Delayed choice experiment affecting the past

[Trollfaz]

I heard many people saying that the delayed choice experiment means that future events can affect the past. I doubt it as I believe that the wavefunction of the particle collapses as soon as a measurement is made. It does not matter where the detector is placed. As long as there is a detector, the interference pattern disappears.

 

[PeterDonis]

I heard many people saying

Do you have a reference? Textbook or peer-reviewed paper? Bear in mind that pop science sources are not valid sources here at PF.

 

[phinds]

I heard many people saying that ...

That's not an acceptable citation for PF and for good reason. Without a reliable citation we have to assume that you are just repeating pop-sci nonsense and/or statements by people who don't know what they are talking about.

For the 437th time, Peter has beat me to the punch.

 

[PeterDonis]

For the 437th time, Peter has beat me to the punch.

 

[Trollfaz]

Wait, does the position of the detector in the setup affect whether or not an interference pattern is seen?

 

[bhobba]

Wait, does the position of the detector in the setup affect whether or not an interference pattern is seen?

No - but of course it affects what you see ie you may place the detector at a point where there is no probability of detecting anything - that's why its best to use a screen or photographic plate so you can see the pattern build up

If you want to understand the double slit please study the following:
https://arxiv.org/abs/quant-ph/0703126

Its not perfect but its way better than pop-sci stuff.

BTW what you are thinking about is the transactional interpretation which is a legit interpretation:
https://en.wikipedia.org/wiki/Transactional_interpretation

If you want to pursue it further then Professsor Kastner, who occasionally posts here, has written a book on it:
https://www.amazon.com/dp/1108407218/?tag=pfamazon01-20

Thanks
Bill

 

[vanhees71]

It's of course utter nonsense that the evaluation of a measurement protocol "changes the past". To the contrary, you have done a measurement and irreversibly stored the result once and (in principle) forever to some storage medium (which can be a simple piece of paper or some digital storage medium).

Take the very nice experiment by Walborn et al. as an example:

S. P. Walborn, M. O. Terra Cunha, S. Pádua, and C. H. Monken, Double-slit quantum eraser, Phys. Rev. A, 65 (2002), p. 033818.
http://link.aps.org/abstract/PRA/v65/e033818
https://arxiv.org/abs/quant-ph/0106078

 

[Trollfaz]

Here is what Wheeler imagined from the experiment
1. A distant star emits a photon many billions of years ago. 2. The photon must pass a dense galaxy (or black hole) directly in its path toward earth. "Gravitational lensing" predicted by general relativity (and well verified) will make the light bend around the galaxy or black hole. The same photon can, therefore, take either of two paths around the galaxy and still reach Earth – it can take the left path and bend back toward earth; or it can take the right path and bend back toward earth. Bending around the left side is the experimental equivalent of going through the left slit of a barrier; bending around the right side is the equivalent of going through the right slit. 3. The photon continues for a very long time (perhaps a few more billion years) on its way toward earth. 4. On Earth (many billions of years later), an astronomer chooses to use a screen type of light projector, encompassing both sides of the intervening and the surrounding space without focusing or distinguishing among regions. The photon will land somewhere along the field of focus without our astronomer being able to tell which side of the galaxy/black hole the photon passed, left or right. So the distribution pattern of the photon (even of a single photon, but easily recognizable after a lot of photons are collected) will be an interference pattern. 5. Alternatively, based on what she had for breakfast, our astronomer might choose to use a binocular apparatus, with one side of the binoculars (one telescope) focused exclusively on the left side of the intervening galaxy, and the other side focussed exclusively on the right side of the intervening galaxy. In that case the "pattern" will be a clump of photons at one side, and a clump of photons at the other side. Now, for many billions of years the photon is in transit in region 3. Yet we can choose (many billions of years later) which experimental set up to employ – the single wide-focus, or the two narrowly focused instruments. We have chosen whether to know which side of the galaxy the photon passed by (by choosing whether to use the two-telescope set up or not, which are the instruments that would give us the information about which side of the galaxy the photon passed). We have delayed this choice until a time long after the particles "have passed by one side of the galaxy, or the other side of the galaxy, or both sides of the galaxy," so to speak. Yet, it seems paradoxically that our later choice of whether to obtain this information determines which side of the galaxy the light passed, so to speak, billions of years ago. So it seems that time has nothing to do with effects of quantum mechanics. And, indeed, the original thought experiment was not based on any analysis of how particles evolve and behave over time – it was based on the mathematics. This is what the mathematics predicted for a result, and this is exactly the result obtained in the laboratory.

 

[Demystifier]

I heard many people saying that the delayed choice experiment means that future events can affect the past. I doubt it as I believe that the wavefunction of the particle collapses as soon as a measurement is made. It does not matter where the detector is placed. As long as there is a detector, the interference pattern disappears.

I think you might be interested in https://arxiv.org/abs/1703.08341 Sec. 5.1.

 

[DrChinese]

... So it seems that time has nothing to do with effects of quantum mechanics. ...

In cases such as this, you are correct. In fact there are a number of experiments - such as the one referenced by vanhees71 - that indicate this. In the experiment, you can see from figures 2 and 8 that the coincidence pattern showing interference does not depend on which measurement comes first (as you might expect). I disagree with his characterization as to whether this indicates the past is changed or not. The conclusion [ordinary causal ordering is NOT respected] is generally considered to be interpretation-dependent, although definitely a minority position.

And your point is generally true with experiments on entangled pairs. (You can even entangle particles after the fact!) Time ordering in QM does not operate the same as in the classical world, and this is evident in the mathematics as you point out. How you choose to make sense of this is, to great extent, a matter of opinion. Wheeler's paradoxical example demonstrates such.

 

[vanhees71]

If you deny my interpretation, are you saying that I change the past by choosing subensembles by looking at a fixed measurement protocol? How can one justify this in a non-esoteric way?

 

[DrChinese]

If you deny my interpretation, are you saying that I change the past by choosing subensembles by looking at a fixed measurement protocol? How can one justify this in a non-esoteric way?

It is no different than asserting an action on a particle "here" has an immediate impact on a distant particle "there" (Bohmian interpretation, which I believe you also deny). All of the standard interpretations are scientifically viable at this time. I am simply point out that yours is one. That's relevant to the OP's question, and is why Wheeler's example is something of a paradox:

Wheeler-Feynman Absorber Theory

 

[vanhees71]

The absorber theory hasn't lead to anything. It's just a dead end in the history of science.

Again, I'm asking, independently of which interpretation of QT you follow, whether it makes sense to say that I change the past only, because I choose some part of the ensemble of photons from a fixed list of measurement results. Imho, it doesn't make sense in any physical theory, because it gives up causality in its weakest form, and without causality we can give up any attempt to describe the world in terms of the natural sciences at all. I don't see any hint yet for such a drastic conclusion.

 

[DrChinese]

Again, I'm asking, independently of which interpretation of QT you follow, whether it makes sense to say that I change the past only, because I choose some part of the ensemble of photons from a fixed list of measurement results. Imho, it doesn't make sense in any physical theory, because it gives up causality in its weakest form, and without causality we can give up any attempt to describe the world in terms of the natural sciences at all. I don't see any hint yet for such a drastic conclusion.

Answering independently of interpretation is a bit difficult, because each interpretation makes a different deal with the devil. But in the spirit of your question: Sure it makes sense to give up causality. Most physicists already believe that there is nothing that determines outcomes in specific individual quantum measurements. Once you have given that up, why is giving up causality a big leap? You have basically already done that.

A time-symmetric class interpretation simply references both the future and the past as part of the context for a measurement. It naturally explains why Bell inequalities are violated because both Alice's and Bob's measurement settings (and no one else's) are components of the overall context. One of the issues with Bohmian logic for entanglement is that it does not explain why ONLY Alice and Bob's measurement settings are components of the overall context - after all, Alice and Bob have no privileged status in a theory in which every object in the universe exerts an instantaneous influence on every other. And I say that without trying to discuss the technical merits of any interpretation, just trying to say why a TS theory could make sense.

Oh my gosh, I've done it again. Drawn into discussiing QM interpretations. Just. Shoot. Me. Now.

 

[vanhees71]

No, what every physicist is giving up is not causality but determinism. Given the state of a quantum system at time $t=0$ and given the Hamiltonian of the system the state of the quantum system at any later time $t>0$ is exactly known due to the dynamical laws of QT. The state, however implies only probabilities for the outcome of measurements, i.e., the observables' values are indetermined (except you have prepared to state such as to make them determined, if this is possible at all).

 

[DrChinese]

Given the state of a quantum system at time $t=0$ and given the Hamiltonian of the system the state of the quantum system at any later time $t>0$ is exactly known due to the dynamical laws of QT.

Thanks for this. You have made my point: those same dynamical laws apply forward and backward in time. A convention is to label t=0 and some point where there is a measurement, but that is strictly arbitrary. At the quantum level, there is no way to distinguish past from future.

For example: Consistent histories is almost the reverse of MWI. What past cases lead us to the present - versus - what future cases result from the present. Those are pretty much the same thing.

OK, I'm sticking my head back in my shell. You get to say whatever you like, and I promise no further engagement. Period. End of subject. Mos def.

 

[vanhees71]

This is besides the point (and in fact there's a direction in time due to the breaking of time-reversal invariance by the weak interaction).

What I don't understand is, how you can think that I change the past of something (in this example the photons used in the experiment) which doesn't even exist anymore, when I evaluate the measurement protocol, which was fixed in the past. If you seriously think that, how can it make sense to you to pursue any natural science since then there'd be no natural laws to study, and thus you'd have no natural sciences at all.

 

[PeterDonis]

the breaking of time-reversal invariance by the strong interaction

Do you mean the weak interaction?

 

[vanhees71]

Do you mean the weak interaction?

argh. Of course. I'll correct it.

 

[Denis]

No, what every physicist is giving up is not causality but determinism.

No, you have to give up causality to save relativity.

Reichenbach's common cause principle is essentially powerful enough to prove Bell's inequality, if you use Einstein causality.

And determinism, as far as it plays a role in Bell's theorem, is derived, not assumed. So, giving it up does not help you at all.

Of course, one can reject Reichenbach's principle of common cause, and continue to name the remains "causality". But I don't understand what would be the point of such a notion of "causality" if there is no longer any need to find causal explanations of observable correlations. One would better name such remains "correlarity" or so. A definition is not really necessary, because its meaning is empty. Correlations are just fine, no need to search for causal exaplanations, astrology becomes a science again.

 

[vanhees71]

This is so far from any practice in contemporary physics that I cannot make any sense of it. Reichenbach is rather more like a philosopher than a physicist, and I feel incompetent to comment on philosophy of physics since it usually lacks clear definitions to get a clear meaning of it.

The point is that modern relativistic local QFT is causal in the usual sense as physicists use the word "causality".

 

[Denis]

It is causal in the weak sense, also known as "signal causality", which forbids to use FTL signals. This is the weakest imaginable sense. If it is the "usual sense" for those using QFT, ok, so be it. But in fact it tells us nothing about causal relations. It is only about correlations.

Reichenbachs common cause principle is quite precise math, so your "usually lacks" is not an excuse. From https://plato.stanford.edu/entries/physics-Rpcc/

[W]hen Pr(A&B) > Pr(A) × Pr(B) for simultaneous events A and B, there exists an earlier common cause C of A and B, such that Pr(A/C) > Pr(A/~C), Pr(B/C) > Pr(B/~C), Pr(A&B/C) = Pr(A/C) × Pr(B/C) and Pr(A&B/~C) = Pr(A/~C) × Pr(B/~C).

And because it is simple and precise math, it is also usable to prove theorems, like Bell's theorem.

 

[vanhees71]

Causality in its weakest form simply means that an effect is caused by other effects taking place temporally before. In other words, the state of a system (described in QT by the statistical operator, in classical physics by a phase-space distribution function, etc.) evolves in time due to the dynamics of the system (defined by a Hamiltonian), such that the state at time $t$ is a functional of the state at all times $t'<t$. The fundamental theories are all local in time, i.e., given the dynamics of the system the state at any time $t$ is uniquely determined by giving the initial state at one fixed time $t_0<t$.

In special relativity you have in addition the constraint that events can only be causally connected if the one is in the forward lightcone of the other, i.e., "signal propagation" is $\leq c$, where $c$ is the fundamental limiting velocity of Minkowski space (to the best of ower knowledge electromagnetic waves travel with this limiting speed, or in other words the em. field is massless).

 

[Denis]

If you interpret QT according to something like the minimal interpretation, you don't have that "weakest form" of causality, because the measurement results pop out into existence without any causal explanation from the past, only the probability of different measurement results is what is caused by the preparation procedure.

But, ok, you have a point that "the weakest imaginable sense" is an exaggeration, let's replace it with "the weakest sense relevant in EPRB discussions". I have never seen anybody to propose signal locality has to be given up because of Bell.

Signal locality will fail in quantum gravity, clearly, and the only way to save causality will be to connect it to a hidden background, where it becomes, indeed, simply related to the temporal order in absolute time.

 

[vanhees71]

There's no "causal explanation" for the outcome of a single measurement. The only thing you know, given the quantum state of a system, is the probability, with which a precise measurement gives one of the possible results. Standard QT in the minimal interpretation is ignorant about what this outcome in a specific case might be, and indeed the confirmation of the violation of Bell's inequality, together with locality, which underlies the most successful theory ever (the standard model of elementary particle physics), in accordance with the predictions of QT tell me that this is the correct picture. There's no "cause" for the outcome of a specific measurement. In other words, QT tells us that nature behaves probabilistic on its most fundamental level. All our debates about "reality", the "measurement problem", and related topics in connection with QT comes from our unwillingness to accept this fundamental "randomness" of nature due to our "classical prejudices".

 

[Denis]

All our debates about "reality", the "measurement problem", and related topics in connection with QT comes from our unwillingness to accept this fundamental "randomness" of nature due to our "classical prejudices".

Not at all. No realist has any problem with accepting randomness, and if this randomness is fundamental or not is an obviously metaphysical hypothesis that you should despise it.
The measurement problem is a problem of some interpretations, which want to present QT as something fundamental, and not a problem at all for realistic interpretations.

A very funny thing, indeed, is that the prejudice against "philosophy" and the rejection of discussions about "interpretations" is very closely correlated with the defense of highly suspect, highly metaphysical ideas, especially with the ideas that those obviously incomplete theories we have today (given that we have not yet a quantum TOE valid at all distances, this failure is obvious) are nonetheless in essential parts fundamentally true.