18 ms·

Can retrocausality solve the puzzle of action at a distance?

113 points by abhishekjha 9y ago

Simon_says 9y ago
No
- ghusbands 9y ago
  Or yes, but only if you accept time travel.
crystaln 9y ago
Also quantum erasure.
analogic 9y ago
Yes
comnetxr 9y ago
They say that action at a distance is "not compatible with special relativity" but retrocausality is because it's not full retrocausality (can't send signals back in time) but just a weaker kind. I think they've missed the point on that as they've just copied the argument for why action at a distance is compatible with special relativity.
- simonh 9y ago
  That's not what they are saying at all. Your pre-judging the outcome by calling the end result 'action at a distance', but action at a distance is a proposed explanation of the outcome, not the outcome itself. That's what they are trying to explain. Retrocausality is an alternative explanation of the outcome to action at a distance.
UncleSlacky 9y ago
Isn't this just Cramer's transactional interpretation of quantum mechanics? https://en.wikipedia.org/wiki/Transactional_interpretation https://en.wikipedia.org/wiki/Transactional_interpretation
- spindle 9y ago
  It's more general than that, and also Price and Wharton don't entirely agree with Cramer about what needs to be proved, but it's similar, yes. It's inferior to the Transactional theory in one way, which is that Cramer's exposition of his theory comes with pictures of elephants.
dvh 9y ago
I don't understand. Isn't action at a distance this: you take red and green marble and randomly put them in two bags. You take one bag with you to Pluto. Then you open the bag and see the red one, now you instantly know what marble is in the bag that stayed on Earth. This doesn't break any law so what's the problem?
- Twisol 9y ago
  In the hidden variables case (yours), it was predetermined which marble was in the bag before you opened it. In the retrocausal case, the information about which marble is in the bag simply isn't part of reality until you've checked. With macroscopic objects like bags and marbles, this is ridiculous, because someone had to put the marble in the bag (a "shared cause"). However, Bell's Theorem shows that entanglement isn't a sufficient "shared cause" to explain his experiment. Thus, we either need superdeterminism (the "shared cause" is the origin of the universe!), spooky action at a distance, or retrocausality.
  - RobertoG 9y ago
    >>" the information about which marble is in the bag simply isn't part of reality until you've checked" It seems to me that the universe uses lazy evaluation (1). (1) https://en.wikipedia.org/wiki/Lazy_evaluation https://en.wikipedia.org/wiki/Lazy_evaluation
    - Twisol 9y ago
      It's even weirder than that, though, because with lazy evaluation you still have a complete description of the state involved; it's just a matter of unpacking it by executing the description. That's more analogous with a hidden variables theory. With retrocausality, it seems that performing an observation fundamentally adds information to that which has been observed. If retrocausality can be likened to a computational mechanism, I'd be really interested to know what it is! Like I said in another comment, it seems like there's some element of monotonic computation involved, but the brand of monotonicity is quite unusual.
Twisol 9y ago
This is a really weird, but really neat concept. Where hidden variable theories assert that the state of a particle is fixed at any given point in time (but that quantum descriptions don't fully capture this state), retrocausality seems to say that the state of a particle is under-determined at any given point in time, and when the undetermined fragment of state is needed to determine the outcome of some future experiment, the result of the experiment adds information to the past in a consistent manner. But this has the side-effect of correlating all experiments that depend on this state! I think this also clarifies why retrocausality doesn't allow information to be transmitted back in time. Every observation on the under-determined state in the past by definition doesn't rely on the undetermined fragment, and hence is consistent with whatever choice occurs in the future. There's some flavor of monotonicity here, but it's pretty bizzare!
- nabla9 9y ago
  It seems to be similar to Murray Gell-Mann's decoherent histories approach (aka consistent histories) where the state of a system can be described as additive sum of all possible histories.
gUXqR8m58CIS 9y ago
This kind of reminds me of something I thought of a while ago when playing around with cellular automata in Golly; With many cellular automata rules, you get chaotic universes, which slowly stabilize into some final state. Frequently there are small active regions which take longer to stabilize, and which may travel across the entire state before they eventually do. So could our 4-dimensional universe be the final stable state in a 5-dimensional cellular automaton (or some similar structure)? This would allow every part of the universe to influence every other part without regard for space/time separation, but the rules could still be set up such that the particles in the final state adhere to causality, don't exceed the speed of light, etc... I think it might only work if the final state represents a fully collapsed universe without superpositions. I'm not sure how realistic that is.
simonh 9y ago
What confuses me about these experiments is that, per special relativity, from the point of view of the photon the moment of it's emission and detection are the same. It's traveling at the speed of light, so no time passes between those events. So talking about it's state at this or that time is not talking about it's intrinsic state, only our external view of it from our frame of reference. But in terms of a description of the photon in it's frame of reference, that's meaningless. EDIT: Or is this what the Paris Zigzag does - treat the state of the photon and it's twin as a single state, regardless of our time-based view of it, incorporating the moment of detection, emission and detection of the entangled twin within the same description?
- jerf 9y ago
  I favor this quantum interpretation a bit for very similar reasons to that. Relativity already tore away our ideas about time being a single steady march forward in an independent fourth dimension, which was a big step. Similarly, I don't consider this "retrocausality"; I consider it a sign that time works differently than we thought yet again, and in fact all qubits and their interactions are probably actually moving forward in time, in what is arguably the most real sense of "time". However, in that view, time is shaped very, very funny relative to what we humans would consider "time", because that means that a photon in the cosmic background radiation emitted that many billions of years ago that hits one of our instruments is still a single atomic event. I consider this view supported by relativity where, as you mention, technically a photon is an atomic event from its own point of view. There isn't any before or after in its own frame of reference. What we call time would then be a secondary phenomenon on top of this much more complicated (from our point of view) time organization of the universe. (However, I reject the idea that that makes "our" time an illusion, or some BS like that. Our human conception of time is perfectly real. It just may not be fundamental. But even if we discover that "real time" is something like what I'm hypothesizing here, human time will still be as real as ever.) What is not clear to me is whether or not this could be set up in a reasonable manner. It's on my long list of things I'd love to have time to try my hand at. It's feasible, you just need to start with maybe a 5 or 6 qubit system that interacts in this "real time" manner in the counterintuitive "retrocausal" manner, but for which there is a point of view in which it all happens forward in this "real time" metric. You don't have to simulate an entire universe to show this is a feasible idea, if you can establish it at that scale, it is merely a matter of extreme complication to assume the universe could work that way. There's also an interesting implication for the "deistic" point of view. If this is how quantum mechanics "really works" then it becomes relatively easy to imagine that the Great Simulator is simulating the universe, but actually has little to no power to influence it, because from his point of view during the computation the entire universe is a tangled mess of everything essentially happening at once (from our point of view). The only thing the Great Simulator might be able to do is examine the final state of the universe after it has finished computing, at which point the entire history of the universe would be open to him, but to affect the state of the universe would be wildly more computationally expensive than the simulation itself, because any attempt to insert or modify even a single qubit's worth of change would essentially require the entire thing to be recalculated again as the changes propagate in (from our point of view) all the way forward and backwards in time again. Depending on the nature of "true time", though, it might be possible, depending on exactly how it ends up working.
EtDybNuvCu 9y ago
All of our reasoning would be much easier if free will were an illusion. Perhaps we should stop fighting it.
- roywiggins 9y ago
  Superdeterminism is even weirder than that though. Why would the universe conspire to make sure that I picked my observation angle just right to maintain Bell's inequality? If you use the outcome of a celestial event to determine the angle, then you've got an absolute enormous conspiracy just to make sure that Bell's inequality isn't violated. It's deeply weird.
- dmichulke 9y ago
  If free will were an illusion, it wouldn't make sense to try to convince anybody of it. IOW, if you're right, your comment is useless :-) But I'd understand why you had to write that anyway :-P
- yters 9y ago
  I always thought the opposite. It seems things like non-determinism and retrocausality made free will more physically coherent instead of less coherent. It is the strictly deterministic universe of Newton that makes free will illusive.
keithnoizu 9y ago
My pet theory is when you make an observation you're narrowing down which subset of the many universes you might reside in. So the particles simply are going to behave the same even though there is no way to know what reality you'll get when you check on the first one. However once you do check on the first one it reduces the subset of possible realities you are in which includes possible outcomes for observation of the paired particle.
ericb 9y ago
We are in a branching simulation. It is lazily loaded and immutable. For undetermined variables, copy on read causes a branch filled in with each possible result. So many-worlds, but within a simulation.
- leggomylibro 9y ago
  Have you ever read the Discworld series? There's a side plot where a group of wizards uses a computer to explore that sort of idea: --- ...The hypothesis behind invisible writings was laughably complicated. All books are tenuously connected through L-space and, therefore, the content of any book ever written or yet to be written may, in the right circumstances, be deduced from a sufficiently close study of books already in existence. Future books exist in potentia, as it were, in the same way that a sufficiently detailed study of a handful of primal ooze will eventually hint at the future existence of prawn crackers. But the primitive techniques used hitherto, based on ancient spells like Weezencake's Unreliable Algorithm, had meant that it took years to put together even the ghost of a page for an unwritten book. It was Ponder's particular genius that he had found a way around this by considering the phrase, "How do you know it's not possible until you've tried?" And experiments with Hex, the University's thinking engine, had found that, indeed, many things are not impossible until they have been tried. Like a busy government which only passes expensive laws prohibiting some new and interesting thing when people have actually found a way of doing it, the universe relied a great deal on things not being tried at all. When something is tried, Ponder found, it often does turn out to be impossible very quickly, but it takes a little while for this really to be the case–in effect, for the overworked laws of causality to hurry to the scene and pretend it has been impossible all along. Using Hex to remake the attempt in minutely different ways at very high speed had resulted in a high success rate, and he was now assembling whole paragraphs in a matter of hours. "It's like a conjurin' trick, then," Ridcully had said. "You're pullin' the tablecloth away before all the crockery has time to remember to fall over." And Ponder had winced and said, "Yes, exactly like that, Archchancellor. Well done." And that had led to all the trouble with How to Dynamically Manage People for Dynamic Results in a Caring Empowering Way in Quite a Short Time Dynamically... - Terry Pratchett, The Last Continent
  - 9y ago
SteveJS 9y ago
How does this interact with pilot wave theory? Is the article saying that the causal direction of time is undefined within a quantum entangled system? Or is that more than what it’s claiming?
tunesmith 9y ago
So is this saying that if retrocausality is possible, then it's akin to the kind of time travel where you may have altered something in the past, but the effects would not manifest until "after" the commencement of the time travel? Like in a person, if you feel some sort of uncomfortableness that leads you to make a choice, maybe it's because a minute ago, some time traveler traveled back a couple of weeks and planted a seed that made you start feeling uncomfortable! Maybe all of those, "You know, I never realized until now, but it turns out I've disliked broccoli for years, I don't know why I eat it every day!" statements are a result of retro causal time travel. (In this case, George HW Bush on an anti-broccoli time-traveling crusade.)
AceyMan 9y ago
I've had this puzzle solved in my own head for quite some time thusly: Distance is a fiction /aka: arbitrary manifold created by our (conscious?) selves interacting with "the world," which is really just a big array of ordered values (N-tuples). To say that "X is N distance from Y" is no more of a valid proposition than "Blue is Louder than Yellow". We attach lots of significance to distance (e.g., 3D space) owing to its utility for "doing science." (Insert "All models are wrong; some are useful", but at a grand, multiversal scale.) My 2¢: YMMV. (me: Philosophy major with a good bit of sci/math as well)
mikhailfranco 9y ago
Perhaps retrocausality is just non-locality in time.
ggggtez 9y ago
"we can see that it couldn’t be used to signal, for the same reason that entanglement itself can’t be used to signal" Nope, you lost me. This seems too hand wavy. Also "This isn’t true of many everyday processes – eggs turn into omelettes, but not the reverse!"... sigh.

113 points by abhishekjha 9y ago

Simon_says 9y ago

ghusbands 9y ago
Or yes, but only if you accept time travel.

crystaln 9y ago

Also quantum erasure.

analogic 9y ago

Yes

comnetxr 9y ago

They say that action at a distance is "not compatible with special relativity" but retrocausality is because it's not full retrocausality (can't send signals back in time) but just a weaker kind. I think they've missed the point on that as they've just copied the argument for why action at a distance is compatible with special relativity.

simonh 9y ago
That's not what they are saying at all. Your pre-judging the outcome by calling the end result 'action at a distance', but action at a distance is a proposed explanation of the outcome, not the outcome itself. That's what they are trying to explain. Retrocausality is an alternative explanation of the outcome to action at a distance.

UncleSlacky 9y ago

Isn't this just Cramer's transactional interpretation of quantum mechanics? https://en.wikipedia.org/wiki/Transactional_interpretation https://en.wikipedia.org/wiki/Transactional_interpretation

spindle 9y ago
It's more general than that, and also Price and Wharton don't entirely agree with Cramer about what needs to be proved, but it's similar, yes. It's inferior to the Transactional theory in one way, which is that Cramer's exposition of his theory comes with pictures of elephants.

dvh 9y ago

I don't understand. Isn't action at a distance this: you take red and green marble and randomly put them in two bags. You take one bag with you to Pluto. Then you open the bag and see the red one, now you instantly know what marble is in the bag that stayed on Earth. This doesn't break any law so what's the problem?

Twisol 9y ago
In the hidden variables case (yours), it was predetermined which marble was in the bag before you opened it. In the retrocausal case, the information about which marble is in the bag simply isn't part of reality until you've checked. With macroscopic objects like bags and marbles, this is ridiculous, because someone had to put the marble in the bag (a "shared cause"). However, Bell's Theorem shows that entanglement isn't a sufficient "shared cause" to explain his experiment. Thus, we either need superdeterminism (the "shared cause" is the origin of the universe!), spooky action at a distance, or retrocausality.
- RobertoG 9y ago
  >>" the information about which marble is in the bag simply isn't part of reality until you've checked" It seems to me that the universe uses lazy evaluation (1). (1) https://en.wikipedia.org/wiki/Lazy_evaluation https://en.wikipedia.org/wiki/Lazy_evaluation
  - Twisol 9y ago
    It's even weirder than that, though, because with lazy evaluation you still have a complete description of the state involved; it's just a matter of unpacking it by executing the description. That's more analogous with a hidden variables theory. With retrocausality, it seems that performing an observation fundamentally adds information to that which has been observed. If retrocausality can be likened to a computational mechanism, I'd be really interested to know what it is! Like I said in another comment, it seems like there's some element of monotonic computation involved, but the brand of monotonicity is quite unusual.

Twisol 9y ago

This is a really weird, but really neat concept. Where hidden variable theories assert that the state of a particle is fixed at any given point in time (but that quantum descriptions don't fully capture this state), retrocausality seems to say that the state of a particle is under-determined at any given point in time, and when the undetermined fragment of state is needed to determine the outcome of some future experiment, the result of the experiment adds information to the past in a consistent manner. But this has the side-effect of correlating all experiments that depend on this state! I think this also clarifies why retrocausality doesn't allow information to be transmitted back in time. Every observation on the under-determined state in the past by definition doesn't rely on the undetermined fragment, and hence is consistent with whatever choice occurs in the future. There's some flavor of monotonicity here, but it's pretty bizzare!

nabla9 9y ago
It seems to be similar to Murray Gell-Mann's decoherent histories approach (aka consistent histories) where the state of a system can be described as additive sum of all possible histories.

gUXqR8m58CIS 9y ago

This kind of reminds me of something I thought of a while ago when playing around with cellular automata in Golly; With many cellular automata rules, you get chaotic universes, which slowly stabilize into some final state. Frequently there are small active regions which take longer to stabilize, and which may travel across the entire state before they eventually do. So could our 4-dimensional universe be the final stable state in a 5-dimensional cellular automaton (or some similar structure)? This would allow every part of the universe to influence every other part without regard for space/time separation, but the rules could still be set up such that the particles in the final state adhere to causality, don't exceed the speed of light, etc... I think it might only work if the final state represents a fully collapsed universe without superpositions. I'm not sure how realistic that is.

simonh 9y ago

What confuses me about these experiments is that, per special relativity, from the point of view of the photon the moment of it's emission and detection are the same. It's traveling at the speed of light, so no time passes between those events. So talking about it's state at this or that time is not talking about it's intrinsic state, only our external view of it from our frame of reference. But in terms of a description of the photon in it's frame of reference, that's meaningless. EDIT: Or is this what the Paris Zigzag does - treat the state of the photon and it's twin as a single state, regardless of our time-based view of it, incorporating the moment of detection, emission and detection of the entangled twin within the same description?

jerf 9y ago
I favor this quantum interpretation a bit for very similar reasons to that. Relativity already tore away our ideas about time being a single steady march forward in an independent fourth dimension, which was a big step. Similarly, I don't consider this "retrocausality"; I consider it a sign that time works differently than we thought yet again, and in fact all qubits and their interactions are probably actually moving forward in time, in what is arguably the most real sense of "time". However, in that view, time is shaped very, very funny relative to what we humans would consider "time", because that means that a photon in the cosmic background radiation emitted that many billions of years ago that hits one of our instruments is still a single atomic event. I consider this view supported by relativity where, as you mention, technically a photon is an atomic event from its own point of view. There isn't any before or after in its own frame of reference. What we call time would then be a secondary phenomenon on top of this much more complicated (from our point of view) time organization of the universe. (However, I reject the idea that that makes "our" time an illusion, or some BS like that. Our human conception of time is perfectly real. It just may not be fundamental. But even if we discover that "real time" is something like what I'm hypothesizing here, human time will still be as real as ever.) What is not clear to me is whether or not this could be set up in a reasonable manner. It's on my long list of things I'd love to have time to try my hand at. It's feasible, you just need to start with maybe a 5 or 6 qubit system that interacts in this "real time" manner in the counterintuitive "retrocausal" manner, but for which there is a point of view in which it all happens forward in this "real time" metric. You don't have to simulate an entire universe to show this is a feasible idea, if you can establish it at that scale, it is merely a matter of extreme complication to assume the universe could work that way. There's also an interesting implication for the "deistic" point of view. If this is how quantum mechanics "really works" then it becomes relatively easy to imagine that the Great Simulator is simulating the universe, but actually has little to no power to influence it, because from his point of view during the computation the entire universe is a tangled mess of everything essentially happening at once (from our point of view). The only thing the Great Simulator might be able to do is examine the final state of the universe after it has finished computing, at which point the entire history of the universe would be open to him, but to affect the state of the universe would be wildly more computationally expensive than the simulation itself, because any attempt to insert or modify even a single qubit's worth of change would essentially require the entire thing to be recalculated again as the changes propagate in (from our point of view) all the way forward and backwards in time again. Depending on the nature of "true time", though, it might be possible, depending on exactly how it ends up working.

EtDybNuvCu 9y ago

All of our reasoning would be much easier if free will were an illusion. Perhaps we should stop fighting it.

roywiggins 9y ago
Superdeterminism is even weirder than that though. Why would the universe conspire to make sure that I picked my observation angle just right to maintain Bell's inequality? If you use the outcome of a celestial event to determine the angle, then you've got an absolute enormous conspiracy just to make sure that Bell's inequality isn't violated. It's deeply weird.
dmichulke 9y ago
If free will were an illusion, it wouldn't make sense to try to convince anybody of it. IOW, if you're right, your comment is useless :-) But I'd understand why you had to write that anyway :-P
yters 9y ago
I always thought the opposite. It seems things like non-determinism and retrocausality made free will more physically coherent instead of less coherent. It is the strictly deterministic universe of Newton that makes free will illusive.

keithnoizu 9y ago

My pet theory is when you make an observation you're narrowing down which subset of the many universes you might reside in. So the particles simply are going to behave the same even though there is no way to know what reality you'll get when you check on the first one. However once you do check on the first one it reduces the subset of possible realities you are in which includes possible outcomes for observation of the paired particle.

ericb 9y ago

We are in a branching simulation. It is lazily loaded and immutable. For undetermined variables, copy on read causes a branch filled in with each possible result. So many-worlds, but within a simulation.

leggomylibro 9y ago
Have you ever read the Discworld series? There's a side plot where a group of wizards uses a computer to explore that sort of idea: --- ...The hypothesis behind invisible writings was laughably complicated. All books are tenuously connected through L-space and, therefore, the content of any book ever written or yet to be written may, in the right circumstances, be deduced from a sufficiently close study of books already in existence. Future books exist in potentia, as it were, in the same way that a sufficiently detailed study of a handful of primal ooze will eventually hint at the future existence of prawn crackers. But the primitive techniques used hitherto, based on ancient spells like Weezencake's Unreliable Algorithm, had meant that it took years to put together even the ghost of a page for an unwritten book. It was Ponder's particular genius that he had found a way around this by considering the phrase, "How do you know it's not possible until you've tried?" And experiments with Hex, the University's thinking engine, had found that, indeed, many things are not impossible until they have been tried. Like a busy government which only passes expensive laws prohibiting some new and interesting thing when people have actually found a way of doing it, the universe relied a great deal on things not being tried at all. When something is tried, Ponder found, it often does turn out to be impossible very quickly, but it takes a little while for this really to be the case–in effect, for the overworked laws of causality to hurry to the scene and pretend it has been impossible all along. Using Hex to remake the attempt in minutely different ways at very high speed had resulted in a high success rate, and he was now assembling whole paragraphs in a matter of hours. "It's like a conjurin' trick, then," Ridcully had said. "You're pullin' the tablecloth away before all the crockery has time to remember to fall over." And Ponder had winced and said, "Yes, exactly like that, Archchancellor. Well done." And that had led to all the trouble with How to Dynamically Manage People for Dynamic Results in a Caring Empowering Way in Quite a Short Time Dynamically... - Terry Pratchett, The Last Continent
- 9y ago

SteveJS 9y ago

How does this interact with pilot wave theory? Is the article saying that the causal direction of time is undefined within a quantum entangled system? Or is that more than what it’s claiming?

tunesmith 9y ago

So is this saying that if retrocausality is possible, then it's akin to the kind of time travel where you may have altered something in the past, but the effects would not manifest until "after" the commencement of the time travel? Like in a person, if you feel some sort of uncomfortableness that leads you to make a choice, maybe it's because a minute ago, some time traveler traveled back a couple of weeks and planted a seed that made you start feeling uncomfortable! Maybe all of those, "You know, I never realized until now, but it turns out I've disliked broccoli for years, I don't know why I eat it every day!" statements are a result of retro causal time travel. (In this case, George HW Bush on an anti-broccoli time-traveling crusade.)

AceyMan 9y ago

I've had this puzzle solved in my own head for quite some time thusly: Distance is a fiction /aka: arbitrary manifold created by our (conscious?) selves interacting with "the world," which is really just a big array of ordered values (N-tuples). To say that "X is N distance from Y" is no more of a valid proposition than "Blue is Louder than Yellow". We attach lots of significance to distance (e.g., 3D space) owing to its utility for "doing science." (Insert "All models are wrong; some are useful", but at a grand, multiversal scale.) My 2¢: YMMV. (me: Philosophy major with a good bit of sci/math as well)

mikhailfranco 9y ago

Perhaps retrocausality is just non-locality in time.

ggggtez 9y ago

"we can see that it couldn’t be used to signal, for the same reason that entanglement itself can’t be used to signal" Nope, you lost me. This seems too hand wavy. Also "This isn’t true of many everyday processes – eggs turn into omelettes, but not the reverse!"... sigh.