Talk:Delayed-choice quantum eraser/Archive 5

Walborn experiment, step by step
Let's get a correct understanding of the Walborn experiment. It's going to be covered in this article or a related article, and the understanding of what it does is currently in dispute.



The important point that needs to be pinned down so there is not interminable dispute on the matter is that after the quarter wave plates (circular polarizers) are installed on the double slits, interference has been destroyed. The contrary circular polarizations applied to all the photons "split in the quantum sense at the double slits" or what I will call the "photon splits," make them unable to get back together again. I hope it is not too crude an analogy, but it's like taking a large breeding population of dogs and locking chastity belts on all the males and on all the females. Until something is done about the chastity belts, productive coupling will cease.

One way to get the two halves back together would be to put a half-wave plate right on top of one of the quarter-wave plates. That would make every "photon split" end up with the same circular polarization, so they can superimpose and exhibit interference. That is not the way chosen by the experimenters who devices this apparatus because they wanted to be able to delay the time at which the polarization conflict was resolved and investigate the possibility of undoing "knowledge of which path a photon had travelled through the double slip diaphragn" until a microsecond before the photon shows up on a detection screen.

They achieved time control by making variable the distance a photon travels to the detector in the signal photon part of the device. Instead of adding a half-wave plate to the idler side, they added a linear polarizer to the signal photon's path, and contrived to make it possible to rotate. Rotating it one way "fixes" the circular polarization problem in the idler pathway in one way. Rotating it the other way fixes things in another way.

My first row above shows what would happen without the double-slit diaphragm or any other additions. There are just two entangled beams of photons pointed at some detectors.

My second row above is equivalent to Walborn et al. Fig 2. "An Interference pattern due to the double-slit is observed." (p. 6) The little graphs should be more like Walton's Fig. 6, i.e. more symmetrical. Feel free to fix my SVG for me.

My third row above is equivalent to Walborn's Fig 3. "Interference has been destroyed."

My fourth row above does not concern what Walborn did. It is just one way of turning interference back on almost as soon as it has been destroyed.

My fifth row above is equivalent to Walborn's Figure 4. "POL1 was set to theta... Interference has been restored to the fringe pattern.

My sixth row above is equivalent to Walborn's Figer 5. "POL1 was set to theta + π/2.... Interference has been restored to the antifringe pattern.

If anybody thinks I am misrepresenting the Walton paper, then let me know now. I am basically paraphrasing what the paper says but in cartoon form. Don't expect my little pictures (some of which I stole from S. aurantiaca) to reproduce the detail shown in the real graphs at the end of the article. P0M (talk) 01:53, 9 February 2014 (UTC)


 * Your hand-generated images might look better if generated by computer. Here is an example of some computer-generated graphics that I put together for the Michelson-Morley experiment article: Illingworth simulation.png Stigmatella aurantiaca (talk) 14:57, 9 February 2014 (UTC)


 * Yeah, if I had the time to learn how to do it I would love to make a moving image that swaps the random inputs that would go to the signal path detector.P0M (talk) 15:16, 9 February 2014 (UTC)


 * Describe to me what you want. Make a few concept sketches and upload to your Google Drive account. I have a week to go before having to go back to work. Maybe I will have time to implement your ideas. I do animations as well. For instance, this one for Michelson-Morley experiment MichelsonMorleyAnimationDE.gif was completely computer-generated. Stigmatella aurantiaca (talk) 15:27, 9 February 2014 (UTC)


 * Here is a short mock-up. I guess I could have put a sort of four-position arrow in one corner to indicate whether the current dot is coming in to match one in detector 1, 2, 3, or 4.

http://www.china-learn.info/MOV%20files/Detector0sums1to4.mov


 * Ideally there would be a random function to determine which of the four detectors would score a hit next, and within that loop another four random choices of what value from under the curve for that set would be mapped on the screen. I've just done my mapping by guess. P0M (talk) 01:44, 10 February 2014 (UTC)


 * Hmmm... For Wikipedia, I would have to use the non-proprietary OGV format. That's no problem with all of the free converters out there. Can't do GIFs because (a) GIF format is a serious memory hog for anything more than toy animations, and (2) there is a serious MediaWiki bug that prevents working thumbnails being created of GIF files with more than a few dozen frames. These are only minor details, though. It would be nice if I were allowed to write Javascript or a Java applet to create the dots on the fly. Have to think about this for a bit. Stigmatella aurantiaca (talk) 05:11, 10 February 2014 (UTC)


 * I used the MOV format because that is what my system turns out and it will play in most web surfers. Judging by the results of the discussions among the several people involved here, it is not easy to see what is going on in the experiment. The article authors sometimes use "geek talk." (E.g., "To recover interference, we measure the polarization of the other entangled photon.") We need to make it as clear as possible to the average well-informed reader both what will be seen in the lab and what experimental conclusions can be drawn from these effects. Probably a complete computer simulation of the experiment would be the most useful even though it is not "encyclopedic." The reader could take POL1 out of gear, put it in "forward," put it in "reverse," and see what happens "on screen.]," pull out the quarter-wave plates and see what that does, and so forth. P0M (talk) 17:20, 10 February 2014 (UTC)

Let's regroup
In Chinese there is a useful description to characterize what we have been doing. People who do it are said to have started out at the open end of a (detached) water buffalo horn and are inexorably striving forth to the most remote, restricted, and probably inextricable dead end at the inside of the point of the horn.

Walborn et al. did an experiment to determine some of the features of time and locality that are weird components of the whole quantum mechanics conundrum. What they wanted, and what they got using their experimental apparatus, was an indication that something very noticeable happened that could be made to disappear by doing something in another part of the apparatus either before or after the tell-tale effect occurs on the detection screen they are watching.

As long as they get their effect and can prevent it or permit it, have it or destroy it, or however you want to describe it in words, they can get the results from their experiment that they want. They have established that one manipulation to entangled photons in the signal path can turn on one observed phenomenon in the idler path, and that reversing that manipulation in the signal path can reverse the observed phenomenon in the idler path. Furthermore, it does not matter whether the manipulation performed in the signal path occurs before or after its effect is observed in the idler path.

We will be doing the average well-informed reader a disservice if we go into the weeds with accusations about the validity of the interpretations used by Walborn et al. to explain why their experiment works as it does. As far as the average well-informed reader is concerned, the mechanism by which they relate time x events in the signal path with time y events in the idler path might as well be a black box. P0M (talk) 16:49, 10 February 2014 (UTC)


 * Maybe another part of Wikipedia policy will help here: Primary, secondary and tertiary sources: "Any interpretation of primary source material requires a reliable secondary source for that interpretation. A primary source may only be used on Wikipedia to make straightforward, descriptive statements of facts that can be verified by any educated person with access to the source but without further, specialized knowledge." Also, Exceptional claims require exceptional sources: "Any exceptional claim requires multiple high-quality sources. Red flags that should prompt extra caution include: ... challenged claims that are supported purely by primary or self-published sources ..." Both of these quotes are from Wikipedia core policy pages. Walborn et al. is a primary source so it cannot be used to support a (disputed) classical interpretation of the delayed choice quantum eraser. RockMagnetist (talk) 17:31, 10 February 2014 (UTC)


 * A further implication of that policy is that this discussion can end unless DParlevliet can find good secondary sources for the classical interpretation. RockMagnetist (talk) 17:33, 10 February 2014 (UTC)


 * Finally, some kind of resolution to this mess. I feel that I have been scheduled to give a talk at Smith Auditorium and went to Jones Auditorium instead. From the stage I looked out at a sea of darkness. The hollow-sounding echoes were subtly trying to tell me something, or so I thought.P0M (talk) 17:42, 10 February 2014 (UTC)


 * I would like to add that everyone involved in this dispute has shown remarkable patience and civility, considering how frustrating it has been for all of you. RockMagnetist (talk) 19:10, 10 February 2014 (UTC)


 * You deserve the thanks of all of us.P0M (talk) 19:14, 10 February 2014 (UTC)

Add text that DCQE is inherently quantum phenomenon
Perhaps one way to resolve this is to add some text to the article saying that the delayed choice quantum eraser is an inherently quantum phenomenon and describing what parts of the experiment cannot be explained by classical mechanics. Can you find sources that say this? RockMagnetist (talk) 17:03, 9 February 2014 (UTC)
 * 1949 may have been the last year that any physicist even thought of using classical mechanics to explain this experiment. It's harder to "explain" using classical physics because there are lots of "Just because, sonny!" things that have to be said when you don't have an explanation for why interference occurs. Sorry. On the other hand, the experiment is regarded as the crown jewel of quantum mechanics.P0M (talk) 18:04, 9 February 2014 (UTC)


 * In the words of Richard Feynmann, ”[The double-slit experiment] is impossible, absolutely impossible, to explain in any classical way, and [it] has in it the heart of quantum mechanics. In reality, it contains the only mystery.”
 * — quoted, in another paper, from R. P. Feynman, Lecture Notes in Physics, Vol. 3 (Addison-Wesley, 1965), p.1-1.


 * I don't have the book, but this looks like what you want. P0M (talk) 21:28, 9 February 2014 (UTC)


 * Unfortunately few sources state the non-classicality of DCQE explicitly. Most sources make more general statements. They typically either state that any experiment using twin photons is non-classical (e.g. Walborn et al., "Spatial correlations in parametric down-conversion", Physics Reports 495, 87 (2010), "These correlations are of fundamental importance in all experiments with twin photons. [...]Both the time and spatial correlations have been proven to be non-classical.") or go the elegant way showing that classical single-photon interference and non-classical two-photon interference using entanglement are mutually exclusive (e.g. Abouraddy et al., "Demonstration of the complementarity of one- and two-photon interference", Phys. Rev. A 63, 063803 (2001), "The conventional theory of optical coherence dictates that the visibility of interference fringes equals the degree of coherence[...]The visibility of two-photon interference fringes in a Young’s double-slit experiment is governed by the degree of entanglement.[...]A basic complementarity between coherence and entanglement underlies the complementarity between single- and two-photon interference"). However, while elegant, I am afraid the latter explanation is far too technical for the average reader. Cthugha82 (talk) 21:55, 9 February 2014 (UTC)


 * Simple precursors to the DCQE experiment, such as the "Do-it-yourself quantum eraser" discussed in Scientific American, have classical-wave explanations. These precursors use single-photon interference. But versions using twin photons are intrinsically non-classical. For example, in Chiao, R. Y., P. G. Kwia, and A. M. Steinberg. Quantum and Semiclassical Optics. J. Eur. Opt. Soc. Part B 7.3:259 (1995), we read the following: "This particular version [i.e. a simple precursor] of the quantum eraser has a straightforward classical-wave explanation when the light source is describable in terms of coherent states. Thus it could be argued that there is nothing particularly quantum about this quantum eraser. Nevertheless, Jordan has proposed a similar Mach-Zehnder version of this experiment [23], in which he has argued on the basis of the correspondence principle that despite the existence of a classical explanation, such first-order interference experiments can be interpreted as true quantum erasers. However, in order to avoid any possible ambiguity concerning the quantum versus classical interpretation, we decided to use the nonclassical two-photon light source described above, in conjunction with the Hong-Ou-Mandel (HOM) two-photon interferometer [24], to demonstrate a quantum eraser with no classical analog." Stigmatella aurantiaca (talk) 06:53, 10 February 2014 (UTC)


 * One could first focus on the basic part below, so without entanglement, because only that is described by the classical explanation. However realize that if this basic experiment gives the same result, then classical no entanglement, second detector or coincidence is needed anymore (but it also does not change its explanation). DParlevliet (talk) 09:07, 10 February 2014 (UTC) [[File:Walborn1a.GIF]]


 * I do not get that point. All the classical stuff is possible only for the (simple) quantum eraser, but not for the delayed choice version. People already mix up these experiments frequently and do not understand that they are fundamentally different, so I advocate not to mix up these two experiments in this wikipedia entry as that will only create confusion for the average reader. Also, note that the citations given beforehand make it clear that the explanations for the simple quantum eraser and DCQE indeed are VERY different. Adding a comment that the DCQE is intrinsically non-classical and/or a comment that DCQE and the simple quantum eraser are very different experiments and should not be confused seems like a sensible thing to me. Cthugha82 (talk) 10:38, 10 February 2014 (UTC)


 * Alright, but is the classical explation I gave for only the QE above right? DParlevliet (talk) 11:10, 10 February 2014 (UTC)


 * I don't see what pedagogical purpose would be served by that. For delayed choice, you need entanglement. So why not start off with the quantum explanation? Stigmatella aurantiaca (talk) 13:49, 10 February 2014 (UTC)


 * The Walborn chapter does start with the full quantum explanation. But if you want to "describing what parts of the experiment cannot be explained by classical mechanic" (RockMagnetist) you must start with the part which can be explained by waves, which is the Quantum eraser according Cthugha82. Then step by step further until the wave has no answer. Then you know which part it is and how to describe. DParlevliet (talk) 14:16, 10 February 2014 (UTC)


 * I strongly disagree with that. Although they have a similar name, the quantum eraser and the delayed choice quantum eraser are completely different experiments and the quantum eraser cannot be considered a part of the DCQE experiment. They rely on a different interference mechanism as has been stated above. The simple quantum eraser is simply a different (and almost unrelated) experiment. The only part which can be explained classically in the DCQE is the coherent pump beam before it hits the BBO crystal. However, that is so trivial that mentioning it does not make sense here. Cthugha82 (talk) 14:36, 10 February 2014 (UTC)


 * But the title of Walborn's article is "A double slit quantum eraser", so no DCQE. So that is possible for classical stuff as you mentioned above? Anyway, my question was if Walborn's basic double slit above can be explained classical. If so, then you can argue why adding the rest of the experiment makes it conflict classical. DParlevliet (talk) 15:30, 10 February 2014 (UTC)


 * From the abstract of Walborn's article: "In addition, we perform the experiment under ‘‘delayed erasure’’ circumstances.". He does both. Whether or not one can explain a simple quantum eraser classically is irrelevant for this discussion as it is about the delayed choice experiment, which cannot by any means and in any part be explained classically. If you want to discuss the simple quantum eraser, you might want to do that on the talk page for the simple quantum eraser experiment. It is completely out of place here. The delayed choice is not simply "the rest of the experiment". It is a fundamentally different experiment as has been pointed out and backed by citations several times now. Cthugha82 (talk) 15:54, 10 February 2014 (UTC)


 * Walborn et al. were very careful to distinguish "delayed erasure" from "delayed choice" in their paper. They are not the same thing. Patrick0Moran, Cthugha82 and I have all told you that we do not accept your attempt to merge the Quantum eraser article with this one. The DCQE experiment is absolutely not "the rest of the experiment." Stigmatella aurantiaca (talk) 16:12, 10 February 2014 (UTC)


 * You are correct. Unfortunately, explaining things and asking for responsive answers to questions seems to be getting us nowhere, sad to say.P0M (talk) 17:08, 10 February 2014 (UTC)


 * Anyway, I will rewrite the wave paragraph, so every word is referenced and it will have no interpretation. DParlevliet (talk) 21:38, 10 February 2014 (UTC)

Discussion subsection titles
The discussion subsection titles seem very unencyclopedic to me:


 * Problems with using retrocausality
 * Details pertaining to retrocausality in the Kim experiment
 * The main stumbling block for using retrocausality to communicate information
 * Yet there are those who persevere in attempting to communicate retroactively

The first one is probably OK, but they all seem a bit wordy to me, almost like a lead-in sentence rather than a section title. "Retrocausality in the Kim Experiment" probably suffices for #2, #3 and #4 can probably be merged (since the 4th section is just a single sentence anyway) and put under some simpler heading like, "Retroactive Communication" or something of that nature. All that said, I'm not sure that encyclopedia articles usually have a "discussion" section - most people won't know what that means. The whole section might be better titled "Implications" or "Interpretation" or something of that nature. 0x0077BE (talk) 18:45, 10 February 2014 (UTC)


 * The subsection titles bothered me, too. Several of us here are currently engaged in lively debate concerning future directions for this article. I anticipate that in a few months it should be greatly improved. Thanks for your comments! Stigmatella aurantiaca (talk) 19:00, 10 February 2014 (UTC)


 * @ IP: Some of the subsections are of questioned utility anyway. There is really no compelling reason to discuss ansibles in this article. Those topics probably crept in because somebody came onto the talk page and asked for recognition that such proposals have been made, e.g., by Professor Cramer.


 * There is a philosophy question involved here, too. Should we be trying to lead inquiring minds by their virtual hands to understand things and supplement the sparse resources of some tiny high school in the hinterlands or the geographical center of the contiguous 48 states, or should we "encyclopedically" represent the short and sweet of what has become known in proper academic language? If the second choice is made we clearly have no business in discussing something that approaches science fiction. Cramer talked about experiments providing he got funding, but nothing has been heard from him on the subject in the several intervening years.


 * There is a huge problem involving "retrocausality," locality, the arrow of time, etc., etc., etc. The experts n the field don't seem to me to be able to deal with the issues in ordinary English. Maybe they can't deal with it in multidimensional math either. Who know? Certainly not I.


 * Maybe we should start with a discussion of what, if anything, the user needs to be told about "retrocausality." By the way, I think that is another dumb term that got into the system by the carelessness of physicists who are better at math than at language and its possible distorting effects on thought.P0M (talk) 19:11, 10 February 2014 (UTC)
 * FYI I'm not an IP user, my chosen username is a 6-digit hex number. Even though I have a physics background, I'm usually mostly disinterested in these sorts of causality/implications type things and generally feel that people are going out of their way to make things seem weirder than they are. Given my disinterest, I'd rather stay out of the discussion about what should be included, but whatever the final decision, I think it should be done the right way. 0x0077BE (talk) 19:36, 10 February 2014 (UTC)


 * That's what I get for trying to beat an ice/snow storm to the supermarket. Sorry. P0M (talk) 22:29, 10 February 2014 (UTC)

Simulations
- Do these work for you? Stigmatella aurantiaca (talk) 21:52, 10 February 2014 (UTC)


 * That simulation is wonderful!


 * Never satiated, I ask whether it would be possible to add one more "window," in which the four existing windows could be summed step by step. Maybe the arrangement could be like the 5 out of 9 black squares in the corner of a checkerboard with the new D0 "detection screen" in the center. That should make it a little clearer to readers why the experiment has to use the coincidence counter to sort out the four pictures.


 * Maybe instead of letters you could use the actual detector identities, i.e., D1 and so forth.P0M (talk) 23:06, 10 February 2014 (UTC)


 * I had labeled the panels A, B, C, D since this first try was only a proof-of-concept. I wanted your feedback before proceeding further. I'm not sure the checkerboard arrangement would really be all that enlightening. The problem is, I don't know what sorts of losses occur in the optical system, so I don't really know what would be visible at D0. My intuition is that, absent filtering by the coincidence circuits, what would be detected at D0 would be dominated by the direct signal. But that's only my guess. I'll redo to show the joint detection rates R01, R02, R03, R04. Give me a few hours. Stigmatella aurantiaca (talk) 00:34, 11 February 2014 (UTC)


 * I suspect that they are not doing this experiment under blackout conditions. On top of that the BBO may not always emit entangled photons. They can filter out extraneous photons by insisting on only "seeing" a photon and its entangled twin that arrive at the (any systematic delay adjusted) same time. So, everything that shows up and gets counted on D1,2,3,4 is matched by an entangled twin that shows up at D0. My mock-up was done by taking a hit from each of the four idler detection screens and painting its copy on the screen that is shown getting speckled by scintillations. (They're all in my mind, of course.) P0M (talk) 01:28, 11 February 2014 (UTC)

OK, I've uploaded new versions of the files labeling with the joint detection rate R notation used in Kim et al. I've lined up R01 and R02 vertically for obvious reasons. Stigmatella aurantiaca (talk) 02:26, 11 February 2014 (UTC)


 * The results are very effective. Compare them to the real video done of a double-slit experiment done with electrons. P0M (talk) 02:58, 11 February 2014 (UTC)
 * I used the graphs themselves as input data for the simulations. The dot densities in each panel reflects the graph amplitudes of the corresponding panel above. Stigmatella aurantiaca (talk) 05:09, 11 February 2014 (UTC)


 * So you have some kind of OCR that can read and interpret graphs? That's rather impressive. It definitely beats doing the job with protractors and graph paper. P0M (talk) 06:43, 11 February 2014 (UTC)


 * No, nothing so fancy. The graphs are just datafiles. Computer programs read and write image datafiles with great ease nowadays. I just wrote code to search for non-white pixels and to note their X-Y positions. Thirty years ago, things were different. Back then, you had to actually know and understand the binary formats of the files to work with them. I remember compressed TIFF files were a real pain. [Imagine my speaking the following in a cracked, whiney, old man's voice] "Young programmers nowadays, they don't know how easy they have it, with full image support built into their software languages..."
 * Of course, before I was a software engineer, I was a molecular biologist, working in the early days of the molecular biology / recombinant DNA revolution. I have some really old tales that would scare the pants off the young-uns, like how much 32P, 125I, 14C, and 3H I was exposed to.
 * Stigmatella aurantiaca (talk) 09:01, 11 February 2014 (UTC)

Administrators' noticeboard
My answer to the command of Stigmatella and POM on the Administrators' noticeboard: All comments above concerns the one disputed paragraph as I mentioned in my report. I have agreed to edit and improve this paragraph. A quick look into the article history shows that until 2 February no edit was disputed nor reverted. Also after the blockage none of the editors above gave an accepted reason for deletion according Wiki rules. Therefore I propose to go back to the last not disputed version of 2 february and start editing from there. I have no problem in deleting the disputed paragraph until I have improved it. DParlevliet (talk) 14:18, 11 February 2014 (UTC)


 * You do not own this article. The decision to revert to the version of 07:41, 1 October 2013‎ was a consensus decision by Patrick0Moran, Cthugha82 and myself. Our consensus decision is that all subsequent improvements to the article should start from the 1 October 2013‎ version. We dispute the 2 February 2013 version. Stigmatella aurantiaca (talk) 15:51, 11 February 2014 (UTC)


 * It is clear from reading Critique and A fresh start that there was indeed a consensus among Patrick0Moran, Cthugha82 and Stigmatella aurantiaca to revert the article to October 1 2013. Your contention that "no edit was disputed nor reverted", is based on article history alone, ignoring the attempts of Patrick0Moran to discuss your changes on this talk page. He was extending a courtesy to you - discuss before reverting - that you were not extending to previous contributors: you made wholesale deletions of fairly good text without any discussion at all. Just look at the increase in size of the article in this diff when Stigmatella aurantiaca reverted to 1 October 2013. RockMagnetist (talk) 17:34, 11 February 2014 (UTC)


 * Just to be clear, I support the revert to 1 October 2013 as well, and judging by the comment by in RockMagnetist position, he does as well. That makes five of us. RockMagnetist (talk) 18:05, 11 February 2014 (UTC)

Comments on the 1 October 2013 version
Although the 1 October 2013 version is a good starting point, it does have some issues, some of which have been mentioned above. It also has a choppy style, with coherent discussions broken up into multiple paragraphs. And it needs more citations. RockMagnetist (talk) 17:39, 11 February 2014 (UTC)
 * Very definitely agree! Stigmatella aurantiaca (talk) 19:35, 11 February 2014 (UTC)
 * Yes, I agree, too. The discussion section really needs some improvement. It might be a good thing if the subsections get merged and the description gets clarified a bit. Do we agree that there is no retrocausality necessary to explain the experiment? I guess so. Cthugha82 (talk) 09:52, 12 February 2014 (UTC)


 * The whole thing about retrocausality gets into physical and philosophical interpretations that I'm not equipped to deal with. For example, while researching secondary sources for this article, I came across various unrefereed articles by David Ellerman dealing with the "separation fallacy" and how, in his opinion, Wheeler etc. have gotten things completely wrong. Brian Green does an informal analysis of Kim et al. in The Fabric of the Cosmos and sides against retrocausality, as do (if I'm reading them correctly) Kwiat, Paul G., Peter DD Schwindt, and Berthold-Georg Englert. "What does a quantum eraser really erase?." Mysteries, puzzles, and paradoxes in quantum mechanics. American Institute of Physics, 1999.


 * Because there has been a lot of debate on retrocausality, the topic does need to be covered, even if our philosophical bent is against it. Lots of users would be accessing this article precisely for coverage of this topic. Stigmatella aurantiaca (talk) 10:26, 12 February 2014 (UTC)


 * Sure, a discussion of that topic is necessary, but I just wanted to ask whether the editors here agree that the article should represent the discussion and come to the conclusion that retrocausality is not necessary to explain DCQE. This is at least the typical contemporary position, expressed e.g. in Phys. Rev. Lett. 107, 230406 (2011) ("ArXiv version") where the authors explicitly disagree with Wheeler in the last paragraph. Cthugha82 (talk) 12:06, 12 February 2014 (UTC)


 * That's a much better reference than any I've dug up on this topic. Thanks! Yes, I'd agree with that. Stigmatella aurantiaca (talk) 12:21, 12 February 2014 (UTC)


 * My own way of understanding the Wheeler experiment is that there is no need, even in the cosmic version, to claim that the history of a given history over millions of years. I hadn't run across anybody that agreed with me until the several analyses were offered here.


 * We have what happens in the lab and we have equations. Anything more is just interpretation, but there is hope of getting rid of some interpretations by showing them not to be self-consistent. I alse agree that "erasure" doesn't destroy anything. It's a very unfortunate term to choose for what in all cases I've studied amounts to marking something and then taking away the effect of that act of marking. I will spare everybody repeating what I've already said on that matter. (P0M (talk) 15:07, 12 February 2014 (UTC)

RockMagnetist position
@RockMagnetist: I see you are part of the "we" which are regrouping, (I suppose against me), but I have not seen what position you have taken and for what reasons. What is your opinion about deleting the few months of editing (the subject of the edit war) and the wave paragraph? DParlevliet (talk) 21:47, 10 February 2014 (UTC)
 * Maybe I'm missing something, but I think 's position was very clear, in that it was taken directly from Wikipedia policy. You need proper, WP:RS secondary sources for your assertions. I think the "regrouping" is to find a clear way forward, which is to rely on Wikipedia policy instead of discussing the merits of the case. Find a secondary source that makes the points you want to make and the discussion can continue, otherwise it's WP:OR. (P.S. I think it's a bit unfair to call out RockMagnetist like this. Let arguments stand on their own merits, it's not about taking sides). 0x0077BE (talk) 22:03, 10 February 2014 (UTC)


 * I am here as a mediator; I know next to nothing about delayed choice quantum erasers and don't have the time to learn right now. My position is that good Wikipedia editors should be improving Wikipedia and not arguing endlessly on talk pages; the policies that I am quoting are designed to facilitate this. As for deleting the last few months of your work - it doesn't really do any harm because it can always be restored if it satisfies Wikipedia policy. Near the bottom of Talk:Delayed choice quantum eraser, I asked the other editors what part of your work should be reinstated. I think the answer by is a good starting point for considering that. RockMagnetist (talk) 22:36, 10 February 2014 (UTC)


 * @DParlevliet. Relax. "Regroup" is what a military squad does when it has gotten separated and squad members may have lost sight of the original objective. We need to ask ourselves question such as, "What is our target audience?" We may want to think about whether to put all double-slit experiments into one huge article, to have a number of separate articles of equal generality, have a general article that functions as a "finding scope" for other, more specific articles, and so forth. P0M (talk) 22:55, 10 February 2014 (UTC)


 * The regroup and attribution of RockMagnetist is about finding Wiki rules to remove my wave part as solution for "how frustrating it has been for all of you". But deleting months of editing surprisingly does not seem to need any Wiki rule. That is not wrong. Everybody has the right to choose how to participate the discussion. As mentioned the wave paragraph needs more detailed references and explanation, so I will add that. About the edit war, we will see how administrators will decide. DParlevliet (talk) 08:47, 11 February 2014 (UTC)
 * This is not a matter for administrative action. In these discussions, we have built a strong consensus against your recent additions, and Wikipedia is built on consensus. Stigmatella aurantiaca (talk)
 * I agree with Stigmatella aurantiaca here. In fact, I think that it would likely be counter-productive to your position that you run to the admins when everyone else is playing nice and trying to build consensus. One reason no one is citing a policy about deleting months of work is that, as far as I know, there is no such policy. We judge things on quality of work, not quantity, so if you spend months working on original research, don't expect it to be published in Wikipedia just because of how long it took. The discussion about inclusion starts with WP:RS - it's the first hurdle to jump over and apparently it has not yet been satisfied in this case. If you don't have that, you're not going to get a lot of support for your position. 0x0077BE (talk) 15:34, 11 February 2014 (UTC)


 * I agree with 0x0077BE. If I wrote something something dumb but persuasive about some experiment that nobody who read it in the following year challenged and then Brian Greene or anybody else with a clear understanding on the topic came along and tore hell out of it, I could hardly argue that the article had been up for so long that it had some kind of invulnerability. P0M (talk) 15:19, 12 February 2014 (UTC)

Text before the edit war
A delayed choice quantum eraser is a quantum eraser equipment in which the information is erased after detection. A quantum eraser is based on a two-path experiment (like double slit experiment or interferometer) in which the wave of (mostly) a photon is split in two waves, which follow different paths. If both waves combine again at a detector, it will measure an interference pattern. Quantum mechanics states that if it is known which path the photon particle followed, the interference disappears. A Quantum eraser first adds the which-path information, showed the disappearing of the interference, and then erases this information, causing the interference to appear again. In a Delayed choice quantum eraser the which-path information is erased after detection of the photon. This experiment was first proposed by Scully and Drühl and performed by Kim e.a.[1] .[2] The experiment was designed to investigate peculiar consequences of the double slit experiment in quantum mechanics, as well as the consequences of quantum entanglement.

Contents

1 Experiments 1.1 Polarizers at the slits 1.2 Interferometer with 50% mirror 1.3 With entangled photons and particle-path detectors (Kim e.a.) 1.4 With entangled photons and circular polarizers (Walborn e.a.) 2 Discussion 2.1 Details pertaining to retrocausality in the Kim experiment 2.2 The main stumbling block for using retrocausality to communicate information 2.3 Yet there are those who persevere in attempting to communicate retroactively 3 See also 4 References 5 External links

Experiments Polarizers at the slits

A Quantum eraser which often used in education is based on placing two orthogonal polarizers at the slits of a Double slit experiment, causing the interference to disappear. Then after the slits a polarizer is placed at 45º, which erases the polarised information of the waves, showing the interference again. This experiment is described in double slit experiment. With classical waves the result can also be explained.


 * While it is true that the above simple experiment can be described in classical terms, this leads to the false expectation that there may be a classical explanation for more sophisticated eraser experiments. Hence this comment is uncalled-for. The same goes for all remaining classical physics analogies introduced in this section. Stigmatella aurantiaca (talk) 14:37, 9 February 2014 (UTC)

Interferometer with 50% mirror Experiment that shows delayed determination of photon path

This Quantum eraser is the Mach-Zehnder interferometer, where a light beam (yellow) is split by a 50% mirror in two beams (red and blue). Both beams are reflected by a mirror and cross each other. Finally each beam ends in a separate detector.[3]

In the top diagram only red or blue photons are detected, so no interference is measured. In the bottom diagram with a second beam splitter both outgoing beams has both red and blue photons, without possibility to distinguish, now showing interference in both detectors. The description is also valid for single photons.

In the classical wave description the wave of the photon is split in two by the 50% mirror. In the upper diagram each detector sees only one wave, so there is no interference with the other wave. In the lower diagram both waves are mixed, so each detector sees both waves, which interfere. The wave determines the possibility of absorption.

The phase difference is introduced along the two paths because of the different effects of passing through a glass plate, being reflected off its first surface, or passing through the back surface of a semi-silvered beam splitter and being reflected by the back (inner side) of the reflective surface.

The result is that waves pass out of both the top upwards exit, and also the top-right exit. Specifically, waves passing out the top exit interfere destructively, whereas waves passing out the upper right side exit interfere constructively. With entangled photons and particle-path detectors (Kim e.a.) Kim EtAl Quantum Eraser.svg

An experimental setup is as follows.[2] A high intensity laser radiates a double slit (vertical black line in the upper left hand corner of the diagram). After the slits a beta barium borate crystal (BBO) causes spontaneous parametric down conversion (SPDC), which generates now and then two identical entangled photons with 1/2 the frequency of the laser photons, in the "red" or "blue" area of the BBO. These photons are caused to diverge and follow two paths by a Glan-Thompson Prism.

One of the photons, the "signal" photon, goes upwards, through a lens, to the target detector D0.

The other photon, the "idler" photon, goes downwards and is deflected by a prism that sends it along divergent paths, depending on whether it came from the "red" or "blue" BBO area.

Beyond each path a 50% mirror acts as a beam splitter (green blocks), resulting in a 50% chance to pass through and a 50% chance reflecting to detectors D3 or D4. The photons which pass through are reflected by 100% mirrors (gray-green blocks) to the detectors D1 or D2. Where both beams cross a third 50% mirror is placed, which reflect 50% of each beam to the other detector.

Because of this arrangement:

If the photon is recorded at detector D3, it can only be a "blue" photon. If the photon is recorded at detector D4, it can only be a "red" photon.

If the photon is recorded at detector D1 or D2, it has a 50% chance to be a "blue" photon and 50% chance to be a "red" one. So when detected by D1 or D2, one cannot know through which slit the photon has traveled. It is said that the red/blue information has been "erased".

Observed results

A coincidence counter selects from the target detector D0 only the events which coincide with the other detectors. This includes a delay of 8 ns to compensate for the 2.5 meter longer path to the other detectors.

When events were counted which coincided with D3 or D4, there was no interference. When events were counted which coincided with D1 or D2, there was an interference pattern.

This is in agreement with the statement in Quantum mechanics that when the path is known, the interference pattern disappears. But most remarkable is that the photons in D0 seem to know that they will not form an interference pattern 8 ns before the path information is detected by D3 or D4. This seems to suggest that the probability of photon detection depends on a future event.

With classical wave physics the absence of interference with D3/D4 can be explained, because when an photon is emitted in the red area of the BBO, there will be no wave in the blue area (and vice versa), so no two waves to interfere. However the appearance of the interference with D1/D2 cannot be explained classical.

Some have interpreted this result to mean that the delayed choice to observe or not observe the path of the idler photon will change the outcome of an event in the past. However, an interference pattern may only be observed after the idlers have been detected (i.e., at D1 or D2).

Note that the total pattern of all signal photons at D0, whose entangled idlers went to multiple different detectors, will never show interference regardless of what happens to the idler photons.[4] One can get an idea of how this works by looking carefully at both the graph of the subset of signal photons whose idlers went to detector D1 (fig. 3 in the paper[2]), and the graph of the subset of signal photons whose idlers went to detector D2 (fig. 4), and observing that the peaks of the first interference pattern line up with the troughs of the second and vice versa (noted in the paper as "a π phase shift between the two interference fringes"), so that the sum of the two will not show interference. With entangled photons and circular polarizers (Walborn e.a.) Quantum eraser Walborn.gif

In this Delayed quantum eraser a beta barium borate (BBO) crystal, radiated by a strong laser, will generate now and then two entangled photons.[5] One photon (yellow path) goes through a double slit to signal detector Ds, with circular polarizer Q1 or Q2 in each path. They are rotated 90° to each other, producing a circular polarization in opposite directions. The other photon (green path) goes to detector Dp, with a linear polarizer cube POL in the path. Only photons in Ds are registered which coincide with photons in Dp. Dp is situated closer to the BBO then Ds, so photons are first detected by Dp. The results were:

Without Q1/Q2 and POL there was an interference pattern in Ds   With Q1/Q2 there was no interference. With Q1/Q2 and POL, adjusted on the Q1 fast axis, there was interference. With Q1/Q2 and POL, adjusted on the Q2 fast axis, there was interference, but 180º shifted with 3. 1-4 with Dp on a larger distance then Ds gives the same result. This is a delayed eraser, because the photon in Dp is detected later then in Ds

The explanation with quantum mechanics is that Q1/Q2 marks the red path and the blue path beams. So one could detect which path the photon went, also when this measurement is not actually done (as in 2). This removes the interference pattern. When placing POL at a suitable angle, the polarization information is erased, so it is no longer possible to know by which path the photon travelled. Therefore the interference pattern is restored. Polarization (green) in both slits in case 3, seen from the detector

In the classical wave description Q1 and Q2 are quarter-wave plates which have orthogonal a "fast" and "slow" axis. The slow axis has 90° phase delay to the fast axis. The axis of Q1 and Q2 are orthogonal (see figure). In 3 and 4 the POL forces the idler wave in a certain direction. Because of entanglement the signal wave will be forced at 90° of the idler wave.[5] In 3 the POL forces the signal wave polarization to be parallel to the fast axis of Q2. Then the output polarisation does not rotate, but is linear, with Q1 having -90° phase shift (slow axis) with Q2.[6] In 4 Q1 is +90° phase shift with Q2. Therefore 3 and 4 gives interference patterns which are 180° shifted, as has been measured. In 2 the incoming polarization can be resolve in two polarizations on the F and S axis, which each give the same result as 3 en 4. So 2 is not an absence of interference, but the sum of two interference patterns which are shifted 180°. With waves the crucial moments are not the detection, but the polarisers. In there is no delayed choice, but the property of entangled photons that if POL changes the polarization of the idler photon. The signal photon will follow directly, also in front of Q1/Q2, although these are situated closer to the BBO then POL.


 * Yikes! I didn't mean copy the whole article. I meant copy the part that you think is not controversial. RockMagnetist (talk) 16:32, 9 February 2014 (UTC)


 * That is all, except the last wave paragraph. Anyway not until begin this month 129.217.159.124 told he did not like the whole article. But also he mentioned only the wave part and some language. DParlevliet (talk) 20:40, 9 February 2014 (UTC)

A question for the others: DParlevliet keeps claiming that some parts of his text was not in dispute and should be left in. Is there any part of his contribution that you consider acceptable? RockMagnetist (talk) 17:00, 9 February 2014 (UTC)


 * In certain ways, DParlevliet's description of Walborn et al.s Quantum eraser experiment is better than the description that is currently in that article. Among other things, his description starts with a better diagram. It doesn't belong here, since that would be merging two articles that most of us don't believe should be merged. However, with proper vetting, it could be used to improve Quantum eraser experiment. The description of the Delayed choice quantum eraser experiment in the current version of this' article could stand a lot of improvement, and with judicious improvement of DParlevliet's English, I think some of his description could be worked in. However,


 * 1) Since this is an inherently quantum mechanical experiment, all references to classical analogies need to be eliminated as being misleading and nonproductive.
 * 2) Certain misconceptions that may have arisen because of DParlevliet's incomplete command of the English language need to be eliminated, for instance the item being discussed in Talk:Delayed_choice_quantum_eraser. Stigmatella aurantiaca (talk) 20:07, 9 February 2014 (UTC)


 * Sounds quite reasonable to me. Would it be better to clean up all of the language first and then decide what to use in addition to or in replace of what is in the article presently in place, or to decide what parts will be useful first and clean them up second? P0M (talk) 15:26, 12 February 2014 (UTC)

How many interference articles and what type
There are three strata of experiments that involve dividing and recombining paths that photons travel with equal probability:

• The Young experiment, which deserves its own article because it is a sort of monument in the history of human thought and theory.

• The fundamental double-slit experiment that Feynmann describes (to paraphrase him) as the crown jewel of quantum mechanics.

• The experiments that go beyond the basic level:


 * ⟡ Experiments that involve electrons, atoms, and molecules


 * ⟡ Experiments that involve entanglement and bring in questions of sequence in time, locality, etc.


 * ̥̊ erasure
 * ̥̊ pre- and post- event erasure
 * ̥̊ delayed choice erasure

Articles we have now grew up without any sort of "project" supervision. Do we need to impose some structure by having a main article with brief leads into each of the above topics, do we need to have separate articles for all three main strata, or what? P0M (talk) 17:47, 11 February 2014 (UTC)


 * It's awfully hard to plan these sorts of things on Wikipedia. We have the WP:WikiProjects of course, which seek to coordinate and improve broad groups of related articles. Our little border skirmish, for example, has shown up on the Wikipedia talk:WikiProject Physics radar, and we can thank RockMagnetist for bringing it to the broader Physics community's attention.


 * Every once in a while, an ad hoc team of editors find that they work well together and work to improve a group of related articles. I was involved in one such ad hoc team. User:D.H and I made major improvements to nearly a dozen articles on the experimental basis of special relativity. D.H was the working experimental physicist with expert knowledge of the history of special relativity, and I was the native English speaker who tidied up his English and added roughly two dozen illustrations to complement his work.


 * If you want to work on refactoring a group of double-slit-related articles, I can see what I can do to help. Be forewarned, however, that after this coming Monday, I won't be able to spend anywhere near as much time on Wikipedia as I've been doing these last few days. Stigmatella aurantiaca (talk) 20:18, 11 February 2014 (UTC)


 * There appears to be nobody else even interested in nixing this proposal. If we don't have consensus on changing this argue to fit into one of the proposed articles (in line with changes in and/or creation of other articles)that will mean that this proposal cannot become part of the way forward.


 * Is the decision of our group to leave the spotted pattern of articles as is? P0M (talk) 14:55, 12 February 2014 (UTC)


 * It seems a reasonable proposal, and like I said, if you begin refactoring the articles, I'll help as much as I can. But I think getting the current individual articles (Double-slit experiment, Quantum erasure, Wheeler's delayed choice experiment, Delayed choice quantum erasure) in good shape has priority over refactoring. Getting the current set of articles in good shape on an individual basis will help us decide how we want to restructure the group of articles as a whole. It's good to have a long term goal, but we have lots to do in the short term first. Stigmatella aurantiaca (talk) 17:22, 12 February 2014 (UTC)


 * That plan is acceptable for me.

Post clean-up details, etc.

 * I notice that you've started working on Quantum eraser. Don't forget to look over what DParlevliet had to say. He had a nice take on various aspects of the subject. Stigmatella aurantiaca (talk) 03:35, 16 February 2014 (UTC)


 * If you mean Quantum eraser experiment, I can't find his changes. They are not in the history to that article. When he was shifting things around I noticed that this experiment and the rather crude diagrams I made for it seemed not to have been moved over, as though he had no thought of including them.


 * Do you mean the interferometer-based experiment instead? Are you thinking of this block of explanatory text?


 * "This Quantum eraser is the Mach-Zehnder interferometer, where a light beam (yellow) is split by a 50% mirror in two beams (red and blue). Both beams are reflected by a mirror and cross each other. Finally each beam ends in a separate detector."


 * "In the top diagram only red or blue photons are detected, so no interference is measured. In the bottom diagram with a second beam splitter both outgoing beams has both red and blue photons, without possibility to distinguish, now showing interference in both detectors. The description is also valid for single photons."


 * "In the classical wave description the wave of the photon is split in two by the 50% mirror. In the upper diagram each detector sees only one wave, so there is no interference with the other wave. In the lower diagram both waves are mixed, so each detector sees both waves, which interfere. The wave determines the possibility of absorption."


 * "The phase difference is introduced along the two paths because of the different effects of passing through a glass plate, being reflected off its first surface, or passing through the back surface of a semi-silvered beam splitter and being reflected by the back (inner side) of the reflective surface."


 * "The result is that waves pass out of both the top upwards exit, and also the top-right exit. Specifically, waves passing out the top exit interfere destructively, whereas waves passing out the upper right side exit interfere constructively."


 * Thanks.P0M (talk) 05:45, 16 February 2014 (UTC)


 * Yes. The Mach-Zehnder interferometer described is basically just the same as the "simple quantum eraser" discussed in this article, but with a different emphasis. What I'm thinking of is a "sauce for the goose, gravy for the gander" situation. The "simple quantum eraser" serves equally well as an introduction to both the Quantum Eraser article and the Delayed Choice Quantum Eraser article, but you don't want to use the same language. What is the best way to avoid using the same language? Start with a different source!
 * I believe that OrbitTheSun, who wrote the nearly undecipherable "Explanation by physical optics" section, may be the same person as DParlevliet, since both have good understanding of the quantum eraser and DCQE experiments and both mangle the English language in a suspiciously similar fashion. If you can successfully translate DParlevliet's description above from Dutch to English, eliminating references to the classical interpretation, then, if you combine it with certain aspects of "Explanation by physical optics", you'll have an independently based description of the simple quantum eraser. The key aspect of "Explanation by physical optics" that is worth considering is an explanation of why the two interference patterns are out of phase by π or 180°. If we assume a 50/50 beam splitter with no losses, conservation of energy demands that the combined probabilities add up to 1 and that the reflected and transmitted beams be out of phase by 180°. (In particular, if the beam splitter is a front surface beam splitter with a dielectric coat, the reflected wave will be out of phase by 180°.) Stigmatella aurantiaca (talk) 08:46, 16 February 2014 (UTC)


 * Wheeler is a really interesting guy. I found a scan of one of this articles. He is much more agnostic about "what is really going on" than are the people who write about his thought experiments. The interesting thing I've been getting in the last couple of days is how valuable these "wacky" experiments have been for making people back down from overly fast-setting opinions.


 * Sadly, most of the things Wheeler wrote are under copyright protection. I don't feel like spending $30 for a twelve-page paper. I guess I wouldn't mind so much if I thought any of these authors get a penny on the dollar.


 * I ran across some things I did on Wheeler that have long since been chewed out of the Wheeler article on Wikipedia but have been retained on some odd-ball website. (Sorry, guys, I guess I should say "non-establishment website.") I think I have a fundamental difference of opinion with Wikipedia. I don't know whether it is "the administration" or whether it is just the preponderant weight of all the would-be experts who write on Wikipedia. Sometimes more is easier to absorb than less. I used both Sears's Optics and the official textbook that we called "Sears and Zemansky." The latter was the compressed version of several MIT textbooks by Sears. It was a big, fat book, and expensive, but it was still smaller than the original four or five volume set. On the other hand I frequently couldn't tell what the short version was trying to say, so I would go look it up in the un-compressed version. Then it was wonderfully clear without sounding pedantic.


 * When I read One, Two, Three... Infinity I had no idea of who George Gamow was. I knew how to pronounce his last name because Willie Ley wrote about such things in Galaxy. But it was clear to me that he could write about deep subjects without losing his audience of dunderheads like me. So, too, could Einstein. Bohr and Heisenberg aren't bad at all. Schrödinger, however, is egocentric, sarcastic, and willing to put both colleagues and readers down. The result is I often don't know how I'm supposed to take some of the things he puts in print.


 * Sometimes I think it may be better to write something that is obviously wrong, rather than to write something that is defensible once the guy explains (for me a five year old) what he was trying to say. That kind of stuff can do anything from mildly irritating a reader to messing somebody up for an extended period of time. Come to think of it, I once recorded a lecture on Dante by a well-regarded professor, thinking I might turn the recorded lecture into something useful for foreign students needing to practice English listening skills before arriving in the U.S. I thought it was a great talk. Then I took it home and tried to transcribe it. I could get the words down on the paper, but I was astonished to discover that for the most part I really couldn't understand what he was trying to say. Sentences melted into nothing but the impression of profundity.


 * Maybe I'll find something in Wheeler that I can at least borrow the structure or the flow from.P0M (talk) 10:20, 16 February 2014 (UTC)


 * P.S. The treatments of the interferometer experiment are indeed basically the same. I think that was due to the way I handled the articles at an early stage. When I was going to write about the Kim article I think I didn't have any easy way to lead the reader in, so I sketched in the simpler experiment. Later on maybe somebody started an article on the simpler experiment. I'm sure I worked on that one too. The Wheeler experiment articles have to deal with it as well. It would have been better to have done the simplest one first, but that was five or ten years ago... P0M (talk) 10:25, 16 February 2014 (UTC)


 * I'm fortunate enough to live near a medium-sized university, so if you need an article and don't mind waiting until I make one of my regular trips to the library, I can save you having to pay $30 for a twelve page paper. Stigmatella aurantiaca (talk) 10:40, 16 February 2014 (UTC)


 * Thank you very much.


 * I've discovered that the Wheeler article had gotten whittled down to much less than it was five years ago, and that Wheeler was a more careful thinker than many of the people who have written about his experiments. I'm still working on bringing that article back. I need to pause, outline, and consolidate some things. I wonder what happened to make all of his stuff so inaccessible.P0M (talk) 19:44, 16 February 2014 (UTC)

Working on the lede, introduction and experiment sections
I've begun work on the lede, introduction and experiment sections. The red marker shows the limits of what I've worked on. Please feel free to correct any mistakes and misunderstandings that I display. Also, feel free to take any material below the red marker that I haven't reached yet, rewording it into something comprehensible, and moving it above the red marker. When we've swept through the entire introduction, we can replace it in one go. Stigmatella aurantiaca (talk) 14:19, 13 February 2014 (UTC)


 * For "However, the interference pattern reappears if the which-path information is erased....," I think it would be clearer to talk about effectively cancelling the marking earlier performed. In two experiments we've worked on, one kind of polarization is used to mark photons that have come through one slit and another kind of polarization is used to mark photons that have come through the other slit. Then there is a "yellow" group with one polarization from the right slit, and a "green" group with the opposite polarization from the left slit. However, to do what is unfortunately called erasure, experimenters do not contrive to somehow return photons to whatever state they individually had before entering the polarizers. Instead, they introduce an additional polarization operation into the apparatus with the result that photons that came through the right slit have an equal mixture of polarizations of the red-coded sort and of the blue-coded sort, and photons that came through the left slit have an equal mixture of red-coded and blue-coded individual polarizations. Left-slit red-coded photons find their right-slit red-coded photons, and left-slit blue-coded photons find their right-slit blue-coded photons. Nothing has been erased the way graphite from a pencil is rubbed off a piece of paper. P0M (talk) 16:24, 13 February 2014 (UTC)


 * Go ahead and fix my wording! Most rapid progress will come if we don't argue about how best to express what we want to say, but if we just go ahead and demonstrate what we want to see in the actual text. We'll come to a consensus quickly enough. I've certainly never claimed to be any sort of expert. As you well know, I've only started educating myself on this topic a few days ago. Stigmatella aurantiaca (talk) 16:40, 13 February 2014 (UTC)

BEGIN EDITED PORTION
The protection has been lifted, so I have moved this text to the article. RockMagnetist (talk) 17:16, 15 February 2014 (UTC)


 * How does it look so far? Stigmatella aurantiaca (talk) 17:30, 15 February 2014 (UTC)


 * You have used extraordinary care to make this article complete, clear, and concise.


 * I have found one place where a word must have dropped out, and one meaningless collection of words that I myself wrote. I think I have fixed both places.


 * Thank you for all your help on this article.P0M (talk) 18:43, 15 February 2014 (UTC)


 * You're welcome! It's been an immensely educational experience working on this article.


 * Sorry, I have been down with the flu for the past week. My two cents on the article: It looks significantly better right now. Do we want to "sharpen" the first section of the article? One could briefly mention the basic issues touched by the experiment like complementarity between particle-like (well defined position/path) and wave-like properties (interference) or something like that if it helps the general reader. However, I am afraid as a working physicist I am not the right guy to decide what helps the average reader. Cthugha82 (talk) 14:01, 21 February 2014 (UTC)


 * I've been looking through many other experiments because I'm editing the Wheeler experiments article. One of the main things to turn up is that many researchers say that there is a "quantum wave–particle superposition state of a single photon" that it does not lose while it transits a double-slit or interferometer experimental apparatus. To avoid giving readers unnecessary preconceptions to root out later on, we may need to be especially aware of our own interpretations.


 * I've done some work on the Simple English Wikipedia where the expectation is that readers may have math and science knowledge available to them in their own language and thoughts, so it is not forbidden to throw in "scary maths" to discourage "laymen." Even with that advantage and with using schematic diagrams that would be nixed on English Wikipedia as suitable only for a book or article written for children, it is quite challenging to put things into English that somebody with a fourth-grade reading level (in English) can handle. (When a laser is aimed at a detection screen, a bright spot will show up on the wall or will be picked up on the CCD and can be seen on a TV screen or something like that that is hooked up to it.) Maybe we could avoid "coherent light" in the present lede. People who know what it means will understand that it could be added into the description, and will understand that it is being left out to keep from distracting the average well-informed reader. Offhand, that was the only place I can remember where I might "dumb things down" a bit.


 * Thanks for your comments and help.P0M (talk) 18:25, 21 February 2014 (UTC)

Faked Wrong results graphic
Im am sorry for the strong words, but this graphics is bullshit.

First: The only recognizable raw D0 result was the ambient noise in the laboratory, which exceeded the signal data by ~ factor 2000. See figure 10 in this document, which gives some additional information over the free publication at arxiv. Without using the coincidence circuit, there was nothing useful to detect at all.

Second: R01 and R02 as disclosed did only show three resp. four stripes.

Third: R03 showed no (light) infringement pattern like in this graphics, but just the "center part" of the gaussian distribution, with amplitudes at the left and right end of the x-scale comparable to the outer stripes of R01 and R02.

Fourth: R04 was not disclosed, they just said that it looked similar to R03. --91.42.106.81 (talk) 19:27, 22 February 2014 (UTC)


 * I think that we need to correct captions to show that this is the idealized information. By the way, a similar simulation was published here:


 * http://strangepaths.com/the-quantum-eraser-experiment/2007/03/20/en/


 * That "faked" image doesn't prove anything. On the other hand I don't think it is appropriate to call something a fake unless it pretends to be something real when it is a fabrication. P0M (talk) 00:27, 23 February 2014 (UTC)


 * I see your point regarding the bottom two images. My ability with SVG is limited. Maybe I can borrow a trick I just learned and make those images more ideal. P0M (talk) 01:16, 23 February 2014 (UTC)


 * We thought it to be totally clear that these are simulated images. Apparently not. We have added the word "simulated" even where we thought it obvious that the image represents a simulation. Stigmatella aurantiaca (talk) 04:59, 23 February 2014 (UTC)


 * I left in some crudities such that even if somebody misses the work "simulated", that the image couldn't possibly be mistaken for as attempting to pass itself off as actual data. Stigmatella aurantiaca (talk) 05:09, 23 February 2014 (UTC)


 * R03 and R04 should be single-slit diffraction patterns, and in an idealized experiment, with sufficient data being gathered, the side bands should be visible. In the actual experiment, insufficient data was gathered to visualize the side bands, as evident from the error bars in the actual data. So the presence of sidebands was not a defect in the graphic being critiqued, and the absence of sidebands in the actual experiment does not represent an idealized result. Stigmatella aurantiaca (talk) 06:00, 23 February 2014 (UTC)


 * Thanks for the explanation and the improvements! So it was not a fake but just misundestanding, sorry for that. --91.42.76.39 (talk) 09:53, 23 February 2014 (UTC)


 * And thank you. If you got the wrong impression from what you said it is sure that lots of other people would have gotten mixed up by the same lack of clarity. It isn't good enough that what we say should be defensible. Ideally it would always be clear enough that it wouldn't need defensive shoring up. P0M (talk) 18:43, 23 February 2014 (UTC)

D0 pattern
Is there any solid source which backs the claim that D0 (less ambient noise) will not show interference, even if retro-causality is forbidden ? I.e. someone who derives / simulates that mathematically? The section "Does delayed choice violate causality?" so far is completely based on Brian Greenes book, but Greene gives no solid explanation on this but just asserts that there is no interference.

I think that the D0 simulations presented in the article are based on mathematics which allow retro-causality, effectively creating the D0 pattern just by adding up D01 to D04. So deriving non-retro-causality from that would be circular. --91.42.76.39 (talk) 10:26, 23 February 2014 (UTC)

A non-retro-causal derivation of the non-interference pattern at D0 can be given based on decoherence: The two superposed states of the signal photons heading for D0 are decoherent because of their entanglement with the idler photons in the other path. (Tracing out the idler photon's state will yield decoherence for the signal photon.) Would be great to find some citable source for this, to put the non-retro-causal explanation on more solid ground. Just as reference, the details of course are too technical for the WP article. --91.42.76.39 (talk) 12:17, 23 February 2014 (UTC)


 * We don't need anything so fancy as an argument based on decoherence. The 180° degree phase difference between the interference of R01 and R02 can be derived as a result of the Fresnel equations (although the Fresnel equations article unfortunately doesn't directly address this). Examine in particular beam splitter BSc.
 * Let us assume, initially, that this beam splitter has a dielectric coat, since that is the simplest case.
 * The Fresnel equations for reflection and transmission of a wave at a dielectric imply that there is a phase change for a reflection when a wave reflects off a change from low to high refractive index but not when it reflects off a change from high to low.
 * No phase shift accompanies reflection of the blue beam from BSc, since the medium behind the beam splitter (air) has a lower refractive index than the medium the light is traveling in (glass).
 * A 180° degree phase shift occurs upon reflection of the red beam from the front of BSc, since the medium behind the beam splitter (glass) has a higher refractive index than the medium the light is traveling in (air).
 * Combination of the red and blue beams at D1 and D2 would therefore result in interference fringes that are 180° out of phase at D1 versus D2.
 * Conservation of energy implies that interference fringes that are 180° out of phase will also result if a metallic coat is used, or if a cube beam splitter is used. Reflection off or transmission through these surfaces results in complex-valued phase shifts, but if constructive interference results in increase in intensity in one image, destructive interference must result in decrease in intensity at the corresponding point in the other image.
 * In other words, the additivity of R01 and R02 to form a featureless sum can be considered to be a result of physical optics considerations.
 * Figure 1 in Mach–Zehnder interferometer and Figure 4 in Michelson interferometer may be useful in helping to understand how this principle works.
 * Mach-Zehnder_interferometer and Michelson_interferometer touch on this subject in a non-mathematical manner.
 * Stigmatella aurantiaca (talk) 12:48, 23 February 2014 (UTC)


 * Thanks for the detailed explanation of the phase shift created by the beam splitter! This helps understanding the experiment. But it does not explain why D0 shows no interference before the idler hits the beam splitter, so it does not help for the non-retro argument.


 * Assume the lower arm of the experiment is a light year away, and we reduce the noise so that D0 shows a signal. The non-retro argument in the article (and in Greenes book) relies on this D0 signal to show no interference, long before any beam splitter gets involved and D1 / D2 signals are measured. So we need a derivation of this D0 non-interference pattern which is independend from the beam splitter, to make a solid argument for non-retro. --91.42.76.39 (talk) 13:24, 23 February 2014 (UTC)


 * Please note that the article mentions - and even pictures - the "raw" signal at D0, i.e. the set of mearsurements at D0 which have not (yet) been correlated to D1..D4 measurements. This is what I am talking about: A non-retro-causal explanation of this "raw" D0 signal pattern is missing, needed to remove circularity from the non-retro argument. --91.42.76.39 (talk) 13:55, 23 February 2014 (UTC)


 * D0 receives two overlapping, non-interfering, single-slit diffraction patterns. Each single slit creates a diffraction pattern. The individual diffraction patterns from the two slits are displaced from each other by the width of the slit, and the diffraction patterns from the two slits does not interfere to create fringes because the beams from the two slits are orthogonally polarized.


 * The interference patterns only show up when we select individual detection events at D0 that correlate with detection events at D1 and D2. Until we receive the detection events at D1 and D2, there is no way of knowing which photon at D0 goes where. This would be true even if we could solve the ambient noise problem to achieve 100% efficiency.


 * Explaining this is probably worth a note. Thanks! Stigmatella aurantiaca (talk) 14:06, 23 February 2014 (UTC)


 * Sorry, the beams from the two slits are not orthogonal polarized. It's the signal and idler photons which are orthogonal polarized, one going to D0 and one to D1..D4, as the article correctly explains. --91.42.76.39 (talk) 14:17, 23 February 2014 (UTC)


 * I will have to defer to what Cthugha82 or P0M have to say. My role in this article was mostly as illustrator and copyeditor. Stigmatella aurantiaca (talk) 14:28, 23 February 2014 (UTC)
 * I was thinking about the Walborn, S. P.; et al. (2002) Quantum eraser experiment. Stigmatella aurantiaca (talk) 14:35, 23 February 2014 (UTC)


 * Yeah, that's a simpler setup - erasure is prefered (via POL1) instead of delayed, so they avoid the retro-causality question. --91.42.76.39 (talk) 14:53, 23 February 2014 (UTC)

I think it is possible to see what is going on without getting too hung-up on time differences. Over and over again different physicists who are familiar with the issues around this kind of thing (starting with Bohr if memory serves) say that we have to look at a completed event, and we don't have a completed event before all of the coincidence counters click in.

If nothing relevant was happening in the idler part of the experimental apparatus, if, for instance, experimenters removed the BBO, then there would be a "red path" and a "blue path" from the double-slit diaphragm, there would be no which-path information at D0 so ambient light excluded all that would show up on that detector would be an interference pattern.

The idea of the basic experiment is to see whether it is or is not possible to get which-path information on the signal photons by getting which-path information on the idler photons. The fact that the experiment worked as planned when the path length for the signal photons is at least as long as the path length for the idler photons means that when a signal photon shows up in a position appropriate to an interference pattern then its entangled twin idler photon shows up in a position appropriate for an interference too, and vice-versa. Its the photons that reach detectors 3 and 4 by single paths that can't show interference. That part of the experiment is so matter-of-course that the article does not even bother to draw in detector 4 and does not provide a graph for it. Rhodes draws it in, and his explanation of how the experiment works has been checked out by Dr. Kim, so it's not like there is anything tricky going on with the experiment or its description on that score. However, what is easily anticipated in the idler section also shows up in the signal section. What makes half of the photons, photons that presumably come by two paths from the double slits, not interfere with themselves and refuse to spread themselves across the screen in interference fringes?

People, myself included, want to say, "What causes half the photons in the signal part of the experimental apparatus to refuse to interfere with themselves? For a physical cause to be involved there must be a force exerted by the idler photon or something that the idler photon "sets off" that reaches the signal photon and changes its behavior. All sorts of fuzzy language is used to describe this supposed physical action. The idler photon "tells" the signal photon that it has just entered a path leading to a single-path detector, and the signal photon "decides" not to interfere with itself. Or, there are "guide waves" that accompany each photon and the "guide waves" communicate with each other somehow so that the signal photon "knows" what is going on with the idler photon and "behaves accordingly." That whole business is puzzling enough to leave thoughtful people unsettled, but then the experimenters throw in another wrinkle. They make the signal photon's path so short in comparison with the idler path that the signal photon shows up as a member of an interference fringe or a member of a diffraction pattern before the idler photon can have had time to reach its detector. So then the "ordinary English" explanation has to be that the idler photon somehow predicts what will happen to it and relays that prediction to the signal photon in time for it to manifest itself appropriately to what is going to happen to its entangled twin.

Let's suppose that I have understood the experiment properly. If that is true then what should experimenters expect to see when they look at entangled pairs and sort them by what happens at D1,2,3,4? How will they vet their experiment? (I am assuming that they are doing the full experiment to try to test Wheeler's ideas.)

Get the lab results and do the discrepancy checks. I'm pretty sure somebody would have noticed by now if Kim et al. had messed up the graphs at the end of their article. Have I misstated anything?

The Kim experiment is a case in which we have both experimental results (reported in the form of graphs at the end of the article) and theoretical discussions about those results. In the case of any doubt the first thing to do would be to check for discrepancies. If, for instance, you don't accept that D0 will not (half the time) show interference, then you must find in the evidence no substantial lack of interference-explainable hits, i.e., no sign of diffraction patterns.

Wheeler does not accept the idea of retrocausality. He proposed a sort of reductio ad absurdum with his stellar example. To him it seemed a little silly, perhaps, to imagine time travel over billions of years done at the whim of an experimenter. But other people looked at the experiments and accepted the idea that it happens. Recent work with Bell Inequalities comes down on Wheeler's side. Photons are in a superposition of wave and particle until they show up somewhere and the superposition "collapses."

OP asked, "Is there any solid source which backs the claim that D0 (less ambient noise) will not show interference, even if retro-causality is forbidden? I.e. someone who derives / simulates that mathematically?"

D0 shows interference half the time (over a large number of runs of the experiment). It fails to show interference half the time. That is the experimental result. If I remember correctly, the whole problem with Entanglement that bothered everybody so much was that time does not enter into the equations. Schrödinger's early hope was that the damned thing would go away somehow, that over time entanglement would just dwindle away. But he couldn't make that work. It was bad enough that correlations would show up even (theoretically) over astronomical distances, but worse when it seemed that correlations might show up without regard to time. So in a sense the whole possibility of entanglement came out of math that says that, retro- or not, time doesn't matter. It's easier to see this, perhaps, in terms of consistency. The idler photon and the signal photon are, in some weird sense, the same thing. So it would be problematical, to say the least, if "the same" photon interfered with itself and did not interfere with itself. It may help to observe that when photons are completely free of experimenter-contrived restrictions they will just do their own thing. When the experimenter forces one twin to do something (e.g., show up as not interfering with itself in a certain run of an experiment) the other twin must obey the same restriction.

OP continued: "I think that the D0 simulations presented in the article are based on mathematics which allow retro-causality, effectively creating the D0 pattern just by adding up D01 to D04. So deriving non-retro-causality from that would be circular." The experimenters have done the experiment and they have published their experimental results in the form of graphs. The graphs have been turned into the pattern-growing simulations of four kinds, so you can see a reverse-engineering of the experimental conclusions to what experimenters might have seen in one run of their experiment. D0 as I drew it is just an artistic representation of what you get if you add the graphs at the end of the Kim et al. paper.

OP continued: "A non-retro-causal derivation of the non-interference pattern at D0 can be given based on decoherence: The two superposed states of the signal photons heading for D0 are decoherent because of their entanglement with the idler photons in the other path." Again, it's too bad that people use the idea of "retrocausality" as though it were a given, and not something that Wheeler wanted to challenge and later researchers have given the business. You are correct to say that there are photons that do not interfere with themselves when they arrive at D0. Experimenters sort those out and find two groups consistent with non-interference. But they also sort out two groups of photons that participate in interference patterns. If experimenters graphed all hits on D0 that were matched by hits on the other detectors, they would add an interference pattern, a second interference with its maxima roughly lining up with the minima of the first (resulting in a general blur), and a couple of diffraction patterns. The resulting graph would be about the same in appearance as the graph they give for D3 except that because interference patterns spread photons much farther along the x-axis, you would see a wider hump. If I had known of your question a few days ago I would have taken note of which of the many paper on this subject have worked the whole thing through mathematically. But the simple answer obtained by looking at the experimental results is that when a photon ends up at a detector that only has one path leading to it, then its entangled twin can't interfere with itself.

DO responded to S aurantiaca: "But it does not explain why D0 shows no interference before the idler hits the beam splitter, so it does not help for the non-retro argument." If anybody could really explain what is going on, and do so well enough to shut everybody else up, then we would be in a different world or at least a different historical period as far as quantum mechanics goes.

It is possible to argue, in the case of Wheeler's inter-galactic thought experiment, that something happens in an observatory today that throws outputs of two telescopes (the red path telescope and the blue path telescope I guess you could call them) into de facto D3 and D4 dead end, no-interference possible, situations. The photon that has been coming in from both virtual images of the Twin Quasar sees that it is coming into two telescopes, one aimed at each image. It says, "Oh, stardust! We can't both be here. One of us twins has be to be retroactively aborted a billion years ago. Drat!" Suddenly a billion years of history is undone, and there is only one photon going into one telescope. That's one possibility, and before John Bell nobody could really find anything against it, time travel aside.

It is also possible to argue that a photon is always in a superposition of wave and particle, that photons always go down both paths that lead from a beam splitter (even if it is a black hole half a billion years away), and that when the two photon-splits that proceed from the gigantic beam-splitter in the sky finally reach the astronomical observatory, then they go down into both telescopes. If there is a sheet of photographic paper at the focus of each of the telescopes then there is a fifty-fifty chance of its showing up on one or the other.

In the first explanation there was a problem of knowing what was going on (whatever that means to photons) over a billion years or so. In the second problem there is an analogous problem of the photon's knowing what is going on over a spatial gap. How do the two photon-splits decide which one of them will turn up on a CCD or other device in the observatory? (Does one play the martyr and throw away its life for the sake of its other self? Do they argue over which shall die? Is this where God get out his dice? Who knows?) Probabilities work themselves out, so if you watched a large number of photons show up you would find half in one telescope's records and half in the other.

So what if the astronomers merge the output of their telescopes on a single CCD? According to the first argument, if a photon had originally "decided" to manifest itself as having traveled all that way as a single particle, and then it saw that it was actually in a double-path situation, it would have to call Mayday and somehow have history rearranged so that for the last billion years it had been travelling as a divided wave that had gone around both sides of the black hole and now would merge with itself on the detection screen. According to the second argument it had always been travelling by two paths and, coming into a real-world situation that created interference between those parts-in-superposition, it would manifest itself by interfering with itself and contributing to an interference fringe.

One of my professors, a prominent neurophysiologist, said that scientific research was literally re- search. You do an experiment or you read about somebody else's experiment. Then you ask yourself, "Did that really happen that way? Did we miss something? Did we misinterpret something?" So you do the experiment over again, or you get creative and put in a wrinkle or two in order to resolve some of your own doubts. Over time you get a consistent picture. You never prove anything, but you have eliminated a lot of things by disproving them. What is left will work for you, help you make rocket ships or whatever, until it fails at some point. Then you've at least got to revise your theory. Sometimes you throw away the old one and start something else.

Researchers don't do one run of an experiment and rush to publication. I don't know how many times Kim et al. went into their lab and made readjustments, checked the accuracy of their coincidence counter, etc., etc. I think they almost certainly had good reason to believe their experimental results. (After all, there were several of them and they all know their craft.) On top of that, there have been other experiments designed to investigate the same general question. Any big mistakes would have created turbulence that people would have noticed.

As to the retrocausality business, I think it is a little like the hidden variable business that was supposed to explain how the arrival of a photon so as to contribute to one or another fringe in a double-slit experiment was not a matter of pure probability, was not "God rolling dice," but was a determinate factor of the photon that decreed early on exactly which fringe it would show up in. Bell came along with his inequality calculations and proved that the hidden variable idea would not work.

Similar calculations have been used to show that the photon leaving a laser or the photon leaving a quasar cannot be pre-determined to participate in an interference or a non-interference situation at the time it shows up one somebody's detection screen. There are two papers that attempt to establish this proof against retrocausality. They are the first attempts I've seen to make a rational choice between ways of accounting for the phenomena investigated in the Kim experiment. P0M (talk) 18:40, 23 February 2014 (UTC)

P.S. I forgot that the references to the Bell Inequality results are in another article.

See: Peruzzo, et al., "A quantum delayed choice experiment," arXiv:1205.4926v2 [quant-ph] 28 Jun 2012. This experiment uses Bell inequalities to replace the delayed choice devices, but it achieves the same experimental purpose in an elegant and convincing way.

As one experimenter explains, "Wave and particle behavior can coexist simultaneously." "Entanglement-enabled delayed choice experiment." by Florian Kaiser, Thomas Coudreau, Perola Milman, Daniel B. Ostrowsky, and Sébastien Tanzilli, in arXiv:1206.4348v1 P0M (talk) 19:03, 23 February 2014 (UTC)


 * After mulling about all this, I come to the conclusion that there are different definitions of causality, and this is the source of all the confusion about it.


 * In the scientific discussion on quantum physics, the relativistic definition of causality has been widely adopted, which says: causality (1) = transfer of information from spacetime coordinate A to spacetime coordinate B. Or in QM terms: influencing the statistical outcome of measurement B by measuring A.


 * Another (common-sense?) definition is: causality (2) = the result of a single measurement A depends on the result of a single measurement B.


 * So I think that this issue may be solved - with great benefit to the article - by explaining and discussing both of this understandings of causality in the light of the delayed choice experiment. Greene may help for (1), while (2) will rather be in the realm of what you called "ordinary English" above (I call it "classical world including the Arrow of time"). Note that a conclusive explanation of (2) needs an interpretation of QM, so one should be aware which interpetation is refered to. --91.42.95.172 (talk) 15:33, 24 February 2014 (UTC)


 * We can discuss the experiment of Kim et al. in this article. It is already a big problem just to lead people through the basic steps involved. Discussions of causality belong in a separate article. Discussions of quantum entanglement belong in a third article. Interpretations of quantum mechanics need another article.


 * Einstein was highly dissatisfied with quantum mechanics because it indicated that "spooky action at a distance" was one of its inevitable consequences. Retrocausality would have made him flip his wig. Prof. Cramer embraces the idea to the extent that he has proposed experiments to establish communication devices that are instantaneous or that work backwards in time. People argued fruitlessly about hidden variables and there are still some holdouts after the Bell inequalities were discovered. We can't pontificate about what the major thinkers of the last 100+ years can't agree upon.P0M (talk) 17:07, 24 February 2014 (UTC)


 * My intention was not to extend the topic of the article, but to fix the already existing section "Implications" which copies the flawed logics of external sources. The flaw consists of answering the reader's natural question "Does this experiment implicate a violation of causality (2)" by "No, it's consensus that it does not violate causality (1)", without cleanly distinguishing between (1) and (2).


 * But I must admit that this misleading reasoning seems to be mainstream, so the WP article may need to stick to it. --91.42.95.172 (talk) 19:52, 24 February 2014 (UTC)


 * We have to be careful to avoid WP:OR. Much of what was discarded from earlier versions of this article in this latest major rewrite was original research. See if you can find WP:RS in support of your reasoning. Stigmatella aurantiaca (talk) 20:08, 24 February 2014 (UTC)


 * I agree. On top of the need to avoid original research, this topic also proved difficult to sort out because just about everything that is said ends up being said in the context of a particular interpretation. Nobody wins in the contest among the many interpretations. P0M (talk) 23:19, 24 February 2014 (UTC)