysician Undercover Reply to Terrapin Station

Fair point, I was only using the theory as an analogy - it does not apply directly to the universe.

The universe's long term behaviour must be characterised by one of the following:

1. expanding
2. contracting
3. static
4. cycling

Time must have a start with [1]. The universe is not one gigantic black hole so [2] is clearly not happening. The equilibrium argument applies with [3] so time would have a start.

That just leaves a cyclic universe [4] as the only possible infinite time configuration:

- The universe's expansion rate is speeding up, counting against the cyclic model
- Unless each crunch is ‘perfect’, entropy would increase on each cycle, which is not what we see
- It seems very unlikely that the 'cycle length' would be constant. It would probably decrease each time (in which case with infinite time, cycling will have stopped by now) or (less likely) it would increase with time (in which case with infinite time, cycle length will be infinite by now)

So time should have a start.
Deleted User April 29, 2019 at 14:17 #283531
This user has been deleted and all their posts removed.
Shamshir April 29, 2019 at 15:25 #283559
Consider the following.

Time has gone through and is always in all states.
It is a finished thing.

But we, you and me, are like the little moving part of the bar of a video player - showing only one frame.
And we keep going and going and going and it keeps coming and coming and coming, but it is already finished.
If the video was not complete, how would you play it?

Everything is set in place - static, and everything is moving because we move.
Devans99 April 30, 2019 at 05:21 #283834
Quoting tim wood
So I ask again. What is your understanding/definition of time?


We don't actually need to know what time is in order to work out it has a start, but I think we can say time is a degree of freedom like space; we are all moving in the time direction. You can slow your movement in the time direction by moving in the space direction.

So time is a degree of freedom so is similar to space (although you can move left/right in space, time you can only move in one direction).

Quoting Shamshir
If the video was not complete, how would you play it?


Future real eternalism is a possibility in my view:

https://thephilosophyforum.com/discussion/5302/an-argument-for-eternalism/p1

So presentism seems impossible for this and other reasons. The main models left are fixed block universe and growing block universe. The future being real goes against common sense but that in itself is not sufficient to rule out the possibility.
Deleted User April 30, 2019 at 05:37 #283840
This user has been deleted and all their posts removed.
Devans99 April 30, 2019 at 05:48 #283841
Quoting tim wood
I suppose if it has a start, then events can be ordered temporally, yes? But relativity says, for a broad class of events, that they cannot be absolutely so ordered.


When adding relativity into the mix, it becomes more complicated. It is the order of the events is dependant on observers speed, that can change. Relativity says no preferred frame of reference.

I don't see observers disagreeing on the order of events as sufficient reason for there to be no start of time? As long as the start of time precedes all other events in all reference frames, there is not a problem?

Quoting tim wood
But the real question is that at there are times and places where time simply appears to be not well-defined, and the same with "places."


- Inside black holes
- Inside the singularity
- Beyond the edge of spacetime

I feel the absence of time in these areas is supportive of a start of time - time can be absent in certain situations suggests time can be absent globally.

Shamshir April 30, 2019 at 05:53 #283842
Reply to Devans99
I understand, but I shall repeat myself.

All you see is what there is.
All you say is what there is.
All you do is what there is.
How do you see, say or do what there is not?
And what there is - is now.
It is always complete, but through manner of division - appears always going.
It is a static, and a man spins inside it - so the static spins.

You wrote, that 'something 'other' than only now exists'.
But if I told you that there is only now and now is eternal, to reconcile the two, would you believe me?
Now, you might ask, but if now came from nothing - nothing predates now, does it not?
It does not. There is no is, do, or be with nothing. Only nothing.
If nothing is, it is not nothing.

You may think time has a start. It does.
So, does it have an end? It does.
Its start is now and its end is now and its present also now.
That's it.
Devans99 April 30, 2019 at 06:02 #283844
Quoting Shamshir
You wrote, that 'something 'other' than only now exists'.
But if I told you that there is only now and now is eternal, to reconcile the two, would you believe me?
Now, you might ask, but if now came from nothing - nothing predates now, does it not?
It does not. There is no is, do, or be with nothing. Only nothing.
If nothing is, it is not nothing.


Not sure I entirely follow, I think eternalism maybe correct (past, present, future all exist). 'Now' cannot exist eternally - if it did, the things within the universe (particles etc...) would have no temporal start and without a temporal start they could have no existence.

Quoting Shamshir
So, does it have an end? It does.


I think so too. I'm a finitist so I would - if eternalism is true, the time dimension is finite. My suspicion is that the time dimension is also circular. So we lead the same lives again and again perpetually (IE the start of time is co-incidental with the end of time: Big Bang meets Big Crunch).
Shamshir April 30, 2019 at 06:16 #283845
Quoting Devans99
Not sure I entirely follow, I think eternalism maybe correct (past, present, future all exist). 'Now' cannot exist eternally - if it did, the things within the universe (particles etc...) would have no temporal start and without a temporal start they could have no existence.

I will try to explain.

Past, present and future are all now. They all exist now. Regardless of their distance.
Like two ends of a rope, past and future, are the same thing viewed from angles.

Imagine there is no now. Then this moment does not exist.
If this moment does not exist, there is no past and future; the whole bridge collapses.
If there is no past, present and future.
There is no time.
And yet there is time.
So there must be now and now is all there is.
The past, present and future exist in the now and because of it.
This is why the now is eternal.
And the temporal start you seek, is 'now'.
Deleted User April 30, 2019 at 15:06 #284014
This user has been deleted and all their posts removed.
Devans99 April 30, 2019 at 15:18 #284023
Quoting tim wood
I'm sure Hawking has been mentioned in your threads. He opined that time is akin to the surface of a sphere in that it is boundless, yet has no beginning or end.


His theory has time as an imaginary variable I think? It sounds a little far fetched, that would make the universe in effect 5D? And I believe even with the no boundary proposal, that real time still has a start (imaginary time does not)?

Quoting tim wood
Democritus, for example, may have given us "atoms," but he couldn't manage, or even imagine, nuclear physics and sub-atomic particles, and that because he was looking in the wrong place (among a lot of other reasons).


I do not believe we need to know the bits and bolts of time before knowing whether it has a start or not - you find out the basic stuff first do you not?

PS Even better argument for time has a start here:

https://thephilosophyforum.com/discussion/5302/an-argument-for-eternalism/p1
Deleted User April 30, 2019 at 15:35 #284032
This user has been deleted and all their posts removed.
Devans99 April 30, 2019 at 15:39 #284035
Quoting tim wood
Let's try this. If something has a start, then there is either something or nothing prior to it.


There is something prior to it; a timeless first cause. There are many arguments for this. A classic from Aquinas is:

1. Can’t get something from nothing
2. So something must have existed ‘always’.
3. IE if there was ever a state of nothingness, it would persist to today, so something has permanent existence.
4. It’s not possible to exist permanently in time (an infinite regress; it would have no start so could not be), so the ‘something’ must be a timeless first cause.

I count this a logical language that leads to a logical conclusion. I think metaphysics has a role to play in guiding science towards the right solutions.
Deleted User April 30, 2019 at 18:00 #284084
This user has been deleted and all their posts removed.
Devans99 May 01, 2019 at 07:19 #284373
Quoting tim wood
Modern physic says not only that you can, but that we have, and do


Modern physics respects the conservation of energy - virtual particles only only transitory effects - if that was not the case, matter density would be infinite by now (with infinite time).

Quoting tim wood
A timeless something is incoherent. Make it coherent.


As far as origins of the universe are concerned, there are only two possibilities:

1. An infinite regress in time.
2. A timeless first cause that created time.

The first has no start so its impossible. The second, we know from relativity that things can be timeless like the photon so it maybe possible. As the first is impossible, it timelessness has to be possible.

tom111 June 09, 2019 at 21:04 #296067
I think the main argument I'd use against the recurrence conjecture is that if the universe was truly repetitive, then surely we would be far more likely to come to being in an isolated pocket of order amongst a lot of disorder. But we see today a large amount of order seemingly with no disorder surrounding. The number of states that entails human beings coming into existence in small pockets of order surely outnumber those of us coming into being in a whole universe of order?

As a side note I don't think the universe not being a static, finite volume is a valid argument against the conjecture. If the universe is expanding like it is today, given enough time (an incredibly large amount of it) it will surely randomly contract again and at points be completely static (assuming theres any fluctuations in the expansion of the universe at all, which it is agreed there probably is).

The only true argument I can conceive is that time must end if the conjecture isn't true. if there were some sort of 'irreversible' change that happened in the universe that meant say, that protons and neutrons can't for atomic nuclei, given an extraordinarily large amount of time surely that rule will be defied at some point or another on a large scale.

Also as one last point, if we define the laws of physics as mathematical rules describing persistent patterns in nature (that probably aren't inherently mathematical but maths is a good tool for measuring things like these), surely given enough time these patterns will also begin to change? Interested to hear what everyone thinks and if this were true, what it would mean for the conjecture.