Is Wave-Particle Duality Really Real? An Analysis of the Double Slit Experiment

[inflector]

You might enjoy this discussion as well:

https://www.physicsforums.com/showthread.php?t=304724

I was simply trying to point out to the OP that there are alternative explanations to the SQM.

 

[DrChinese]

Just so there is no misunderstanding, I think someone should point out that even in classical physics, the interference pattern disappears if you have crossed polarizers in front of the alternate slits. There's nothing really mysterious about it.

I am not sure about that actually being the case. What element of a classical explanation leads to that conclusion? Let's consider the wave perspective as being the operative picture, since that has the potential for interference. The source is vertically polarized (0 degrees) and the polarizers are either +45/+45 (Parallel) or +45/-45 (Crossed). You get full interference in the Parallel mode and NO inteference in the Crossed mode. I say any classical explanation will predict SOME interference in the Crossed mode. Now, why do I claim that?

Clearly, the polarizer stops half of the waves coming through. And clearly, the Parallel mode should transmit the light through both polarizers under both the quantum or classical views. But, as with Bell's Theorem, the classical picture adds an extra assumption not present in the quantum view. That assumption being realism. That assumption means that there must be a real value for that wave passing through the polarizers in either the Parallel or Crossed modes - even if it is only tested in a single mode at a time.

A little thought will give you these statistical requirements of such a theory.

a. The usual cos^2(theta) rule - i.e. Malus. This applies to each slit separately, and must apply at other angles than +/-45 degrees such as +/-30 degrees etc.

b. You also must have the prediction of no interference when the polarizers are Crossed. And this must apply at ALL angle settings, not just +45/-45! For example, it must true at +30/-60, +40/-50, ... i.e. whenever the difference is 90 degrees.

There are NO datasets which can meet both of these requirements consistently. You can make it appear to work for +45/-45 alone but then it will not work for +30/-60 etc. I.e. it will not work for all possible values for the polarizers simultaneously. And that violates the realism requirement, which QM does not need to respect.

If there were light waves, the classical picture will end up predicting (once you allow for realism) that sometimes there would be light that would pass through BOTH slits even in the Crossed mode. Such waves would be polarized out of phase coming out of the slits. But I don't believe you would have complete cancellation in that situation. There would be some interference effects leftover. My thinking could be wrong on that, but there is definitely no dataset that meets the requirements a. and b. above, i.e. a classical realistic description for passing through the slits.

 

[DrChinese]

Just to add to my prior post: I think the situation could probably be formulated into an inequality using logic from Bell's Theorem. Basically it is now the same situation, in which the 2 slits are the analog of Alice and Bob.

And of course we are completely disregarding the obvious contradictions in the purely wave picture involved with a classical description. Such as the fact that the photon can only be detected at one slit or the other, and never both as you would expect.

 

[Frame Dragger]

Just to add to my prior post: I think the situation could probably be formulated into an inequality using logic from Bell's Theorem. Basically it is now the same situation, in which the 2 slits are the analog of Alice and Bob.

And of course we are completely disregarding the obvious contradictions in the purely wave picture involved with a classical description. Such as the fact that the photon can only be detected at one slit or the other, and never both as you would expect.

Not to mention experimental evidence showing that while light has wave-like properties, it is subject to the effects of gravity as predicted by GR. There really is a fairly impressive near intersection between QM and GR, which promises to breed something new if they can be reconciled. Classical Mechanics consitantly takes it in the theoretical pants from QM. Who knows, we could get lucky and confirm The Higgs Mechanism through the LHC and end this debate. That would be... refreshing. Then again, Classicists might still find a theory sufficiently contorted so as to be familiar while competing with QM (in the minds of some at least). Such is the case, in my view, with dBB, and the issue we're having here re: the DS experiment!

 

[conway]

If there were light waves, the classical picture will end up predicting (once you allow for realism) that sometimes there would be light that would pass through BOTH slits even in the Crossed mode. Such waves would be polarized out of phase coming out of the slits. But I don't believe you would have complete cancellation in that situation. There would be some interference effects leftover.

There are no interference effects of that kind between light waves that are polarized 90 degrees to each other. Classical EM predicts that the pattern on the screen will simply be the sum of the patterns from the individual slits. The same as QM predicts.

 

[Frame Dragger]

There are no interference effects of that kind between light waves that are polarized 90 degrees to each other. Classical EM predicts that the pattern on the screen will simply be the sum of the patterns from the individual slits. The same as QM predicts.

Cite?

 

[DrChinese]

There are no interference effects of that kind between light waves that are polarized 90 degrees to each other. Classical EM predicts that the pattern on the screen will simply be the sum of the patterns from the individual slits. The same as QM predicts.

As Frame Dragger asks, I too would be interested in a citation - or at the very least a description of a dataset that has the properties I detail above. I don't think there will be any on this kind of experiment other than a quantum explanation. I can imagine that a classicist might think the predictions would be the same, but I just don't see how. I see a lot of hand-waving that realistic theories yield similar predictions to QM, but these often don't hold up when analyzed in detail.

 

[conway]

As Frame Dragger asks, I too would be interested in a citation - or at the very least a description of a dataset that has the properties I detail above. I don't think there will be any on this kind of experiment other than a quantum explanation. I can imagine that a classicist might think the predictions would be the same, but I just don't see how. I see a lot of hand-waving that realistic theories yield similar predictions to QM, but these often don't hold up when analyzed in detail.

I have to confess that I didn't read your scenario carefully enough to be sure that I understood it. I simply replied to some specific things you said in the body of your posts that seemed to go against what I think I know about classical e-m theory. Perhaps I'm not understanding your point at all. Are you really saying that there are experiments you can do with an ordinary laser, a pair of slits, some polarizers and a photographic plate, where you end up with an interference pattern different than what would be expected from garden-variety e-m theory?

If that's your claim then I disagree with it and am willing to try and explain my objections. If that's not your point then I've simply misunderstood what you were saying.

 

[Frame Dragger]

I have to confess that I didn't read your scenario carefully enough to be sure that I understood it. I simply replied to some specific things you said in the body of your posts that seemed to go against what I think I know about classical e-m theory. Perhaps I'm not understanding your point at all. Are you really saying that there are experiments you can do with an ordinary laser, a pair of slits, some polarizers and a photographic plate, where you end up with an interference pattern different than what would be expected from garden-variety e-m theory?

If that's your claim then I disagree with it and am willing to try and explain my objections. If that's not your point then I've simply misunderstood what you were saying.

I give you 10 out of 10 for honesty when you could have evaded. There is a terrible TERRIBLE wikipedia article about this very scenario... but it had a great picture. In it, the contrast between the expected linear trajectory of the photon as predicted in a classical model vs SQM was shown. If I remember correctly even at 90 degrees there is still interference. I don't know if it's visibile however, and that is pissing me off! Now I should go and try and cite MY sources... oy!

 

[Mentz114]

This may be of interest. See page 41 for an experimental setup similar to the two-slits, with tourmaline polarisers.

 

[DrChinese]

Are you really saying that there are experiments you can do with an ordinary laser, a pair of slits, some polarizers and a photographic plate, where you end up with an interference pattern different than what would be expected from garden-variety e-m theory?

What I am saying is that once you know the outcome of an experiment, it may seem to match a classical expectation - but that the same classical expectation may not be supportable from theory alone. I am certainly advocating the idea that whenever the polarizers are in the Crossed mode, there is NO interference. That would be the expected result as far as I know from anyone's prediction. I mean, who wants to predict something different than experiments show?

So the question is: would a classical (realistic) wave explanation *actually* predict this? And I am saying: NO, there is no realistic dataset which supports that result. As with all quantum experiments, the context of the observation is important. Classical (realistic) theories are context-free. So there MUST be a difference somewhere. Clearly, the Crossed polarizers are a specific context and are fully independent of the source - just as in standard Bell tests. So I am asking: what theoretical basis is there for asserting that the quantum result can be obtained from a realistic perspective? Because I don't believe there is one.

And just to be clear: I recognize that this question is in some ways moot. After all, the (classical) wave picture of light is refuted by papers such as "Observing the quantum behavior of light in an undergraduate laboratory". And further, I know that most folks just assume that there is close correspondence between the classical picture and the quantum picture in many cases. When there is a difference, the quantum picture rules. But I think that this is a case in which the classical picture does not apply at all.

 

[DrChinese]

This may be of interest. See page 41 for an experimental setup similar to the two-slits, with tourmaline polarisers.

That is a cool paper. Where/how did you find it?

 

[Frame Dragger]

Wow.. that's a copy of a (partial apparently) dissertation from the MIT library that's been added to their digital archives. That IS a great paper... and you've got the very issue we're talking about... aaaaannnndddd the matrices are pretty clear. Of course, that doesn't help Dr. Chinese in terms of eliminating the Classical point of view as a valid description of these events. Then again, I'm not sure anything out there does one way or the other. There you go doc, if you didn't already have the doctorate you would have material for another dissertation. Fond memories, right? ;) Yeah...

Oh... http://libraries.mit.edu/archives/index.html I think that might be a good starting point for lots of goodies.


EDIT: I should add... purely as a historical document I really enjoyed reading that paper. Thank you VERY much Mentz114, for sharing it!

 

[conway]

I'm sorry I have to repeat myself but I still haven't understood what you are trying to claim. Do you mean to suggest there a difference between what I've called "garden-variety e-m theory" and what you are calling "classical (realistic) wave theory"? I don't like to push people for a "yes-or-no" answer but in this case I honestly don't see any way around it. Specifically:

Are you really saying that there are experiments you can do with an ordinary laser, a pair of slits, some polarizers and a photographic plate, where you end up with an interference pattern different than what would be expected from garden-variety e-m theory?

 

[conway]

Okay, I think I now know what you mean by "classical (realist)" vs "garden-variety". There is a line of argument whereby people say "obviously you cannot explain such-and-such with the classical wave theory, but if you modify the theory by assuming that for example conservation of energy may be temporarily violated..." etc. So I'm thinking what you mean to say is "Of course the standard wave theory doesn't explain the interference experiment, but beyond that, even if you modify the standard theory by the addition of some bizarre ad-hoc assumptions, you still cannot explain the facts.

I think that's what you're getting at. In other words, you are taking my statement and making it even stronger by expanding it to cover the scope of all possible wave theories.

If that's what you mean, I'd just as soon simplify the discussion by stipulating that I'm not interested in invoking exotic modifications of the wave theory. When I said "garden-variety", that's what I meant.

 

[DrChinese]

I'm sorry I have to repeat myself but I still haven't understood what you are trying to claim. Do you mean to suggest there a difference between what I've called "garden-variety e-m theory" and what you are calling "classical (realistic) wave theory"? I don't like to push people for a "yes-or-no" answer but in this case I honestly don't see any way around it. Specifically:

Are you really saying that there are experiments you can do with an ordinary laser, a pair of slits, some polarizers and a photographic plate, where you end up with an interference pattern different than what would be expected from garden-variety e-m theory?

No, the experiments come out as you would expect.

The authors of EPR were local realists (there were a lot around then), and didn't realize that the assumption of realism added severe constraints to classical theory. Bell later showed everyone what some of those assumptions were.

So just saying that "everyone knows" that the classical result is the same as the quantum prediction won't work. Since entanglement gives us a tool, we can probe all kinds of cool things today. But even the old double slit has things to tell us that are still relevant.

 

[DrChinese]

Okay, I think I now know what you mean by "classical (realist)" vs "garden-variety". There is a line of argument whereby people say "obviously you cannot explain such-and-such with the classical wave theory, but if you modify the theory by assuming that for example conservation of energy may be temporarily violated..." etc. So I'm thinking what you mean to say is "Of course the standard wave theory doesn't explain the interference experiment, but beyond that, even if you modify the standard theory by the addition of some bizarre ad-hoc assumptions, you still cannot explain the facts.

I think that's what you're getting at. In other words, you are taking my statement and making it even stronger by expanding it to cover the scope of all possible wave theories.

If that's what you mean, I'd just as soon simplify the discussion by stipulating that I'm not interested in invoking exotic modifications of the wave theory. When I said "garden-variety", that's what I meant.

I follow your point, and it fits with what I am claiming. Naturally, anyone today is going to be on board with the quantum treatment and so there is not much serious debate. The "garden variety" perspective works for most situations. And if everyone thinks it is identical in all respects to a classical (local realistic) perspective, then that is OK for most any discussion. But if you try to get "rigorous" about the point (not that it is necessary), the argument cannot hold up.

The double slit is a wonderful experiment to show the wave-particle duality of both light and matter. The polarizers are a nice twist to show that the which path information can be learned without even disturbing the light as it travel to the screen. ZapperZ often posts a reference to a paper by Marcella that ties the double slit directly to the Heisenberg Uncertainty Principle. So there is plenty to be learned from this one experiment.

 

[DrChinese]

Consider a change to the setup I mention in post#18:
-------------------------------------------------------------
To make this clear, consider the following two setups A and B, which are identical except for the settings of the polarizers in front of each slit. In each case, the source beam is polarized to 0 degress and the L (left) slit has a +45 degree polarizer. The R slit has a ++45 degree polarizer in the A setup but has a -45 degree polarizer in the B setup. Sorry for the crude drawing...

A. Inteference IS seen
...======
... Source
...== | ==
... |
... V
== /L/ = /R/ ==

.===Screen===B. NO Interference seen
...======
... Source
...== | ==
... |
... V
== /L/ = \R\ ==

===Screen===

--------------------------------------------------------
New version:

...======
... Source
...== | ==
...| = |
...| = |
.../ = \ <---- Polarizers are here now
...| = |
...| = |
...| = |
...| = |
...| = |
...| = |
...v = v
=== L = R ===

===Screen===i) the distance from the source to the slits is increased substantially;
ii) there is a barrier (marked as "=" above) between the source and the slits that separates the 2 side, but only up to the point where the slits are... after that, the light passing through the 2 slits can interact/interfere as usual;
iii) the polarizers are moved to a spot in between the source and the slits, close enough to the source that the polarizers can be considered spacelike separated;

Now, assuming that the barrier in between the 2 sides does not upset the experiment (just a guess really), we can see that the experiment resembles a Bell test with spacelike separated Alice and Bob (the Left and Right polarizers being the analog to Alice and Bob). And we know what that means: a Bell inequality. If it were possible to make a sensitive test with a double slit, then I would expect the result to shown explicitly that the old-style classical prediction cannot hold. We know the angle settings because they are the same as a traditional Bell test exactly.

 

[conway]

I'm sorry for your wasted effort but I no longer have any idea what we're talking about. I'm going to have to drop out of the discussion.

 

[DrChinese]

I'm sorry for your wasted effort but I no longer have any idea what we're talking about. I'm going to have to drop out of the discussion.

I guess "No, the experiments come out as you would expect" might be ambiguous.

 

[conway]

Yes, I generally had trouble understanding what you meant with that kind of statement.

 

[Phrak]

Just to add to my prior post: I think the situation could probably be formulated into an inequality using logic from Bell's Theorem.

You do realize that Bell's inequality refutes quantum mechanics, right?

 

[jtbell]

Bell's inequality refutes quantum mechanics

How so?

 

[Fredrik]

"Contradicts" would probably be a better word than "refutes".

 

[Phrak]

Yeah. 'Refutes' is not the right word.

How so?

Should Bell's inequality be true, some predictions of quantum mechanicanics are incorrect.

Violations of Bell's inequality could support quantum mechanics.

To be sure, quantum mechanics predicts statistical results that would violate Bell's inequality. Quantum mechanics predicts that Bell's inequality should be violated, and with particular statistical results.

And, to be sure, it's a common error to invert the meaning of the inequality. Bell could just as easily have inverted the inequality to it's inverse, and we'd be free of confusion--though I think he initially believed it would be found experimentally supported and 'action at a distance' found false.

 

[mintparasol]

There is NO chance that a particular photon polarized at 0 degrees could pass through both a +45 polarizer and a -45 degree polarizer in the quantum mechanical view. The operative formula is the COS^2(L-R) rule, where L-R=90 degrees so that the result is 0 and there is no interference.

On the other hand, in the classical perspective, there IS a chance that any particular photon polarized at 0 degrees could pass through both a +45 polarizer and a -45 degree polarizer. Do you see why? The rule is different because the probability is resolved independently for each slit, unlike in the quantum view in which it is the relative angle of the L and R slits is important. So now you get COS^(L-0)*COS^(R-0) and there should be some interference because the result is .25 which is >0.

Well at least you are thinking about it...

I would be interested in a demonstration of a classical wave effect which eliminates all interference when partial polarizers (or classical equivalent) are present in a setup similar to the double slit. You use the example of audio, and I think a careful consideration of your analogy will demonstrate that the interference will NOT be eliminated after all - as it would need to be to match light in a double slit.

Of course there is a lot more to the story on the quantum side anyway. If light were waves (only) then a lot of things would be different (in contradiction to experiment - see for example: http://people.whitman.edu/~beckmk/QM/grangier/Thorn_ajp.pdf" ). If light polarization operated as you imagine, things would be different with entangled particle pairs (in contradiction to experiment).

You may not be aware of all of the experiments out there (who is?), but you might want to at least ask before talking about the emperor's clothes. The double slit is just one piece of the puzzle.

Ok, I've printed off that paper for further reading, thanks!

I'm just wondering if we could arrange some mechanism to put our bass wave +45 degrees out of phase at the left slit and -45 degrees out of phase at the right slit, would the math work in the same way for soundwaves? Would there be any interference detectable rear of the slitted screen?

 

[conway]

But sound isn't polarized. Its a longitudinal wave while light is a transverse wave. In any case, there is no interference with light if the waves from the two slits are polarized at 90 degrees to each other. This is predicted by classical physics exactly the same as QM and does not represent any kind of mystery.

 

[DrChinese]

You do realize that Bell's inequality refutes quantum mechanics, right?

I am sure we agree that a violation of a Bell inequality refutes classical realism, and is consistent with QM. I simply think of the inequality as a boundary condition on (local) realism.

 

[Frame Dragger]

I am sure we agree that a violation of a Bell inequality refutes classical realism, and is consistent with QM. I simply think of the inequality as a boundary condition on (local) realism.

I have to say that's a good way of putting it. I've been doing a good deal of reading on deBB, and a refresher on The TI, and more and more it seems that dBB and a few other theories are simply the ones that have not been swept aside by modern experimental evidence. dBB basically hinges on the QM being in violation of Bell's Inequalities, and not classical realism. The argument of (mechanical) bias in experiments such as a "2 channel" test, keeps dBB alive unlike most other theories which required LHV's.

However, the switch from locality to non-locality in dBB seems logically contrived for the sole purpose of maintaining a particular interpretation of reality as something other than emergent phenomena from a quantum soup of superpositioned probabilities. That doesn't make it wrong, and from everything I'm reading dBB can't be refuted right now and if nothing else it does a very good job of highlighting the huge gap in explaining how the microscopic and macroscopic worlds combine to form what we experience as reality in SQM. I think that the TI is probably on the right track, but it's hard to say that those theories, or MWI are more contrived than dBB.

Given all of that, I still fall on the side of SQM for the sake of the results it provides and predictions it makes. Bell expected "Spukhafte Fernwirking" to be experimentally disproved, and instead it is now experimentally observed on a regular basis.

In my view, one of the ways this debate can be settled would be the emergence of even a VERY rudimentary computer capable of using qubits to complete an algorithm. Failure would mean little, but success in line with expectations of SQM would be pretty damned confirmatory. The problem with modern approaches is that they still use Classical mechanics that exploit quantum behaviour... which can be explained by theories such as dBB. Go figure.

The bottom line is that Bell kind of draws the line between the Classicist and SQM view of reality. It's a boundary in the common usage and not just the scientific term of art.

 

[DrChinese]

I have to say that's a good way of putting it. I've been doing a good deal of reading on deBB, and a refresher on The TI, and more and more it seems that dBB and a few other theories are simply the ones that have not been swept aside by modern experimental evidence. dBB basically hinges on the QM being in violation of Bell's Inequalities, and not classical realism. The argument of (mechanical) bias in experiments such as a "2 channel" test, keeps dBB alive unlike most other theories which required LHV's.

However, the switch from locality to non-locality in dBB seems logically contrived for the sole purpose of maintaining a particular interpretation of reality as something other than emergent phenomena from a quantum soup of superpositioned probabilities. That doesn't make it wrong, and from everything I'm reading dBB can't be refuted right now and if nothing else it does a very good job of highlighting the huge gap in explaining how the microscopic and macroscopic worlds combine to form what we experience as reality in SQM. I think that the TI is probably on the right track, but it's hard to say that those theories, or MWI are more contrived than dBB.

...

The bottom line is that Bell kind of draws the line between the Classicist and SQM view of reality. It's a boundary in the common usage and not just the scientific term of art.

I agree with what you are saying. dBB gets by because the context is non-local. Another interesting set of interpretations is the Time Symmetric group (including Relational BlockWorld RBW), in which a future context is allowed to influence the present. These have the "benefit" of being local, contextual (non-realistic) and time symmetric. Of course you swap one assumption for another, so whether the result is net better is a matter of preference.