# Retrocausality in action

Discussion in 'Physics & Math' started by RJBeery, Apr 24, 2012.

Messages:
4,136

3. ### Aqueous Idflat Earth skepticValued Senior Member

Messages:
6,152
Imagine how problematic retrocausality becomes if Victor's decision is delayed for years, or billions of years, or until the last Planck time just before the Big Crunch.

Another problematic scenario is one which creates a feedback loop in which Victor's delayed decision depends on a measurement from Bob and/or Alice and/or the random number generator.

Another scenario is one in which Victor makes a delayed decision, then later reverses his decision.

Messages:
31,550
What is MWI?

7. ### SyneSine qua nonValued Senior Member

Messages:
3,515
Many worlds interpretation?

8. ### BelieveHappy mediumValued Senior Member

Messages:
1,194
What does it have to do with this again? (honestly don't get it)

9. ### przyksquishyValued Senior Member

Messages:
3,171
Just to be clear about two things with this sort of experiment, which I'm not sure are made clear in the article cited in the OP:

1) Alice and Bob don't see anything unusual and can't tell just from their data when Victor chose to entangle his particles and when he didn't - i.e. they have nothing that would allow them to predict Victor's actions in advance. They just get a message from Victor some time later telling them that a particular subset of their data exhibits nonlocal correlations.

2) Victor has no control over which particular subset of Alice and Bob's data shows nonlocal correlations. In all experiments of this type, Victor's ability to entangle the photons is nondeterministic: ideally what Victor would do is perform a Bell state measurement which randomly projects his photons into one of four different entangled states. Then Victor knows that Alice and Bob's results could show nonlocal correlations in those four subsets of cases (provided Alice and Bob are making the right measurements). In reality, real world implementations of Bell state measurements aren't successful a significant fraction of the time (on the order of 50% to 75%).

So overall, what Victor has is a way of identifying (but not choosing or controlling) a subset of Alice and Bob's measurement results that exhibits nonlocal correlations, without actually having to look at their data.

10. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
If by problematic you mean paradigm-shifting for the way that most view the Universe, then I would agree. If you mean problematic in that you feel retrocausality is incapable of explaining this experiment then I think you're confused on what the term means.

11. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
Hi Przyk, I'm glad you're chiming in. Whether or not Alice and Bob can receive useful information "from the future" here is not relevant. What I'm asking is, how can this undeniable demonstration of the future affecting the past be encapsulated in an MWI interpretation? If Alice's and Bob's worlds are split at the time of their respective measurements, the "weighted branch" variable would still need to be determined prior to Victor making his choice.

I mean, you could always do away with Free Will but then MWI becomes redundant anyway, right?

12. ### przyksquishyValued Senior Member

Messages:
3,171
Can you explain, in your own words, what you think this "undeniable demonstration of the future affecting the past" is? I'm asking because I happen to have some familiarity with this type of experiment, and I don't see it that way. Nothing about Alice or Bob's results is necessarily affected as a result of anything Victor does. Victor just has a way of being able to identify a subset of Alice and Bob's data that exhibits nonlocal correlations without actually looking at their data.

13. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
That phrase is a rewording of what I said earlier, isn't it? That Victor, from the future and in the context of a Many Worlds Interpretation, can determine the branch-weight of particular measurement outcomes of Alice and Bob in the past? I thought that one of the supposed "benefits" of MWI is that it allows for unidirectional evolution in the description of a system.

In terms of describing the retrocausality here, I'm confused why you don't see this as obvious. Note, I'm not saying that sending information back from the future is a requirement or even a certain potentiality of retrocausality (although I personally believe it's possible), what about this:

Alice and Bob are local to each other. They each measure their first respective photons while sending their second respective photons off into the ether of space in the same direction (with the traveling photons they also include instructions for their future recipient). One hundred years later Victor receives these photons and, after reading the instructions, decides whether or not to correlate Alice and Bob's past measurements based upon whether the North or the South* won the Intergalactic Super Bowl of 2112. They can repeat this process to achieve an arbitrarily high degree of confidence about the Super Bowl of 2112. Details aside, if nothing else I can't understand how the validity of this influence can be denied!

*Why the hell is there a North and South in an intergalactic game of anything? Because I'm late for lunch and trying to crank this out quickly

14. ### przyksquishyValued Senior Member

Messages:
3,171
That's the whole point: there isn't really anything tangible that Victor does determine. The article you linked to is a bit misleading in this regard, when it says:
But when Victor chooses to "entangle the photons" as this article puts it, what he actually does is perform a Bell state measurement which, in the case of the experiment cited in the OP (arXiv preprint available here), could distinguish two of the four Bell states and gave an inconclusive result the rest (50%) of the time. In the subset of cases where he gets a given Bell state, he can know that the corresponding subset of Alice and Bob's results will show nonlocal correlations (provided Alice and Bob performed the right measurements, which is arranged beforehand). But he has no control of what that subset will be: he only learns it by looking out for certain results to measurements he performs, which occur randomly.

Not really, since MWI isn't any different to the standard way QM is presented in textbooks in that regard. The main benefit of MWI is as a possible solution to the measurement problem that allows measurement to be interpreted as a physical interaction like any other.

15. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
Przyk, you didn't address my thought experiment..?

16. ### Jarek DudaRegistered Senior Member

Messages:
212
There is a decade old quantum erasure configuration which seems to literally allow to send information back in time(?):

We produce two entangled photons - first spin up, second spin down or oppositely.
Photon s comes through double slit on which there are installed two different quarter wave plates changing polarization to circular in two different ways.
Finally there are two possibilities:
x y R L
y x L R
where columns are: linear polarization of p, initial linear polarization of s, circular polarization of s after going through slit 1, circular polarization of s after going through slit 2.
So if we know only the final circular polarization of s, we still don't know which slit was chosen, so we should get interference. But if we additionally know if p is x or y, we would know which slit was chosen and so interference pattern would disappear.
So let us add polarizer on p path - depending on its rotation we can or cannot get required information - rotating it 45deg we choose between classical and interfering behavior of s ... but depending on optical lengths, this choice can be made later ...

discussion: http://www.thescienceforum.com/physics/27354-controlled-delayed-quantum-erasure-where-causality.html

What is the problem with something like that?
Time paradox? But we have to remember that it would control just few degrees of freedom. There have left huge amount of thermodynamical dof - they seem random for us, but physics could easily bend them to fulfill some additional time loop-like constrains ... for example making that such channel would lie.

17. ### przyksquishyValued Senior Member

Messages:
3,171
I did, indirectly:
Victor doesn't actually do this. As I explained in my previous post, Victor has no control over which of Alice's and Bob's measurement results are correlated.

In any set of data that Alice and Bob collected, they could always look at the data and pick a subset that showed nonlocal correlations, just by selecting the correlated looking outcomes more often and tending to discard the rest. In this scheme, Victor just has a way of detecting such a subset without actually looking at Alice's and Bob's data.

18. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
I'm not talking about single points of data. I'm talking about conclusions made over the aggregate samples of data. If the instructions given to Victor were [South wins = no correlation, North wins = corrrelation], and the North wins the 2112 Intergalactic Super Bowl, then Victor could correlate all subsequent photon pairs that he receives from Alice and Bob. Meanwhile, 100 years earlier, Alice and Bob are measuring their respective photons and deciding whether or not they are correlated. If not, they will exhibit pure random behavior; if so, they will tend towards a behavior from which they can deduce with increasing probability that the North will win the 2112 Super Bowl.

19. ### przyksquishyValued Senior Member

Messages:
3,171
Neither am I, and so am I. That's the whole point of what I've been trying to get across to you: Alice and Bob do not see anything different in the statistics of their results. They have no way of telling what Victor did.

What happens is this: Alice and Bob get a lot of results that they can use to build up a probability table $P(ab \mid xy)$, where x and y are respectively Alice's and Bob's measuring device settings and a and b are their measurement outcomes. For this experiment, if they use all the data, they get local correlations (the results don't violate a Bell inequality). But it is always possible for Alice and Bob to meet up afterwards and "cheat" by going through their data and keeping some of it and throwing some of it away, in such a way that they deliberately construct a subset of their results which is characteristically nonlocal. What's remarkable about the experiment you linked to in the OP is that Victor has a way to identify such a nonlocal subset in Alice and Bob's data without actually having to look at the data. But again: he cannot control what that subset is, and it isn't even due to him that there exists such a subset. You expect such subsets anyway.

20. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
It's not that complicated Przyk. Alice and Bob measure their respective series of local photons, collusion permitted because they are together in the lab, to determine whether they all appear pairwise to have correlated polarizations. They should not need any information from Victor to determine this!

21. ### Aqueous Idflat Earth skepticValued Senior Member

Messages:
6,152
No, I was imagining that it's either problematic to construct the experiments I mentioned, or else to get predictable results. Some of them would seem to set up oscillations. The first one, delaying Victor's measurement until the Big Crunch, would seem to allow for creation and annihilation of the universe to become mutually causal. If that's not best described as problematic, it would at least make for an interesting religion.

As far as where my understanding trails off, it's with the leftward track of the positron

which is just as mystifying as the arrival of the anti-quark from the future.

22. ### RJBeeryNatural PhilosopherValued Senior Member

Messages:
4,136
"Mutually casual"...you might find my views on this subject interesting. Przyk and I have had it out on this quite a few times and it is my opinion that MWI seems to be inadequate to explain our observed behavior of Nature.

Last edited: Apr 26, 2012
23. ### przyksquishyValued Senior Member

Messages:
3,171
And my response is similarly not complicated: if Alice and Bob look at all their collected data in this experiment, they don't see any nonlocal correlations, regardless of what Victor does. Your thought experiment is perfectly clear, but it is also based on a misunderstanding of the experiment that was actually performed, due to the rather misleading way the article you cited describes it.

Please bear in mind I'm not getting my information by gleaning stuff out of the article you linked to or what people are saying about it. I actually have prior experience with this type of experiment and the theory behind it (I did a theoretical analysis of a similar entanglement swapping experiment a couple of years ago), and I looked up the actual research paper for some of the specific details (like some of the details regarding the Bell state measurement they performed).