On Thu, Dec 23, 2021 at 10:01 AM Jesse Mazer <[email protected]> wrote:
> On Wed, Dec 22, 2021 at 4:54 PM Bruce Kellett <[email protected]> > wrote: > >> On Wed, Dec 22, 2021 at 10:12 PM smitra <[email protected]> wrote: >> >>> On 21-12-2021 22:48, Bruce Kellett wrote: >>> > >>> >> > Rubbish. If there were a common cause, then that would have to depend >>> > on the final polarizer orientations. And those are not known at the >>> > time of creation of the entangled pair. You are, then, back with some >>> > non-local influence (or retro-causation). >>> >>> The setting of the polarizers will be the result of some physical >>> process. Whatever you specify for that process should be included in the >>> analysis of the problem. But when you do so, it's inevitable that in an >>> MWI analysis, there is not going to be any nonlocal effect other than >>> trivial common cause effects. >>> >> >> I see. So in desperation you resort to the superdeterminism escape. >> > > I don't think Saibal was referring to superdeterminism? Or are you > suggesting the MWI version of locality involves superdeterminism? If so > that's wrong, superdeterminism involves some special constraint on initial > conditions such that variables associated with the entangled particles > (hidden or non-hidden) at the moment they are sent out in opposite > directions are correlated in advance with the future choices of detector > settings by the experimenters. > > MWI is not necessary for the understanding of the correlations of >> entangled particles, as my simple example shows. In an actual experiment, >> the analysis is identical in many-worlds and collapse models. The >> additional worlds in MWI add nothing to the explanation. >> > > They allow it to be local without superdeterminism, because the "matching" > of local worlds can be done at a point in spacetime that has both > experimenter's measurements in its past light cone, I gave you a toy model > demonstrating how this can work in my post at > https://www.mail-archive.com/[email protected]/msg91022.html > > > One way to think about local vs. non-local explanations is to imagine > running a *simulation* of a Bell type experiment, using three or more > separate computers that are each responsible for simulating what's going on > in a localized region of space, say the location of experimenter A > ('Alice'), the location of experimenter B ('Bob'), and the location of the > emitter midway between them. The computer simulating the location of the > emitter has to run some algorithm that assigns states to the two emitted > particles > The whole point of the entanglement is that there are no separately assigned states to the two particles. They are in an entangled, non-separable, state. So that the particle that goes to A is non-locally linked to the particle that goes to B. There is a simple rule here. If the particles interact only locally, then the joint state is separable. If the state is non-separable, the interactions are non-local. All local states are separable. Non-separable states are non-local. Modus tollens. (the algorithm is allowed to involve something like a random number > generator, it need not be deterministic), and then it can transmit some or > all of that information to the computers simulating the locations of Alice > and Bob. Then once the computer simulating Alice's location receives that > information about the state of the simulated particle arriving there, it > runs some algorithm to decide what detector setting Alice selects, and what > happens in that local region when she measures the particle with that > detector setting (again we are allowed to use a random number generator), > and the computer simulating Bob's location does the same. If we want to > simulate a model of physics that obeys locality, then computers simulating > events with a spacelike separation, like Alice performing her measurement > and Bob performing his, cannot be in communication at the moment they each > compute the local outcome at their own location. And if we want to avoid > superdeterminism, the computer simulating the emitter cannot have any way > to predict in advance what measurement setting Alice and Bob are going to > use at their own locations--over many trials, the states it assigns to the > particles on each trial cannot be statistically correlated with the future > choices of detector settings by Alice and Bob on that trial. > > A simulation based on a MWI style toy model could respect both of these > conditions, locality (no communication between computers when they are > determining the results of events that are supposed to be at a spacelike > separation, like Alice's measurement and Bob's measurement) and non > superdeterminism (the computer simulating the emitter generates the states > to send to Alice and Bob with no advanced knowledge of what detector > setting they are going to choose in the future). The twist with the MWI is > that the computers simulating Alice and Bob's locations don't generate > unique outcomes, but rather a collection of different outcomes for > different simulated copies of Alice and Bob. If all the copies then send > signals reporting their results back towards the location of the emitter at > the midpoint between them, then the "matching" between copies of Alice and > copies of Bob > Again, the trouble here is that your model requires this "matching" operation, which is not present in the actual experiments, such as that of Aspect. In the real-world experiments, there is no further interaction at the point where Alice and Bob meet. All they have to do is calculate the correlations implied by the data in their respective lab books. The Alice and Bob that meet up (and there are 2^N such Alice-Bob pairs in MWI) they simply compare notes. The non-locality of the entangled state means that they were always both in the same world, so no "matching up" is required. There is no work for the third central computer to do in the real-world experiments. can be done in the computer simulating the location of the emitter, when it > has had time to receive signals traveling at the speed of light or slower > from the Alice-computer and the Bob-computer, thus there is no violation of > the locality condition. If for example there is a third observer, Carla, > who is at the location of the emitter and waiting to receive signals from > Alice and Bob, Carla only gets split into different copies receiving > different possible messages from Alice and Bob at point when Alice's > measurement and Bob's measurement are both in her past light cone. > > In contrast, no simulated rules for a *single* world could reproduce Bell > inequality violating statistics in a situation like this with 3 distinct > computers for each local region, not if we imposed the constraints of > locality and non-superdeterminism described above. I think you claimed > earlier that your fGRW idea is supposed to be local, but if you tried to > formalize it sufficiently to build rules for a simulation subject to these > constraints, it wouldn't be able reproduce Bell inequality violating > statistics either. > The trouble with these computer simulation models is that the conditions are generally set in a way that does not reflect the real-world experiment. Given a simulation, you can set it up to prove anything you want. The requirement is that you provide a local explanation of something like the actual Aspect experiment -- where there is no third computer or further interaction when the data from the separate ends of the experiment are used to calculate correlations (or probabilities). Incidentally it has never been claimed that fGRW gives a local account of these experiments. fGRW is Lorentz invariant and fully relativistic. But it also respects the non-locality of quantum mechanics. I think that since you have not refuted my no-go theorem for locality preservation in MWI, there is no real point in my finding all the errors made in the endless attempts to give local MWI accounts of Bell inequality violations. Bruce -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/CAFxXSLS92zf3uhGP9Dh0Ny0d37haS3kas6JifFs5xDnc1JUtYQ%40mail.gmail.com.
