Recent Entries: Beslan Terrorist "The Colonel" Identified as Khochubarov Ruslan Tagirovich; Shamil Basayev Claims Responisibility for the Attack | Photos of Beslan's Dead Terrorists Published | Marching Morons |

July 22, 2004

Shahriar S. Afshar: Quantum Rebel

10/25/04: See the preprint of Afshar's paper on IRIMS here.

Shahriar S. Afshar: Quantum Rebel

I've got the New Scientist material about Afshar and his experiment, but in awkward formats. And I'm in a huge hurry. (Here are links to all my posts about the Afshar experiment.) So here's the editorial from the July 24th issue of New Scientist (p. 3) along with a sidebar about the experiment from the cover article (p. 34).

Buy the magazine. Read the whole thing.

UPDATE (July 29th): Welcome Metafilter readers! This is only one of a number of posts I've made about Afshar's experiment. The first was Quantum Mechanics: Not Just a Matter of Interpretation on April 26th. This post described the experiment and its implications for various interpretations of quantum mechanics.

It has been widely accepted that the rival interpretations of quantum mechanics, e.g., the Copenhagen Interpretation, the Many-Worlds Interpretation, and my father John Cramer's Transactional Interpretation, cannot be distinguished or falsified by experiment, because the experimental predictions come from the formalism that all such interpretations describe. However, the Afshar Experiment demonstrates in an interaction-free way that there is a loophole in this logic: if the interpretation is inconsistent with the formalism, then it can be falsified. In particular, the Afshar Experiment falsifies the Copenhagen Interpretation, which requires the absence of interference in a particle-type measurement. It also falsifies the Many-Worlds Interpretation which tells us to expect no interference between "worlds" that are physically distinguishable, e.g., that correspond to the photon's measured passage through one pinhole or the other [the word "measured" added 4/28. -KC].

The Transactional Interpretation, on the other hand, has no problem in explaining that Afshar results. "Offer waves" from the source pass through both pinholes and interfere, creating a condition in which no transactions to the wires can form. Therefore, no photons are intercepted by wires, as Afshar observes. The quantum formalism makes the same predictions.

More recently, the post prior to this one gave links to the downloadable edition of the issue of New Scientist. Afshar will appear on NPR's Science Friday tomorrow, July 30th.

A FURTHER UPDATE: Analog says it will have my father Joh Cramer's forthcoming popular science column on the Afshar experiment up on their site by tomorrow morning. (Analog has a long lead time. So the column won't be in print for several months.) 6/30: It's up. Read it here!

Kathryn Cramer at July 22, 2004 01:33 PM | Link Cosmos | Purple Numbers | Edit

thank you!

Is it possible to put up scans that are larger for those with difficulty reading small type can read it?


Posted by: redjade at July 23, 2004 07:42 AM

Those are all I have at the moment. I can't do anything to improve it until Tuesday.

Posted by: Kathryn Cramer at July 23, 2004 08:46 AM

You can post the image as the background to an html page and it will be large enough to read. You also have a security breach to your site. You might want to lock it. If you need I can make the page for you and send it to you for posting of the two images as html's. Just link to the pages once uploaded.

Posted by: Spacedust at July 24, 2004 12:50 AM

Let say I am dummy!

How do you explain Shryar theo. to someone to layman's word

Posted by: hesam at July 24, 2004 11:29 AM

IMO this is a very good lay interp. posted above.

Posted by: Al at July 24, 2004 12:01 PM

Until I get time to post something higher res, here's a suggestion for those wanting bigger type: bump the resolution of your monitor down. It will make the letters bigger.

Posted by: Kathryn Cramer at July 27, 2004 02:55 PM

Afshar's claim seems to be that even though they knew which slit the photon went through, they still got interference, violating the principle of complementarity. However, reading various responses online it seems likely that the presence of the wires themselves introduces some new uncertainty about which slit the photon went through, because the wires can scatter photons so that a photon going through the left slit can end up in the right detector and vice versa. One poster on sci.physics.research explains this with a nice thought-experiment about enlarging the size of the wires until they are almost touching, forming, in effect, a new series of "slits":

Now I haven't done any calculations or read the New Scientist article except looking at the lab setup graphics, but if I would hazard a quick guess, it would be that it will turn out that even if the wires are placed in the interference fields valleys, the finite width of the wires will diffract just enough photons to erase the which-way information that was gained by focusing the detectors at the holes in the wall through the lens.

Consider the limiting case with wires placed with their centres in the interference fields valleys as before, but expand their width so much that they almost touch each other. What you have now is yet another wall with a bunch of slits in! Obviously, almost all which-way information is lost after the wavefronts pass these almost infinitesimal slits since they will diffract the photons equally no matter from which hole in the *first* wall they originated, so any detector placed after this obstacle will be like running a new multiple-slit interference setup (although with the lens now severely defocusing the too-closely placed new slits). And since the which-way information from the first wall is erased, interference is free to happen between the first and the second wall. After the secondary wall the detectors can pick up which-way information causing them to behave as if there was little subsequent interference.

Conversely, the other limiting case is with no wires (or secondary wall) present. Then all which-way information is present and again the detectors behave as if there was no interference.

The experiment shows a case in between these limits and the effect I guessed at above could (and should, according to traditional QM) turn out to always cancel any attempt to find both 100% interference and 100% which-way information. This would be better showed with some calculations of course...

Kathryn, would you be willing to ask your father if the finite width of the wires does indeed erase some of the which-path information by causing a nonzero amplitude for paths where the photon goes through one slit but scatters against the wire and thus ends up in the opposite detector? If so, couldn't this show how complementarity is preserved in this experiment?

Posted by: Jesse at July 28, 2004 04:06 PM

I have forwarded your question.

Posted by: Kathryn Cramer at July 28, 2004 04:12 PM

Thanks! It might be a moot point though, I just looked at the diagram and accompanying text more carefully and I realized the poster I was quoting may have misunderstood what happened in this experiment--he seemed to be under the impression that the "interference" in this experiment was found at the detectors, and the scattering of photons against the wires could explain this. But the diagram seems to say that no interference was found at the detectors--the interference Afshar is talking about was just in the fact that no photons were scattering against the wires because they were all placed in the interference valleys. So the idea seems to be that interference is the explanation for why no photons scatter against the wires, but the focusing lens behind the wires makes sure that photons from the left slit always go to the left detector and the photons from the right slit always go to the right detector--this is the "violation of complementarity", that the photons behave like a wave in avoiding the wires but behave like particles when arriving at the detectors. I'm not sure that the notion of "complementarity" has ever been sufficiently well-defined to say that this experiment violates it though, and in any case, as long as the results of the experiment match the predictions made by the standard theory of quantum mechanics, it cannot be taken as a falsification of the Everett interpretation, since the basic idea of the Everett interpretation is to keep the standard rules for wavefunction evolution but just to drop the "collapse" idea (the projection postulate).

Posted by: Jesse at July 28, 2004 04:42 PM

Jesse: Here's the response to your question from my dad:

Several people have asked me similar questions, usually centered on various effects of the wires in the two-pinhole measurement. Several months ago our UW Physics Department Chair, Prof. David Boulware, pointed out to me that Afshar's wires are placed in just the positions of the opaque stripes in a diffraction grating, which would produce multiple images of the pinhole and, in particular, would put flux from pinhole 1 at image 2, etc. This would compromise the purity of the "which-way" measurement. This point was very troubling, so I immediately asked Shahriar Afshar about grating effects, wire scattering, etc.

His reply was to send me some data from single-wire measurements. He has done measurements in which he uses ONLY ONE wire placed at one of the interference minima and measures the flux everywhere, not just at the image sites. He did such measurements in three situations: (1) wire in & both pinholes open, (2) wire out & both pinholes open, and (3) wire in & only one pinhole open.

Measurements (1) and (2) show very clean images of the pinholes, with essentially nothing detected between them. The two measurement plots were so similar that they were essentially indistinguishable. Measurement (3), on the other hand, is very different and shows considerable flux outside the image positions from light that is scattering from the wire. In other words, many photons are hitting the single wire with one pinhole open, but essentially nothing is hitting the wire with both pinholes open. No light is scattered because the single wire is placed in an interference minimum where there is no flux of light. Further, with only one wire present there should be no diffraction grating effect (which depends on the coherent interference of waves passing through multiple slit openings). Therefore, I find that the single-wire measurement defuses any claim that the Afshar measurement is not purely "which-way" because of grating effects or wire scattering.

Posted by: Kathryn Cramer at July 28, 2004 06:32 PM

What's the cover-date on that issue of New Scientist?

Posted by: James at July 29, 2004 11:26 AM

July 24th.

Posted by: Kathryn Cramer at July 29, 2004 11:31 AM

Just a quick question, as I have only tonight become aware of this work. Are the two mirrors used simply to scan images of the respective pinholes across the detectors to build up an image of each at the (stationary) detectors?

Two comments perhaps. Einstein of course proposed at least one experiment to challenge the idea of complimentarity (recoiling screen) but was countered by Bohr, in that particular case probably correctly. Second, would Bohr have had a problem with the fact that the interference pattern is never actually observed but (very cleverly) inferred?

Finally Kathryn, thanks for this blog. How wonderful it is that our understanding can be challenged so cleverly at such fundamental levels.

Posted by: David Herne at July 29, 2004 01:00 PM

John Cramer (via Catherine Cramer) says:

Further, with only one wire present there should be no diffraction grating effect (which depends on the coherent interference of waves passing through multiple slit openings).
This is not quite right. Assume the wire is vertical. Then the photons that pass on the left will interfere with those passing on the right.

Consider a shallow pool of water with a brick sticking out of it. When a wave hits the brick, the parts that pass on the two sides will interfere with each other beyond the brick.

What bearing this has on the experiment, I do not know.

Posted by: MonkeyBoy at July 29, 2004 03:18 PM

Being a particle physist I am somewhat hesitant to support Mr. Afshar's theories. The very idea that particles are illusions is absurt for countless reasons, one being that they are the very foundation of Quantum Physics, as shown by Planck's constant. I am also a firm believer in a somewhat unpopular belief that is none the less, supported by the current laws of physics, that is the belief that we may travel through time. Kip Thorne showed that relativity allows it, and yet there are also the paradoxs, such as killing ones own grandfather, however the Everett Many Worlds theory of quantum physics makes sense of this by allowing for alternate realities to be created. I will keep an open mind however.

Posted by: Motumbo Amechanichan at July 29, 2004 04:54 PM

See my new post. They've done the experiment for a single photon.

Posted by: Kathryn Cramer at July 29, 2004 07:58 PM

I just read the full article in the current New Scientist and I want to throttle the journalist who cleverly left the all-important hook (for non-specialists, anyway) until the last paragraph. At least I know why it matters now.

Posted by: Feòrag at July 29, 2004 08:37 PM

What happens when the detector for pinhole 1 is covered up? moved?

Covering up the detector for pinhole 1 should increase the absorbance to 6% (the interferenceless absorbance, as opposed to 0%, which is claimed as observed when the detector is uncovered). Also, if one moves detector 1 "vertically" (meaning say, farther from detector 2), and this experiment is working as it claims, one should see the absorbance oscillate.

Is this what happens? Could you perhaps ask Afshar?

Posted by: Anonymous at July 30, 2004 06:34 AM

Has anyone tried this with single photons? David Deutsch bases his theory of the "multiverse" on interference expriments with single photons, in which interference patterns are produced by interactions with ghost photons in other universes.

It would be interesting if the wave/particle duality was preserved in single-photon experiments.

Posted by: Geoffrey Bennett at July 30, 2004 07:01 PM

The single photon results came through yesterday: see Afshar Experiment for a Single Photon.

Posted by: Kathryn Cramer at July 30, 2004 08:49 PM

Regarding the single photon experiment, is there still an interference pattern? If so, what is the explanation, if one does not accept the notion of multiple universes?

Posted by: Geoffrey Bennett at July 31, 2004 10:02 AM

Sorry to post this here but I haven't been able to link to Rowan University or Shahria's email address for over 12 hours. Anyone know of problems?

All the best.

Posted by: David Herne at August 1, 2004 12:34 AM

Sadly, there seems to be a sharp increase in the attacks to Rowan University's main server, as well as Verizon's regional hub after my NPR interview announcement. We will attempt to post through mirror sites if the problem persists. Keep checking, and thank you for your patience! Best regards.

Posted by: Shahriar S Afshar at August 1, 2004 01:07 PM

Solution to scan size problem:
Firefox and Image Zoom extension:

Posted by: underlyingreality at August 1, 2004 05:45 PM

First, I want to thank you for your excellent site Kathryn. And I'm wondering what this experiment will mean to quantum computing. Don't they rely on superposition of states and collapse of wave function? Also, how does the quantum computing research corroborate (or not) Afshar's results? In particular, I'm thinking of a spin-linkage experiment that was reported a month or two ago. Thanks.

Posted by: Eric Farnsworth at August 1, 2004 10:21 PM

send me report with details

Posted by: mostafa basirian at August 2, 2004 02:46 PM

This experiment "demonstrates" the wave-like quality of light by the _absence_ of an effect of wires placed where the waves would cancel. It is a "gotcha" similar to the winning of a shell game by lifting two of the shells and claiming the pea is under the third. I think Bohr would be philosophical enough to be pleased to see this complement of complementarity.

Posted by: David Landowne at August 5, 2004 01:46 PM

See the URL
for my comments on this experiment.
(In short, it is entirely consistant with Complementarity and Bohr would have had no problems with it).

Posted by: W. Unruh at August 7, 2004 04:24 PM

W. Unruh is exactly right. The main problem with Afshar's interpretation, as I see it, is that the attempt to pick out which slit the photons went through comes AFTER they have had a chance to interfere with each other, at which point it is no longer possible to say with certainty which slit they came from. The indistinguishability of the photons becomes an obvious problem for Afshar if one considers Feynman's interpretation that the photons may traverse all possible paths (with differing probabilities) from the slits to the detectors. Afshar's experimental results are entirely consistent with modern quantum mechanics.

Posted by: D. Knapp at August 7, 2004 06:18 PM

Not impressed by this.

A lot seems to be getting read into the supposed information about which slit the single photon went through. But according to strict CI, this decision isn't made until the moment of measurement -- the system of experiment plus detectors are in superposition given by:

|left-flash> + |right-flash>

and only collapses into a single state at some point where the quantum world becomes the classical world. Far from contradicting complementarity, this is the very definition of it.

Posted by: Skeptic at August 7, 2004 09:24 PM

Thank you all for taking your time to think about and address my findings and conclusions. I am setting up a FAQ section in my web-page, answering as many criticisms as I can gather, so please check my page often.

To D Knapp: Please address my experimental setup, rather than discussing an alternative experiment, which although simpler, is an inaccurate analogy nonetheless, and incapable of refuting my arguments. There is a fundamental difference between a wavefront-splitting interferogram and an amplitude-splitting one like the Mach-Zehnder you have used in your thought experiment, which is often ignored by most physicists. I will be addressing your setup/arguments in my FAQ page.

To Skeptic: It is true that before the measurement is made at the image plane, one can use the superposition, however, upon the measurement, the which-slit information is established and retrodictively, one must assign a particular slit to the photon. When you collapse the superposition state, only one of the slits will be found as the origin, (that is to say the Cat is found either to be dead or alive...) I will address this issue in a theoretical paper on the theory of measurement in a couple of months. At issue is the law of conservation of linear momentum, but this is not the right forum for the discussion. See my FAQs.

We are thinking about putting together a seminar at Rowan University to discuss my experiment sometime this Fall. So please stay tuned, and let me know if you'd be interested to participate.

Best regards.

Posted by: Shahriar S Afshar at August 7, 2004 10:55 PM

How can you "retrodictively" assign a value to something that never had a determinate value in the first place? This is exactly what the CI tells us *not* to do. The "which-way" information is an illusion arising from putting the system into an eigenstate of position at the level of the detector. Prior to measurement, the system is not in an eigenstate of position, and you can't have positional information about something which has no determinate position to begin with.

I do agree though, that your experiment is bascially a Schrodinger's cat situation. But CI tells us that systems do not have determinate properties until measurements put those sytems into eigenstates of those properties, and I think this is the fundamental feature of the CI which you and your supporters are missing.

And while I'm no supporter of the MWI, I have to say I don't think you've falsified that either. Like the CI, systems in the MWI only acquire determinate properties when measurements are made on them. So the "world branching" occurs at the level of the detector, not at the level where the photons pass through the slits.


Posted by: Skeptic at August 8, 2004 06:03 AM

Dear Kathy,

The issue of retrodiction and so forth has its origins in Heisenberg's own writings, later expanded by Einstein, von Neumann and many others, (lately Griffiths, Gel-Mann etc.) Again at stake are the conservation laws...

I (and every quantum theorist) agree that the quantum formalism in my experiment predicts an earlier interference and a later which-slit determination at the image plan. As much was agreed upon even between Einstein and Bohr (See Wheeler ref.s on his delayed-choice experiments). The issue is "measurement". According to CI, "measurement" of the which-slit information for a photon, makes it (non-locally) to have behaved like a particle, (i.e. a BB), thus incapable of having interfered.

The source of the contention is not the quantum-mechanical formalism; it is the ad-hoc assumptions injected to the QM by the measurement theorists, the first of whom were Bohr, Heisenberg and Pauli.

Posted by: Shahriar S Afshar at August 8, 2004 08:20 AM

Sorry, I think this is simply wrong.

Apparently you're claiming that the CI predicts something that the quantum formalism does not. But by definition, the CI and the quantum formalism are identical up to, but not beyond, the point at which the world starts to behave according to classical rather than quantum rules.

The issue of retrodiction, in my opinion, is a red herring. The formalism, and the CI, are both quite clear -- the photon has no determinate position until it is forced into an eigenstate of position. As for the stuff about measurement forcing the past history of the photon into a "particle" state, well this is a piece of metaphysical baggage which has never been a part of the CI as I've understood it, and as far as I know, never has been. In fact, the original two-path experiment itself contradicts this.

As time goes by, my admiration for Bohr has increased. He seems to have predicted, well in advance, exactly what would happen if physicists started trying to turn themselves into amateur philosophers -- they'd end up spending their entire lives chasing metaphysical cats up blind alleys. How right he was.


Posted by: Skeptic at August 8, 2004 08:58 AM


I too admire Bohr, for his sincere and valiant attempts to rescue human logic, as well as his dabbling with philosophy. One cannot blame physicists to do some philosophizing, given most of them have doctorates in philosophy (PhD)!

Here's a quote from Bohr himself discussing the welcher-weg experiments:

�...we are presented with a choice of either tracing the path of the particle, or observing interference effects�we have to do with a typical example of how the complementary phenomena appear under mutually exclusive experimental arrangements�[1]

You be the judge of what he meant!

[1] N. Bohr, in: Albert Einstein: Philosopher-Scientist, P. A. Schlipp, Ed. (Library of
Living Philosophers, Evanston, Illinois, 1949).

Posted by: Shahriar S Afshar at August 8, 2004 11:12 AM

We don't really want to play the game of, "what did Bohr really mean?", do we? ;-) After all this is supposed to be science, not the exegesis of sacred texts. But, I'll have a go with the comment you mention.

It seems perfectly clear to me that Bohr is right, given that he is describing the traditional two-path experiment. One can either perform a position (particle) measurement, to trace the path of the particle, or one can elect not to perform such a measurement at this stage and observe the inerference effects.

As I see it, you're in effect agreeing with this, but arguing that you can indirectly infer the path
by adding additional apparatus to the experimental system. But the CI specifically *rules out* this
sort of indirect evidence to establish the value of determinate properties in evolving, unobserved systems, because the CI quite explicitly and unambiguously says the system doesn't actually have any determinate properties until a specific measurement puts the system into an eigenstate of one of the properties. The so-called "eigenvalue-eigenstate link" is absolutely fundamental to the CI, and is what the delayed choice experiments are supposed to illustrate.

In fact, the classical two-path experiment itself makes it quite clear that performing a particle measurement *after* the interference has occurred does not somehow endow every point on the particle's path with a determinate position, because the final distribution of particles is emphatically not consistent with such a hypothesis. If it were, then presumably there'd be no need for quantum mechanics in the first place.


Posted by: Skeptic at August 8, 2004 12:28 PM


I suggest you familiarize yourself with the literature on Complementarity. Here is a good paper which was later published in PLA with many good ref.s:

I also suggest that you write a theoretical paper after my paper appears in PRL (and you've had a chance to read it) to refute my arguments. This forum is simply not capapble of addressing detailed discussions.

Good luck!

Posted by: Shahriar S Afshar at August 8, 2004 03:07 PM

Dr. Afshar asks me to discuss his experimental setup rather than the alternative one introduced by W. Unruh. I emphasize that my comments do not depend on the specifics of the experimental setup (whether one is doing Unruh's or Afshar's experiment) but on the moment of the measurement of position of the photons. Afshar detects the photons AFTER they have had a chance to interfere, AFTER they have travelled through the same lens. The photons from slits 1 and 2 become indistinguishable upon interference at the lens. The photons reaching one of the two detectors could have come from EITHER slit; it is not possible to claim that all of the photons detected by detector 1 came from slit 1. This is the same point being made by Cathy/Skeptic. Quantum mechanics predicts Afshar's results; there is no violation.

I'd also like to briefly comment on Afshar's use of the Schrodinger cat analogy. His claim that the which slit information is preserved is equivalent to stating that if one sees a dead cat upon opening the box then one can infer that the cat must have been dead from the start. However, one doesn't know how long the cat has been dead, only that it was dead when the box was opened. Before that the cat was in a superposition of alive/dead and there is no way of knowing its history because one was not watching it until the box was opened. Because the position of the photon is measured at the end of the experiment, there is no way of knowing which slit the photon came from.

Posted by: D. Knapp at August 8, 2004 06:21 PM

Has anyone any views on how Dr. Afshar's experiments sit with David Bohm's hidden variables/implicate order theory?

Posted by: Peter Fyfe at August 8, 2004 10:54 PM

Dear Mr. Knapp,

Very briefly, if you can write the wavefunction of EACH of the images as a 50%/50% superposition of the wavefunctions originating from the two pinholes, then I would concede that there is no which-way information in the images, and you will be eligible to win my $1000 prize for this feat. However, I am confident that you--or anyone else for that matter--won't be able to do it without breaking the quantum-mechanical rules of calculation!

Dear Mr. Fyfe,

Please check my web-page in about a week, I'll try to address your questions there.

Best regards.

Posted by: Shahriar S Afshar at August 8, 2004 11:47 PM

Why does it have to be a 50-50 superposition? Since the deviations from 50-50 will be symmetrical, they should cancel each other out.

(who still thinks this is all irrelevant anway)

Posted by: Skeptic at August 9, 2004 01:17 PM

Hi all,

The way I see quantum mechanics is this. The properties (energy, charge, field, etc) of the quantum system are distributed over all the states defined and described by the wave function. I believe that this is a continuous distribution, with actual quantum or sub-quantum connectlvity between the eigenstates. This implies a completely continuous wave function. When an interaction ( of sufficient strength and asymmetry) occurs, a localization of the property distribution of the system takes place and the wave function appears to collapse. This apparent collapse is just a reorganization of the property distribution of the system, due to the exact nature of the interaction. Since we due not Know the exact nature of the interaction, it appears probabilistic, but is actually totally determined, within the limits of the uncertainty principle.

As an example, consider a quantum system that can described as being in state A or B with equal probability. This would mean that the properties of the system are equally distributed over states A and B, with a very small amount constituting the state corresponding to the connectivity. When this interacts with another non-quantum (already localized) system whose exact structure is unknown, then there is a 50/50 chance for the localization of the original system to occur in either state A or B. If the quantum properties of the second system are non- localized and known, localization probabilities for the first system will be different.

In the double slit;it experiment, the probabilities are totally dependent on the photon's property distribution at the point of interaction with the screen, since the exact structure of the screen is unknown and it presents an equal opportunity for interaction with the various portions of the photon state distribution.

In this view, there is no abrupt collapse of the wave function. There is just a continuous change in the wave function due to the localization and de-localization of the property distribution of the systems involved.

Entanglement implies a partial or complete overlap of the property distribution and/or the connectivity distribution. It could also imply a secondary connectivity distribution.

Posted by: z at August 9, 2004 04:13 PM

An addition to my last comment.

In the double slit experiment the photon always goes through both slits (properties distributed in a two state form).The property distribution then takes on the normal (interference} density form as the two state form expands from the slits.. This always disallows a which-way determination since it does not exist.

Posted by: z at August 9, 2004 06:05 PM

To Dr Knapp:

I'm afraid I don't understand your comment about Feynmann path integrals.

Are you suggesting that the wavefunction at each detector contains components from both slits? As far as I can see Dr Afshar is clearly right, this cannot be the case without violating both unitarity and the laws of optics.

My point was a slightly different one -- the problem I have is with the idea that one can determine the path of a particle which everyone agrees doesn't actually exist.

But of course as soon as one starts getting into the metaphysics, then all bets are off. How is one to ascertain the metaphysics of mingled wavefunctions? Is there any point in doing so?


Posted by: Skeptic at August 10, 2004 02:09 AM

Since you have demonstrated, albeit indirectly, that interfernce is taking place during the experiment, then this implies that the wavefunction at each detector MUST contain components from both slits.

If it did not contain components from both slits, then you could not have demonstrated interference.

The wavefunction at the detectors is just the spacially extended wavefunction that provides for the indirect evidence of interference.

Posted by: z at August 10, 2004 06:55 PM

Addition to my previous comment.

The focusing lenses return the wavefunction to a rwo state condition with each state containing elements from each (both) of the two slits.

The detectors then have a 50/50 chance of recording the photon.

This is similar to what happens in a Rydberg atom when pumped up to a large superposition. The decay of the superposition often results in a two state function before settling into its final state.

Posted by: z at August 10, 2004 07:04 PM

Another addition:

It is also possible that the wavefunction is completely reduced to a single state at the focusing lenses. This however, does not change the fact that elements of both slits are contained in the wavefuntion before its localization and NO amount of conceptual argument will change the fact that the detector CAN NOT unambiguously determine the which-way information.

Posted by: z at August 10, 2004 07:24 PM


Since the evolution of the wavefunction must be unitary (according to QM) interference in this case must arise from simple addition of wavefunctions. The wavefunctions themselves however remain separate (by definition, according to unitarity) which means that each wavefunction behaves independently according to the laws of optics. In other words, the wavefunction from each slit behaves the same regardless of whether the other slit is open or closed.

The wavefunction at each detector is therefore unaffected by the interference, according to the rules of quantum calculation. Whether it's also unaffected in terms of the metaphysics is a question which, I admit, I don't know how to answer (if indeed there is an answer.)

Anyway, I'm now dropping out of this discussion. It may be the case that I didn't fully understand complementarity to begin with -- although I still have doubts about how much this conception of complementarity is due to Bohr. In this case Shahriar Afshar may well be right that the contemporary version of complementarity has been falsified. I still remain unconvinced that this experiment has any implications for the Copenhagen Interpretation (which admittedly is as much due to Heisenberg and Dirac as to Bohr).

Anyway, many thanks to Dr Afshar for his good-natured contributions to this discussion.



Posted by: Skeptic at August 10, 2004 10:20 PM


Although the wave funtions behave seperately not depending on whether one or two slits are open, for each photon the property distribution (real or imagined) is different. The same at the detectors.

Since it is the additive property of the wave functions that produce interference, these must be included at the detectors if interference is happening.

Posted by: z at August 11, 2004 04:47 PM


I thought of another way to look at this. The wave function of a sinmgle photon after passing through the slit(s) is a superposition of the photon passing through slit A and the photon passing through slit B.

The wave function does not collapse at the wire. This shows that it is still intact.

When the wavefunction reaches the len(s) and or detector(s) it is still a superposition of the photon passing through slit A and the photon passing through slit B.

Therefore, at each detector there is information from each leg of the wave function and you can not unambiguously determine which leg of the wavefuntion has facilitated the collapse at the detectors.

Posted by: z at August 11, 2004 05:32 PM


I thought of another way to look at this. The wave function of a single photon after passing through the slit(s) is a superposition of the photon passing through slit A and the photon passing through slit B.

The wave function does not collapse at the wire. This shows that it is still intact.

When the wavefunction reaches the len(s) and or detector(s) it is still a superposition of the photon passing through slit A and the photon passing through slit B.

Therefore, at each detector there is information from each leg of the wave function and you can not unambiguously determine which leg of the wavefuntion has facilitated the collapse at the detectors.

Posted by: z at August 11, 2004 05:33 PM

While it was interesting to learn that results exist for the single-photon version of this experiment, the actual results were not reported, and I am still interested to know what happened.

The statement was made that this experiment rules out the many-worlds interpretation. If interference still occurs with single photons in this experiment, how is the interference explained? How is it that a single photon can interfere with itself if not with photons in a parallel universe?

Posted by: Geoffrey Bennett at August 11, 2004 09:00 PM

I am quite confident in the result of Shahriar Afshar on an experimental point of view. There is no a priori problem for the one photon experiments and G. Bennett will certainly no surprised if
the result would be the same ( of course it will be interesting to know which kind of source was used: laser with low intensity, atomic source , non linear crystal... ). There is however some questions concerning the experimental system in it self. I understand that if his work is submitted to PRL he wants to be prudent and that he can not reveal the details. But after all this mediatization it would be good for the credibility to reveal certain details on the set up
him self ( distance, dimension,....)in order to allow other physicists to validate at least theoretically his models. I propose to him that he give us on this web page some precise informations.... (the page of his university is quite dead for the moment)
I want to precise that I am not convinced on his interpretation
and that his definition of a which path experiment is not the one
generally used in the ''quantum'' community. However the result is sufficiently interesting to be now discussed
on a rigorous level.

with best regards Aurel.

Posted by: realist at August 13, 2004 10:06 AM

Simple questions, to Dr. Afshar.
If you erase the information about the "welcher weg" did you expect any change at those wires? (I suppose not, you still have that interference, or, to say it better, that "negative" measurement of interference). And if you "restore" the information about the "welcher weg" are there any changes at those wires? (No, I suppose). So, can we say that the information (about the which way) has nothing to do with the interference at the wires?

Posted by: s at August 14, 2004 12:44 PM

Dr. Afshar believes that particles are illusions, but this makes little sense if you look at what started the principals of quantum physics in the first place. It was realized that light can't be made up of only waves because wavelengths would be infinite i.e. the ultraviolet catastrophe. Planck calculated the size of quanta, or the tiny packets of energy that create limits to wavelengh, therefore I believe that particles DO exist, and cannot be just "illusions".

Posted by: Dr. Motumbo Amechinamican at August 15, 2004 11:52 AM

the concept of photon is not easy to eliminate . I Think in particular to a old experiment made by Roger Aspect and Grangier 1n 1986 ( published in europhys. lett. ) . In this experiment, which used correlation between two entangled photons, it is proven that any semiclassical model a la lamb and Afshar is too naive. The idea of this experiment is to consider a wave packet going on a beam splitter ( semi transparent glass). the wave packet will be separate in to two parts but if the wave packet contain only one photon we expect to detect the photon in only one side . This old gedanken experiment proposed by Eintein and made by Aspect is a proof of the photon existence because a continuous classical wave packet can not justify this anticorrelation. With a pair of photons we use one photon to trigger and count events at the exit channel of the beam splitter and we can see the anticorrelation of arrival on the two detectors A and B associated with the reflected and transmitted beam.

I would like to know if any body can explain this experiment without photons.

Posted by: realist (in polland ) at August 15, 2004 03:14 PM

please send me report with details

Posted by: limoo at August 16, 2004 09:15 PM

In his 1994 book "Schrodingers Kittens" John Gribben describes a real experiment whereby wave and particle behaviour were shown to exist at the same time. even though it is a laymans book the experiemnt was real. It was done using two prisms with a very small gap between them inserted in one of the possible photon paths after the photon has passed through a half silvered mirror so that you could not tell which was the photon had gone. The gap was very narrow (10^-9 meter)and as only waves can tunnel across air gaps it showed waves and particles at the same time. however it did not show which way information but it immediately showed the CI to be on very shaky ground.

John Gribbens book also looked at other recent experiments (at the time) and concluded that the CI of the conscious observer being required and the principle of complementarity is basically incorrect.

after much sole searching and discussion he states that J Cramers transactional interpretation of QM best fits the known facts of QM but it makes no different predictions to standard QM itself !!??

Now that a experiment has shown wave particle behaviour at the same time and not destroyed the interference pattern then surely this must show the differences between stanadard QM and other hidden varible theories (still possible as Von Neuman made a mistake in his calculations).

Surely we will shed some light on QM now ??

Posted by: quantum fanatacist at August 24, 2004 11:14 AM

I'm coming into this discussion a little late, so if my questions have already been answered somewhere else I'd appreciate someone letting me know. Thanks in advance!
1) Precisely what happened when the experiment was conducted using a single photon, and have any of these results been duplicated yet?
2) Does this really invalidate the old CI notion that a wave collapse occurs at the moment of measurement, and how exactly is "measurement" to be defined (or not defined/addressed) in terms of the wires and the rest of the set-up? I'm not at all clear on this, despite the diagrem from NS. This is a tricky epistemic problem, anyway --I'm not sure that any experiment could decisively overcome it (although I am quite willing to keep an open mind).
3) A transactional interpretation of QT seems, at least to me, to be every bit as counterintuitive as the notion of "many worlds" or even the radical idealism suggested by the theory that consciousness itself is what causes waves to collapse. I'm not sure that the notion of different parts of the universe shaking hands really does much to restore "classical reality." This isn't really a question, I suppose -- more of an observation. I guess what I'm trying to articulate is the question, "Precisely how does this experiment affect the philosophical/cosmological issues that have given rise to so many competing interpretations?"

Posted by: J.J. Stambaugh at August 31, 2004 03:48 PM

Further thought about question (2) in my previous post:
I guess what I'm really expressing is some scepticism that the experiment actually shows photons behaving as both particles and waves at the same time. If true, of course, this is a revolutionary experiment. But are there any other intepretations of the results?

Posted by: J.J. Stambaugh at August 31, 2004 03:59 PM

After my holliday I found on internet the description of the afshar experiment done by Unruh (see the web page of Afshar). I confirm the point of view of Afshar the experiment of unruh is not pertinent. Indeed Unruh supposed (in a certain sense) that in the set up of Afshar (with young double slits) the presence of the wires with only one slit open modify strongly the propagation of light . In other term following unruh a wire will scatter so much photons that we can never be sure if a photon will go on the good detector. this is true in the unruh experiment using a Mach Zehnder interferometer but in the Afshar experiment (as he said to me in a email) the fraction of light scattered in the wrong detector is only 6 percents and not 50 percents as in the problem proposed by unruh.

However this doesn't prove that the result of afshar is correct but only that critics concerning his works are very often not complete (or even wrong ). a good list of wrong answer is present in new scientist letters (i like this journal more and more): some of them made the same assumption that unruh (or me at the beginning ) some readers pretend that wire are not usefull or that the two informations (wave and path ) are not obtained at the same time . These replies are all incorrect or not pertinent but it will be too long here (for the moment) to explain why.

I still wait with impatience the publication of the afshar works.... to reply

Posted by: realist at September 1, 2004 08:42 AM
The Comment feature of this page no longer works. To leave a comment, find the new version of the page in the Archives Index, or leave your comment on one of the newer blog entries. -K

Powered by Movable Type 2.63