Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Afshar 2 slit experiment-peer review

  1. Feb 3, 2005 #1
    Afshar 2 slit experiment--peer review

    I have looked all over the web--I still have not seen any peer review of this experiment. According to Afshar, it should have been done 3-4 months ago.

    Anyone know anything about this?
  2. jcsd
  3. Feb 3, 2005 #2

    Hans de Vries

    User Avatar
    Science Advisor
    Gold Member

    It still wasn't done at Jan 13th according to his blog:

    http://irims.org/blog/index.php/2004/09/25/questions_welcome [Broken]

    There is a thread on this site where you can find what we think
    is the likely explanation of his experiment (at the end):


    Regards, Hans
    Last edited by a moderator: May 1, 2017
  4. Feb 4, 2005 #3
    Thanks--will read the thread
  5. Feb 5, 2005 #4
    I had a question about Afshar's experiment : a lens is a momentum analyzer, it is a device that serves to measure impulsion. Indeed, the refraction indice depends on the wavelength. Therefore, the photon entering the lens can be in a well-defined position state, but the measurement of its impulsion must put him into a well-defined impulsion state. Impulsion and position states belong to incompatible basis according to Heisenberg. Any information about position of the photon must be erased by this measurement of impulsion.

    Is this argument against Afshar's experiment incorrect ?
  6. Feb 6, 2005 #5
    It has been a long time since I studied Fourier optics,so looked up a book---it says that that given a field distribution f(x,y) on one plane,you can find the distribution on another parallel to it at a distance d by a convolving f(x,y) with exp(-ia(x^2+y^2) where a =k/2d.The effect of a lens is to multiply the incident phase distribution by exp(ik(x^2+y^2)/f),where f is the focal length.Using the above one can easily show that given some field distribution g(x,y)in the front focal plane of the lens,the distribution on the back focal plane is simply the Fourier tranform of g(x,y).Now in Afshar's experiment,neither the wire grid nor the slits are at a focal distance from the lens.Besides what you obtain on the image plane is not simply the image of the wire grid----you have to distinguish between the case when one slit is open from the case when both are open.So one could follow the following approach:-a single slit gives a wavefunction of the kind \psi(x)=constt. over the slit width,0 elsewhere.Fourier transforming this you get \psi(p)=constt. sin(ap)/p(where a=slit width) in the momentum space.Now the wire grid and lens are some kind of momentum filters---simple multiplication in x,y domain now gets replaced by convolutions in momentum domain.One should work out the whole thing in this manner for the cases of one slit open and both open.I haven't done this but I guess what Afshar gets from simple ray optics should come out---but his conclusions are contestable.
  7. Feb 26, 2005 #6


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Not to drag this any further, but it might be useful to add to the Afshar saga by pointing out that Kastner has uploaded his take on this:


    This thing sure has a lot of "baggage" already for something that hasn't officially appear in print.

  8. Feb 26, 2005 #7


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Afshar was in my department last week... but I couldn't make it to his talk. :frown:
  9. Feb 27, 2005 #8


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    You mean this guy actually exists, and not some illusionary particle like the Higgs? :)

    Oh wait, no. The "illusionary particle" is his manuscript, the way it is going. :)

  10. Feb 27, 2005 #9


    User Avatar
    Science Advisor
    Gold Member

    Note: he wasn't ACTUALLY observed...

    If his comments on Wikipedia are any indication, I think he is awaiting his Nobel as we speak. Countering another contributor, he wrote:

    "As for my arguments showing a violation of Complementarity being self-contradictory, you may benefit from a short course on the history of physics, which will clarify to you that such "paradoxes" have invariably led to new chapters in physics, e.g. Schrodinger's cat, EPR paradox, etc. "

    He also decided to re-write the Double Slit section to include himself in as a key source LOL. I removed his self-references, as I sometimes do others who write themselves into the history of the science of quantum mechanics before even being published.
  11. Feb 27, 2005 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Oh my! This is worse than the Podkletnov effect!

  12. Feb 28, 2005 #11

    Hans de Vries

    User Avatar
    Science Advisor
    Gold Member

    Kastner's paper is more like a promotion for Cramer's Transactional
    Interpretation. Sort of taking advantage of Afshar's suggestion that his
    experiment is compatible with this interpretation but avoiding the heat of
    Afshar's complementarity principle violation claim.

    Kastner's interpretation is now placed in Afshar's Wikipedia page. I think it should
    rather be placed in a similar Transactional Interpretation Wikipedia "promotion

    John G Cramer himself on proving TI with Afshar's experiment:

    http://www.analogsf.com/0409/altview2.shtml [Broken]

    It seems that the claims that Copenhagen and MWI are disproved and TI
    is confirmed came originally from Cramer instead of Afshar.

    Regards, Hans
    Last edited by a moderator: May 1, 2017
  13. Mar 6, 2005 #12
    Kastner on Afshar

    To those of you interested in the Afshar experiment:

    If you read the latest version of my paper, at the philsci archive (pitt.edu) or at arxiv.org, I think you will see that I do directly address Afshar's claim to have refuted Complementarity (indeed the title of my revised version indicates that I think he hasn't).

    I am just using TI as a tool to provide physical insight into what is going on in the Afshar experiment. I don't see Afshar as refuting Bohr or necessarily providing evidence for TI over other interpretations. So, while I am obviously sympathetic to TI and think it deserves serious consideration, my paper on Afshar is primarily about critically analyzing his claims, not about promoting TI.

    Ruth Kastner
  14. Mar 7, 2005 #13

    Looking at the single photon at a time case, all the wire grid does is confirm that the superposition is in place before being modified by the lenses. It is a null measurement that actually confirms the existence of the superposition without an interaction.

    In this is the case there is an effective energy/momentum/spin distribution to the superposition that will disallow which-way information, since this distribution can produce probabilities for either detector, since as far as I can see the detectors, not the lenses, are what collapses the superposition.

    Last edited: Mar 7, 2005
  15. Mar 7, 2005 #14
    to juju

    That's more or less the conclusion reached in my paper...

  16. Mar 7, 2005 #15


    User Avatar
    Science Advisor
    Gold Member

    Welcome to PhysicsForums, Ruth. That is an interesting paper you have written, I will definitely look at it! I like the perspective you have taken: what would Bohr say?

    One of the mistakes some people make in advancing a position is failing to give enough credit to the other side. The Afshar experiment more or less assumes its interpretation as being acceptable without giving too much weight to the opposition view. I doubt that many folks are going to agree that Cramer's Transactional Interpretation (TI) carries less baggage than the Copenhagen Interpretation (CI).

    According to Cramer: "The problem encountered by the Copenhagen and Many-Worlds Interpretations is that the Afshar Experiment has identified a situation in which these popular interpretations of quantum mechanics are inconsistent with the quantum formalism itself." Yet I don't see where Cramer himself ever specifically proposed such an experimental test prior to Afshar (although perhaps I missed that).

    Personally, I think there is plenty of merit in consideration of Cramer's basic ideas. (If time is symmetric, why not?) You never know where it could lead. But it hasn't really produced anything concrete yet. Ditto with MWI, of course.

    And for reasons I seriously question, the Afshar experiment is closely connected to TI. If the experiment is good, it will stand regardless of the theory. Why confuse the two? It's not like the formalism is likely to change much anyway. In other words, it seems like Afshar has gone out on a limb for little advantage.
    Last edited: Mar 7, 2005
  17. Mar 16, 2005 #16
    On Afshar's response to my paper

    Thanks, Dr. Chinese.

    I notice that Afshar has posted some thoughts in response to my paper. Briefly, I should reiterate that the point of my paper is not to make a perfect analogy (obviously position,x, is a different kind of operator from spin) but only to point out that the essence of the Afshar experiment is to demonstrate the coexistence, in a single experiment, of a good interference measurement (at the grid) with a good "slit-observable" measurement (at the final screen). Some people have disputed that he has even done that; I concur with him that he has in fact done that. The question is: does this violate Complementarity? If it does, then so does the spin experiment in which we combine a good measurement of "spin along x" at a given time with a later, good measurement of "spin along z" at a later time. That is the basic point of my paper. If Afshar's experiment violates Complementarity, then so does this spin pre- and post-selection measurement, because it combines sharp measurements of complementary observables at two different times. However, I don't think people have worried about the latter as a threat to Complementarity. If Afshar is correct about his interpretation of what Complementarity means, then he should consider the spin experiment to also be a violation of Complementarity.

    Afshar has argued that his experiment is not perfectly analogous to the spin experiment because he can introduce an asymmetry in his "which-slit" measurement that is not present in the spin example. This misses the point, because you don't need to combine spin-x and spin-z to make the point I'm making. You can combine, say, spin-x with spin along any other direction, which would also introduce an asymmetry. All you need is any two noncommuting spin measurements.
  18. Mar 16, 2005 #17
    The wires don't diffract the combined fields in the dark fringes.

    But they DO diffract each individual wave coming from from their
    respective sources. This mandates that some scattering by the wires
    take place to the "sides" to complement what is blocked by forward-scattering from the other slit.

    Particularly, there is a non-zero probability that the photon would show up
    at the WRONG detector even with one slit only because the wires would
    scatter it there.

    This means it's a perfect slit-randomizer, not an analyzer. The wires
    destroy your chances of knowing which slit the particle came through.
  19. Mar 16, 2005 #18
    "slit observable" not the same as "which way"

    If you read my paper at arxiv.org, you will see that I address this issue. Specifically, I point out that when you prepare a source in the "both slits open" state, you never really get "which way" information by later doing a "slit observable" measurement, whether or not the grid is in place. A good measurement of the slit basis observable, which Afshar does perform, is not equivalent to the disclosure of a trajectory through a particular slit. Thus the terminology "which-way" is misleading. It is not legitimate to "retrodict" that a photon "went through" a particular slit when it is prepared in a two-slit state.

    Indeed, "preparing a photon in a both-slit state" means that the "photon" is not localized. It should simply be thought of as an excitation of the em field. When you introduce a screen at the end, you force a spatial localization of this field excitation.
  20. Mar 16, 2005 #19
    Yes, Rkastner I agree with you. I meant that when the experiement is performed
    with only one slit, you will still register readings on both detectors BUT in a different
    proportion than when the wires are not there. It is NOT true that the wires do
    not scatter in the two slit case.

    A classical analysis of the reflection of a distant light from a metal sphere can provide some insight into this wire business.

    -Just because you see an image of a distant source reflecting from a single point
    on a shiny sphere it does NOT mean that the entire sphere does not scatter light toward you.
    Indeed, if you "removed" all the dark portions of the surface, the edge of the cut which was dark before will light up via diffraction.

    The wires DO scatter even though they are in a node of the diffraction field.
    They are arranged to scatter as if they weren't there but ONLY when both slits
    are open is this achieved.
    Last edited: Mar 16, 2005
  21. Oct 2, 2007 #20
    It seems to me Afshars experiment demonstrates no more than does Youngs slit experiment. That is to say there is an interference pattern with both slits open & no interference pattern when one slit, or pinhole in Afshar's case, is covered. The light scattering off the wires in Afshar's experiment because there are no dark bands when you uncover one slit, is simply another way of demonstrating what Young had already done. Namely that you need two slits to obtain an interference pattern. The only difference in Young's experiment is that light scatters off the screen where previously there had been dark bands, towards the experimenter's eye. In Afshar's experiment the light scatters off wires instead where there had been dark bands towards a solid state detector.
    The question of how is it that a photon is able to go through both slits or pinholes to interfere with itself, remains the same. Afshar's experiment poses no new questions & answers no questions.
    Surely the only reason some semblance of an interference pattern remains even when one pinhole is covered, is because the wires act like a diffraction grating.
    Last edited: Oct 2, 2007
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook