PDA

View Full Version : Uncertainty and Macroscopic Past

moozoo
2012-Feb-27, 12:11 PM
I hope this makes sense to others.

The past only exists in the state of the present, since you can only directly observe the present.
If there is a fundamental uncertainty in measuring the state of the present (Heisenberg uncertainty principle) doesn't that imply that the past is also fundamentally uncertain.
Is the past in a superposition of states?
Does the act of recording the past and the level of detail in the recording (writing it down, recording a video) affect this uncertainty/possible states that are in superposition?

For example consider a frictionless billiard table.
The balls are setup in the standard triangle and then hit with the cue ball.
At some point in the future I assume the Heisenberg uncertainty principle will limit the ability to predict back in time to the break off.
At that point the only way of knowing that there was a break off is via a detailed recording of it. e.g memory of the start, video recording, writing it down.

Is it correct to interpret a record of macroscopic historical events as points though which a single set of unrecorded events passes or as an infinite number of possible pasts that must pass though those fixed points.

Strange
2012-Feb-27, 12:48 PM
For example consider a frictionless billiard table.
The balls are setup in the standard triangle and then hit with the cue ball.
At some point in the future I assume the Heisenberg uncertainty principle will limit the ability to predict back in time to the break off.

Interesting idea. But that isn't really the way we understand the past. Imagine the balls had left tracks on the table; that is a better analogy for how we understand the past. Also, if you look at a complete system (a star, say) then even though you can't track back the individual interactions that led to each separate photon or particle you can still accurately describe the reactions on the macroscopic scale.

Andrew D
2012-Mar-02, 02:07 AM
I hope this makes sense to others.

The past only exists in the state of the present, since you can only directly observe the present.
If there is a fundamental uncertainty in measuring the state of the present (Heisenberg uncertainty principle) doesn't that imply that the past is also fundamentally uncertain.
Is the past in a superposition of states?
Does the act of recording the past and the level of detail in the recording (writing it down, recording a video) affect this uncertainty/possible states that are in superposition?

For example consider a frictionless billiard table.
The balls are setup in the standard triangle and then hit with the cue ball.
At some point in the future I assume the Heisenberg uncertainty principle will limit the ability to predict back in time to the break off.
At that point the only way of knowing that there was a break off is via a detailed recording of it. e.g memory of the start, video recording, writing it down.

Is it correct to interpret a record of macroscopic historical events as points though which a single set of unrecorded events passes or as an infinite number of possible pasts that must pass though those fixed points.

I have thought about this a lot. Even on macroscopic scales, natural processes are chaotic in nature (ex, the solar system (http://www.pnas.org/content/98/22/12342.full)). Assuming time symmetry, these processes are just as unpredictable in reverse as they are moving forward, so it makes sense that the past states of a system are just as unpredictable as the future states. I like to imagine a 'certainty cone' similar to a light cone, but where distance in the 'space' dimension is some measure of certainty; deterministic laws certainly have limits, and you can probably find a system-dependent space-time 'interval' (cognoscenti should think multidimensional ball) outside of which classical predictions 'almost certainly' fail. I don't think this can be thought of as an intrinsic quantum uncertainty though, just a consequence of using simple classical approximations to model complicated non-deterministic systems.

John Mendenhall
2012-Mar-02, 04:49 AM
I hope this makes sense to others.

The past only exists in the state of the present, since you can only directly observe the present.
If there is a fundamental uncertainty in measuring the state of the present (Heisenberg uncertainty principle) doesn't that imply that the past is also fundamentally uncertain.
Is the past in a superposition of states?
Does the act of recording the past and the level of detail in the recording (writing it down, recording a video) affect this uncertainty/possible states that are in superposition?

For example consider a frictionless billiard table.
The balls are setup in the standard triangle and then hit with the cue ball.
At some point in the future I assume the Heisenberg uncertainty principle will limit the ability to predict back in time to the break off.
At that point the only way of knowing that there was a break off is via a detailed recording of it. e.g memory of the start, video recording, writing it down.

Is it correct to interpret a record of macroscopic historical events as points though which a single set of unrecorded events passes or as an infinite number of possible pasts that must pass though those fixed points.

Your question has deep philosophical implications. There is no 'present'. What you or I or any other observer observes has happened in the past, even if no more than the time for photons to fly from the screen to your eyes. And for each observer, from their own reference frame, with all that that implies.

Regards, John M.

Andrew D
2012-Mar-02, 08:04 AM
Your question has deep philosophical implications. There is no 'present'. What you or I or any other observer observes has happened in the past, even if no more than the time for photons to fly from the screen to your eyes. And for each observer, from their own reference frame, with all that that implies.

Regards, John M.

I don't think the OP suggests that "there is no present." It does say that the past configurations can only be determined insofar as they evolved into the current state, but that we can use measurements of the current state (and the rules which govern its time evolution) to establish bounds of probable past and future configurations. The fact that we can only make a small amount of measurements means the confidence in predictions of future states should be relatively low. However, the fact that we are able to make measurements in the present suggest that the present exists. Relativity describes the notion of how the present at one point in space-time is related to the present in another, and while it does say that the reality of simultaneity is not exactly how we had thought it was, we are still able to define a meaningful notion of 'present'.

Back to the OP, a problem with the pool table analogy is that there is a small amount of energy in the system, all applied to the system within a relatively small space and time. Dynamical systems in nature are much more complicated, since all of the bodies are interacting in multiple ways on multiple scales, and are possibly influenced by factors outside of the system. For example, the distribution of mass in the early solar system depended on many more factors than just the mass of the Sun, especially before the interplanetary space was cleared, but there's no reason to think that all of those factors don't contribute to the current structure and evolution of the solar system. And again, if you're really talking about macroscopic systems here, the Heisenberg uncertainty principle is not the source of the unpredictability.

moozoo
2012-Dec-17, 09:53 AM

Apparently Steven Hawking and Thomas Hertog have considered this.
From the pdf linked to http://arxiv.org/abs/hep-th/0602091v2
"Top down cosmology is a framework in which one essentially traces the histories backwards, from a spacelike
surface at the present time. The no boundary histories of the universe thus depend
on what is being observed, contrary to the usual idea that the universe has a unique,
observer independent history."

And again, if you're really talking about macroscopic systems here, the Heisenberg uncertainty principle is not the source of the unpredictability.

At what point do quantum level effects perturb a macroscopic chaotic system such that it can not be seen as deterministic.
Do those quantum level effects over time cause the system's past to become a superposition of states? Think of this in terms of the present looking back into the past without any recording.
How does the act of recording (samples of) the past define it?

Hlafordlaes
2012-Dec-17, 01:53 PM
Ken G had a thread going for a while on the butterfly effect and classical uncertainty, in which statistical mechanics was used, not quantum uncertainty. Good background read germane to this discussion.