Quantum Entanglement: Explaining the Logic Behind It

[Thomas T]

Hi,
I read some about quantum entanglement but I don't really understand the logic behind it.

If you quantum entangle 2 particles by for example colliding them and then separate them by a long distance they are both in superposition and you do not know the spin of them.

Then you separate them by a long distance and measure one of them and thereby force it to show its spin, in the exact same moment the other will fall out of its superposition and take a spin in the other direction than that of the measured particle, right?

So (one of the) theories is that the particles exchange information by sending information back in time and therefore they change at the exact same moment...

I have very difficult to believe this mechanic. To me it sounds too far fetched and humans like to believe in ridiculous explanations like if you would see a weird light in the sky one think of oh that must be some alien spacecraft instead of oh maybe its a helicopter or airplane or maybe some reflected light..

First of all: Could someone please explain to me how you can quantum entangle 2 small particles and keep track of them when separating them over a long distance? I have problem tracking my dog.. but tracking 2 so small particles and measure them seems very very hard..

Second: They are always of opposite spin. Wouldnt it make more sense of just assuming there is some mechanic that makes them always take opposite spins when they collide so when they are separated they already have that spin from the start even if we for the moment do not know which spin they have since we didnt measure it yet.

If you measure particle 1 and it spins in 1 direction you know the other particles spin even without measuring and if you measure its spin you will only confirm what you already know and by measuring it you force it out of superposition.. It doesn't matter which particle you measure first..
I don't really see why they need to exchange information back in time unless you can separate them by a really long distance and continiously measure them and forcechange the spin of one of the particles and see that the other particle change spin at the exact moment you force change the spin of the other particle.. (maybe you can do this and its been verified and I missed this part?) But unless this is possible I don't really see what evidence there is the particles exchange information back in time?

Can someone please explain this for me. Thanks.
(Sorry bout my english, its not my primary language)

 

[DrChinese]

1. If you quantum entangle 2 particles by for example colliding them and then separate them by a long distance they are both in superposition and you do not know the spin of them.

2. Then you separate them by a long distance and measure one of them and thereby force it to show its spin, in the exact same moment the other will fall out of its superposition and take a spin in the other direction than that of the measured particle, right?

So (one of the) theories is that the particles exchange information by sending information back in time and therefore they change at the exact same moment...

3. I have very difficult to believe this mechanic. To me it sounds too far fetched and humans like to believe in ridiculous explanations like if you would see a weird light in the sky one think of oh that must be some alien spacecraft instead of oh maybe its a helicopter or airplane or maybe some reflected light..

4. Could someone please explain to me how you can quantum entangle 2 small particles and keep track of them when separating them over a long distance? I have problem tracking my dog.. but tracking 2 so small particles and measure them seems very very hard..

5. They are always of opposite spin. Wouldnt it make more sense of just assuming there is some mechanic that makes them always take opposite spins when they collide so when they are separated they already have that spin from the start even if we for the moment do not know which spin they have since we didnt measure it yet.

If you measure particle 1 and it spins in 1 direction you know the other particles spin even without measuring and if you measure its spin you will only confirm what you already know and by measuring it you force it out of superposition.. It doesn't matter which particle you measure first..

A lot of questions!

1. Yes, this is close enough to start the discussion.

2. No one knows the exact "mechanics". This is one possible method.

3. There is nothing ridiculous about entanglement itself. There have been thousands of experiment confirming it.

4. The most common way is to use Parametric Down Conversion. Laser light shines into a special crystal, which occasionally emits entangled pairs of photons. These are collected and can be manipulated a number of ways, including routing them by fiber. Google it to learn more.

5. That would be logical! And in fact Einstein and others believed much the same. However, in 1964 John Bell proved that this assumption would actually lead to a conflict with experiment. Google Bell's Theorem, or go to this page (written by me):

http://www.drchinese.com/Bells_Theorem.htm

 

[phinds]

@Thomas T quantum entanglement is just one of the many things in cosmology and quantum mechanics that are completely counter intuitive and are things for which our human "common sense" and "intuition" are not only useless, they are in fact counterproductive. When I first started learning about a lot of these things, I would stomp around my room shouting "THAT CAN'T BE RIGHT !" and pulling out hair. Well, OK, I never actually got any hair to come out but it wasn't for lack of trying.

You'll need to suspend your sense of disbelief as you study these things and just listen to what experiments have shown.

 

[stevendaryl]

Second: They are always of opposite spin. Wouldnt it make more sense of just assuming there is some mechanic that makes them always take opposite spins when they collide so when they are separated they already have that spin from the start even if we for the moment do not know which spin they have since we didnt measure it yet.

Yes, that makes a lot more sense, except for the fact that it's not true. What you've described is a "local hidden variable" theory, which assumes that each particle already has an unknown value for its spin before it is measured, and the two particles have opposite values.

That explanation doesn't work. It's a little complicated to explain why not.

First, spin is not just "spin-up" or "spin-down". It's spin-up or spin-down relative to a particular orientation of the measurement device. To make it simple, let's suppose that the orientation can be one of three possibilities in the x-y plane: (1) In the y-direction. (2) At an angle 120 degrees clockwise away from the y-direction. (3) At an angle 240 degrees clockwise away from the y-axis. If one particle is measured to be spin-up in direction 1, then the other particle will be spin-down in that direction. Similarly for directions 2 and 3.

So if you're going to try to explain it by saying that the two particles have pre-determined spin outcomes, that doesn't just mean "spin-up" or "spin-down", it means that for every possible direction (I'm only considering 3 for simplicity), it must be pre-determined. That means that there are 8 possible states for the first particle:

1. UUU: Each of the three directions gives spin-up for the first particle. Corresponding to this, state of the second particle must be DDD: all three give spin-down.
2. UUD for the first particle and DDU for the second.
3. UDU for the first particle and DUD for the second.
   
4. UDD for the first particle and DUU for the second.
5. DUU and UDD
6. DUD and UDU
7. DDU and UUD
8. DDD and UUU

Let's assume that each of these states is equally likely. Then that shows that if the first particle is measured spin-up in direction 1, then its state must be one of the first 4 possibilities. For each of those, look at the corresponding state for the second particle. It must be DDD, DDU, DUD, or DUU. Now, let's look at the spin value for the second direction. We see that there are two possibilities with the second result D and two possibilities with the second result U. So the equal-likelihood assumption would lead you to the conclusion that there is a 50/50 chance of the measurement of the second particle along the second direction will result in U (spin-up).

However, quantum mechanics does not predict this. It predicts that if the first particle is measured to be spin-up in direction 1, there is actually a 75% chance that the second particle will be spin-up in direction 2. The mathematical details are these: The probability that the two particles will have the same measured spin (up or down) is $sin^2(\theta/2)$, where $\theta$ is the angle between the two directions that the particle spins are measured relative to. For $\theta = 120^o$, this gives $sin^2(120/2) = 0.75$

You could try something else, with different probabilities for the 8 states, but it's provable that there is no way to make it work. The results being predetermined does not agree with experiment.

 

[Thomas T]

A lot of questions!

1. Yes, this is close enough to start the discussion.

2. No one knows the exact "mechanics". This is one possible method.

3. There is nothing ridiculous about entanglement itself. There have been thousands of experiment confirming it.

4. The most common way is to use Parametric Down Conversion. Laser light shines into a special crystal, which occasionally emits entangled pairs of photons. These are collected and can be manipulated a number of ways, including routing them by fiber. Google it to learn more.

5. That would be logical! And in fact Einstein and others believed much the same. However, in 1964 John Bell proved that this assumption would actually lead to a conflict with experiment. Google Bell's Theorem, or go to this page (written by me):

http://www.drchinese.com/Bells_Theorem.htm

Wow, thanks for the fast reply. I tried to understand Bells Theorem on your page but I need to take another look tomorrow after I had some sleep haha.

3. No, I was unclear. I believe in the entanglement as 2 particles created or entangled trough collision (or whatever) in such way they become mirrors of each others. (Like if you have a magic box where you collide 2 green spheres the magic box change the colour on one of them white and the other black). Then the magic box wrap a paper around each sphere so you can't see the colour of them and you send one to China and open the wrapping on the one you kept. If it is white you know immediatly that the one in China is black and vice versa. But that doesn't mean your sphere sent some mystical message to the sphere in China to say I am white so you need to be black..and I have even harder time to believe the said message is sent back in time like "ok in 30 milliseconds I will reveal myself as white so we need to synchronize our watches so you can turn black in the same time"...

4. Ok will do, its just hard to imagine how one would grab 2 entangled photons and put them into fibres and then keep track of them when they move at the speed of light.. I need a rested brain for this :)

5. Well Einstein is pretty good company I guess :)

6. Is it possible to force a particle into opposite spin? Like with my sphere example is it possible for you to dye your sphere from white to black and in the same moment have a chineese observer magically see "his" sphere turn white? This would to me indicate (if the distance was long enough and the clocks would be so accurate you can be sure the change occurs at the exact same time) some information might be sent back in time.. otherwise the idea information is sent back in time is as absurd as the big bang teory. That all matter in the universe was concentrated in 1 really really small point and then expanded from there :)

@phinds yeah I know but I don't like when common sense can't explain things so I opt for the model is incomplete or we lack fundamental understanding of one or more parts that will give a more reasonable explanation :) I say go back to basics and do it again and do it right this time :D No I studied elecronics 20 years ago and back then I also studied some quantum physics and at the time I could get a grip around it and it made sense at the time. Now I haven't studied it in 20 years so I forgot a lot and now I just can't get it to make sense anymore.

 

[phinds]

... I just can't get it to make sense anymore.

Well, see, that's your problem. You expect it to make "sense" but the universe is not obligated to make what WE consider "sense" and in fact fairly it often doesn't, it just does what it does and it's up to us to accommodate it into our belief system.

 

[Thomas T]

Yes, that makes a lot more sense, except for the fact that it's not true. What you've described is a "local hidden variable" theory, which assumes that each particle already has an unknown value for its spin before it is measured, and the two particles have opposite values.

Ok I really need some sleep soon.
I didn't really get your point. Is it false that if particle 1 have U spin in direction 1 particle 2 will always have a D spin in direction 1? Is it false the value is unknown before measured? At best I can get your post to that if you know the spin in direction 1 you can predict to 75% what the spin will be in direction 2 (dependant of the angle between direction 1 and 2) but you won't know for sure before you measure (there is still 25% chance it will be the other direction than you predicted).

I don't have a problem with if you know the spin in 1 direction in one particle you can predict the spin in other angles with different probabilities for the same particle (and if the other particle is an exact mirror of the first particle you can also predict the probability of spin in that particle as well for different directions).

I guess the above is not at all what you are saying. I have to read it again tomorrow, haven't slept for a long time now heh.

@phinds well I know it made more sense to me back then that it does today and I agree the universe is not obligated to make sense to me..I don't know I just have a feeling it made more sense to me 20 years ago than it does now. I am sure I forgot cruicial pieces of the puzzle along the way but I also seen and experienced more so I don't know what is what.. that I forgot or the more I learned about other things made what I once took for granted (because the teacher said so) make less sense.

 

[DrChinese]

3. No, I was unclear. I believe in the entanglement as 2 particles created or entangled trough collision (or whatever) in such way they become mirrors of each others. (Like if you have a magic box where you collide 2 green spheres the magic box change the colour on one of them white and the other black). Then the magic box wrap a paper around each sphere so you can't see the colour of them and you send one to China and open the wrapping on the one you kept. If it is white you know immediatly that the one in China is black and vice versa. But that doesn't mean your sphere sent some mystical message to the sphere in China to say I am white so you need to be black..and I have even harder time to believe the said message is sent back in time like "ok in 30 milliseconds I will reveal myself as white so we need to synchronize our watches so you can turn black in the same time"...

4. Ok will do, its just hard to imagine how one would grab 2 entangled photons and put them into fibres and then keep track of them when they move at the speed of light.. I need a rested brain for this :)

5. Well Einstein is pretty good company I guess :)

6. Is it possible to force a particle into opposite spin? Like with my sphere example is it possible for you to dye your sphere from white to black and in the same moment have a chineese observer magically see "his" sphere turn white? This would to me indicate (if the distance was long enough and the clocks would be so accurate you can be sure the change occurs at the exact same time) some information might be sent back in time.. otherwise the idea information is sent back in time is as absurd as the big bang teory. That all matter in the universe was concentrated in 1 really really small point and then expanded from there :)

3. There is a specific analogy just like yours (except about socks), and you can read about it. It is no coincidental it was written by John Bell. He explains in the most basic terms why this analogy doesn't work.

Bertlmann's Socks and the Nature of Reality
https://cds.cern.ch/record/142461/files/198009299.pdf
https://hal.archives-ouvertes.fr/jpa-00220688/document

But there is no good way to understand why the analogy fails *without* reading and understanding the math of Bell's Theorem itself. It is NOT hard at all. Here is another page I created to explain that:

http://drchinese.com/David/Bell_Theorem_Easy_Math.htm

The super short version is: The "mismatched socks" analogy works great when you measure the 2 particles at the same angle settings. But at most other settings, everything falls apart. Believe it or not, no one noticed that for decades prior to Bell.

4. You keep track of them by extremely accurate timing of their arrival at a detector. Technology has advanced quite far on that score, to the level of femtoseconds (10^-15).

5. Indeed. He died before Bell's paper appeared.

6. Yes and no, and answering that properly is a bit too complicated to explain to you yet. The key thing is that you cannot do something "here" to cause a specific change "there". There is no possibility of faster than light signaling.

 

[stevendaryl]

Ok I really need some sleep soon.
I didn't really get your point. Is it false that if particle 1 have U spin in direction 1 particle 2 will always have a D spin in direction 1?

After you measure both, you will find that they have opposite values. However, it is wrong to assume that that means that they had those values before they were measured.

The heuristic way to understand it is in terms of "measurement collapsing the wave function". Before you measure particle 1, it has a 50/50 chance of being spin-up or spin-down in every direction. After you measure it and find that it is spin-up in direction 1, that "collapses" the other particle to be in the state spin-down in direction 1. If the second particle is spin-down in direction 1, then it has a 25/75 chance of being spin-down in direction 2.

However, there is no way to make it work by assuming that each particle has a definite spin direction at all times.

 

[StevieTNZ]

However, there is no way to make it work by assuming that each particle has a definite spin direction at all times.

What about Bohmian Mechanics? (I did see your post in another thread about different 'interpretations' of QM are distinguishable prediction wise)

 

[stevendaryl]

What about Bohmian Mechanics? (I did see your post in another thread about different 'interpretations' of QM are distinguishable prediction wise)

As I understand it, Bohmian mechanics assumes that particles have a definite position at each moment, but does not assume that they have definite spins. But I'm not familiar with the version that incorporates spin.

 

[David Lewis]

...in the exact same moment the other will fall out of its superposition and take a spin in the other direction than that of the measured particle, right?

No. The spin of the other quantum object is undefined until it is measured.

 

[1977ub]

Any experiment such as the Aspect experiment fits inside some larger light cone. We are not in a position to disprove that local hidden variables spreading out "conspiratorially" through this light cone locally influence the measurement devices and measurement-chosing individuals to create seemingly non-local correlations, true?

 

[PeterDonis]

Any experiment such as the Aspect experiment fits inside some larger light cone. We are not in a position to disprove that local hidden variables spreading out "conspiratorially" through this light cone locally influence the measurement devices and measurement-chosing individuals to create seemingly non-local correlations, true?

False. Violations of the Bell inequalities are violations of the Bell inequalities, regardless of what region of spacetime the measurements occupy. The actual derivation of the inequalities makes no assumptions whatever about the spacetime relationships of the measurements.

If you think it's important that the measurements are spacelike separated (in order to rule out various loopholes that people have proposed), then the fact of them being spacelike separated is invariant; it holds no matter what other "larger light cone" the whole experiment fits inside.

 

[zonde]

Any experiment such as the Aspect experiment fits inside some larger light cone. We are not in a position to disprove that local hidden variables spreading out "conspiratorially" through this light cone locally influence the measurement devices and measurement-chosing individuals to create seemingly non-local correlations, true?

True. But conspiracy loophole can't be closed in any real experiment so the only strategy is to ignore it.
And if you think about it, there is no flaw in such strategy. If nature conspires against us about using some model then this model is indeed useless for us and there is no point in keeping it.

 

[DrChinese]

Any experiment such as the Aspect experiment fits inside some larger light cone. We are not in a position to disprove that local hidden variables spreading out "conspiratorially" through this light cone locally influence the measurement devices and measurement-chosing individuals to create seemingly non-local correlations, true?

Echoing PeterDonis' comment: Your statement misses some important experiments. Entangled particles need NOT ever have been in a common light cone. I.e. they do not need to originate at the same place (have a common source) as in most entanglement experiments.

Obviously, you can hand wave away any experiment you don't like. So what constitutes a disproof to you may not match the opinion of others. However, clearly Bell is something to contend with in any scenario, and coming up with a counter-explanation (along lines you suggest) is quite a challenge.

 

[1977ub]

True. But conspiracy loophole can't be closed in any real experiment so the only strategy is to ignore it.
And if you think about it, there is no flaw in such strategy. If nature conspires against us about using some model then this model is indeed useless for us and there is no point in keeping it.

This approach admits we can't rule out local determinism, and that seems to make more sense than spooky spacelike action-at-a-distance.

 

[PeterDonis]

This approach admits we can't rule out local determinism

We can't if "local determinism" means that not just the states of the entangled particles are determined in advance, but which measurements are going to be made on the entangled particles are also determined in advance, so that the experimenters don't really have any choice about which measurements to make. This view is generally called "superdeterminism" in the literature.

But if the experimenters are free to choose which measurements to make--if that is not determined in advance in coordination with the states of the entangled particles--then no, "local determinism" cannot explain results that violate the Bell inequalities.

 

[bhobba]

No, I was unclear. I believe in the entanglement as 2 particles created or entangled trough collision (or whatever) in such way they become mirrors of each others. (Like if you have a magic box where you collide 2 green spheres the magic box change the colour on one of them white and the other black). Then the magic box wrap a paper around each sphere so you can't see the colour of them and you send one to China and open the wrapping on the one you kept. If it is white you know immediatly that the one in China is black and vice versa. But that doesn't mean your sphere sent some mystical message to the sphere in China to say I am white so you need to be black..and I have even harder time to believe the said message is sent back in time like "ok in 30 milliseconds I will reveal myself as white so we need to synchronize our watches so you can turn black in the same time"...

You are facing the same quandary Bell did - see his original paper:
https://hal.archives-ouvertes.fr/jpa-00220688/document

I can tell you my view but best to glean what you can from the paper first.

Thanks
Bill

 

[zonde]

This approach admits we can't rule out local determinism, and that seems to make more sense than spooky spacelike action-at-a-distance.

Not exactly. We can rule out local determinism as a part of any scientific model for entanglement phenomena, as hypothetical models exploiting superdeterminism (conspiracy loophole) are usually considered non-scientific.

 

[DrChinese]

This approach admits we can't rule out local determinism...

What "this approach" (superdeterminism) admits is that the result of *no* scientific experiment - anywhere, anytime - is an representation of the physical laws of nature. The speed of light is not a constant... it only comes out that way because we can only robotically choose to measure those particular examples that point us to that.

Clearly, this defeats the point of science. That being to describe patterns (and exceptions). Superdeterminism describes nothing. Zip. It is an anti-description.

 

[1977ub]

What "this approach" (superdeterminism) admits is that the result of *no* scientific experiment - anywhere, anytime - is an representation of the physical laws of nature. The speed of light is not a constant... it only comes out that way because we can only robotically choose to measure those particular examples that point us to that.

Clearly, this defeats the point of science. That being to describe patterns (and exceptions). Superdeterminism describes nothing. Zip. It is an anti-description.

Is MWI better in this regard? If science's successes have taught us anything, it is to look for chains of cause-and-effect in the context of fully local causation.

 

[stevendaryl]

What "this approach" (superdeterminism) admits is that the result of *no* scientific experiment - anywhere, anytime - is an representation of the physical laws of nature. The speed of light is not a constant... it only comes out that way because we can only robotically choose to measure those particular examples that point us to that.

Clearly, this defeats the point of science. That being to describe patterns (and exceptions). Superdeterminism describes nothing. Zip. It is an anti-description.

I think you've expressed this a number of times before. But it doesn't seem correct to me. Yes, if you explain some phenomenon by saying that it happens because the initial conditions of the universe were set up to make it happen that way, you haven't actually explained anything UNLESS you have a theory about those initial conditions. So the superdeterministic explanation is not the end of the story; you have to also develop a theory of those initial conditions. So the endeavor isn't vacuous. For example, suppose someone could come up with some not-too-ridiculous differential equation describing the distribution of matter/energy in a spacelike hypersurface, and then proved that the deterministic evolution of this distribution would show EPR-like correlations. To me, that would be a reasonable theory. It would be explanatory and falsifiable.

To say that superdeterministic theories are unscientific because you can explain anything at all by picking initial conditions appropriately is to me comparable to saying that Newtonian physics is unscientific because you can explain anything at all by making up special purpose forces. Newtonian physics as a framework is not falsifiable, but a specific collection of forces is falsifiable.

In my opinion, there is not a big distinction between superdeterminism and retrocausality. If the behavior of a particle depends on both conditions in the past and conditions in the future, then in a certain sense, it's not a causal theory. The laws of physics instead would serve as constraints on the history of the universe. But whereas causal theories allow one end of history (the initial conditions) to be chosen freely, and then the rest of the history is constrained by that choice, in a retrocausal theory, both ends are constrained in some way. If there is only one possibility that makes the whole history consistent, then I don't see how it is any different from a superdeterministic theory: things just evolve the way they have to evolve to make things work out. The retrocausality then would just be a heuristic of calculating this one possibility.

It's a little complicated to try to make this intuition precise. But let's take the EPR experiment. A retrocausal explanation might say that when a twin electron/positron pair is created, it has a hidden variable whose value is influenced by the future measurement choices made by the two experimenters Alice and Bob. But Alice and Bob are themselves physical systems, so their choices are actually determined by facts in their causal past. So the hidden variable created at time $t_1$ depends on facts about Alice and Bob at times $t_2 > t_1$ and $t_3 > t_1$ through retrocausality, but those facts about Alice and Bob in turn depend on facts about the universe at times $t_1$ through ordinary forward causality. Put them together, and you get a constraint about what is happening at time $t_1$. That's superdeterminism. Retrocausality is a way to calculate what the constraints on initial conditions at time $t_1$ must be.

 

[zonde]

Yes, if you explain some phenomenon by saying that it happens because the initial conditions of the universe were set up to make it happen that way, you haven't actually explained anything UNLESS you have a theory about those initial conditions. So the superdeterministic explanation is not the end of the story; you have to also develop a theory of those initial conditions. So the endeavor isn't vacuous. For example, suppose someone could come up with some not-too-ridiculous differential equation describing the distribution of matter/energy in a spacelike hypersurface, and then proved that the deterministic evolution of this distribution would show EPR-like correlations. To me, that would be a reasonable theory. It would be explanatory and falsifiable.

The point is that these initial conditions influence outcome through very complicated processes. Very small and seemingly insignificant changes in initial conditions can lead to wrong result. Another thing is that measurement choices in different experiments are made based on very different processes. In one case it's PRNG, in another it's QRNG, in yet another it's digital version of old movie from which binary information is extracted, and so on. That theory of initial conditions should be so universal that it covers all the possible versions of how experimenters have designed determination of measurement settings.

To say that superdeterministic theories are unscientific because you can explain anything at all by picking initial conditions appropriately is to me comparable to saying that Newtonian physics is unscientific because you can explain anything at all by making up special purpose forces. Newtonian physics as a framework is not falsifiable, but a specific collection of forces is falsifiable.

If you have no control over initial conditions then it might be quite hard to test Newtonian physics. But experiments are designed in such a way that we have some control over initial conditions. In other words we don't judge Newtonian physics just by looking at end result only.

 

[DrChinese]

1. I think you've expressed this a number of times before. But it doesn't seem correct to me. Yes, if you explain some phenomenon by saying that it happens because the initial conditions of the universe were set up to make it happen that way, you haven't actually explained anything UNLESS you have a theory about those initial conditions. So the superdeterministic explanation is not the end of the story; you have to also develop a theory of those initial conditions. So the endeavor isn't vacuous. For example, suppose someone could come up with some not-too-ridiculous differential equation describing the distribution of matter/energy in a spacelike hypersurface, and then proved that the deterministic evolution of this distribution would show EPR-like correlations. To me, that would be a reasonable theory. It would be explanatory and falsifiable.

To say that superdeterministic theories are unscientific because you can explain anything at all by picking initial conditions appropriately is to me comparable to saying that Newtonian physics is unscientific because you can explain anything at all by making up special purpose forces. Newtonian physics as a framework is not falsifiable, but a specific collection of forces is falsifiable.

2. In my opinion, there is not a big distinction between superdeterminism and retrocausality. If the behavior of a particle depends on both conditions in the past and conditions in the future, then in a certain sense, it's not a causal theory.

1. True. I say superdeterminism is neither a theory, nor is it science. In some variation, it says:

a) Bell's Inequality is violated because experimenters do not select a fair sample.
b) The reason the sample is not fair is because experimenters do not truly have the free will to select the sample (or the sample items are biased for reasons other than experimenter sample selection).
c) The bias is *exactly* that which bring the "true" universe in line with local realism; while the biased sample is exactly in line with the predictions of QM.
d) This convoluted mechanism (a/b/c), as far as we know, only manifests itself in tests of local realism. In all other areas of science, a sample is representative of the full universe.

Yes: we agree there is nothing falsifiable here, and therefore there is no true current theory of superdeterminism. You believe a "superdeterministic explanation is not the end of the story; you have to also develop a theory of those initial conditions". My issue here is your assertion this is not the end of the story. You don't know that a serious true theory could exist, that is speculation. When one is offered, then perhaps I might agree with you. However, the obstacles that would need to be overcome are so great, I don't see how any future theory could address those without a complete re-write of most of the standard model. Oh, and probably a re-write of standard cosmology as well.

No: I don't believe there are ANY initial conditions of the universe that would lead to the Bell test experimental bias a la superdeterminism. That is not the specific objection I have. My objection is that superdeterminism was invented simply as a possible way to get around the implications of Bell's Theorem. I don't believe the early originators of the idea (probably including Bell himself) ever took this seriously. But any serious consideration of this leads to anti-scientific conclusions. Basically, we can say that the laws of physics appear to be X in experiments, but are actually Y (where Y is anything I care to hand pick). Why not assert that c is NOT actually a constant? It's just a sampling issue. Perhaps the true value of c varies from .5c to 1.5c.

This is hand waving at its very worst,. I say again: there is no science here. Not yet anyway. When there is some, we can debate that. Personally, I would be happy if superdeterminism were banned from the quantum physics forum until there were a suitable published reference cited for us to consider. (And please: don't mention the Gerard 't Hooft papers; these don't even touch the surface of the objections to superdeterminism.) 2. I don't get the comparison to retrocausality. I would agree that a retrocausal theory is not causal, but that is hardly more than a personal preference on your part. True, some retrocausal theories (Wheeler's or Cramer's Absorber Theory) barely are more than a framework for an interpretation. But there are acausal theories (Relational BlockWorld being one) that are being well developed and are quite serious. The difference is that these seek to explain orthodox QM at some level. They do not propose that QM is actually wrong.

And superdeterminism does say QM is wrong: it says that the true correlation rate for a suitable photonic Bell test (angle settings for Alice and Bob 120 degrees apart) is .333 (at the local realistic limit), while QM says it is .25 (which is observed in test samples).

 

[stevendaryl]

1. True. I say superdeterminism is neither a theory, nor is it science.

Superdeterminism is not a theory, but it may be a FEATURE of a theory.

Yes: we agree there is nothing falsifiable here, and therefore there is no true current theory of superdeterminism. You believe a "superdeterministic explanation is not the end of the story; you have to also develop a theory of those initial conditions". My issue here is your assertion this is not the end of the story. You don't know that a serious true theory could exist, that is speculation.

Of course. But it seemed to me that you were saying that there can't possibly be a satisfactory superdeterministic theory.

 

[stevendaryl]

2. I don't get the comparison to retrocausality.

Retrocausality is a way of computing superdeterministic initial conditions.

And superdeterminism does say QM is wrong: it says that the true correlation rate for a suitable photonic Bell test (angle settings for Alice and Bob 120 degrees apart) is .333 (at the local realistic limit), while QM says it is .25 (which is observed in test samples).

What? That doesn't make any sense. You're saying that a superdeterministic explanation of EPR correlations doesn't explain EPR correlations? Your complaint about superdeterminism is BOTH that it is unfalsifiable, AND that it is falsified?

If you know ahead of time that Alice is going to measure her photon's polarization at angle $\theta_A$, and that Bob is going to measure his photon's polarization at angle $\theta_B$, then you could certainly choose polarizations such that Alice and Bob have the correct correlations. With 50% probability, you either:

1. Send photons polarized at angle $\theta_A$ to both Alice and Bob, or
2. Send photons polarized at angle $90^o + \theta_A$ to both.

where's the $.333$ coming from?

 

[DrChinese]

Of course. But it seemed to me that you were saying that there can't possibly be a satisfactory superdeterministic theory.

I have thought about that point numerous times. I keep thinking I should put together a list of the hurdles one would need to jump over to end up with local realism and superdeterminism as a combination.

Basically, all entangled particle pairs that are ever created would need to know how they are going to be measured in advance so they can - if they are being detected - give the correct statistics (per QM predictions, presumably the "true" statistics otherwise). Since you can entangle particles that have never coexisted, that adds a whole new set of requirements too. Each source must be passing on the needed information to every particle it creates. And so macroscopic systems (lasers) need to know what information to impart to the particles created.

Seems unlikely to me, but like Lloyd in Dumb and Dumber: "So you're telling me there's a chance..."

 

[DrChinese]

1. Retrocausality is a way of computing superdeterministic initial conditions.

2. What? That doesn't make any sense. You're saying that a superdeterministic explanation of EPR correlations doesn't explain EPR correlations? Your complaint about superdeterminism is BOTH that it is unfalsifiable, AND that it is falsified?

If you know ahead of time that Alice is going to measure her photon's polarization at angle $\theta_A$, and that Bob is going to measure his photon's polarization at angle $\theta_B$, then you could certainly choose polarizations such that Alice and Bob have the correct correlations. With 50% probability, you either:

1. Send photons polarized at angle $\theta_A$ to both Alice and Bob, or
2. Send photons polarized at angle $90^o + \theta_A$ to both.

where's the $.333$ coming from?

1. I don't understand this statement, and don't see how one relates to the other.

2. Normally, superdeterminism goes with local realism. We know from Bell that the local realistic limit for your example* (Type I PDC) is .333 for Theta(A-B)=120 degrees. We know that the observed value is .250. Ergo the superdeterministic explanation covers the difference, refuting the so-called "fair sampling" assumption of most Bell tests (never mind that "loophole" has been closed). In superdeterminism, the observed sample yields an average that is different than the total universe. I call it the "Unfair Sampling Assumption". That unfair sample can be achieved by taking away (apparent) freedom of choice, and/or by

Superdeterminism is designed for one and only one purpose: to provide a local realistic explanation of observed results. It is not just a question of initial conditions, because you need physical laws to go with it that guarantee that the quantum predictions are met on each test run.

*Of course since there is realism, the photons cannot be polarized at anyone specific angle. That would lead to different statistics entirely, product state statistics. Those are quite different than the entangled state stats.

 

[DrChinese]

stevendaryl (and all): If we have veered to far away from the OP, I accept responsibility for that. If needed, we could break this off to a new thread. Somehow I think our positions are not that far apart. We may define "superdeterminism" in different terms. Considering there is no formal definition anywhere that I know of, that wouldn't be surprising.

-DrC

 

[zonde]

2. I don't get the comparison to retrocausality. I would agree that a retrocausal theory is not causal, but that is hardly more than a personal preference on your part. True, some retrocausal theories (Wheeler's or Cramer's Absorber Theory) barely are more than a framework for an interpretation. But there are acausal theories (Relational BlockWorld being one) that are being well developed and are quite serious. The difference is that these seek to explain orthodox QM at some level. They do not propose that QM is actually wrong.

Retrocausality and global constraint of RBW is the same superdeterminism when analyzed by observer who experiences time ordered events.
It's a bit paradoxical that you have such a strong position against superdeterminism but at the same time claim that global constraint of RBW is "quite serious".

 

[DrChinese]

Retrocausality and global constraint of RBW is the same superdeterminism when analyzed by observer who experiences time ordered events.

I have no idea what you are talking about. There is no similarity in RBW - a theory with published papers and a book on it - and superdeterminism (something which defies any meaningful description whatsoever).

1. What "global constraint" are you referring to in RBW?

2. And time ordering in RBW is simply a small slice which precisely corresponds to a quantum context. That should hardly be controversial, since that context is part of the orthodox quantum prediction (since QM is contextual).

 

[zonde]

1. What "global constraint" are you referring to in RBW?

How do you think RBW explains EPR-type correlations? According to RBW past, present and future co-determine each other via "adynamical global constraints", that's how it explains them. (See description of the book "Beyond the Dynamical Universe: Unifying Block Universe Physics and Time as Experienced")

 

[DrChinese]

According to RBW past, present and future co-determine each other via "adynamical global constraints", that's how it explains them. (See description of the book "Beyond the Dynamical Universe: Unifying Block Universe Physics and Time as Experienced")

Good book.

Yes, that is the set that makes up a block. It would include the entangled particle source, and Alice and Bob's measurement choices. That is the exact same context you have in orthodox QM, and because they are in "contact" (as part of the block) they can explain the EPR correlations. Of course, causality is sacrificed, which is why it is described as "adynamical".

What does any of that have to do with superdeterminism? Where there is no book to reference? It is just made up, not even worthy of a napkin that I have seen.

 

[DarMM]

Yes, that is the set that makes up a block. It would include the entangled particle source, and Alice and Bob's measurement choices. That is the exact same context you have in orthodox QM, and because they are in "contact" (as part of the block) they can explain the EPR correlations. Of course, causality is sacrificed, which is why it is described as "adynamical".

I assume by EPR correlations you mean CSHS correlations or similar as the EPR correlations have a local classical model.

Anyway what I really mean to ask is how does replicating nonclassical correlations work in RBW? I've heard the interpretation has ontological emergence, what does this mean?

Sorry to ask, the only proper source seems to be the new book, which I don't have yet. Also it was superdetermined that I would ask.