Is there a 'memory-effect' of measuring devices?

[entropy1]

If I'm correct, a series of with respect to some property entangled particles exhibits a correlation between several measurements of that property by means of two measuring devices.

My question is: is it possible that between measurements the physical constitutions of the measuring devices communicate their previous measured results with each other, thereby creating a 'memory' of the previous measurement of the other device, thus making it possible to create the appropriate correlation over several measurements?

I recall vaguely that this has been examined (theoretically or experimentally).

 

[phinds]

Sounds like magic to me. Do you have a reference?

 

[entropy1]

Sounds like magic to me. Do you have a reference?

Unfortunately not.

 

[phinds]

Unfortunately not. Since I am not really an expert so to speak I am afraid I'll get lost in all possible info on the web. I don't know how to quickly scan the technical papers...

Well, I don't think you'd find anything anyway. I mean, seriously, doesn't this sound like magic to you? It sounds like you are suggesting that disparate measuring devices could become entangled in some way just by virtue of having measured the same thing. How do you reckon that could happen?

 

[entropy1]

Well, I don't think you'd find anything anyway. I mean, seriously, doesn't this sound like magic to you? It sounds like you are suggesting that disparate measuring devices could become entangled in some way just by virtue of having measured the same thing. How do you reckon that could happen?

I imagine that by coming into contact of the two measuring devices by starting to share the same lightcone(s) (for instance, if measurement results are being compared on macro-level, the devices have a long time to exchange information by, for instance, EM radiation), that the measuring devices assimilate, among other information, also information about their respective measurements. I don't know the details. I may, of course, be mistaking!

All has to do with quantuminformation converservation laws.

You can call it magic if you want. Whether that is appropriate, I don't know haha

 

[Nugatory]

My question is: is it possible that between measurements the physical constitutions of the measuring devices communicate their previous measured results with each other, thereby creating a 'memory' of the previous measurement of the other device, thus making it possible to create the appropriate correlation over several measurements?

You only get one measurement on each member of each pair, and no about information about one pair can affect the correlation of the next pair (as long as the preparation of each pair is truly independent) so such an effect cannot reproduce the observed correlations.

 

[phinds]

You only get one measurement on each member of each pair, and no about information about one pair can affect the correlation of the next pair (as long as the preparation of each pair is truly independent) so such an effect cannot reproduce the observed correlations.

Nugatory I certainly agree w/ what you have said but it doesn't seem to directly address the question of whether measuring devices communicate with each other (which is what I called magic).

 

[entropy1]

You only get one measurement on each member of each pair, and no about information about one pair can affect the correlation of the next pair (as long as the preparation of each pair is truly independent) so such an effect cannot reproduce the observed correlations.

Has this been officially demonstrated?

 

[Heinera]

If I'm correct, a series of with respect to some property entangled particles exhibits a correlation between several measurements of that property by means of two measuring devices.

My question is: is it possible that between measurements the physical constitutions of the measuring devices communicate their previous measured results with each other, thereby creating a 'memory' of the previous measurement of the other device, thus making it possible to create the appropriate correlation over several measurements?

I recall vaguely that this has been examined (theoretically or experimentally).

If I understand you correctly, you are referring to something called the "memory loophole". There is some literature on this loophole, e.g.:
http://arxiv.org/abs/quant-ph/0205016

 

[entropy1]

If I understand you correctly, you are referring to something called the "memory loophole". There is some literature on this loophole, e.g.:
http://arxiv.org/abs/quant-ph/0205016

That looks a bit like it. I am not sure though if that paper reflects what I mean... Did you google this?

 

[Truecrimson]

My question is: is it possible that between measurements the physical constitutions of the measuring devices communicate their previous measured results with each other, thereby creating a 'memory' of the previous measurement of the other device, thus making it possible to create the appropriate correlation over several measurements?

I recall vaguely that this has been examined (theoretically or experimentally).

You are right. This has been examined before both theoretically and experimentally. It was taken into account in the loophole-free Bell test from the Delft group last year.

Quantum nonlocality, Bell inequalities, and the memory Loophole
http://arxiv.org/abs/quant-ph/0205016

In the analysis of experiments designed to reveal violation of Bell-type inequalities, it is usually assumed that any hidden variables associated with the nth particle pair would be independent of measurement choices and outcomes for the first (n−1) pairs. Models which violate this assumption exploit what we call the memory loophole. We focus on the strongest type of violation, which uses the 2-sided memory loophole, in which the hidden variables for pair n can depend on the previous measurement choices and outcomes in both wings of the experiment.

Time, finite statistics, and Bell's fifith position
http://arxiv.org/abs/quant-ph/0301059

Experimental loophole-free violation of a Bell inequality using entangled electron spins sperated by 1.3 km
http://arxiv.org/abs/1508.05949

 

[Nugatory]

Has this been officially demonstrated?

It follows from the definition of "independent" preparation - you'll notice that I managed to sneak that term into my answer. If you demonstrate any pair-to-pair influences, you've shown that we don't have independent preparation.
This is a well-understood "loophole" in Bell's theorem - and the only one that cannot be closed by experiment. If you google for "superdeterminism quantum mechanics" you will find some discussion of the limitations of the approach.

but it doesn't seem to directly address the question of whether measuring devices communicate with each other (which is what I called magic).

We've done experiments in which the not only the measurement, but the the choice of which measurement to make, on one particle lies outside the past lightcone of the choice of measurement and the measurement for the other particle - so we've solidly excluded the possibility of magic influencing the correlation of the measurements on anyone pair. However, all parts (pair generation, two choices of measurement to make, two measurements made) of the measurement of one pair will be in the past light cone of all parts of the measurement of some future pair, so we cannot exclude magic. What we can say is that if the magic creates correlations in violation of Bell's inequality, it must bias the one or more of the ostensibly independent events in the future. That's what takes us to superdeterminism - and you can form your opinion on the merits of that approach.

 

[entropy1]

If I understand you correctly, you are referring to something called the "memory loophole". There is some literature on this loophole, e.g.:
http://arxiv.org/abs/quant-ph/0205016

In the excerpt, it reads:

We show that the 2-sided memory loophole allows a systematic violation of the CHSH inequality when the data are analysed in the standard way, but cannot produce a violation if a CHSH expression depending linearly on the data is used. In the first case, the maximal CHSH violation becomes small as the number of particle pairs tested becomes large.

So, it seems like the amount of violation of the Bell-inequality decreases as the dataset, and thus the accuracy of the determination of the correlation, grows?

 

[Heinera]

So, it seems like the amount of violation of the Bell-inequality decreases as the dataset, and thus the accuracy of the determination of the correlation, grows?

That is correct, when one uses the standard CHSH inequality. But there are modified versions of the CHSH inequality that eliminates this loophole completely.

 

[entropy1]

If I understand you correctly, you are referring to something called the "memory loophole". There is some literature on this loophole, e.g.:
http://arxiv.org/abs/quant-ph/0205016

We show that the 2-sided memory loophole allows a systematic violation of the CHSH inequality when the data are analysed in the standard way, but cannot produce a violation if a CHSH expression depending linearly on the data is used. In the first case, the maximal CHSH violation becomes small as the number of particle pairs tested becomes large. [emphasis added]

I estimate reading and understanding the paper in detail is a rather optimistic goal for me. So, just to make sure, I dare to pose yet another question. It is about the boldface text I emphasized in the quote from the excerpt of the paper. Does 2-sided mean: the two measurement devices exchange information about previous measurements of each other (being part of each others' lightcone), or both memorizing their own measurements, independent of the other device (not in each others' lightcone)?

Thanks.

 

[entropy1]

We[/PLAIN] focus on the strongest type of violation, which uses the 2-sided memory loophole, in which the hidden variables for pair n can depend on the previous measurement choices and outcomes in both wings of the experiment.

It looks like they examine the possibility of local hidden variables that depend on the outcomes of the measurements. That may be not exactly what I mean. I was thinking of global hidden variables that 'store' the 'outcomes' of all 'measurements' made (in the universe), in this case influencing the behaviour of the measurement devices in subsequent measurement(s). This still means the particles do have to have local hidden variables to make this work. Anyone on this?

 

[entropy1]

I have to think this one through. Thanks for the replies.

 

[Heinera]

It looks like they examine the possibility of local hidden variables that depend on the outcomes of the measurements. That may be not exactly what I mean. I was thinking of global hidden variables that 'store' the 'outcomes' of all 'measurements' made (in the universe), in this case influencing the behaviour of the measurement devices in subsequent measurement(s). This still means the particles do have to have local hidden variables to make this work. Anyone on this?

If the detector settings that Alice and Bob choose are completely random and independent of anything else in the experiment, including past settings and results, then the mechanism you suggest cannot lead to a violation of Bell's inequality. The proof of Bell's theorem is still valid with detectors that change properties from run to run, as long as this change is not influenced by the current detector setting chosen in the other wing of the experiment.

 

[entropy1]

I was thinking of the conservation of information law (eg. momentum, spin, polarization). Does this law exist and if so, does it impose related properties/measurement results? Thanks for your patience y'all!