# Two experiments on the Wave-Particle duality

Gold Member

## Main Question or Discussion Point

Two new papers on beautiful experiments were published this month
concerning the Wave Particle duality of the photon:

Paradox in Wave-Particle Duality
S.Afshar et. Al. in Foundations of Physics.

Experimental realization of Wheeler’s delayed-choice
GedankenExperiment

J-F. Roch, A. Aspect, P.Grangier in Science.
http://arxiv.org/PS_cache/quant-ph/pdf/0610/0610241.pdf [Broken]

What these experiments have in common is that they seem to produce
exactly as one would expect from classical optics / EM radiation theory,
and both do so one photon at a time. An effect which continues to
puzzle physicist.

In the above sense these experiments do not bring anything new but
the discussion is all about the interpretation of this effect and more
concrete: The validity of Bohr’s Principle of Complementarity:

Niels Bohr said:
A single quantum mechanical entity (= photon, electron...) can either
behave as a particle or as wave, but never simultaneously as both.

http://en.wikipedia.org/wiki/Complementarity_(physics)
So, Bohr said in 1927 that it's either wave or particle but never both
at the same time. At this time this was a change of mind for Bohr who
had opposed Einstein's idea of the photon as a quantum particle for
more than a decade:

http://arxiv.org/PS_cache/physics/pdf/0212/0212090.pdf [Broken]

The two papers should be seen in the light of this. The authors make
diametrically opposed claims:

The authors of the first experiment claim to show that the photon can
have simultaneous wave and particle properties and disagree with Bohr.

The authors of the second experiment hold on to Bohr and claim that
their experiment demonstrates that "In the present, one can change
something that has already happened in the past" Namely the decision
of the photon to behave as a particle, or, as a wave.

All authors agree that Bohr's principle of complementarity disagrees
with Einstein's ideas and his work on relativity. Both groups quote
Wheeler as if he agrees with their (opposite) claims in the conclusion
of their papers.

Regards, Hans

Last edited by a moderator:

## Answers and Replies

Related Quantum Physics News on Phys.org
ZapperZ
Staff Emeritus
Last edited by a moderator:
Our realization of Wheeler’s delayedchoice GedankenExperiment demonstrates beyond any doubt that the behavior of the photon in the interferometer depends on the choice of the observable which is measured, even when that choice is made at a position and a time such that it is separated from the entrance of the photon in the interferometer by a space-like interval.
http://arxiv.org/PS_cache/quant-ph/pdf/0610/0610241.pdf [Broken]
Does this mean we can influence the past?

Last edited by a moderator:
Gold Member
Does this mean we can influence the past?
Well, That's a claim the authors make.

"We, now, by moving the mirror in or out have an unavoidable effect on
what we have a right to say about the already past history of that photon".
(authors quoting Wheeler)
I would reason that, in theory, one could expand their claim into the extreme
with polarized Cosmic Background Radiation. The same experimental setup
would then be able to change certain events at the time of the Big Bang.

Now do we want to believe this? It's really all about the general validness of
Niels Bohr's 1927 principle of Complementarity. If it's not strictly valid then
the extraordinary claims go away and the universe "returns normal".

Without this principle the photon does not need to make a choice of acting
either like a wave or as a particle, nor does it need to revise this choice
somewhere in the future depending on some observer.

In my opinion, people like Aspect and Grangier may well contribute more to
the diminish of the strict interpretation of the principle of complementarity,
by adhering to these extraordinary claims, than Afshar who's campaigning
directly against it with his experiments.

This, for the people following the field, would be somewhat amusing from
an historical viewpoint.

Regards, Hans

Last edited:
Two new papers on beautiful experiments were published this month
concerning the Wave Particle duality of the photon:

Paradox in Wave-Particle Duality
S.Afshar et. Al. in Foundations of Physics.

Experimental realization of Wheeler’s delayed-choice
GedankenExperiment

J-F. Roch, A. Aspect, P.Grangier in Science.
http://arxiv.org/PS_cache/quant-ph/pdf/0610/0610241.pdf [Broken]

I draw your attention that the date of receive is Dec.2005 http://www.springerlink.com/content/q110r82074w03277/fulltext.pdf
and the date of publish is Jan. 2007!

Last edited by a moderator:
DrChinese
Gold Member
Does this mean we can influence the past?
You cannot influence an observation made in the past (or at least you can't prove you have).

DrChinese
Gold Member
In my opinion, people like Aspect and Grangier may well contribute more to the diminish of the strict interpretation of the principle of complementarity, by adhering to these extraordinary claims, than Afshar who's campaigning directly against it with his experiments.
Extraordinary? The results seem (to me) to be exactly in keeping with what you would expect from QM. Which is weird enough... and of course that is why it is such a cool experiment. Thanks for the reference.

Last edited:
DrChinese
Gold Member
1. I think that Afshar is overstepping the bounds of the experimental results in his conclusions. He says:

"...the applied technique appears to allow us to circumvent the limitations imposed by Heisenberg’s uncertainty principle..."

However, nowhere (that I see) is the HUP shown to be violated. We would need something like the following to make a convincing case:

$$\Delta p \Delta q < \hbar$$

2. Also, as to the principle of complementarity (BPC in the article) itself: nothing says that you cannot measure something "a little" as a particle and "a little" as a wave. Clearly, the referenced experiment purports to show this within error constraints that are not particularly strict. I would have expected a conclusion like "violated by 3 standard deviations" or similar if the point is to be convincing.

3. At this point, I assume the publishing was allowed to go forward because the results themselves are reproducible. I do not doubt that there will be debate on what the results themselves say. I think the quote from Wheeler (". . . for quantum theory to say in one breath ‘through which slit’ and in another ‘through both’ is logically inconsistent...") can be construed many ways. I certainly don't see this experiment as particularly supporting it.

Last edited:
many worlds?

Gold Member
Part I

1. I think that Afshar is overstepping the bounds of the experimental results in his conclusions. He says:
"...the applied technique appears to allow us to circumvent the limitations imposed by Heisenberg’s uncertainty principle..."

However, nowhere (that I see) is the HUP shown to be violated. We would need something like the following to make a convincing case:
$$\Delta p \Delta q < \hbar$$
The experimental result is entirely explained by Classical Optics/
EM Radiation, There's no need to use $\hbar$ or Heisenberg's uncertainty
principle.

"Circumventing" Heisenberg’s uncertainty principle here means that
HUP is simply not applicable to the experiment. It doesn't imply that he
claims that he found some way to change $\Delta p \Delta q \geq \hbar$ to $\Delta p \Delta q < \hbar$.

He could (should) have been clearer, certainly.

Regards, Hans

Last edited:
Gold Member
Part II

2. Also, as to the principle of complementarity (BPC in the article) itself: nothing says that you cannot measure something "a little" as a particle and "a little" as a wave. Clearly, the referenced experiment purports to show this within error constraints that are not particularly strict. I would have expected a conclusion like "violated by 3 standard deviations" or similar if the point is to be convincing.
Niels Bohr's principle of complementarity is just a subset of the
Copenhagen interpretation of Quantum Mechanics. So, the
probability aspect of the interpretation may turn out to be right
while BPC may turn out not.

BPC says that a single photon can only behave either as a wave
or as a particle, but not both a the same. It also can't be "half"
particle, "half" wave. Other interpretations make other claims.
For example the Bohmian QM interpretation assumes that a photon
is both a particle and a guiding wave at the same time.

So for me we're talking about a QM interpretation issue here.

Regards, Hans.

Last edited:
Gold Member
Part III

3. At this point, I assume the publishing was allowed to go forward because the results themselves are reproducible. I do not doubt that there will be debate on what the results themselves say. I think the quote from Wheeler (". . . for quantum theory to say in one breath ‘through which slit’ and in another ‘through both’ is logically inconsistent...") can be construed many ways. I certainly don't see this experiment as particularly supporting it.

A good reviewer looks at the merits of the experiment itself.

I think the reviewer has been convinced that there must be an
interference pattern and that the light through the two holes is
focused on two different locations. I'm convinced as well. It's
just what you expect from classical optics. I would have liked
to see the figure on page 299 earlier. It nicely addresses the
diffraction grating ideas Vanesch and I discussed here:

Now is BPC (Bohr’s principle of Complementarity) in conflict
with Classical Optics?

BPC says that a single photon can only behave either as a wave
or as a particle, but not both a the same time. Classical Optics
says that photons ALWAYS behave as waves, (until they're
finally absorbed) So yes, I think there is conflict indeed.

For me the essential point is that the wave can go anywhere,
follows all possible paths, all over the place, while the absorption
only occurs at one single place.

Now which path was followed by that what caused the absorption?
(that what we call the "particle") Why does the remainder of the
wave lead NOT to an absorption elsewhere? What happens at all
with the remainder of the wave? This is the mystery of Unitariy.

BPC wants to remove the wave from all paths that were unlikely
to be followed by that what caused the absorption. It suggests
a partial remedy for the "Collapse of the wave function":

(1) If a photon is split by a beam splitter and detected Left and
not Right then BPC says that nothing went Right, neither particle,
nor wave. So we don’t need to explain what happened to the
wave at the Right side.

(2) If a photon is split by a beam splitter and both sides are made
to interfere then BPC says that the photon went both ways, So in
this case the wave follows both paths.

Now, In case of the experiment under discussion, BPC would,
when recognizing the detection at Left, want to remove everything
from the other path like in (1) However, this would also remove
the interference and thus the result of the experiment.

So BPC is in conflict with Classical Optics and thus with Bohr’s
other principle: That of Correspondence.

Bohr’s Principle of Correspondence (1923)
Bohr’s Principle of Complementarity (1927)

http://en.wikipedia.org/wiki/Correspondence_principle
http://en.wikipedia.org/wiki/Complementarity_(physics)

Bohr’s Principle of Correspondence states that if the light consist
out of sufficient numbers of photons we should get the Classical
Optics result back. No quantum mechanical interactions took place
other than the final absorption in the detectors.

Both the experiments discussed on this thread produce the results
that would be expected from Classical Optics.

Regards, Hans

Last edited:
DrChinese
Gold Member
The experimental result is entirely explained by Classical Optics/
EM Radiation, There's no need to use $\hbar$ or Heisenberg's uncertainty
principle.

"Circumventing" Heisenberg’s uncertainty principle here means that
HUP is simply not applicable to the experiment. It doesn't imply that he
claims that he found some way to change $\Delta p \Delta q \geq \hbar$ to $\Delta p \Delta q < \hbar$.

He could (should) have been clearer, certainly.

Regards, Hans
Well, I don't see how one can say "HUP doesn't apply" and "limits of the HUP were exceed" both without demonstrating the violation. I would agree that their language about the HUP could be dropped altogether and it would be more accurate.

DrChinese
Gold Member
Niels Bohr's principle of complementarity is just a subset of the
Copenhagen interpretation of Quantum Mechanics. So, the
probability aspect of the interpretation may turn out to be right
while BPC may turn out not.

BPC says that a single photon can only behave either as a wave
or as a particle, but not both a the same. It also can't be "half"
particle, "half" wave. Other interpretations make other claims.
For example the Bohmian QM interpretation assumes that a photon
is both a particle and a guiding wave at the same time.

So for me we're talking about a QM interpretation issue here.

Regards, Hans.
I don't see the BPC as fundamental anyway. In my view: the HUP is more precise, while BPC is more of an approximation. The HUP says that an observation CAN be both wave-like and particle-like as long as the $$\Delta p \Delta q \geq \hbar$$ relation is satisfied. It is certainly possible to construct a setup in which there is some interference and some which-way information, but not enough of either to be absolutely certain in specific cases. As best I can tell, this is what the experiment is doing.

But I am having a little difficulty understanding their argument anyway. Do you understand it well enough to help me? Specifically, I am trying to understand the 4 diagrams a/b/c/d and how they get to the idea that the wave nature is being demonstrated. I follow that the wire is placed at minima of the expected wave pattern (going from a to b) and it is only slightly destructive. But how does that provide a convincing argument for the wave nature being observed?

Gold Member
Specifically, I am trying to understand the 4 diagrams a/b/c/d and how they get to the idea that the wave nature is being demonstrated. I follow that the wire is placed at minima of the expected wave pattern (going from a to b) and it is only slightly destructive. But how does that provide a convincing argument for the wave nature being observed?

One can discuss these 4 diagrams from a Classical Optics point of view:

The bottom two diagram show one hole closed, one hole open. The
wave interacts with the thin wires. These will scatter the wave and
produce a diffraction pattern. (The series of little bumps in the output)

If both holes are open then there will be an interference pattern with
the wires carefully placed in the lows of the pattern. The wires don't
scatter the wave and no diffraction pattern will occur. Most of the
energy will get through unscattered.

So it's the absence of the diffraction pattern in the output which
demonstrates best that there must be interference. To see more how
diffraction patterns from an array of thin lines (or thin slits) should look,
go here:

http://www.msm.cam.ac.uk/doitpoms/tlplib/diffraction/convolution.php [Broken]

Regards, Hans

#### Attachments

• 45.4 KB Views: 262
Last edited by a moderator:
Two new papers on beautiful experiments were published this month concerning the Wave Particle duality of the photon:

Paradox in Wave-Particle Duality
S.Afshar et. Al. in Foundations of Physics.

Experimental realization of Wheeler’s delayed-choice
GedankenExperiment

J-F. Roch, A. Aspect, P.Grangier in Science.
http://arxiv.org/PS_cache/quant-ph/pdf/0610/0610241.pdf [Broken]

What these experiments have in common is that they seem to produce
exactly as one would expect from classical optics / EM radiation theory,
and both do so one photon at a time. In the above sense these experiments do not bring anything new. The authors make diametrically opposed claims:

The authors of the first experiment claim to show that the photon can
have simultaneous wave and particle properties and disagree with Bohr.

The authors of the second experiment hold on to Bohr and claim that
their experiment demonstrates that "In the present, one can change
something that has already happened in the past".

Both the experiments discussed on this thread produce the results
that would be expected from Classical Optics.
1.S.Afshar et. al paper is not new. If the presented result is correct, the entire reformulation of QT is required. Since the Classical Optics (Electrodynamics) is in compliance with the standard QT (P.A.M. Dirac), it means that the entire reformulation of classical electrodynamics is required in addition.

2.I consider the A. Aspect et al paper the outstanding experimental achievement. The first time the single photon real life realization of Wheeler’s delayed-choice experiment is performed. However, the “explanation” have nothing to do with the obtained result.

Regards,Dany.

P.S. DrChinese, sorry in advance for the “demons”.

Last edited by a moderator:
DrChinese
Gold Member
If both holes are open then there will be an interference pattern with
the wires carefully placed in the lows of the pattern. The wires don't
scatter the wave and no diffraction pattern will occur. Most of the
energy will get through unscattered.

So it's the absence of the diffraction pattern in the output which
demonstrates best that there must be interference. To see more how
diffraction patterns from an array of thin lines (or thin slits) should look,
go here:
Thanks, that helps. So here is my next question. I am guessing that they are asserting that the path in the B picture is left to left and right to right, and they base that on the idea that the B is "built up" from the C and D pictures. So that is how they know the "which way" information. Is that correct?

DrChinese
Gold Member
P.S. DrChinese, sorry in advance for the “demons”.
OK, I guess an apology is in order. :uhh: You obviously were not trying to criticize professionals in general, which is what I had thought from my (incorrect) reading of your earlier post.

I don't see how QT needs to be reformulated if the Afshar experiment stands. One would first need to demonstrate that the Heisenberg Uncertainty Principle is violated - I think. How otherwise is the BPC included in the formalism of QT?

I don't see how QT needs to be reformulated if the Afshar experiment stands. One would first need to demonstrate that the Heisenberg Uncertainty Principle is violated - I think. How otherwise is the BPC included in the formalism of QT?
I do not identify the difference between your and my arguments. May be only that you use “q-bits” for your arguments and I use “c-bits”. I refer to the QM and Classical Physics formalism without interpretation. I describe the N.Bohr and A. Einstein statements as following:

1. The measurement equipment are the macroscopic devices; therefore, they obey the laws of Classical Physics.
2. The Classical Physics is the theory of certain events.
3. The Classical Electrodynamics require a special relativity space-time geometry.
4. The Quantum Theory is the theory of uncertain events (in general, every object is the extended object: HUR).
5. The results of measurements are given in terms of the eigenvalues (spectrum) of a complete set of mutually commuting self-adjoint operators (observables).

The special relativity require the collapse of wave packet in time (delta t=0, instant) when the measurement performed (A. Einstein, 5-th Solvay conference). BPC require the collapse in space (delta x=0, delta p=0, particle) when the measurement performed. Unitarity require the number of particles to be a conserved quantity. In both experiments above N=1. The operator of number of particles in that case has two eigenvalues 0 or 1 (“which way”). The beam-splitter with the counter is the macroscopic measurement device. The beam-splitter without the counter maintain the HUR delta N>0 (selfinterference).

Check my statements above. If you agree that they provide the adequate verbal description of CED and QED formalism then S.S. Afshar result inconsistent with it.

Regards, Dany.

P.S. Perhaps, our wording lead to confusion: everybody agree that without the measurement the unitary evolution of the QM system is deterministic. Through the unitary evolution the Heisenberg Uncertainty Relation holds with certainty. It is better to say: Heisenberg Dispersion Relation.

Last edited:
DrChinese
Gold Member
Check my statements above. If you agree that they provide the adequate verbal description of CED and QED formalism then S.S. Afshar result inconsistent with it.
The words are probably OK, and may be a good verbal description. But... I would not agree that the BPC is part of the QM formalism. I did a quick check of a couple of older books (Heisenberg, Dirac) and this was not present as a formal step. I will look some more.

I believe this experiment may be a which way test to a certain degree (let's say 90%), and a wave test otherwise (say 10%). I think this would fit nicely with the HUP. It would also be in the ball park of the observed results, I would guess.

I certainly don't think it is both at the same time - although I could be wrong. It reminds me of good magic... we have been misdirected and don't see the real explanation for the results.

Last edited:
I would not agree that the BPC is part of the QM formalism. I did a quick check of a couple of older books (Heisenberg, Dirac) and this was not present as a formal step. I will look some more.
Larmor (1931): “ So perhaps with our friend Bohr: he might want to instruct us about the correlations of too many things at once…”

I believe this experiment may be a which way test to a certain degree (let's say 90%), and a wave test otherwise (say 10%). I think this would fit nicely with the HUP. It would also be in the ball park of the observed results, I would guess.
The collapse is the experimental declaration that the transition from Quantum World to the Classical World took place (E. Schrödinger Cat= HUR).

Together with my son I prepare now the paper where we will try to discuss coherently double-slit, Stern-Gerlach, etc. We will suggest the experimental test of our statements in the single photon/electron setups. If our prediction will turns out to be correct, the result obtained by A. Aspect et al will be also clear.

Regards, Dany.

Last edited:
Gold Member
Thanks, that helps. So here is my next question. I am guessing that they are asserting that the path in the B picture is left to left and right to right, and they base that on the idea that the B is "built up" from the C and D pictures. So that is how they know the "which way" information. Is that correct?
Yes, although one has to be very careful in saying that picture B is "build up"
from C and D. Note that the central peak in C and D is only 85%. The missing
part has been scattered away into the diffraction pattern.

If you do a linear add of C and D then one might cancel the two diffraction
patterns because they have opposite phase add, but, there's nothing to
ADD to the 85% peak of picture C to bring it to 98% except for a little
0.46% bump in D which would be subtractive (?) rather than additive.

So the conclusion would be that the 98% in the peak is very likely to be
as good as all from the corresponding hole.

Regards, Hans

Gold Member
1.S.Afshar et. al paper is not new. If the presented result is correct, the entire reformulation of QT is required. Since the Classical Optics (Electrodynamics) is in compliance with the standard QT (P.A.M. Dirac), it means that the entire reformulation of classical electrodynamics is required in addition.
It's likely the earlier consternation around this experiment but it in
fact produces what one would expect from Classical Optics, see some
of the previous post.

2.I consider the A. Aspect et al paper the outstanding experimental achievement. The first time the single photon real life realization of Wheeler’s delayed-choice experiment is performed. However, the “explanation” have nothing to do with the obtained result.

It should be pointed out that the measured experimental results are as
well explained by Classical Optics, which holds, as usually, also in the ‘single
photon at a time’ case. The experimental setup allows one of two choices
to be made: (see figure 5)

Measurement Choice 1: No interference (<1%).
Light on one path is horizontally polarized, while that on the other path
is vertically polarized, Indeed: Classically, H polarized light does NOT
interfere with V polarized light. (look at the EM components)

Measurement Choice 2: Full Interference (>94%)
Light on both paths are essentially equally polarized (at the Wollaston
prism, both 50% H and 50% V). Classically, equally polarized light
DOES interfere.

The paper however claims that in Choice 1, no interference constitutes
a proof that there is no wave behavior. Instead, the photon is assumed to
behave like a particle, which follows only one of the two paths.

Following Bohr's 1927 Principle of Complementarity, the paper comes
to the claim that by taking either Choice 1 or Choice 2 we can change
the outcome of an event which occurred in the past: The decision of the
photon to behave either as a wave or as a particle which comes down
to the decision to take only one or both paths.

The validness of this claim requires the proof that, in case of Choice 1,
(No interference), the wave behavior of the photon is excluded. This
proof is not given since the absence of interference is in fact predicted
by the wave behavior of polarized light.

Formally there is no proof if the wave/particle associated with the photon
went only one way or both ways. In neither case would there be any
interference.

Regards, Hans

Sorry for the delayed response (I had no choice).

It should be pointed out that the measured experimental results are as well explained by Classical Optics, which holds, as usually, also in the ‘single photon at a time’ case.

Both the experiments discussed on this thread produce the results that would be expected from Classical Optics.”

“Classically, equally polarized light DOES interfere.”
I consider a single photon wave packet a pure QM object. I consider a macroscopic measurement device as a statistical ensemble of microscopic objects. I studied from P.A.M. Dirac that equally polarized light DOES NOT interferes classically. Please define what you mean “Classical Optics”.

Formally there is no proof if the wave/particle associated with the photon went only one way or both ways. In neither case would there be any interference.
???

Similarly to DrChinese I do not understand N.Bohr wording. For me, QT is the local field theory of massive/massless waves. Localities (interactions) represent the particle behavior and field means mathematically as well as physically the extended object which propagates as a wave.

Regards,Dany.

Last edited:
SF
It got on PhysOrg now: http://www.physorg.com/news92937814.html

“Afshar’s experiment consists of the clever idea of putting small absorbing wires at the exact position of the dark interference fringes, where you expect no light,” Knoesel said. “He then observed that the wires do not change the total light intensity, so there are really dark fringes at the position of the wires. That proves that light also behaves as a wave in the same experiment in which it behaves as a particle.”

Last edited by a moderator: