Official Afshar Experiment Blog Questions 1

Dear professor Afshar,

I was fascinated by the recent experiment made recently by you. Afshar and by its controversial interpretation. I would like to comment that for me there are two fundamental problems in this
experiment. The first problem concerns the use made by you of the Bohr complementarity principle and the second the definition of what we call generally a ''which path experiment'' in quantum
physics.

Let me first explain what is the meaning of the Bohr principle. Quantum mechanics is a statistical theory and by
definition this theory can only give us informations about averages and probabilities obtained by experiments made on many identical systems. In classical physics we can build such models easily for example if we play with a dice. However in classical physics we know that in principle we could describe exactly the motion of the system studied (the dice). In quantum physics we have no such fundamental theory and we must only consider statistical predictions. The problem is that these probabilities are connected to a wave function introducing a strong dualism in the formalism. How indeed to justify the double slit experiment made by Young with particle going to only one slit? It seems that the particle knows magically the presence of the second slit. Bohr and Heisenberg realized that if you would like to build such a theory of what happen in the slits region you would need observations which would disturb or destroy your original system prohibiting then any positive conclusions. Imagine now a photon crossing a double slit you can naturally put a detector just after the slits to see the photon but you will then absorb it and you
will not be able to use it again. Oppositely you can detect the particle on a screen after a lens and observe after statistical accumulation of identical event the famous fringes. You have then a choice between measuring the distribution of particle in the slit screen or to observe the distribution of particle in the fringes plane. But this two choices exclude each others. This is the idea of the principle of complementarity : with a same
particle you can not extract sufficiently of information to reconstruct in a statistical way all the predictions potentially contained in a same wave function. Einstein didn't like very much dices (he prefered cats) and imagined experimental protocols to use two time the same photon and to break down the principle of Bohr. Unfortunately for him and for us all of them failed. The fundamental reason for that is that a quantum system is very fragile by nature and will be immediately disturbed by the measuring apparatus. Now considering the experiment of Afshar we
see that we can observe after the lens two separated spots A' and B' which are images of the two pinholes A and B. Naturally neglecting some losses if you count the number of particle in A' you will know how many particles crossed A and similarly for B' and B. But do you know something about fringes before the lens? Clearly it is not the case. If you want to know something about them you would have to introduce a fringes detector which will disturb the system. Now the wires introduced by you are located at the minimum of the fringes so that the perturbation is very
weak. But in counterpart the information that these wire give us is very small. Imagine that your photon survive to the wire and that it is detected in A' or B' you will be able, as previously, to say by a statistical accumulation of similar events how many particles came from A and B. The spots however will be slightly modified by the wires and the distribution of particle in the slits plane will be slightly uncertain. But now with the same surviving particles you can not reconstruct completely the fringes you will just know that if your particles survive to the wire it it is for sure that the density of particle at the minimum of the fringes is closed to zero. As you see you can not have completely the two informations and the more you know one the less you know the other. This is contained in the principle of Bohr and far to be a counter example your experiment is just one other
proof of its general validity.

The second problem concerns the concept of potential trajectory followed by a photon between the pinholes and the imaging plane. Indeed you not only claimed that he can disprove the complementarity principle but the added that he can effectively determine the path followed by the particle. Following here a old assumption originally enounced by Bohr, Heisenberg and Wheeler,you accepted that even with the two pinholes open a photon trajectory (if trajectory there is) will necessarily connect a pinhole to its optical image as it is if the other pinhole would be closed. However, as is was realized by numerous physicists as de Broglie, Bohm, Bell and Scully, this is a free metaphysical
assumption which depends of our model of reality and which can not in general be experimentally tested [read for example " Surrealistic Bohm trajectories", Englert et al in Z. Naturforsch. Vol 47a , p 1175.]. Effectively nothing in your experiment forbids a photon coming from one pinhole to go in the 'wrong' detector for the second pinhole. This is the case for example in the d eBroglie-Bohm's theory which is an alternative conception of
quantum mechanics.

My conclusion is that in the experiment made by you and presented in New scientist not only the principle of complementarity is still valid but it is not necessary a which path experiment.

Prof. Afshar,
Congratulations on coming up with such an elegant experiment that challenges long accepted concepts in physics.

My question is somewhat similar (I think) to the second question in the 1st comment(although I can't put it in so technical terms :)). how can one prove that the photon coming from one pin hole will definitely end up at the corresponding detector. Is it possible that it gets scattered at the wires and end up at the opposite detector? The absolute dark region of the interference pattern has zero width, so one can not put a wire precisely at that point without blocking some light. There will be some light at the vicinity of the absolute dark point bouncing off the wire. What happens to this light? does this introduce an uncertainty into the detection scheme?
- Mehrdad

Dear Prof. Afshar:
I am in awe of the simplicity of your experiemnt. And, I am very enthusiastic about your results! It has changed my thinking entirely about quantum physics!

Question:
Was the experiment repeated with a single-photon source? If so was the result the same?

Thank you for any information....Kristian

Dear everybody,

Please take a look at my preprint:

http://www.irims.org/quant-ph/030503/

It answers many of your questions. But generally speaking let me provide you with my starting assumptions and some comments:

1) Quantum-Mechanical formalism is completely valid, and should be used to make the calculations for the the amplitudes, as well as the observables in any experiment.

2) According to QM, an image produced by a lens contains the "which-way" information. More correctly, in a double-slit type of experiment, the images of the slits contain the "which-slit" information, given the images are well-resolved and have virtually no overlap. This is based on a long list of literature starting with the original "Heisenberg's Microscope" proof of the Uncertainty Principle, in which the image prodoced by the lens provides the information on the spatial origin of the photon. Wheeler specifically uses this method in his delayed-choice experiment. See the above preprint for more references and detailed argument.

3) The visibility of an interference pattern is defined as : V= (I_max - I_min)/(I_max + I_Min). Notice that when I_min = 0, the V==1; (== means "is identical to"). That means if you find directly, or indirectly that I_min=0, you NEED NOT measure I_max. As long as you know that there is light passing through the wire grid region, and its total flux is not affected by the wires, you have the NECESSARY AND SUFFICIENT criteria for establishing a perfect interference. (See the above preprint for detailed QM analysis).

4) Having establshed that V=1, Bohr's Complementarity demands that K=0 [1], that is the which-way information must be completely lost. For an imaging system like the lens, it means that the images of the two pinholes must overlap completely due to diffraction and scattering caused by the wires. If we were to obtain such a result, Complementarity would have survived, beacuase we could no longer tell which of the two slits the photons are coming from. The actual experimental result is the same as the case in which both pinholes are open and there are no wires: Two crisp images.

5) Some have argued that the wires destroy the which-way information and somehow conspire to produce the same two well-resolved images. As I have indicated in my talks at Harvard, Princeton, TAMU, and elsewhere, I would award a $1000 prize to the first person who can show that the wavfunction of EACH image in the image plane in the case that both pinholes are open, and the wires are present, is a 50/50 mixture of the two wavefucntions emerging from EACH pinhole. I claim that it is impossible to achieve this feat using QM formalism.

6) Please notice that I do not claim that the which-way information in the last case is still 100%, just that it is not 0% as Bohr claims. Mehrdad is correct in that a risidual amount of light hits the wires and is scattered into the wrong image, even when we have a perfectly visible intereference pattern. This is due to the finite thickness of the wires. However, the reduction in the which-way information is quite small (so that which-way information is about 99.99% or better)comapred to the case when only one pinhole is open and the wires are present, simply beacuse in that case the incident flux on the wires is relatively large, and thus the scattering caused by the wires is also considerable. At any event, Complementarity demands that for the case where V=1, the which-way information is 0%. This is clearly not the case in my experiment.

7) Finally, the single-photon experiment has produced the same exact result as my original one.

Best regards.

Shahriar S Afshar

[1] Englert et al in Z. Naturforsch Vol 47a , p 1175. (among many other ref.s listed in my preprint).

Dear Professor Afshar,
I would like to comment your precedent vision.
First I would like to thank you for the online preprint. I Neverthelles remark that the date of this preprint is 2003 in may... Does it means that this article was not accepted or that he was resubmitted??? in addition the name seems to be connected to the los alamos preprint web page. However I can not find it on it.....

Concerning the different points of your comment:

I agree with 1) which is obvious in the context but I desagree completely with 2).
It is not because Heisenberg and Wheeler or even God used this old argument concerning trajectories that it is a good one ...
This idea was criticized by Scully and other as I told you already.

Concerning 3)I desagree too. You think as it will be possible to construct fringes without observing them. With the same kind of reasonning I can prove in basic geometry that by one point only one line goes. Such postulate is obviously false because we need at least a second point. You say that you dont need the complete recording supposing that fringes are here even when you can can not observe them . But your reasoning is circular because it is just what you want to prove.

Naturally I desagree with 4) too because your visibilty is close to zero not to one as you claimed.

I agree with 5) even if it is sad for the 1000 dollars. Obvioulsy I can not prove that 1+1=3.... perhaps it is a good sign for me.

The point 6) is simply wrong because as I said before V=0 (practically). In addition you use the relation of duality introduced by greenberger but I want to mention that
on a quantitative point of view this relation is not sufficiently general to be applied here . D and V refere to a particular class of experiments which seems not to contain your experiment. indeed your knowlege concerning the interference is so weak that the visibility V can not be experimentally defined. Similarly the modification of the spots is not so trivial that the one used in the Greenberger relation. Prudently I will consequently reformulate my sentence : The point 6) is simply wrong because as I said before we know practically nothing about the fringes. (The same remark concerns naturally my discussion of the point 4)....


The point 7 is quite nice for an experimentalist as me. What was the set up?


with best regards

your best reader Aurelien

Dear Aurelien,

The paper was finalized in 2003. However, it was submitted to AJP late September 2004. Since the subject matter is highly controversial, it was necessary to ensure that a few high-ranking experts read, understand and approve of the content of the paper before making it publicly available. I hope you take a deep look into its contents, and respond as rigorously.

Further, the paper is posted on an internal irims.org archive, and not the lanl or arXiv. If you wish to refer to this paper, please make sure you add www.irims.org/quant-ph/xxxxxx

1) Now, are you saying that the wires do not establish V=1?
2) Am I correct in saying that you believe V=0 at the wires?
3) If so, why aren't the wires blocking the decoherent distribution of light at the wire grid?

Regards.
Shahriar S Afshar

Dear Shahriar,

thank you very much for your informations
perhaps i will write a reply but i need at least a reference to your paper so I will probably wait.... I will see

In fact I considered seriously your article before to write my last comment.

There is some point in the text that i could criticized as for example the axiomatic structure based on theorems 1 ,2 and 3 and the fact that this seems not to be connected with the rest of the text but this is for me a detail that the referee will probably see and I hope that he will not refuse the paper for this aspect.

I think that the essential point of conflict between us come from the fact that you believe that there is fringes in front of lens and that you dont need to observe them to prove their existence. I know that you believe in a 'all is wave ' model (me too in fact: funny isn't it ? ) and in this context it is clear that fringes must be here even if I don't observe them. But here we speak about experimental fact and not metaphysical structures. the unical way to see fringes in your experiment is to construct them using a lot of photons as you did in the first experiment. but if you don't put the screen and use your lens with the grid you dont see the fringes and you have not sufficiently of experimental information to conclude that these fringes exist.

A last comment but why did you choose AJP??? this journal is cetainly a very good one but your experiment show something of new or controversial and this is not really the politic of this journal.
(naturally there is a lot of exeptions).

whatever your answer good luck for your article.

best regards
Aurelien

Dear Aurelien,

You can use http://www.irims.org/quant-ph/030503/ as a permanent reference for my manuscript. There are other manuscripts under peer-review right now, but I prefer this one, due to the fact that it is written in a pedagogical sense (hence the submission to AJP.) There maybe some redundancies in this paper, but they are aimed at clarifying experimental and theoretical facts. The theorems are trivial, but one is amazed on how many notable physicists forget to recall them during the discussion of this experiment.

I do not have any particular fondness of the "all is wave" model, but that SEEMS to be the best way to explain the nature of light at this moment. I have designed a series of new experiments, which will directly probe this question... Let Nature reveal herself!

Best wishes.
Shahriar S Afshar

Dear Shahriar Afshar
Thanks for your time,
I would like to know how you can justify the old experiment made by Grangier et al. [1] on the basis of a ''all is wave '' model . by such model I mean a theory where a photon is a continuous and macroscopical train wave which can have a coherence length of several meters (and I dont speak about radio waves). The old semiclassical model of Lamb can not indeed explain all. In particular he fails to justify the coincidence experiment of Grangier. Scully(who worked with lamb as you know) says directly in his book [2] that a semiclassical model fails to justify entanglement. If i detect a photon here this means that the rest of the wave desapears instantaneously .... this old paradox was already pointed out by Einstein.

Personally(because I am old fashioned) I guess that we need at least something as a singularity in the wave to justify all ....


For the rest I think that you are right, I can use your preprint as a reference.

With best regards from Aurelien Drezet

[1] P.Grangier G.Roger A.Aspect europhysics letters 1 pp 173-179 1986.
[2]Quantum optics but you know this reference .....

Your recent experiment is presumed to violate the principle of complementarity by observing both interference and non-interference simultaneously in the same experiment. This contention rests on the belief that there are locations of zero intensity because an interference pattern is present.

We construct an 'Afshar-like' experiment for which this premise is false.

Consider the double slit experiment with circular polarizers covering each slit. Photons passing through slit 1 are now right circularly polarized and photons passing through slit 2 are left circularly polarized.
Each position state is now entangled with a polarization state that acts like a tag attached to each photon. We now know, in principle, which slit the photon passed through. There is no interference.

We now place an x-polarizer, immediately in front of the photon detector. Now, every photon passing through the apparatus is x-polarized before detection. There is no way to determine which slit a photon went through, and we see interference in the angular distribution of scattered photons. The polarization tags have been "erased" by the presence of the x-polarizer.

We now place eraser-detectors at the minima of this interference pattern and no particles are ever detected, as long as the erasers are in place. But, none of the photons passing through the slits ever passes through any of the erasers and no photon tags are ever erased. The position states of these unerased, undetected photons are still entangled with polarization and, when detected, they do not exhibit any interference. This is confirmed when these previously undetected photons impinge on a detector screen.

We now have the essential elements of the Afshar experiment:
a) non-interference at a detector screen and
b) no photons detected at the previously measured positions of an interference minima.

There are locations of zero intensity because the erasers are in place, not because an interference pattern is present. If an interference pattern were present, we would be able to see it on the observation screen. But, we don't; no interference pattern is ever observed anywhere. Your (and Cramer's) claim that if there are locations of zero intensity then there must be an interference pattern present is disproved.

We emphasize that quantum mechanics does not describe the behavior of undetected photons. It only gives us the probability of getting through the apparatus and into the detector. As John Archibald Wheeler has said, "-------, we have no right to say what the photon is doing in all its long course from point of entry to point of detection".

The existence of points of zero intensity does not constitute an observation of an interference pattern for the photons that pass by the wires undetected. Only the actual detection of an ensemble of these previously undetected photons can do that. No experiment can tell us anything about the probability distribution for undetected photons, and quantum mechanics doesn't tell us anything either.

It is meaningless to ascribe an interference pattern to undetected photons.

You can detect both interference and non-interference only if you do two different experiments. In the first experiment, place the observation screen in front of the lens and you will, indeed, observe an interference pattern. But, of course, you are now unable to see non-interference because all photons are intercepted by the screen. In the second experiment the detection screen is placed behind the lens. Here, there is no interference, but now you have no knowledge of what the undetected photons in front of the lens are doing. You cannot perform both experiments simultaneously. Complementarity remains intact.



thomas_marcella@uml.edu

Dear Aurelien,

The "all is wave" approach used in the semi-classical electrodynamics is rather different from my approach. For the past 18 years I have been developing a Theory of Everything in which elementary particles are nothing but "sources" of electromagnetic radiation and thus their dynamics is described by wave mechanics . This is a highly speculative theoretical work and it is not my intention to advertise it here, but it is quite successful in providing the correct relativistic results in calculating the inertial mass of such particles, resolving the dark matter/energy mystery in cosmology, etc. For a primer on this theory of mine, take a look at this link to the abstract

https://commerce.aip.org/jsp/Lookup.jsp?item=APCPCS000458000001001033000001&src=PFtop

I will be posting this paper on my archive shortly.

The main point of my “which-way” experiment has been to shed doubt on the validity of Complementarity and show that insisting on Einstein's definition of photons leads to contradictions in an experiment like mine. What it ultimately means as to the true nature of light, we just simply have to wait for the next set of experiments.

Regards.

Dear Thomas,

As the name implies, the Principle of Complementarity is supposed to be universally valid, i.e. there can be no exceptions where the Principle fails. The experiment proposed by Unruh and (perhaps yours) are two examples of the great many experiments in which Complementarity does survive. It would be very helpful if instead of suggesting your own version of the "Afshar Experiment" and then defeating it, you simply address my arguments in the real "Afshar Experiment," which to my knowledge, is the only one thus far in which Complementarity clearly fails. To be fair, I wish to better understand your thought experiment and perhaps you could kindly e-mail me a short paper with a few equations, and the schematics of your experimental setup. Meanwhile I would appreciate if you could take a look at my preprint and let me know exactly where I have erred:

http://www.irims.org/quant-ph/030503/

In the above manuscript, I clearly mention that "perturbative" means of measurement in which "tagging" and direct destructive detections are made (such as your scheme,) are certainly in agreement with Bohr's Complementarity. It seems however, that Bohr and his colleagues did not recognize that ensemble properties such as the interference pattern can be measured nondestructively. Using nondestructive measurement schemes, one can indeed demonstrate a failure of Complementarity Principle as delineated in my paper.

Regards.

I would like to comment your last message concerning the experiment of Marcella. I agree completely with his last part when he says that we can just speak about what we effectivelly detect. It was what I was saying to you when I was speaking about the problem of the visibility. If you observe your photon in the interference plane you will find fringes with V=1 but of course you dont see the images. If you see the spots you can say nothing about the fringes and V=0 ON an experimental point of view (but ok as i said already it is perhaps nor a good way to speak). It is only when you do the two different experiments that you can controle the relation D^2+V^2 =1 (of course to apply such relation you need to have the right to do it : It is however a good approximation if the wire are not to thick) .

Concerning his proper gedanken experiment I have two remark . First (and this remark is applied to your experiment too) you can not speak of which path without reference to a particular model of reality. In the Bohm model it is not a which path experiment and for this reason I contest the generaltiy of such name (but I will use it now not to confuse you).
Secondly, I dont think that in thisexperiment it is just to tell that there is no fringes at all. There is fringes in this experiment but there are not observed completely. Clearly you see a minimum when you project your state on |R>+|L> or a maximum if you project on |R>-|L>. However this is a proof that it is the eraser interacting with the incoming wave which produce fringes and not as it is claimed that there is no fringes at all. You just select one result or the other by changing of detector. Of course for the photons ouside of the eraser the fringes are not here but it is normal ....

This experiment is then a mixture beween a measurment of the fringes and of the so called which path measurment. For a particular photon you will made one or the other experiment depending in which location it is detected on the screen: mimimum or maximum= fringes and for the rest of the screen =''which path''.

In the Afshar experiment too you need two experiment too in order to have all the interference or the nice spots. But if you put absorbing wires you will have the same situation that in the marcella experiment: photon absorbed = fringes , photon no absorbed = images (''which path'').

These two experiments are then equivalent and both satisfy complementarity. consequently none of them can be used to refute the interpretation of the other.



with best regards
Aurelien

Oh!!! a last thing . what do you think about the experiment propsed by Agarwal [1,2]?
this experimend pretended toot to see something concerning the wave nature and on the particle nature which seem forbidden by complementarity . However as in your experiment thex can just see these two behavior in 2 differents experiments. you see the wave in the bema splitter if you loofk what is happening inside or you see the path if you look oustide but one exclude the other. for you it is the same : you see fringes or you see 2 spots : the rest is metaphysical.

aurelien (bis repetita)

Oh I forgot

[1] Physics letter A 168, 95 (1992)
[2]Physics letter A 328, 313 (2004)



aurelien (again me)

Dear Aurelien,

You haven't really read my manuscript have you? If you had, you would have noticed that I have mentioned Home/Agarwal experiment. Please take a look at paragraph 3, page3, on the Home/Agarwal experiments, and also read Ref. [20] papers on the topic. Their experiment does not measure Complementary wave-particle measurements. Complementary measurements have non-commuting projection operators, i.e. one measurement rules out the other. Interference and which-way are two such measurements, but tunneling and subsequent detection of the photon imply each other not negate one another, which is the requirement of Complementarity. Thus their experiment does not test Complementarity, and is not in the same category as mine.

Please also, read section 3. "Measurement revisited" of my manuscript and critique that. http://www.irims.org/quant-ph/030503/
Regards.

Dr. Afshar,

I know that your currently available paper does not cover intermediate conditions, where V and or K take values between zero and one. I was wondering: What have you found when you do address these conditions? Do any of your other submitted papers deal with this issue?

Regards,

David Strayhorn, MD, PhD
Dept of Neurology
Wash U School of Medicine
Saint Louis, MO

Dear Dr. Afshar.
In your interesting paper you state (p. 2), in logical terms, the complementary principle, for 'sharp' complementarity observables. I think this principle - as stated - must be reformulated, since it is well known that you can get interference, on the screen of a Young two-slit set-up, even when the quantum passed - for sure - through 'only one' of the slits. See the references below.
- L. Janossy, and K. Nagy, Annalen der Physik, 17, (1956), p. 115-121;
- Leonard Mandel, J. Opt. Soc. Amer., 49, (1959), p. 931;
- R.M. Sillitto and Catherine Wykes, Physics Letters, 39-A-4, (1972), p. 333, who performed the Janossy and Nagy experiment and found a good interference when just one photon was present in their interferometer, at a time, and when an electro-optic shutter was switched several times during the time travel of each photon, closing one of the two slits, and then the other (but reopening the first), so that it was sure that the photon went through 'only one' slit.
In terms of photons the condition for interference was that the two paths lead to the same cell of phase space, so that the path of each photon was intrinsically indeterminate. Of course, if you close one slit, and then the other (and the first reopens), the shutter must be switched in a time which is less than the
uncertainty in the time arrival of the photon at the screen.
I'm inclined to think that the principle of complementarity (for 'sharp' observables) must be reformulated in a more general way, in terms of 'indistinguishability' of paths.
I have many more questions to ask but, for now, I thank you very much.
Regards,
Serafino Cerulli-Irelli (Rome)

dear Shahriar,

I read your paper but i didint watch all the references. in addtion I wrote my letter at midnight (I have not a lot of time to consider your experiment ) so please : mea culpa.

However this doesnt change any thing and I just give the Home/Agarwal experiment to point out some similitudes. first I agree that their definition of what they call complementarity is wrong : they see a wave going through a beam splitter and they deduce that there is an evanescent wave. for me it is not in contradiction with Bohr because the wave particlwe duality is a characterist of ALL experiments. In this sense they confuse wave particle duality and complementarity .
However we can analyse the complementarity with this experiment too

Indeed the key point is that they never observe this evanescent wave as. If they want to observe this wave they will change the experiment drastically . I recall you that the transmission of the beam splitter increase when the two prisms are closer . you change then at the same time the reflection. This prove that the two surface constituiting the gap are fundamental and that there is a superposition of different ''counter propagative''waves in the cavity (in the sense of evanescent waves). Consequently your argument is wrong : the two projector of the wave in the cavity and ouside doesnt commute at all ...

the rest of my argumentation is still true : you can not see a photon without eyes and the rest is theology

regards
aurelien

Is Serafino Cerulli-Irelli saying that we can see interference with only one wave ? Can we hear the clap of one hand too ? I have some doubt.... but science is funny .

If you want to see interference you need a contribution from each apertures a think that this is clear .... I hope

Because Mandel was a solid physicist perhaps I forgot something but what .....

aurelien

Yes, I'm saying exactly that. Of course here the 'indistinguishability' is (perhaps) related to the time-dependent Schroedinger equation. So it is a sort of time-energy 'indistinguishability'. Please read - if you can - those papers by Mandel, and Sillitto and Wykes. I read those papers years ago, and my memory is a bit chaotic! (Janossy & Nagy old paper was just a gedanken experiment, pointing out that if we use a random macroscopic shutter/opener, or something like this, we could be aware of the 'welcher weg', standing outside the set-up, thus our 'consciousness' should collapse the superposition ...).
Regards,
s.

Sorry. I forgot to say that, in principle, it is not so different from the 'one single photon - emitted by two different sources' interference, in which you combine two amplitudes (giving interference) one related to a source, and the other related to the other source. Here (Sillitto and Wykes exp.) there is just one source, but the shutter/opener, moving very fast, gives a similar effect: 'indistinguishability'.
s.

Dear David,

I am working on a theoretical paper to address the intermediate stages in V and K. I'll post it here as soon as it is finished (in a couple of months.)

Regards.

Dear Serafino,

QM is weird, but not that weird! It is impossible to produce interference effects for a photon the wavefunction of which does not receive-at least partially- a contribution from both slits/paths. So, I am positive you have misunderstood the papers' conclusion. I'll try to take a look though...

Regards.

Dear Aurelien,

I am starting to feel like I am in a Seminary! Your "mea culpa", "you can not see a photon without eyes and the rest is theology" and Serafino's "thus our 'consciousness' should collapse the superposition"!!!

But seriously, I certainly agree that if one does not detect a photon, talking about its behavior is simply beyond what physics allows, however, I do detect all the photons in my experiment in the image plane. Once you detect a photon, it's past behavior can be discussed within the limits allowed by the laws of physics as we understand.

All the photons arrive at the two images having passed the wire grid region. So physically-speaking, we know that (i) they could not have come from any other region, and (ii) there was no significant reduction in the total flux. The only way such an attenuation-free passage is possible is to say that there where no photons incident on the wires (physics demands that if photons were incident on the wires, they should be blocked). The only way to achieve no incident flux on the wires, is to use QM and figure out a process by which it can happen. That process is interference and no other process exists to explain it. In fact if there is another means by which the photons can avoid hitting the wires, then we would have a discovery on out hands, something our known laws of physics do not allow!!! But I highly doubt that: Interference happens before the lens, and Fourier does his magic with the lens to give us which-way information, all within the known laws of physics. We are forced to wake up and face the fact: the only bit of discrepancy and "theology" (or even the occult) here is Complementarity and it needs to go back where it came from: a Buddhist monastery.

Regards.

Dear Shahriar ,
It is true that the quantum physics comunauty looks very similar to a sect but I dont want to enter in this debate here.

The problem in your argumentation seems to be again connected with the fact that you know that a wave function is necessary to explain all evolution of the probability in space in time and that neverthelles you dont observe these fringes. I think that you are more religious that Bohr on this point :

YOU believe that fringes are Here even when we dont observe them and then YOU believe that we don't need to measure them ( a minima of absorbtion at the wire position is in fact a partial measurement of the intrterference fringes). This is the exact opposite of the point of view of Bohr.
Following Bohr the wave is hereas a tool but not effectivelly realized (i.e it is just a potentiality) and the the unical way to transform potentiality into a fact is to detect your photon. The rest is really metaphysic theology or science fiction.....

Dont think however that I like Bohr so much ( I prefer Einstein and deBroglie) but I just defend his point of view effectivelly because I dont believe in it (in order to clarify my proper conceptions I mean).



alea jacta est

aurelien

5) Some have argued that the wires destroy the which-way information and somehow conspire to produce the same two well-resolved images. As I have indicated in my
talks at Harvard, Princeton, TAMU, and elsewhere, I would award a $1000 prize to the first person who can show that the wavfunction of EACH image in the image plane in the case that both pinholes are open, and the wires are present, is a 50/50 mixture of the two wavefucntions emerging from EACH pinhole. I claim that it is impossible to achieve this feat using QM formalism.

I believe there is a very simple argument which provides the proof required.

The wavefunction at the lens is a sum of the wavefunctions originating from each pinhole. Under certain assumptions (which apply in the case of your experiment) the two wavefunctions can be said to have the same amplitude at each point in the plane of the lens (although the amplitude varies in this plane). The only difference between the two wavefunctions is the way the phases vary in the plane. In some places the phases are such that the total amplitude is zero (where you place your wires), and in other places the phases are the same giving maxima in the interference pattern. As the amplitudes from the two pinholes are the same, the total amplitude at any point can be said to be a 50/50 mixture of amplitudes orignating from the two pinholes.

So far so obvious. Now, the Schrodinger equation is a wave equation and therefore obeys Huygens Principle: The amplitude of a wave at any point in space can be treated as a new source of a wave. Therefore the plane at the lens can be treated as a new source of waves. It is clear that the total amplitude at every point in this plane has a 50/50 contribution from each pinhole and so, when we propagate it forwards, any future amplitude will be made up of a 50/50 mixture from the two pinholes. Therefore when a photon is detected at one of the two detectors it can be said to be made up of a 50/50 combination of photons from each pinhole.

The complementarity principle is an attempt to describe the behaviour of the quantum formalism. If the complementarity principle says something which disagrees with the quantum formalism then the complementarity principle is an incorrect statement and needs to be amended or discarded. In this case the complementarity principle says that, since you can indirectly detect the interference pattern you cannot know any which way information about the photon. The above argument shows the complementarity principle still holds.

Please reply to my e-mail address and we can sort out the transfer of the $1000.

Dear Dr. Afshar.

"It is impossible to produce interference effects for a photon the
wavefunction of which does not receive -at least partially- a contribution
from both slits/paths."

I agree with that, of course. The trick is to get two different paths having
just one slit opened for a very short time, and then the other for a very
short time. Now imagine a *single* photon entering this peculiar
'interferometer'. The condition for having the interference (tested also in
case of *single* photons) is that the two possible and different paths
(first slit closed, second opened + second slit opened, first closed) can
lead to the same point of phase space. This occurs if the shutter operates
its switching in a time (Heisenberg's delta t) which is *less* than the arrival
time of the photon at the screen. In this case the 'which path' the photon
took is uncertain (like in the usual two-slit) but it is also evident that
the photon passed through 'only one' slit, because only one slit was opened.

Regards,
s.

Correction, of course!

The condition for having the interference (tested also in
case of *single* photons) is that the two possible and different paths
(first slit closed, second opened + second slit CLOSED, first OPENED) can
lead to the same point of phase space. This occurs if the shutter operates
its switching in a time WHICH IS LESS THAN THE UNCERTAINTY of the arrival time of the photon at the screen (Heisenber's uncertainty).

s.

Dear Aurelien,

Am I right in saying that according to your view, there is negligible flux on the wires? If yes:

1) Then we have no choice but to declare that the cause of this lack of incident flux is destructive interference. There simply is no other process by which a certain regions of space (wires) do not receive flux, while we KNOW that regions close to the wires do (we see all the photons at the image plane).

2) If you dont agree with (1), then please tell me how you can use QM to show a particular region CONTINUALLY receives no flux, without assuming interference as the cause.

Regards.

Dear Morgan,

"The only difference between the two wavefunctions is the way the phases vary in the plane. "

Bzzzzz... Wrong answer!

The two wavefunction also have an ANGULAR separation due to the fact that they are coming from two distinct spatial locations. The wavefunction is (at least) a 3D VECTOR entity and not a vector in certain dimensions (so you can have interference) and a scalar in the other. As any good advanced physical optics textbook can demonstrate, a lens operates as an ANGULAR discriminator, and the linearity of the Fourier transform ensures that a suitably placed lens can resolve two sources of certain minimum angular separation. That minimum ANGULAR separation is the lens' resolving power. If you doubt this, then you must boldly announce that Heisenberg’s Microscope proof of the Uncertainty Principle is also wrong, because he uses the lens to obtain the “which-way” information for the photon that scatters off an electron, and the spatial resolving power of the lens enters the delta_P . delta_x >h relation as delta_x…

Sorry, NO $1000 prize for you!!!

Regards.

Regards.

Dear Serafino,

I'll get back to you after I read the papers you mentioned.

Regards

Several people have claimed that your experiment falsifies the many-worlds interpretation, but I falied to find any statement to that effect in your paper. Have you made this claim, or is it just others?

Dear Jim,

I claim that my experiment rules out Complementarity as a “Principle” that has no exceptions.
Generally, I am not very fond of "Interpretations", but from my limited exposure to MWI, it does not even qualify as a viable theory because it seems to violate conservation of linear momentum in interference experiments, as I have pointed out to David Deutsch in the past. One does not need my experiment to question the validity of MWI, but if others think my experiment rules it out, I have no quarrel with that!

Regards.

On Aurelien's theological/metaphysical point of the need to validate fringes, can you replace the wires by optical fibres (they may need to be roughened to transmit through their walls) and tap at least some proportion of photons which may impinge (when not destructively interfered away)?

Regards.

Dear Tom,

Yes, you can. Better still, you can plcae a thin array of photodetectors that form a line not thicker than 1/10 of the peak to peak distance of the fringes. I predict that the results WILL BE identical to mine.

Regards.

Dear Shahriar,

I am a little confused by your response, and I don't think you have discredited my proof.

"The wavefunction is a 3D VECTOR entity and not a vector in certain dimensions (so you can have interference) and a scalar in the other."

No it isn't, and I'm not sure how you got this from my previous post. The wavefunction is a complex scalar field, which means at each point in space there is a complex number. This complex number has an amplitude and a phase.

Perhaps I can clarify some of the words used in my proof. Wherever I say "the plane", "this plane", or "the plane of the lens" I am referring to the physical plane parallel to your sigma1 plane which goes through the lens. The argument is just as valid for your sigma1 plane too, however. Wherever I say "amplitude" I am refering to the magnitude of the wavefunction. So wherever I talk about amplitudes or phases varying in the plane I am referring to the magnitude and the phase of the wavefunction being different at different points in space(specifically in the specified plane).

"As any good advanced physical optics textbook can demonstrate, a lens
operates as an ANGULAR discriminator, and the linearity of the Fourier transform ensure that a suitably placed lens can resolve two sources of certain minimum angular
separation. That minimum ANGULAR separation is the lens' resolving power."

I agree with this entirely, and so Heisenberg's Microscope proof is safe!

Coming back to the statement that you disputed,

"The only difference between the two wavefunctions is the way the phases vary in the plane. "

The wavefunctions from the two pinholes can be calculated explicitly, but it really isn't necessary for the argument I provide. These two wavefunctions are identical except that their centres are separated by the distance between the two pinholes. This separation persists as we get further from the pinholes and the two wavefunctions start to overlap significantly. By the time we reach the lens we can see that the radii of the Airy discs from the the two pinholes are much larger than the separation of their centres (using numbers from your paper about 20mm compared to 2mm). Therefore the magnitudes of the wavefunctions from each slit are always roughly the same, and the approximation that they are the same gets better the further the lens is placed from the pinholes.

Also the distance between neighbouring maxima in your interference pattern is much smaller than the total size of the Airy disc because the phases from each of the two pinholes are varying faster than the amplitudes as you move around the plane of the lens.

Since the amplitudes from the two pinholes are approximately the same at each point in the plane of the lens, the natural conclusion of my argument using Huygens principle is that approximately half the photons coming from each pinhole end up at each detector.

I still think I am due the $1000.

Morgan

Dear Morgan,

"These two wavefunctions are identical except that their centres are separated by the distance between the two pinholes."

No! The two wavefunctions are NOT identical (regardless of their spatial separations.) Crucially, they have a different ANGLE of INCIDENCE. That angular separation is what allows the lens to form two distinct images of the sources. The purpose of this blog is not to teach Physical Optics. I suggest that you formally study the subject [1] or ask an expert in the field to explain coherent imaging theory for you. But to quickly remind you of why you are wrong, think of two distant stars with a small angular separation. The physical lateral separation of the stars may be many hundreds of light years, but the lens could care less about that spatial separation, because all it does is to separate the two sources based on the slight difference in the angles of incidence of the two beams. To think otherwise simply demonstrates lack of knowledge in the field.

"I agree with this entirely, and so Heisenberg's Microscope proof is safe!"

Well if the imaging theory used in Heisenberg's proof is safe, so is mine, which is exactly the same argument. If you reflect on your statement a little, you would realize that your above statement undermines your original claim.

I am sorry for the emphatic tone in my response, but this experiment is far too important to be dismissed by patently erroneous arguments. Simply repeating a false claim does not make it true!

Regards.

[1] For starters, please take a look at the Fourier Optics section (p. 512) of Hecht’s book called Optics (3rd Edition).

Dear Prof. Afshar!

Will You agree that standard quantum theory brings formal procedures to calculate probabilities of elementary processes only. Quantum mechanics does not supply an EXPLAINING MODEL of "weird" behaviour of quantum objects, does it? And this is what distinguishes QM from ALL the other scientific traditions. Science should provide EXPLAINING MODEL (underlying picture, in John Cramer's words).

I have read John Cramer's paper, saying Your experiment seems to verify his transactional interpretation of QM (TIQM). I believe that TIQM is one step to pecome a MODEL, not simple an INTERPRETATION. You can view the MODEL I propose via the reference I supply.

I would be happy to know Your (and anybody's else) criticism. Please email me: kurakin.pavel@gmail.com. Thank You.


-----------
P.S. For those who are interested.

I proposed a conjecture that Cramer's transactions can happen in HIDDEN TIME, not equivalent to physical time. One may treat hidden time as true physical entity or as a mathematical object (like wave functions) only, I don't care. I only proposed an formal algorithm, how hidden time must relate to physical time ("sewing procedure").

I unmodestly think that "transactions in hidden time" concept is the only serious PHYSICAL hypothesis of last decades. Not an overcomplicated mathematical trics, which govern quantum physics for last decades, but a PHYSICAL idea. Three basic ideas bring a radically new concept in quantum physics:

1) Feynman's many-path formulation of QM
2) John Cramer's TIQM
3) When amplified by HIDDEN TIME, (1) + (2) make a new PHYSICAL MODEL of quantum phenomena (NOT an interpretation).

Dear Dr Afshar,

I am very interested in interpretations of QM. As soon as I heard about the availability of your preprint, I downloaded it.

I am having difficulties understanding the aim of your experiment. Only your Figure 1 shows a classical interference setup between the two pinholes, and the scan of the field intensity, essentially at infinity, is the classical interference pattern.

In Figs 2 and 7, the scans of field intensity are not taken at infinity, but the other side of a lens, in the plane of the image of the two pinholes. Hence a full-blown classical interference pattern with pronounced minima is not to be expected.

In media reports of your work, emphasis has been placed on the lack of difference between Figures 2 and 7. Your wire grid is well clear of the plane of the pinholes or of the image, so it is totally out of focus, and it would not be expected to lead to large differences in the shape of the profile in the two cases. However, on close examination, there are differences in the shoulders of the images of the two pinholes. You are working close to the diffraction limit, and I would expect that the wires would lead to increased oscillatory behavior of the sidelobes of the resolving function. This is quite consistent with what you have observed.

In fact you appear to agree that your results are in perfect accord with the quantum formalism. This is just what I might expect. The analysis of your set-up is a lot simpler than the calculation of the Lamb-Rutherford splitting or of the g-factor of the electron.

Bohr is now dead; he set the quantum revolution in motion but was not totally omniscient. His simplistic version of quantum theory is now just of historical interest. By all accounts he was a bit of a mystic and enjoyed the blockage of thought his complementarity principle led to. The sound of one hand clapping. Now, half a century later, even though we still don't understand completely what QM means, we have a lot more experimental results to draw on. Bohr thought that single atoms were not "things" and could not be directly observed, but now we have a number of techniques at our disposal that allow us to do just that. It seems to me that you are setting up a straw man and knocking it down. Is there anyone out there who seriously supports the narrow interpretation of Bohr's which-way argument? Any serious physicist would do the calculations and predict the result which you have observed. There are much more important aspects of the Copenhagen interpretation to be clarified, in particular the existence of wavefunction collapse, and the questions of non-locality and causality.

Cheers,

Zigoteau.

Dear Pavel,

The primary goal of Science is the discovery and categorization of Natural phenamena, discovery of interconnectedness of such phenomena, and testing of hypotheses that predict a certain relationship between these phenomena. Therefore, as long as a theory is successful in providing the correct answer for a given initial condition, the theory is a successful model. Modern physical theories have mathematical structures, and are precise within the confines of the mathematical limits. As to why a certain set of mathematical relations happens to have a physical manifestation, one has to resort to philosophy, or even metaphysics. As an example, Newton's law of gravity was a perfect solution to the apparent backward motion of planets in the line of site of observers on earth, but so were the Ptolemaic epicycles. Newton's ideas were accepted because it explained a lot of other seemingly disparate phenomena like the tides, the orbit of the Moon, falling of an apple, (as well as established evidence for the heliocentric worldview such as the moons of Jupiter and the phases of Venus, etc.) But when ultimately Newton was forced to give an answer as to what lies behind his theory, and what "causes" gravitation, he avoided that question and said "I propose laws, and cannot be bothered with hypotheses." Einstein answered that question partially, and we ended up with a powerful new model the GR.

So, as an experimentalist, I only wish to test validity of most commonly agreed upon theories and concepts to advance science. As a theoretician with a philosophical bent, I enjoy trying to answer question like why is there inertia (e.g. see my paper: https://commerce.aip.org/jsp/Lookup.jsp?item=APCPCS000458000001001033000001&src=PFtop ), or why does QM work. As more and more phenomena are explained better and better, the role of physicists as Natural Philosophers becomes critical in finding out the answers to these "hidden" mechanisms, and I wish to do my little part in that realm as well.

I will visit your site and get back to you soon.

Regards.

Dear Zigoteau,

"...you are setting up a straw man and knocking it down."

This "straw man" has scared generations of physicists away from asking some of the most fundamental questions about QM long before I was even born, so I can hardly be blamed for his appearance. Bohr's Complementarity is the rite of passage for new initiates, and almost all textbooks on QM mention it. Therefore, it was essential that an experiment like mine clearly demonstrate that Complementarity is not a part of the QM formalism and is in fact incorrect. My experiment validates the QM formalism and invalidates Bohr's Complementarity.

The question of wavefunction collapse will be investigated in the next set of experiments I am currently preparing.

Regards.

Dear Shahriar,

Thanks for your response. The relevance of what you have achieved in your experiment is now sinking in gradually.

In a way, it is related to Hund's paradox, to which I have given a certain amount of thought recently. That's the Hund of Hund's rule for the spin occupation of degenerate orbitals, amongst others. Anyway Hund pointed out way back in the 1920s (I think) that a chiral molecule cannot possibly be an eigenstate of the Hamiltonian, which essentially commutes with reflection in any plane (coupling via the weak interaction with nuclear spins is negligible for organic compounds and does not invalidate the paradox). The true eigenstate must be a superposition of the two enantiomers, so that according to CI, observation of chirality should lead to immediate racemization. Experimentally, it does not.

Good luck and best regards,

Zigoteau.

Dear Prof. Afshar!

Thank You so much for Your reply.

1) I would be happy to get aquainted with Your electromagnetic theory of inertia, because I believe that in SOME SENSE such theories must be valid. As A. Polyakov, a distinguished specialist in superstrings pointed, "any untrivial idea is valid in some sense."

But unfortunately, Your paper on is not free-- and online-- accessible at the URL You pointed.

2) Yes, I do agree that laws of physics are laws of physics, regardless how easily they are viewed and conceived. And laws (equations) are not obliged to seem "explaining" and easily imaginable. As Heisenberg once noticed, for ancient Egyptians a river flowing NOT in meridianal direction was unimaginable.

Still the science, I believe, does move to more and more clarity, simplicity and visuality in its formulations. Say, Feynman's many-paths formulation of QM is VERY visual and conceivable, compared to Heiseberg's formulation in matrices.

Please note, that A. Einstein's parer "Zur elektrodynamik bewegter korper" DID NOT provide "pridictions, distinguishable from current knowledge" and thus today's "Physical Review" would NOT accept that paper for publishing.

What Einstein's classical paper DID supply, was a new kind of logical ordering of previous knowledge, and nothing more. So, logical unity and clarity (including, if possible, visuality), I believe, has scientific validity in its own.

3) Dear Prof. Afshar, please do believe I imagine a huge busyness of a scientist like You. I kindly ask You not to be too polite. If You are not interested in a new model, take it easy.

But if You do move towards my proposal, I wanna move towards too, and I wanna propose You significantly shorter paper I proposed at first:

http://time2050.narod.ru/texts/dettime.html

It is the heart of our approach, currently developed at Keldysh Institute of Applied Mathematics, Russian Academy of Sciences.

Thank You for Your attention and sorry for trouble.

Dear Dr. Afshar,

Thank you for your reply regarding intermediate values of V and K. I suppose I'll go ahead and post my thoughts on the issue that prompted my question. I have been contemplating how V and K might vary as functions of the thickness t of the wires: V(t), K(t). If t = 0, then this is basically the same as no wires at all, so V(0) = 0 and K(0) = 1. Following the reasoning of your paper, if we set t to some arbitrarily small but nonzero value, then K(t) remains (approximately?) 1, whereas V(t) jumps from 0 to 1. This change in V(t) from 0 to 1 appears to indicate a discontinuity in V(t) at t = 0. Do you agree with this line of reasoning? Do you think there is a theoretical lower limit to the size of the wires? What if they are no thicker than an atom? If you think there is no lower bound, then how would you justify this sudden "jump" in V(t)?

One further thought. On page 9 you make the argument that "all we need to establish perfect visibility is to determine I_min = 0." Are you arguing that it is sufficient to show that I_min is less than some very small value epsilon, or -- alternatively -- does your argument rely on showing that I_min equals zero *exactly*? (I suppose that the former is sufficient, *if* it is possible to calculate a lower bound on I_max based on the total flux at the detector, in which case all you need to observe is that epsilon in which all you need to observe is that epsilon is much much less than the lower bound of I_max, therefore (by eqn 2 of your paper) V is very close to 1.

Regards,

David Strayhorn

Dear David,

You are correct. The physical limit is that the wires must be “opaque”, i.e. their thickness must be at least ~5 lambdas. Also, note that there are no real zeros in experimental physics, and the "epsilon" you mentioned depends on the margin of error (MOE) of your photo-detection process (for a given wire thickness, the higher the signal/noise ratio, the lower the epsilon). So essentially, the upper bound for the thickness is determined by the MOE, and the lower bound is determined by the opacity requirements.

Regards.

Dear Pavel,

Please give me a couple of months for things to slow down a bit. By then I will have had a chance to think about your approach. But I genrally agree with you that ultimately the TOE will be a readily visualizable theory, so obvious that most people will understand its foundational concept(s).

Regards.

Dear Shahriar ,
So many comments during this week end!!!
(who is this zigoteau ???? I like his name a lot probably french I suppose )
I will just concentrate on our discussion.

You justify your point of view in such a way that I can only agree with you: the unical known quantum process which can explain the presence of zero on the wires is fringes. This is clear and I never claimed that the formalism is wrong. If you put your detector in the interference plane you will see for sure these fringes there is no doubt . But Bohr was the most prudent physicist of his time and he prefered to say that when you close your ''eyes'' you are sure of nothing in particular in quantum mechanic. It was for him the unical way to solve the wave particle duality. I like particulary this sentence of him:

''since it is only the circumstance that we are presented with a choice of either tracing the path of a particle or observing interference effects, which allows us to escape from the paradoxical necessity of concluding that the behaviour of an electron or a photon should depend on the presence of a slit in the diaphragm through which it could be proved not to pass. We have here to do with a typical example of how the complementary phenomena appear under mutually exclusive experimental arrangements and are just faced with the impossibility, in the analysis of quantum effects, of drawing any sharp separation between an independent behaviour of atomic objects and their interaction with the measuring instruments which serve to define the conditions under which the phenomena occur.''
Bohr [1949]

On this point he was unfortunatelly wrong because hidden variable models exist which predict with succes all the effect of the quantum universe (see de Broglie and Bohm with some problems for the relativistic model , and the many world model of everett that i interprettas a hidden variabel theory where the the the dynamical value of the system ar just the observables of the formalism) and in particular the wave particle duality. However his point of view concerning complementarity is absolutely non refutable as well that we accept the validity of quantum mechanic.

You can said that there is fringes or that there is 2 spots but YOU will never be able to USE THE SAME PHOTON in the two experiments.

You never effecvelly reconstruct experimentally the fringes point per point ( your prefer to deduce them on the basis of the theory )so i am right when I say that your fringes are metaphysical and that the unical part of the firnges which is experimentaly real is the one absorbed by the wires ..... and i am right when i said that your reasoning is circular.


with best regards


Aurelien Drezet

Dear Shahriar,

"this experiment is far too important to be dismissed by patently erroneous arguments. Simply repeating a false claim
does not make it true!"

I agree. One can't use forcefulness in a scientific argument. One has to address the issues with clarity and an open mind.

"No! The two wavefunctions are NOT identical (regardless of their spatial separations.) Crucially, they have a different ANGLE of INCIDENCE. That angular separation is
what allows the lens to form two distinct images of the sources."

I don't need to read an optics text book, thanks. The optics I learnt at high school is sufficient for this problem: Regardless of the presence or absence of an angular difference between the two wavefunctions, the lens will focus the light from each pinhole to a separate point. That's what lenses do, map one point onto another.

"But to quickly remind you of why you are wrong, think of two distant stars with a small angular separation.
The physical lateral separation of the stars may be many hundreds of light years, but the lens could care less about that spatial separation, because all it does is to separate
the two sources based on the slight difference in the angles of incidence of the two beams. To think otherwise simply demonstrates lack of knowledge in the field."

I think I see the point of confusion. The lens discriminates the angular separation of the stars. The angle in question being defined by star1--centre of lens--star2. In your experiment the angle is pinhole1--centre of lens--pinhole2. The only difference between the two cases is that a star is (usually) an isotropic emitter of radiation whereas a pinhole is not. A pinhole emits more strongly in some directions than others: it is not these directions that the lens is discriminating. I believe that when a pinhole in a plane is illuminated from behind, most of the light is emitted perpendicular to the plane in something called an Airy disc (there are also higher order diffraction rings at other angles). Hence my assertion that the wavefunctions from the pinholes should be identical apart from a lateral translation.

As I read your paper you have a plane wave impinging upon a pair of pinholes in a flat barrier. The wavefunction from one will be identical to the other, apart from a translation. Perhaps you have curved the barrier such that the two wavefunctions come out at different angles, but this is not stated in the paper. Also, it seems unlikely because eq 7 in your paper can only have been obtained by making the assumption that the two wavefunctions are essentially the same (ie same amplitudes but different phases). This is the exact same assumption that you have been disputing.

Anyway, the presence or absence of an angular difference is also entirely irrelevant to my argument which you still have not addressed.

Morgan.

Hi! I'll respond as soon as I have the time. Meanwhile, please feel free to talk amongst yourselves...--S

Aurelien Drezet writes:
'You can said that there is fringes or that there is 2 spots
but YOU will never be able to USE THE SAME PHOTON
in the two experiments. '

I would say that it is possible that a photon is a carrier of a limited amount
of information, say 1 quantum of information, say 1 bit. So it is difficult
indeed to get from a photon (= a quantum of information) information
about *both* non-commuting (sharp) observables, since this information
is expected to require more than 1 bit. It is, of course, possible to get, from
a photon, information about *both* non-commuting observables, but this
information is a probabilistic one (Greenberger, YaSin, Englert, Zeilinger,
Wootters, Zurek, etc.).

Now there are two points.

Afshar's 'little wires' give us direct information about the absence (nearly) of photons, in those points. This is not a *direct* information of the presence ('visibility') of 'fringes' elsewhere. Nevertheless it is *indeed* information. More or less, it appears like the 'negative measurement' information gained by Renninger in his gedanken experiment. (Does a 'negative measurement'
collapse, according to the orthodox QM, a superposition of states?).

The second (naive!) point is this. What would happen if we use, instead of photons, entangled photons (bi-photons)? A bi-photon, because of non-separability, appears to be - according to orthodox QM - a single, unique quantum system (with a different de Broglie wavelenght I can imagine). But it is difficult, for me, to believe that also a bi-photon is a carrier of such a limited information (= 1 bit). Thus, it seems to me that *in principle* a bi-photon
(or a multi-photon!) could provide *more* information about *both* non-commuting (sharp) observables. (I'm not thinking here necessarily that it is possible to measure 'fringes' with one entangled photon and 'which way' with the other, since I know there is a strict correlation of of both momenta, in entangled pairs).

But it is too late now. And I've written enough stupidities.

Au revoire,
- serafino

Dear Morgan , I imagine that S. Afshar will give you an other argument but it seems that I can explain in ''few ''worlds why your are wrong. This just come from the linearity of the wave equations and it doesn't realy matter that you think with electromagnetic wave or with scalar wave .

When you close one pinhole (say 2) practically all the light goes to 1' . Except that you have some few percent of light diffracted by the wires but this is far of 50 percent. Similarly if you open 2 and close 1 you will see 2'. Now the linearity of the theory tell you that if you open the 2 holes you will have a linear superposition of the wave at every point of space. You like the Huygens-Fresnel principle so think that in reality you can apply this principle to the total wave by summing the cases of the two single pinhole experiment. Then clearly the surface integral made on the sum can be decomposed in two part which separate completely the evolution of the two waves.

Now if you look in the detail: if you want to treat rigourously the problem of the lens you need at least the Fresnel approximation this means that all will come from the difference in the phase taking into account not only plane waves but parabolic waves too . The amplitude supposed constant will play no role in the paraxial approximation and all the separation of the two waves will comes naturally from an integration of the phase factor (I ensure I made the calculation).

in other terms you can not mathematically says that only because the amplitude is constant that the law ''50/50'' is valid (an integration of a phase factor can generate all the effect).

The real problem is not when you think into wave but when you think into particle. A simple model of photon can be constructed by imagining that the photon follow the line of energy flow . The photon path will depend of the densitiy of energy which means on the square of the field (more exactly the square of his norm) . then the principle of supperüpostion is broken for the trajectory of the photon: The photons coming from 1 will go to 2' and the ones coming from 2 will go to 1'.

I t is not a good idea to create a cocktail with the two representationsof the phenomena....


with best regards

(PS:as you ''few'' is a relative notion)
Aurelien

To serafino ( i am tired too : probably we hare crazy in Europa)
Your argument concerning information is pertinent and I never say that we have no information at all. If the wire are absorbing the surviving photon plus the absorbed one will give us the initial flow . However even in the Renninger or in the Dicke experiment we can not violate complementarity : the information on the total number of particle will not give you an information on the shape of the wave function. To have such information you need a precise measurment which in the configuration of afshar is not compatible with the detection of the two images.

Anton Zeilinger (and others) wrote a lot concerning this problem and a lot too concerning the entanglement problem : You can not have all the information with entanglement too and the complementarity will be still respected for sharp or not so sharp measurements.


Ok tomorrow is another day


Tchao

Aurelien

Dr Afshar,

I was wondering if you have any views on how your experiments sit with David Bohm's hidden variables/implicate order theory?

Please be "gentle" - I'm a quantum outsider (i.e. non-physicist)

kind regards and congratulations - a wondrous and important debate rages.

Peter FYFE
Playwright
Sydney, Australia

Mr Afshar

Congratulations on creating such a stir on the philosophy and interpretation of QM. I have been following this field for 15 years now and although we allmust have read about the double slit (and variations) experiment 100 times in 100 different books I have only seen one other experiment that shows waves and particle behaviour at the same time. The other experiment is described in John Gribbins (the laymans best friend in QM) book entitled "in search of schrdingers kittens" which describes a experiment by which particles at the start of the experiement are subsequantly seem exhibiting wave like behaviour (quantum tunneling effect as only waves can tunnel or so I have been told) through a xystal placed in one of the which way paths with a narrow air gap between them. No one knows what the implications are for quantum phyics but I agree that the Copanhagen interpretation is and always was on shaky ground.

Let us hope that your paper is peer reviewed successfully and no flaw in your arguments is found.

"Simon Hopkins" wrote in message news:...
> Ok I'm an amateur just beginning to train myself in QMs so what I am about
> to say is probably going to sound quite dumb. Surely Afshar's experiment is
> simply measuring properties of different photons at each stage of the
> experiment - those that pass prior to the lens, those after the lens and
> those ejected from atoms in the mirror. So in this sense Afshar is not
> really measuring complementarity of the same "object". Only if probability
> functions on the same photon pass though the lens can true complementarity
> be assured.
>
> I have a crude thought experiment to answer this question. We place a lens
> between the slits and the detector and pass photons one at a time through
> the experiment. The closer the lens is placed to the slits the smaller will
> be the overlap of the two wavefunctions. If the probability functions are
> truely refracted then we would expect an interference pattern no matter how
> close to the slits we place the lens. However, I expect that the photon
> would be we absorbed by an atom in the lens. At this stage the multiple path
> info would be lost (wavefunctions collapse and new ones are created). Thus a
> photon emerging from the lens would not"know" about the slit arrangement so
> no intereference pattern would be seen (instead we see two clusters).
> Simon.

Contrary to stereotyped physical interpretation of an interference
phenomenon of a light on two slots:

The experimental fact of existence of a "virtual interference"
in VLBI basically excludes ANY POSSIBILITY of simultaneous
passing of a photon through both slots (separate magnetic tape
for each antenna/slot). There is no ANY POSSIBILITY
in VLBI for a photon to pass simultaneously through both
slots/antennas at all!!!

============
The events happening on _slots of an interferometer have
primary significance, all _other _events happening in an
interferometer have the status secondary.
============

In a case of VLBI an interferometer, we have:

The absence of influence of a state of an electromagnetic field
in space of one slot (antenna) on a state in other one becomes
perfect obvious, since a limit of a distance between slots
(antennas) experimentally is not reached, and this distance can be
made _physically vast_ on a comparison with a wavelength.
(Earth diametr or many more)

See additional INFO:
http://groups.google.com/groups?selm=e16a4a22.0409210207.2c8d0f34%40posting.google.com

The events happening on _slots of an interferometer have
primary significance, all _other _events happening in an
interferometer have the status secondary.
The experimental fact of existence of a "virtual interference"
in VLBI refutes physical interpretation of an interference on two
slots of "particle - photon", which one you have described ABOVE!!!

The experimental fact of existence of a "virtual interference"
in VLBI basically excludes ANY POSSIBILITY of simultaneous
passing of a photon through both slots (separate magnetic tape
for each antenna/slot). There is no ANY POSSIBILITY
in VLBI for a photon to pass simultaneously through both
slots/antennas at all!!!

The VLBI interferometer is an interferometer
with independent registration of signals in shoulders and the
process of summation of signals is carried out in the computer.
The phrase " process of summation of signals is carried out in
the computer " allows clearly to seize essence " concepts of an
interference pattern " and source of an origin of this concept.
In the given type of an interferometer there is some
arbitrariness in choi