Physics gurus: Quantum Erasor Experiment. New commentary. and My thought experiment.

Factual Questions
1

Please visit the following page and review the commentary. Let me know if you see any mistakes.

“A Delayed Choice Quantum Eraser Experiment” with Commentary

My comments and questions on this experiment:
The beam splitter BSA and BSB are where the action is. If the beam splitters send the idler photon toward D3 or D4, we have which path information, and the interference pattern will disappear. If the beamsplitters send the idler photon toward Ma or Mb, we do not have which path information, and the interference pattern will appear.

What makes this really amazing, is that the signal photons “decide” to interfere or not before the idler photon reaches BSB or BSA. In fact the “choice” of whether the idler photon gives which path information can be arbitrarily delayed depending upon the distance to the beam splitters. Right? So in effect the signal photons know the future choice of the idler photons?

Now PLEASE correct me if I am wrong here: If we replaced the beam splitters with a switch which could be turned on or off by a human operator, the same results would occur. If the switch made the idler photon go to D3/D4, no interference pattern. If the switch made the idler photon procede toward the quantum eraser, yes interference pattern. And this would be true even if the switch decision was made after the signal photon made its choice to interfere or not? (am I wrong here?)

If this is true, then doesn’t this enable us to predict the future?

I would propose the following thought experiment. Please tell me if I make an error in my expectations of this experiment.

Replace the beam splitters with such a human controlled switch. Then increase the distance the idler photons travel before hitting the switch, preferably a lot. (I would think we could do this with a big bundle of fiber optic cables or some other clever technique.) Let’s say somehow we increase the time delay to 20 minutes. And just for elegance sake, lets suppose only one person has access to the switch, so only that one person could ever turn it on or off. Let’s call that person Dr. John.

Then do the same experiment.

The interference pattern should appear or disappear depending upon whether or not the switch will be triggered 20 minutes in the future by Dr. John? So the photons in effect should know whether or not Dr. John will or will not flip that switch in the future, right?

If this really worked, then we would have created a disturbingly accurate means of predicting the future actions or non actions of Dr. John. Not only that, we would have a means of predicting the future location of Dr. John. If Dr. John was in the bathroom, and the interference pattern suddenly appeared or disappeared, then we would know that in 20 minutes Dr. John would flip the switch in the control room.

Or for an even more useful example, imagine we increased the time delay for flipping the switch to a full 24 hours. And Dr. John made sure to flip that switch at precise times every day. Every day we would see the interference pattern switch on or off in accordance with the flipping of the switch 24 hours in the future. If one day the interference pattern did not flip on or off, Dr. John would know something prevented him from arriving the following day to flip it. In such a way, someone could even anticipate their death or other such severe events.

I could go on and on with how this could be used for predicting more and more meaningful events about the future…so I must be wrong somehow.

Please let me know.

2

Just wanted to correct something. I think bouncing the photons off of a mirror in Geosyncronous orbit would be a better means of gaining significant time delay for the switch.

3

No physics gurus available? :frowning:

4

I’m not a physics guru but if I understand the article correctly then what you are saying would certainly SEEM to logically follow from the article’s conclusion. It would seem to imply that a kind of “time radio” could be built which could be used to send binary info back in time. The implications of this would be staggering and it would be fun to speculate about what effect this would have on human society. Like I said, IANAPG, but I’m sure that someone who understands the physics better will come along and burst this bubble.
I will say that it does seem like it would cause some paradoxes. What would happen if I decided to be perverse and flip the switch to the opposite of what is indicated? For example, I see an interference pattern but then flip the switch so that I receive the which-path information? Would I slip at the last moment and flip the switch to the opposite of what I intended? Would the switch break or get stuck or would some other malfunction occur that would cause it to act consistently with the previously received result?
Is it possible to change the future which has been predicted? If not, then the device’s usefulness would be limited to only being able to act in ways which would not change the predicted event. For example, you might be able to use it win the lottery since placing the bet would presumably not alter the winning number. But could you use it prevent a future terrorist attack? You might be able to get around this by being careful about what info you send. For example, if you received a time-radio message stating that Joe Blow intends to commit some act at some date and time, you could detain Joe (I’m ignoring the legal complications) without changing the fact of his intent and therefore not making the message invalid. But it seems odd to me that the mere phrasing of the message would make a difference.

5

I don’t have the time to read the link in detail, but here’s my WAG as to where your thought experiment would fail. (N.B. IANAP)

By replacing the beam splitters with a human you are eliminating the probabilistic 50-50% effect of the beam splitters with a deterministic effect. The human operator would flip the switch forcing the beam to take one path or the other. This would destroy the quantum information which is the basis for the experiment.

There is a comment in your link that says that in some quantum eraser experiments the choice of the path is made by the experimenter instead of a beam splitter, but I am sure that the setup would not be compatible with your thought experiment.

6

ready, faster than light consequences of entanglement like that which you are talking about DO occur and seem at first glance to weirdly predict the future. Only they don’t. This is because no information has been transferred.

In order for this experiment to work, we have to set up something that has a length that is long enough so that the time delay between each detection is noticeable. This means we have to have a huge experiment. If we increase our time delay by 20 minutes as you suggest, it would have to be roughly as large as the distance from Jupiter to the Earth.

What happens? You make your measurement. “Aha!”, says you, “Dr. John will flip the switch in 20 minutes.” You send a message to Dr. John, “Did you flip the switch? I say you are going to!”

Dutifully, the message travels to Dr. John at the speed of light. 20 minutes later, just after Dr. John flips the switch your message arrives. “Why yes, I did!” says Dr. John. Ooops, the event HAS ALREADY occurred in the past. No information was transmitted faster than the speed of light and there was nothing you could do to “predict” Dr. John’s flipping of the switch. In reality his flipping of the switch did not occur in your future light cone, it took place at a space-like separation from you and therefore there was no real time-like separation between the two events. Therefore no prediction of the future occurred!

This makes some free will theorists go bonkers with anger.
Except on close analysis, as there really is no violation of causality going on, in a sense Dr. John made the decision and you measured it AFTER he made it. They are both equally valid interpretations as no information was passed faster than the speed of light.

Isn’t entanglement a joy?

7

P.S. It seems you wish to get around this paradox by adding in mirrors and other “detectors” along the trips of these photons. You cannot do that because any detector like a mirror or a fiber optic cable will collapse the wavefunctions of the photons you are trying NOT to measure. You will find them better determined than what you need in order for an interference pattern to show up and your experiment will always give one result instead of the desired “two” in order for “choice” to occur.

8

Does this mean that I couldn’t place a mirror 10 light minutes away, thus getting a 20 minute delay, while keeping Dr. John and the switch right next to me? Does bouncing the photon off of that mirror constitute a measurement? What is it measuring? (I’m not disputing you, I’m trying to learn.)

9

It looks like my response to you, David M. was eaten in the latest board death. Basically the answer is that the mirror acts just as a detector. You can’t simply bounce the photons off a mirror or send them through a fiberoptic cable that loops back without destroying the interference pattern you need (and the associated indeterminate entaglement). What the mirror measures is basically the photon’s state. It acts a detector the same way a screen acts as a detector in the traditional double slit experiment.

10

What if we were to drap a small black hole out to where Jupiter is and shoot a photon so that it gets bent 180 degrees by the black hole. I am pretty sure that no observation would occur yet the phton comes back to the original position 20 minutes later whilewe can communicate the results in far less than 20 minutes.

11

No, a black hole as you describe its purpose, Shalmanese, still acts as a detector. This is because in order to gravitationally lens coherently (and get to Dr. John) you need to have exactly the right trajectory for your photon beam. This acts as a photon selection, the very thing you cannot have in the experiment in order for it to work.

You can, in fact, use large masses as a sort of absurdly large double slit experiment as proposed by John Wheeler. If it happens that a beam of particles coming from a distant quasar is lensed, the various microlensing events could conspire just right to create alternating interference and non-interference patterns, dependent upon how “determined” the photons were during each lensing event.

12

Well, I’m not a physics guru, but I am a physicist. I have to disagree with much of what’s been said so far. First, though, I agree that of course you can’t use the quantum-eraser experiment for faster-than-light communication, let alone to predict the future. But the reason has nothing to do with mirrors.

JSPrinceton claims that (1)your measurement allows you to determine whether Dr. John will flip the switch at some point within the next 20 minutes, but that (2)this does not cause a paradox because you can’t tell him your result before he actually does the action. But both of these statements are false.

If you could determine whether Dr. J would flip the switch 20 minutes from now, you would have achieved one bit of communication, from Dr. J to you, outside Dr. J’s future light cone (this is easy to see by drawing a spacetime diagram; I’ll try to ASCII one up if you want), violating causality. If quantum mechanics obeys relativity, you could use two such experiments to allow communication into your own past. (Any relativistically-invariant physical property which can be used to transmit information outside the future light cone can be used to transmit information into the past light cone, something else which can be seen pretty easily using spacetime diagrams.)

But in fact your measurement cannot determine this; your measurement results will be identical, whether or not Dr. J has flipped his switch. The interference patterns that appear in the quantum-eraser experiment only appear in the correlated detection rate between D[sub]0[/sub] and D[sub]1-2[/sub], and this of course can’t be determined until you and Dr. J meet and compare notes. (If you read the paper, you find that the joint detection rates R[sub]0i[/sub] satisfy R[sub]01[/sub]+R[sub]02[/sub] = R[sub]03[/sub]+R[sub]04[/sub]. For your proposed experiment, if Dr. J has flipped the switch, the detection rate you measure at D[sub]0[/sub] is R[sub]01[/sub]+R[sub]02[/sub]; if not, it’s R[sub]03[/sub]+R[sub]04[/sub], the same rate.)

The deal with the mirrors is mostly a red herring, I think, though I am not sure about the details. It’s true that a mirror can, in principle, be used to detect photons bouncing off it, by measuring the momentum transferred. This is also true of the beamsplitters in the experiment, and of any other method you might want to use to redirect your photons. But the experimental results shown in the paper demonstrate that this measurement can be made weak enough that it does not destroy the interference effect, at least for the wavelengths of light (~700nm) they used–otherwise, whenever the photon reflected off BSA or BSB it would be “measured” and there would never be an interference pattern.

I should also point out that there’s nothing magical about using entangled photons, as described in this paper, to design a quantum-eraser experiment. You could achieve the same results by simulating the system with a small quantum computer, for example. (Similar experiments have been done, for example, using an NMR quantum computer.) In this case, the whole setup is localized in a small space, and Dr. John could be sitting right beside you; you don’t need any mirrors or black holes orbiting Jupiter to reflect your photons back to you. :slight_smile:

13

From my quick jot of a spacetime diagram, Omphaloskeptic, this is what I see.

  1. the separation between the events is spacelike not timelike.

  2. The event of Dr. John “deciding” will always occur before the photon arrives.

  3. The best you can do is speed of light communication. This is because Dr. John’s decision event in spacetime is elsewhere and not in either lightcone. So there is no communication of information (as the experiment is one-way only). The events of detection and measurement must necessarily occur simultaneously in the photons “rest-frame”.

I guess I read the paper differently than you. The measurements look like they are different depending upon what the set-up looks like. Of course, you need to ask Dr. John what he did, which implies, at best, speed of light verification. This is just like the spin-measurements made of entangled electrons down under.

14

  1. Yes, the separation between the events is spacelike. But if communication is possible between two spacelike-separated points in spacetime, and if relativity holds and space is isotropic, then this communication mechanism can be used to communicate back in time. (Sketch of method: Let communication be possible from point A to point B, where A and B have spacelike separation. Then there is an inertial frame with A at the origin in which the time coordinate of B is negative. So communication is possible from (0,0) to a point (r,-t) (t>0). Now communicate from (0,0) to (r,-t) and from there to (0,-2t) (this step requires relativity and sufficient symmetry of space).)

  2. Yes, Dr. John’s decision always occurs before the photon from you (point B) to Dr. John (point A) arrives. But that’s not important. The point is that if you could see the interference pattern appear or disappear based on Dr. John’s actions at a spacelike-separated point, Dr. John would be able to communicate with you, outside his future light cone, and outside your past light cone. (That is, you could tell what he was doing before a photon could be sent from him to you.)

  3. It’s true that relativity (together with causality) limits communications to the speed of light. But seeing the interference fringes appear or disappear based on actions of Dr. John outside your past light cone (not just outside your future light cone) would be faster-than-light communication, and (as I sketched in (1)) allow communication back in time.

The “measurements” (in the paper, figures 3 and 4 vs. figure 5) do look different based on the setup. I put “measurements” in quotes, though, because these are not the results of local measurements that you make. They’re the results of correlated measurements (the “joint detection rates” R[sub]0i[/sub]), where you make a measurement with D[sub]0/sub and Dr. John makes a measurement with D[sub]i[/sub], and you only plot the results when you both measured a photon.

If, instead of the joint detection rates, you just plot your overall detection rate, there’s no change in what you see whether or not Dr. John has flipped the switch. (If Dr. John has flipped the switch, your detection rate is R[sub]01[/sub]+R[sub]02[/sub]; if you look closely at figures 3 and 4 and imagine adding the two curves up, you can see that the peaks in one are troughs in the other. The sum will be indistinguishable from the sum of figure 5, showing R[sub]03[/sub], and a similar, but slightly shifted, plot showing R[sub]04[/sub].

So, the local measurements look the same, independent of the setup. As you say, you need to ask Dr. John what he measured, and correlate his measurement results with yours, to see any interference effects at all. The theory behind this experiment is, as you said, quite similar to the theory of spin measurements on entangled electron states: The results of a local measurement on one of the individual electrons are random and convey no information about any measurement which is made on the other electron; only when the two measurement records are brought back together can correlations be seen.

15

I think, Omphaloskeptic we are actually in agreement. No communication can take place between two events with space-like separation. There is no means of transferring any information about whether or not Dr. John pulled the switch because the measurement is done in a reference frame that is necessarily simultaneous in the two photons “rest frame”. Therefore there is no causality paradox.

Where I like a slight-rewording is on this point,

because of the way the experiment is set up, Dr. John really ISN’T communicating with you. You cannot see the interference pattern in this way. This is just like any other entanglement experiment. We separate the entangled electrons and then take two event measurements that will always be correlated such that spin is conserved. This spooky action of the distance SEEMS like ftl, but it is not for the reasons you and I both outline. In a way, Dr. John really doesn’t make a decision as much as he is making a selection of certain photons that are measured (just like any given entanglement experiment). The experiment will show correlation only for the photons that are properly selected. This is why comparing notes is needed.

I think

bears repeating. It really is the crux of my argument and yours.

This is why the mirror becomes problematic, because it’s conspiratorial. Either you get no signal because of detection effects at the mirror’s end or you get no signal because of the addition of the detection rate components. I don’t know that we can say for sure which answer is actually correct (I think they are actually both saying the same thing).

16

OK, it’s possible that I just misunderstood your earlier post. When you said

I read this as meaning that you thought you could make some measurement allowing you to distinguish the cases {Dr. J will flip the switch} and {Dr. J will not flip the switch}. The ability to make such a determination would violate causality, which is what I was trying to explain.

I agree with everything you say here (and I think it’s a good explanation of what’s happening in the experiment, and more concise than mine). My quote there was, again, under the assumption that you thought you could see an interference pattern appear and disappear as Dr. John flips the switch. It wasn’t meant as an explanation of what actually happens, just of why what I thought you were saying didn’t make sense.

Hmm, I’m pretty sure these are not the same thing. Empirically, the setup for this experiment (figure 2) shows two mirrors (M[sub]A,B[/sub]) at which detection would provide which-path information for the photons before they reach D[sub]1-2[/sub] (the detectors which preserve the interference pattern), so if the mirrors provided a measurement then we would never see an interference pattern in the joint detection rates. (The diagram is a “schematic,” so it may not be entirely accurate, but the two mirrors are mentioned in the text as well, so they at least exist in the beampath.) I think that except in pretty exceptional cases (and of course in the much more common cases where the photon is absorbed or scattered instead of reflected specularly), reflection of photons off mirrors will not cause an effective decoherence to occur.

17

Thanks JS Princeton and Omphaloskeptic, very intersting.

18

I see! Well, the confusion in my lines lies in talking about “events” and observations. I didn’t say which way Dr. John flipped the switch for good reason, I just said he flipped the switch. That’s a weird distinction, I know, but it’s basically what we have to drive at in order to understand the way the experiment is set up. We should be able to know that Dr. John has been able to make some selection measurement since we made a measurement, but not which way the selection goes until after the inevitably “Did you flip the switch?” Of course, Dr. John could have blown the experiment to smithereens at his end, and we won’t know about it until after the question gets asked. It is important that the interval be light-like in separation otherwise you will not see any measurement, as you pointed out.

These mirrors, as far as I can tell, do not provide detection because they are “blind” to the effects of the reflection. There is no interruption of the double-slit effect whether they are there or not. Putting a mirror at the D[sub]1-2[/sub] would however be a way of describing the revised experiment.

That’s right. It’s not decoherence that’s occurring but a selection of a collapsed “detected” wavefunction that is dependent upon the actual way the mirror is used in the experiment.

I think this is mostly semantics at this point.

19

https://youtu.be/2Uzytrooz44

So check it out. Four years after I first published this post, PBS spacetime does an awesome show about this exact setup i outlined. I am sure others had the idea of course, but it is cool to finally have a mainstream show demonstrate why this is such a tantalizing experiment, as Matt would say.

I haven’t gone back to read if the SDMB replies/solutions are the same as the current physics experimental evidence. Anyways. Interesting.

20

Four?