Is this credible: can the future affect the past

[member 529879]

http://physicsworld.com/cws/article/news/2012/aug/03/can-the-future-affect-the-past

Is this article credible? If it is credible can somebody simplify it a little?

 

[atyy]

Yes, it is credible. The Two-State-Vector Formalism is simply quantum mechanics formulated in a different way. The part that is probably wrong in http://arxiv.org/abs/1206.6224 is the statement that only the Two-State-Vector Formalism is adequate to explain the results, whereas the Two-State-Vector Formalism does not differ from standard quantum mechanics (as far as I understand).

So I would agree more with Charles Bennett's point of view as reported in the article: "Charles Bennett of IBM’s T J Watson Research Center in Yorktown Heights, New York, who is a specialist in quantum-information theory, is not convinced. He sees TSVF as only one way of looking at the results, and believes that the findings can be interpreted without any apparent "backward causation", so that the authors are erecting a straw man."

 

[wabbit]

Seems reasonable. Despite the mysterious sound of it, I am sure this provides no way to actually change the past.
The experiment is built from constituents that do no allow this and there is no magic.

 

[craigi]

http://physicsworld.com/cws/article/news/2012/aug/03/can-the-future-affect-the-past

Is this article credible? If it is credible can somebody simplify it a little?

Yes. It was published in the Annals of Physics in February of this year.

You can read the paper here http://arxiv.org/abs/1206.6224

 

[paisiello2]

I think it's in-credible.

 

[wabbit]

It is credible, the experiments would surely work, but the conclusions are all interpretation.

My summary would be :

If we define the word "influence" in such a way that we can influence something even though we cannot change it, and if we define the word "measurement" so that we can measure a quantity without actually learning is value, then it is possible to build a subtle experiment where a future measurement influences a past one.

Maybe I didn't get that exactly right, but I think it's close.

I suspect we could even play that interpretation game in classical physics, without QM. Pretty sure we can do that for the wonders of action at a distance from entanglement.

 

[stevendaryl]

I've read a little bit about the two-state formulation of quantum mechanics, which puts quantum mechanics on a time-symmetric basis. The rough idea is to erase the distinction between state preparation (which is done in the past) and measurement (which is done in the future). The wave function is causally influenced equally by boundary conditions in both the past and future.

I actually don't have any problems with back-in-time influences. That's no weirder than any other interpretation of QM (they're all weird). However, I just don't see how it can work. There is, it seems to me, a potentially vicious circle created by allowing causality to go in both directions in time. The propagation of the wave function is assumed to depend on boundary conditions in the future. But the future in turn depends on how things play out, quantum mechanically (that is, future boundary conditions might depend on quantum mechanical events such as whether or not Schrodinger's cat dies). Somehow, nature finds a self-consistent solution.

In general, backwards causality can be paradoxical. For example, if in 2020 I send a message to myself in 2015 and say: "Don't send this message", then that's a paradox. Obviously, in time-symmetric quantum mechanics, causal influences sent into the past are fuzzy or uncertain enough that one can't send a coherent message. So consistency is saved by uncertainty. That's similar to the case of the Bohm-DeBroglie interpretation of QM--in that interpretation, there are faster-than-light influences, but they are microscopic and cannot be used to send FTL messages.

 

[wabbit]

I haven't read this version, I but I think the authors claims is very different from that - actually they do not claim any real backward causality, they only claim that using a notion of causality which includes "A causes B but A cannot actually change B".

In a time reversal, of course if we start from a different future we get a different past. But this is not mysterious at all.

I apologize here, I realize that I am discussing those same issues in at least two threads at the same time, and it may be tangential to both. I should probably start a fresh thread instead asking about causality and time reversal, excluding collapse and similar things since such interpretations do not help me understand things better at all.

 

[stevendaryl]

I haven't read this version, I but I think the authors claims is very different from that - actually they do not claim any real backward causality, they only claim that using a notion of causality which includes "A causes B but A cannot actually change B".

Actually, that's the delicate balancing act that several different interpretations of quantum mechanics have to perform. How can you have causal influences that propagate, but make sure that those influences don't actually affect anything?

 

[wabbit]

Yes. My non-specialist impression is that the issue doesn't come from QM (not only at least) but from the difficulty of intuitive reasonning with conditional probabilities that such interpretations often include in a somewhat implicit way, which is why I usually try to steer clear of them.

 

[HuskyNamedNala]

How would one test an event that has not taken place yet that will affect events that have or already are taking place? Wouldn't you have to know the future even prior to the experiment?

 

[wabbit]

How would one test an event that has not taken place yet that will affect events that have or already are taking place? Wouldn't you have to know the future even prior to the experiment?

I agree. We know that if A is true then B is true. So we say "A causes B". But this is equivalent logically to "not B causes not A". I suspect that if we do not decide first that by causality we mean forward causality, the exclusion of backward causality will not magically appear by itself.

I am reading Rovelli's papers on this topic, which seem to gradually lift the fog in my mind, though I am still confused:)

 

[atyy]

How would one test an event that has not taken place yet that will affect events that have or already are taking place? Wouldn't you have to know the future even prior to the experiment?

In Newtonian physics, the future causes the past. This can be shown to be equivalent to the past causing the future. So to test whether the future causes the past, you just need to test whether the past causes the future.

 

[atyy]

Here is an introduction to the Two State Vector Formalism.

http://arxiv.org/abs/0706.1232
New Insights on Time-Symmetry in Quantum Mechanics
Yakir Aharonov, Jeff Tollaksen

Please remember that it is just a reformulation of quantum mechanics, not much different from Newtonian mechanics being equivalent to Lagrangian and Hamiltonian mechanics. The value of equivalent alternative formulations of a theory are that some things might be easier to see in another formulation. Historically, weak vales were seen by Aharonov via the Two-State Vector Formalism. However, weak values can be understood in the usual formulation of quantum mechanics also. It is a matter of personal taste whether one prefers to understand weak values in the Two State Vector Formalism or in the standard formalism.

 

[Demystifier]

Please remember that it is just a reformulation of quantum mechanics, not much different from Newtonian mechanics being equivalent to Lagrangian and Hamiltonian mechanics.

That's a good analogy. For example, Lagrangian mechanics could also be interpreted as influence of future to the present time, because particle accelerates as it accelerates now in order to attain a given final position with minimal action during the future.

 

[craigi]

That's a good analogy. For example, Lagrangian mechanics could also be interpreted as influence of future to the present time, because particle accelerates as it accelerates now in order to attain a given final position with minimal action during the future.

I think there is an important difference here in that, in classical mechanics, time is fundamental. Conversely, the arrow of time can be seen as an emergent feature of quantum theory.

 

[Demystifier]

I think there is an important difference here in that, in classical mechanics, time is fundamental. Conversely, the arrow of time can be seen as an emergent feature of quantum theory.

I don't think that there is such difference. Yes, the arrow of time is emergent in quantum theory, but it is also emergent in classical theory. The time itself (not the time arrow) is fundamental in both classical and quantum theory.

 

[PeroK]

That's a good analogy. For example, Lagrangian mechanics could also be interpreted as influence of future to the present time, because particle accelerates as it accelerates now in order to attain a given final position with minimal action during the future.

If you look at Lagrangian mechanics that way, how would you explain?

Light travels so as to minimise the time taken between A and B. If you shine a beam of light off a mirror, then the angle of incidence equal angle of reflection rule minimises the time between points. So, when it gets to B, it has done so along the path that minimised the time.

But, if light knows that it will end up at B, it could get there faster by veering off immediately and heading to B in a straight line from A. Why go via the mirror?

 

[wabbit]

If you look at Lagrangian mechanics that way, how would you explain?

Light travels so as to minimise the time taken between A and B. If you shine a beam of light off a mirror, then the angle of incidence equal angle of reflection rule minimises the time between points. So, when it gets to B, it has done so along the path that minimised the time.

But, if light knows that it will end up at B, it could get there faster by veering off immediately and heading to B in a straight line from A. Why go via the mirror?

Why indeed, when it goes from A to B, does it go via the mirror and not directly? Because you picked those photons or light rays that do so and determined their path conditional on that - others do take the direct route.

 

[craigi]

I don't think that there is such difference. Yes, the arrow of time is emergent in quantum theory, but it is also emergent in classical theory. The time itself (not the time arrow) is fundamental in both classical and quantum theory.

You're correct, but I don't think we can expect time to retain its fundamental status in a successful theory of quantum gravity, whereas it's reasonable that the emergent arrow of time would still be relevant.

 

[Demystifier]

If you look at Lagrangian mechanics that way, how would you explain?

Light travels so as to minimise the time taken between A and B. If you shine a beam of light off a mirror, then the angle of incidence equal angle of reflection rule minimises the time between points. So, when it gets to B, it has done so along the path that minimised the time.

But, if light knows that it will end up at B, it could get there faster by veering off immediately and heading to B in a straight line from A. Why go via the mirror?

Because light is not smart enough to know the global minimum of all possible paths. It only knows one of the local minima.

If that does not look intuitive enough, here is a simple metaphor from everyday life. A man wants to marry the most beautiful girl available in the world. But he is not smart enough to do that, simply because he does not know all the available girls in the world. He only knows the girls in his neighborhood. So he chooses a local maximum, by picking up the most beautiful girl from his neighborhood.

 

[DrChinese]

How would one test an event that has not taken place yet that will affect events that have or already are taking place? Wouldn't you have to know the future even prior to the experiment?

Not really. You can entangle photons AFTER you measure them! This has actually been done.

http://arxiv.org/abs/quant-ph/0201134

There is no way to use this technique to signal from the future to the past (or other crazy ideas).