Talk:Wheeler's delayed-choice experiment/Archive 1

Removing a section
I propose removing the entire section titled: "Working out implementation of the experiment," for the reasons given below. Alternately the section should be rewritten and significantly shortened; it is entirely too long.

The problem it the section seems to have numerous fundamental errors. In a number of places it refers to determining whether an interference pattern exists from a single photon. This is both impossible and meaningless. You need to examine a collection of photons to determine an interference pattern exists. Another fundamental error is that it talks about detecting photons from a double slit experiment at a specific locations, or expecting specific trajectories for the photons. For double slits, this is false for either interfering particles, or for which-way particles. The confusion might be for the interferometer experiments, in which the paths are always end up in one of two final paths. Interferometer and double slit are different experiments.

One line encompasses both errors; "Since only one photon is to be examined at any one time, the interference effect must be determined by noting the trajectory of the single photon." Other errors include talk about the binocular telescope observing one photon (impossible), and the assertion given in the double slit diagram in the section is false (all detectors will detect some photons if they are behaving like particles/bullets).

I propose removing the entire section. --Lostart (talk) 04:29, 29 December 2007 (UTC)


 * I now have a copy of Wheeler's article on which much of the discussion of his proposed experiment is presumably based. It does not have the business about gravitational lensing at all. But it may be that it will be possible to get a clearer explanation of what he actually proposed.


 * Many of the delayed choice experiments involve the use of single photons or single pairs of entangled photons. What many of them do is to put a narrow aperture detector on a movable base that is drawn across the scope of a regular detector screen. The number of hits at each position is counted up, so a picture of interference fringes is built up over time. If interference does not occur then the hits occur in the central range where a diffraction pattern might be expected. The reason that they do this is that they want to be able to associate, temporally, the arrival of photons with other, changeable parts of the experiment. P0M (talk) 05:40, 29 December 2007 (UTC)


 * See "Experimental realization of Wheeler’s delayed-choice

GedankenExperiment," V. Jacques1 et al. (Charts at the end.) I think the gravitational lensing, use of "binocular telescopes," etc. is an improbable experiment as it has been described. On the other hand what I have seen of Wheeler's writing indicates that he would not be the one to suspect of making any mistakes. P0M (talk) 05:55, 29 December 2007 (UTC)

The method you describe above is correct, but the way it's described in the section is wrong, and there are other errors in the section, as I mentioned. I will go ahead and remove the section until it can be redone without errors.--Lostart (talk) 17:30, 30 December 2007 (UTC)

Change of mind--rereading the section, I realize the errors are more about descriptions of things; I think the editing of the section in place is doable.--Lostart (talk) 18:03, 30 December 2007 (UTC)

Section still needs to be reduced, as does the following section, "Wheeler's own discussion of these problems," and written more clearly, concise, and in the style and tone of a Wikipedia article. It should only have the minimum necessary to explain the experiment and it's major points. Each section, and each paragraph should be clear on it's own; it shouldn't be written like a narrative.

The diagram "Image:Wheeler experiment problems.svg" is confusing to me. The detectors should be shaped like the tubes, or telescopes. BTW, I found the use of 'binocular' confusing, and not a good analogy. A binocular is a single unit, with both sights focused in the same (parallel) direction; neither of which is true for what you are describing.--Lostart (talk) 20:26, 30 December 2007 (UTC)


 * I agree. I would never have used the expression myself. I have been trying to get a clear idea of what Wheeler himself said. The original article was a mess IMHO. But I have tried to work through it step by step and work out what he may actually have been talking about. I am now pretty sure that his point was actually that whatever is observed here on Earth depends on which observation is made. Period. He is not asserting some kind of backwards-in-time determination.  I think the following is probably wrong:
 * "The idea behind the Wheeler experiment is that it might be possible to determine which side of a double-slit experiment a photon traveled through without destroying the interference pattern that occurs when the two versions of its probability wave interact on the detector screen of the typical double-slit experiment."
 * Wheeler says very clearly that one cannot assert anything about a "photon" in between measurement/observations. The idea in the original Wikipedia article, and, I think, in some of the secondary stuff on-line, is that if you replace a detection screen with a couple of tightly focused telescopes then, magically, all the energy (all the photons) that were originally going to be spread out among the interference fringes are suddenly going to funnel down one or the other telescopes. If that were true then one ought to be able to set up a laser, a double slit, and a detection and aim the whole thing out toward the "empty" sky. Then you would presumably get some interesting results when you blocked each of the slits in turn and placed a single narrow piece of paper at the center of each of the diffraction patterns. When both slits were open again, all the light would presumably strike these two strips of paper. Or, if you wanted to really duplicate the Wheeler experiment in miniature, you could put soda straws there so that the light would go down the soda straws and hit photographic emulsion or something at the far end. I haven't tried the soda straws, but just to test my intuition (and in the forlorn hope of coming up with a killer experiment) I cut the center out of my detection screen and let the "real photons" go out the back end of the apparatus, just as they would have gone into two mouse telescopes. Unfortunately for my plans, the rest of the interference patterns were fine, and what was missing appeared in the distance on the side of my refrigerator.  I am pretty sure that if I lined up a bunch of little telescopes all the way across they would just pick up their share of the interference pattern.  As long as wave fronts (from the same emission) are in the same physical space they are going to interfere.  Unfortunately I can't write "personal research" into the article. That's why it is important to find what Wheeler said. I've asked Dr. John Cramer who says he's seen the astronomical version somewhere but can't give me a definite citation. So it's probably out there somewhere.


 * Thanks for helping with this article. You may have more time than I have at the moment, and maybe you have access to some physics professor who has this information at his/her fingertips. P0M (talk) 00:14, 31 December 2007 (UTC)


 * I fixed the diagram (telescopes) and cut one paragraph. P0M (talk) 02:01, 31 December 2007 (UTC)

Major change
I found a quotation that made it clear what Wheeler intended in his experiment. I am afraid that he did not work out the details of his thought experiment, but maybe I am wrong. Anyway, I have tried to make the substitute "telescope" diagram in accord with what he claimed. The problem that I see with what he said should be apparent from the diagram. He would need to put a diverging prism near the double slits to make the telescopes get the kind of image conditions he wants, but the same prism would destroy any chance of interference. That's what I think, but who cares. The point is to describe the experiment as he wanted it done. If this new part looks o.k. it should take a lot of slack out of the rest of the article and I can begin to tighten it up. P0M (talk) 06:43, 31 December 2007 (UTC)

The astronomical part of it has the opposite problem. The beams are already diverged and would have to be brought together in the observatory. If you brought them together you would then get interference (assuming that the path lengths were equal), but you could not pick out where any given photon came from. If you wanted to tell which beam the photon came from you would just look through one telescope or the other -- but while you were doing that you couldn't get any interference. There is nothing very profound about any of this stuff in itself. "You see what you want to see, and you hear what you want to hear," as an old popular song said. But there may be a catch in all of this. Who is to say that the photon does not indeed travel in both paths and then "decide" which telescope to show up in on the basis of "chance?"

I don't want to write my conclusions into the article, but I do want to ferret out my own mistakes and misconceptions that might cause me to let something in or keep something out on false grounds. So thinking clearly abut what (IMHO) isn't very clearly written about anywhere will probably be helpful in selecting what things to quote and/or to emphasize. P0M (talk) 06:59, 31 December 2007 (UTC)

Very Difficult Run-On Sentence:
The fundamental lesson that Wheeler enunciates in accord with these experiments is that whether phenomena are observed that indicate that light has a wave nature or phenomena are observed that indicate that light has a particle nature depends on whether the experiment is set up to detect waves or particles.

How about:

The double-slit experiment is meant to observe phenomena that indicate whether light has a particle nature or a wave nature. The fundamental lesson of Wheeler's delayed choice experiment that Wheeler enunciates in accord with these experiments is that the result whether phenomena are observed that indicate that light has a wave nature or phenomena are observed that indicate that light has a particle nature  is that the result  depends on whether the experiment is set up to detect waves or particles. Originalname37 (talk) 18:55, 3 January 2008 (UTC)
 * It looks o.k. to me. Thanks. P0M (talk) 19:39, 3 January 2008 (UTC)

The delayed choice experiment as a beam splitter interferometer modification
In Wheeler's books "At home in the universe" and "Geons, Black Holes, and Quantum Foam" he describes the delayed choice experiment in terms of a modified beam splitter interferometer. Light from the source is split by one beam splitter and recombined at an observation location by use of mirrors. One can choose either to leave the output beam splitter out, and thus obtain "which way" detection by means of which detector goes off, or insert the output beam splitter, in which case interference will cause the beam to recombine and exclusively excite one detector.

Can anyone confirm what the relationship of this is to the modified double-slit experiment that this page currently describes? Wheeler has a note on p.322 of "At Home": "'The center of discussion in the Bohr-Einstein dialogue was more often the so-called double-slit experiment than the beam splitter depicted in Figure 4. The latter is made the focus of attention here because it presents the central point without getting into the physics of interference patterns.'" I remember seeing the modified double-slit in an earlier Wheeler book, but can't remember which.

Furthermore, Wheeler explicitly describes the cosmic version in both books mentioned. His proposal is that similar to the normal delayed-choice experiment, the two images of a gravitationally-lensed quasar (he cites 0957 + 561 A,B as a candidate in "At Home") may be detected at two crossed detectors, or a beamsplitter may be inserted between the detectors and interference will cause exclusively one detector to receive the output beam.

What's strange is that none of this seems to have been mentioned in any way, either on Talk or in the main article. Am I missing something? Shernren (talk) 14:06, 28 May 2008 (UTC)


 * A citation to Wheeler's book would be quite useful. Please find it for the article.


 * The article, as I found it, was in pretty bad shape. The problem I see with the experiment is that the lab on earth is dealing with a couple beams of light containing a continuous stream of photons, so the experiment is not anything like one done with single photons. If one could isolate one photon, then it would make sense to ask whether it had come by the left path or the right path. But the essential element of all these experiment is that there is initially one psi wave that has extension, and it divides due to either a double slit or a wave splitter or something of that sort. In the double slit experiment, some "thing" has to become superimposed on the detection screen, and both the left path and the right path are significant because the interference pattern is a function of the distance between slits. So, to me, the gravitational lens serves as a beam splitter. Each photon's psi function has to go by both paths.  If the astronomer merges those psi functions appropriately then they may interfere with themselves. If the astronomer "runs them down separate pipes" so to speak, a photon will show up at the end of one or the other path.  If your books can verify and/or clarify, or even straighten out this interpretation, please provide citations. Maybe I can find the book locally.  P0M (talk) 21:21, 28 May 2008 (UTC)


 * The books are:


 * At Home In The Universe, by John Archibald Wheeler, 1996, Springer-Verlag New York Inc. pp. 118-119 discusses the delayed-choice experiment, pp. 124-127 talk about the cosmic-scaled variant. pp. 288-290 talk about it being a modified double-slit experiment.


 * Geons, Black Holes, and Quantum Foam, by John Archibald Wheeler with Kenneth Ford, 1998, W.W. Norton and Co. New York. pp. 335-338 discuss both the delayed-choice experiment (in terms of an interferometer set up on a baseball diamond) and its cosmic variant, although the language is much more reader-friendly.


 * Your interpretation seems right based on the original Wheeler materials. However, more likely than not the astronomical sources used in the cosmic version would be so faint that the photons will indeed arrive one at a time instead of in a constant beam; Wheeler certainly assumes so in his discussions in both books. Shernren (talk) 06:43, 29 May 2008 (UTC)


 * Interesting. Good point about the single detections of photons. How does one properly conceptualize a wave-function? It's a non-thing, but it "goes." It would probably be worthwhile to work on Loglan so one could talk about these subjects without unintended meanings sneaking in.  I believe it is meaningless to talk about whether "the" photon went by path A or path B.  I'll try to find time to locate the books. Thanks. P0M (talk) 04:26, 31 May 2008 (UTC)

Work in progress
I've discovered that some important points have been lost from this article, little by little, so I am in the process of revising it. Right now I need to outline and reorganize some parts. P0M (talk) 21:24, 16 February 2014 (UTC)

I have now cut the lede paragraph down to an appropriate size and have done about half of the introduction to the two main types of "delayed choice" experiments. P0M (talk) 08:28, 17 February 2014 (UTC)

I've started to combine two sections of closer explanation that pertain to interferometer experiments. P0M (talk) 18:26, 17 February 2014 (UTC)

I've removed parts of the article that need to be reconsidered, leaving what I hope is coherent and correctP0M (talk) 01:45, 18 February 2014 (UTC)

Perhaps no duality at bottom?
Consider another interpretation: the photon is always both a wave and a particle. Namely, it is a so-called "particle" when it interacts with something else. It is a so-called "wave" when considering its wavefunction; and it always passes as a wave either by one slit or two. The fact that interference occurs could be explained by closer examination of what is an "interaction" (e.g. with the screen). Perhaps with one slit the interaction will occur just once, and with two slits it can occur in several places. Note that when assuming a deterministic interpretation like the Bohm's one quantum interactions can be explained by a kind of "potential" (scalar field) which is dependent on the wavefunction(s) (probability density); perhaps the interaction occurs deterministically if a certain "critical level" is exceeded, realistically thinking. Thus one might assume that due to the presence of two slits these critical levels exist in multiple places, and with just one slit the interaction is centered in one place; but this might be a phenomenon well proceeding from a deeper theory (and maybe even intermediate states, like centered interactions but not single, would be possible) ... Certainly there is an interaction, even though in multiple places, with e.g. electrons, so perhaps it is better to say about interacting with just one or with several particles at a time (and a certain Feynman diagram might perhaps be constructed also for the latter). 178.42.114.180 (talk) 00:48, 18 October 2014 (UTC)