Why is entanglement necessary for understanding quantum mechanics?

[DrChinese]

And if those photons are really unpolarized or "undefined" upon emission, then why don't we simply make them polarized first and then see what is really going on and how it actually works?

Once a non-reversible measurement is performed (ie it is polarized), it is no longer entangled. Please keep in mind that the Heisenberg Uncertainty Principle (HUP) is at work, even with entangled particle pairs. You cannot beat the HUP! So when you observe Alice, you obtain identical information about Bob (and vice versa). That is the limit.

 

[DrChinese]

That's a very good point. I think there have been attempts to explain EPR using local means by exploiting the differences between actual experiments and the idealization presented in most theoretical discussions of Bell's Inequality. For example, I think that someone named "Dereiter" or something like that? The idea is that Alice and Bob don't necessarily always measure the corresponding photons. If you assume that the likelihood of getting a mismatch is correlated with their filter settings, then maybe it's possible to reproduce the QM predictions for EPR.

I'm not sure if all such loopholes have been closed by experiments, but it's a little puzzling to think that errors in interpreting data would just happen to reproduce the predictions of QM.

De Raedt et al is the team you are referring to. Very complex stuff and far out of the league of this discussion. That's another thread. If we try to go down that path here, I will seek to close this thread.

Not trying to shut down discussion but this thread is already nearing its useful limit with over 200 posts now. Jabbu has barely moved, and should be doing more homework before posting additional comments.

Jabbu, I hope you are reading this and listening carefully.

 

[Jabbu]

No, the raw data is very difficult to decipher and is not in a format that can be readily analyzed. It would almost be easier to do the experiment yourself. Of course, there is no real reason to see the raw data unless you simply don't believe the results. The experiments are described in plenty of detail for those who are interested.

According to how you said correlation is experimentally calculated, the raw data must be in this format:

theta_A = +30, theta_B = +30
A: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0
B: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0

That's the only actual information. What does it mean depends on which interpretation you like the most, but some interpretations make more sense than other. The question is which one makes the most, and hopefully, total sense.

 

[DrChinese]

According to how you said correlation is experimentally calculated, the raw data must be in this format:

theta_A = +30, theta_B = +30
A: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0
B: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0

That's the only actual information. What does it mean depends on which interpretation you like the most, but some interpretations make more sense than other. The question is which one makes the most, and hopefully, total sense.

Bell test data must be time-stamped and "lined up" (Alice's timestamps matched with Bob's) after the fact. It takes computer programming to analyze and make sense of. All of this is far past where the current thread sits. Until you understand the basic science, none of that will make much sense.

Yes, there are a variety of interpretations, and it is good to understand those: orthodox QM, Bohmian Mechanics, Many Worlds, and Time Symmetric types are good ones to be aware of.

 

[Jabbu]

I should highlight that in actual experiments it is impossible to know that a photon did not go through, let alone that both of them did not go through. The typical setup is usually quite different than the (0,1) values being discussed. Typically you have a beam-splitter with two arms at each station. One of them labelled +1 and other labelled -1. a (+1,+1) or (-1,-1) result is a match and a (+1, -1) or (-1, +1) result for each pair is a mismatch.

1/0 is more readable and easier to type than +1/-1.

The way correlatiosn (actually expectation values) are calculated in the experiment is also quite different from the equations being discussed. They are calculated as <AB>, ie the average of a product of results on both sides for the given pair of angular settings. It is this <AB> value that matches the QM expectation value.

How do you apply that to this sequence:

A: 1 0 1 0 1 0 1 0 1 0 1 1 0 0 0 0 1 1 0 1
B: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0

...what's correlation percentage?

But it is even worse than that. Experiment do not give you pairs. Rather, you have a random series of time-tagged +1/-1 results at Alice, and another random series of time-tagged +1/-1 results at Bob corresponding to when the various detectors clicked. Then after the experiment you try to find pairs of clicks close enough in time which you *assume* belong to the same pair. Any unpaired value is discarded. There are no one sided (or does not pass through) values in the calculation.

I'll keep that in mind.

 

[Jabbu]

I don't think you followed my statement. They can be entangled AFTER they are observed. They no longer exist. That is because time ordering is very different than you might expect.

If they no longer exist how do you confirm they are entangled now and were not entangled before?

Once a non-reversible measurement is performed (ie it is polarized), it is no longer entangled.

They can be entangled from the beginning if we polarize them both at the same angle. Has no one attempted the experiment with constant photon polarization?

 

[Jabbu]

It takes computer programming to analyze and make sense of.

I'm counting on it, I've written programs to analyze binary sequences before.

 

[billschnieder]

How do you apply that to this sequence:

A: 1 0 1 0 1 0 1 0 1 0 1 1 0 0 0 0 1 1 0 1
B: 1 0 0 1 0 0 1 0 0 1 1 0 1 0 0 1 1 1 1 0

...what's correlation percentage?

Like I said earlier, there are no 0 events in actual experiments, because only detected photons count towards the expectation value. So the sequence above seems easier to type and discuss but bears no resemblance to anything that an experiment produces, at least any experiment that has ever been performed. If you think about it, that sequence is pretty incredible. To get a sequence like that, you need to know that two photons were emitted at the same time but neither was detected in some cases, and in others that two were emitted but only one was detected. In an actual experiment you only know what was detected. If you try to measure at the source, you destroy them and invalidate the experiment. If you detect only one, you can't be absolutely sure that another one was emitted but did not go through. So the 11 case is the only certain case for matches that you can have in any experiment. And even then you have time-tagging and matching to deal with, which complicates matters even further. If you want to see some actual experimental data, see http://people.isy.liu.se/jalar/belltiming/ which has a portion of the Weihs data.

But that was for EPR, even if you think about Malus, how do you know that a photon did not go through to be able to record a 0 data point? The expectation values are obtained experimentally using very different methods, first you measure total intensity without filter, then you insert filter without changing the source and measure total intensity again. This is how you obtain cos^2 relationship experimentally. That is why you will never see experimental data for Malus, similar to your sequence above.

It is easy to discuss thought experiments and fictitious data but when you get down to actual experiments, most of that makes no sense.

 

[DrChinese]

1. If they no longer exist how do you confirm they are entangled now and were not entangled before?

2. They can be entangled from the beginning if we polarize them both at the same angle. Has no one attempted the experiment with constant photon polarization?

Are you even reading the references I am posting? You have been going around in circles stuck with the same ideas for a long time.

1. Obviously, they violate a Bell Inequality. That is the standard for entanglement. Since they are actually made to be entangled about 50 nanoseconds AFTER they are detected, I am not sure what you are looking for. Non-entangled pairs do NOT violate a Bell inequality.

2. Have you not been listening to what everyone has been telling you? Photons with a specific polarization are not entangled. Once they are given a specific polarization, entangled photons are no longer polarization entangled.

-------------------------

Again Jabbu: I am urging you to think carefully before you post. You should not be a) repeating questions you have already asked; and b) you should not be making speculative statements such as "They can be entangled from the beginning if we polarize them both at the same angle." That is completely wrong as has been repeatedly explained.

There are many others reading (and answering on) this thread. It is normal for you to ask a question, get an answer, then review relevant material before asking another question. That is how you learn, which is what we assume you are here for. If you are here to debate the subject, this thread will likely come to a very quick close. This is not that kind of forum.

 

[DrChinese]

If you want to see some actual experimental data, see http://people.isy.liu.se/jalar/belltiming/ which has a portion of the Weihs data.

Thanks for the link, Bill.

 

[Jabbu]

Like I said earlier, there are no 0 events in actual experiments, because only detected photons count towards the expectation value. So the sequence above seems easier to type and discuss but bears no resemblance to anything that an experiment produces, at least any experiment that has ever been performed. If you think about it, that sequence is pretty incredible. To get a sequence like that, you need to know that two photons were emitted at the same time but neither was detected in some cases, and in others that two were emitted but only one was detected. In an actual experiment you only know what was detected. If you try to measure at the source, you destroy them and invalidate the experiment. If you detect only one, you can't be absolutely sure that another one was emitted but did not go through. So the 11 case is the only certain case for matches that you can have in any experiment. And even then you have time-tagging and matching to deal with, which complicates matters even further.

Then we shall talk only about splitter polarizers where photons are detected for both 1 and 0.

If you want to see some actual experimental data, see http://people.isy.liu.se/jalar/belltiming/ which has a portion of the Weihs data.

Looks good, but they are asking for $25. Could you copy-paste some of the data for 0 degrees relative angle?

 

[Jabbu]

2. Have you not been listening to what everyone has been telling you? Photons with a specific polarization are not entangled. Once they are given a specific polarization, entangled photons are no longer polarization entangled.

I'm just asking if anyone has performed the experiment with constant photons polarization. Do you know?

 

[Jabbu]

If you want to see some actual experimental data, see http://people.isy.liu.se/jalar/belltiming/ which has a portion of the Weihs data.

Found it in "belltiming.tar.gz". That's exactly what I need. Thank you!

 

[DrChinese]

I'm just asking if anyone has performed the experiment with constant photons polarization. Do you know?

People generally don't publish null results. I am sure everyone who has done a Type I polarization experiment has seen this, since using only 1 crystal gives the effect I described (and you are asking about, which I have explained previously several times). It takes 2 crystals precisely aligned to get entanglement in the first place, else you get polarized pairs and those are not polarization entangled. I am not sure if I or anyone else can explain it any further. If you are still unsure, do the experiment yourself.

At this point, I should ask: what about "the need for entanglement" (the title of this thread) do you not understand? Entangled pairs describe a non-classical behavior as has been discussed to the Nth degree. We have explained in detail what statistics are evident and the differences with non-entangled particles.

Jabbu, my intuition tells me that we are reaching the end of the line for this thread if you are reduced to skipping study of ground-breaking experiments (numerous references provided for Weihs, Zeilinger, etc) and instead you asking about experiments providing no insight at all. (You may as well ask whether anyone has published if entanglement is different during leap years.) If you need to take this thread in a new direction, such as the discuss of raw experimental data, a new thread is in order.

 

[bhobba]

After photon A goes through polarizer A, how can photon correlation make photon B go through polarizer B with 100% chance?

The Born rule applied to Bell states:
http://en.wikipedia.org/wiki/Bell_state
http://www-inst.eecs.berkeley.edu/~cs191/sp05/lectures/lecture2.pdf

Its a logical consequence of the axioms of QM. So your question is why are the axioms of QM true? Well of course the answer to that is correspondence with experiment. But modern research has distilled it to one key assumption - see post 137:
https://www.physicsforums.com/showthread.php?t=763139&page=8
An observation/measurement with possible outcomes i = 1, 2, 3 ... is described by a POVM Ei such that the probability of outcome i is determined by Ei, and only by Ei, in particular it does not depend on what POVM it is part of.

Why is photon entanglement not part of the equation?

What do you think a Bell state being maximally entangled means? And, that being the case, why do you think photon entanglement isn't part of it?

Did you read Bell's Bertlmanns socks paper?

Thanks
Bill

 

[Jabbu]

and even then you have time-tagging and matching to deal with, which complicates matters even further. If you want to see some actual experimental data, see http://people.isy.liu.se/jalar/belltiming/ which has a portion of the weihs data.

You were right, it's like fishing with dynamite. Here's a little sample of unmatched raw data:

Alice TIME      Bob TIME        A-B
1. 0.00111168   0.0011126       0-0
2. 0.0011236    0.00114889      0-3
3. 0.00114997   0.00116779      1-0
4. 0.00115135   0.00120643      0-0
5. 0.00120123   0.00129848      1-0
6. 0.00122004   0.0013142       0-0
7. 0.00122526   0.00137652      1-0
8. 0.00124101   0.00141566      0-1
9. 0.00124709   0.00151352      3-0
10. 0.00125743  0.00152207      0-2
11. 0.00128748  0.00154191      2-0
12. 0.00131066  0.00157764      0-0
13. 0.00133157  0.0016791       1-0
14. 0.00137614  0.00177568      0-0
15. 0.00142413  0.00178678      2-0
16. 0.00147483  0.00189219      0-0
17. 0.00148506  0.00191145      2-0
18. 0.00150237  0.00196715      0-2
19. 0.00150441  0.00197128      2-0
20. 0.00151065  0.00199187      0-0
21. 0.00151289  0.00202538      2-0
22. 0.00151499  0.00205014      0-3
23. 0.00152948  0.00210897      0-0
24. 0.00155419  0.00211805      0-2
25. 0.00158218  0.00213923      2-0
26. 0.00161245  0.00216286      0-1
27. 0.00162987  0.00219944      1-0
28. 0.0016477   0.00220125      0-3
29. 0.00165599  0.00220627      2-0
30. 0.00169664  0.00225132      0-3
31. 0.00169747  0.00226738      1-0
32. 0.00171081  0.00236668      0-3
33. 0.00173233  0.00237884      3-0
34. 0.00181079  0.00241838      0-0
35. 0.00184828  0.00250185      0-0
36. 0.00187782  0.00250618      0-1
37. 0.0019019   0.00260884      0-0
38. 0.00190597  0.00261734      0-1
39. 0.0019224   0.00264896      1-0
40. 0.00197127  0.00265759      0-3
41. 0.00198946  0.00284926      2-0
42. 0.00202538  0.00287557      0-0
43. 0.00203134  0.0029057       0-0
44. 0.00203239  0.00291423      0-3
45. 0.00208354  0.00293723      0-0
46. 0.00209534  0.00294736      0-2
47. 0.00210002  0.00300586      0-0
48. 0.0021307   0.00303086      0-3
49. 0.00213906  0.00313415      1-0
50. 0.00215645  0.00315502      0-1

In about 0.001 second Alice recorded 41 detections, and Bob only 20, huh? How can there be matching photon pairs if one stream is constantly ticking at a higher rate than the other one? Also, out of hundred detections, for 0 rotation Alice recorded 66 + and 13 - detections, and for 45 rotation there was 15 + and 6 - detection. This doesn't look like 50%-50% chance to me. Perhaps I'm misinterpreting something, would you know?

 

[Jabbu]

It takes 2 crystals precisely aligned to get entanglement in the first place, else you get polarized pairs and those are not polarization entangled.

Do you know what type of crystal are those and what is it they do to make photons entangled?

 

[Nugatory]

Do you know what type of crystal are those and what is it they do to make photons entangled?

Google for "spontaneous parametric down-conversion". Typically the crystals are beta-barium oxide.

Before this (already highly derailed thread) becomes even more derailed, consider that there are several ways of creating entangled particle pairs (not necessarily photons) suitable for Bell-type experiments, multiple ways of creating polarization-entangled photon pairs, all of the experiments differ in how they're designed and analyzed, and the technology used for gathering and analyzing the data has changed greatly over the past 40 years. If you want to know exactly how a given experiment was conducted... You have to find the paper and read it.

 

[Nugatory]

In about 0.001 second Alice recorded 41 detections, and Bob only 20, huh? How can there be matching photon pairs if one stream is constantly ticking at a higher rate than the other one? Also, out of hundred detections, for 0 rotation Alice recorded 66 + and 13 - detections, and for 45 rotation there was 15 + and 6 - detection. This doesn't look like 50%-50% chance to me. Perhaps I'm misinterpreting something, would you know?

You need to look at a sample that's large enough to pick the signal out of the noise. Often only one member of the pair will be detected, and not all of the photons that make it into the detector are members of entangled pairs.

Have you tried playing around with the software that came along with that data sample?

 

[DrChinese]

Do you know what type of crystal are those and what is it they do to make photons entangled?

This is answered by the references I provided in posts #21 and #29.

As promised, Jabbu, your post has been reported.

-DrC

 

[Doc Al]

This thread has long been derailed from the initial question on entanglement. Now it seems to be going in circles regarding Bell's theorem and tests, rife with mistaken ideas, with no clear end in sight. It is time to close it.

Lots of good information (and suggestions for study) have been provided. I hope interested parties take full advantage.