Nonlocality: correlation vs causation

[Ken G]

You are saying we cannot distinguish between cause and effect precisely because we cannot measure an arrow of time, correct?

I'm saying the distinction we draw between cause and effect comes from how we process the information around us, it's not in any way "embedded" in that information. We learned to think a certain way, we have no way of knowing if time really flows in the direction we tell ourselves it does, it's just a story we create, untestable. A construct of language that serves us in some way, a convention of mutual exchange.

Consider a model of physics that places the causal relationship between a muon and an electron and two nutrinos as "backwards". That is, an electron and two neutrinos cause a muon.

So far so good-- indeed, due to the reversibility of the elementary interactions, this is a perfectly mainstream idea.

While this causal relationship satisfies and accommodates what is observed, as a condition statement: if there is an electron and two neutrinos, then there will be a muon, is false (you can just have an electron and two neutrinos that have nothing to do with each other). That is, this physical model only accommodates the causal relationship, while, the classical physical model predicts it. That is: if there is a muon, then there will be an electron and two neutrinos (there are two decay modes but you can see that is irrelevant here).

There is no difficulty repackaging that logic to make it consistent with the laws of physics. The story sounds weird, because it is not our convention for speaking and does not gibe with our interpretations of our elementary perceptions (not the elementary perceptions themselves, but the stories we build up around them), but it does not violate the laws, which are themselves time reversible.

Let's take your example. We would say a muon is created, and it is inevitable that in some short time with a certain expectation value, it will cause the creation of an electron and two neutrinos. Here's my story. We have an electron and two neutrinos on a convergent path. That's just our starting point, we are not asking why they are on the converging path any more than we ask why there was a muon in your story, or why there was a Big Bang for that matter. But given that we have an electron and two neutrinos on a convergent path, this fact compels that there will be a muon created when they come together. The muon is not a stable particle, so its existence now compels that something must shortly happen that will destroy that muon (which we were interpreting as the event that created it). Nothing in that story will be violated by any law of physics that we have tested, because those laws are time reversible.

So I guess while it's possible to interpret effect as being the cause and visa versa, it would not be a strong physical theory in some cases...

This depends on what one means by a "strong physical theory." It will agree with all the laws of interactions that we know of, because those laws are time reversible, so in that sense it is a perfectly good physical theory. It's problem will appear at a much higher level, the level of intelligent processing, the level of telling stories that work for us. The story I'm telling will fail to satisfy some purpose we are satisfying with that story, it's just not the way we think. Maybe we'll have difficulty attributing blame to a murderer if they claim their act was compelled by the death of their victim. For whatever reason, the story won't survive how we think about the world, but that's a different kind of failure-- it's a failure that comes from us, or our interactions with our reality, but not from a reality that doesn't include us.

 

[Ken G]

"Sequencing" and "ordering" mean the same thing.

It depends on how they are used. For example, in astronomy we have the "main sequence." This is just distributing stars along a line. If we distribute them by mass, the high mass ones are at one end and the low mass ones at the other. If we distribute them by surface gravity, the high surface gravities are at the other end from the high masses. So we call it a "main sequence", with no implication of any sign to it-- the sign depends on whether you are talking about mass or surface gravity. The fundamental meaning of a sequence is just a linear arrangement, the sign is really a separate property, so we need a separate word if we are also implying a sign to the distribution. For that purpose, I'm using "sequence" as the signless linear distribution, and "ordering" as that plus a sense of sign. Any words will do, as long as we recognize these are the meanings we need to distinguish.

 

[jfy4]

There is no difficulty repackaging that logic to make it consistent with the laws of physics.

How would you do this?

But given that we have an electron and two neutrinos on a convergent path, this fact compels that there will be a muon created when they come together.

Does it? My point from the last post is that this is not assured, nor-even probable in many cases. I'm confident there are many situations (in this backwards world) where an electron and two neutrinos are on convergent paths, yet no muon is created. However, as I stated, the causal relationship is allowed. Yet once again consider our current model. The decay of a muon into an electron and two neutrinos is predicted and some might say, with the standard model's current success, guaranteed!

I'm not arguing that particular physics is not time-reversable, but that a model that runs backwards will not have the same predictive power our current model has. With this being said, I'm extremely interested in analyzing this more in detail, I have not thought of this before so I'm quite excited about it.

 

[Ken G]

How would you do this?

I thought I just did!

I'm confident there are many situations (in this backwards world) where an electron and two neutrinos are on convergent paths, yet no muon is created.

It depends on "how convergent" they are, but yes, given the HUP we can't say if they will exactly meet or not. But that's no matter, the principle of reciprocity handles those kinds of things without difficulty. We can still tell the stories, it just becomes "when an electron and two neutrinos are on convergent path, there is X probability they will create a muon." And that same X is going to appear in your story about a muon creating an electron and neutrinos, just in a very different guise because the information is being packaged in the opposite order.

However, as I stated, the causal relationship is allowed. Yet once again consider our current model. The decay of a muon into an electron and two neutrinos is predicted and some might say, with the standard model's current success, guaranteed!

That isn't saying anything fundamentally different here. You predict a muon can make an electron and two neutrinos, which then diverge, I predict that a converging electron and neutrinos can create a muon. The same numbers, the same matrix elements, are going to matter in both those stories. The laws themselves are just the same, they're reversible. Only the context of the story changes, and I can modify that context to make it work just as well either way. It just won't serve the special needs we have for the story, which is where the story comes from, not from the interactions themselves.

I'm not arguing that particular physics is not time-reversable, but that a model that runs backwards will not have the same predictive power our current model has.

Yes it will, but only if we approach the experiment in reverse. We do this all the time, we might have a final exam question that asks "a particle has speed v and location x under acceleration a, where was it five seconds ago?" That's a prediction that would take a different kind of experiment, involving different conventions, to check, but it's perfectly good physics.

With this being said, I'm extremely interested in analyzing this more in detail, I have not thought of this before so I'm quite excited about it.

Excellent, I do think there are some lessons here, though exactly what they are is not so clear!

 

[jfy4]

@KenG

Sure, but your story is out to match a world where the data doesn't run backwards. The given correlation is: If a muon is observed, then an electron and two neutrinos will be observed. Your story is such that, an electron and two neutrinos that will be observed, prompt the existence of a muon, and like I said, you can certainly claim this as the causal relationship. But, what does your story say when I detect a muon? Nothing. it says, "the muon does not cause the existence of two neutrinos and an electron." So my point is still that your story is fine as a setting up a causal relationship, however, I think you will be hard pressed to make predictions for new phenomena.

I have to leave for a number of hours, don't have any great insights without me

 

[ThomasT]

@ Ken G

In support of the contention that causation isn't a physical principle (and indeed that we can speak of no physical principles or dynamical laws governing the evolution of our universe), you seem to be presenting the view that we have no apprehension of a reality 'out there' functioning independent of the way we process sensory data -- that, for example, the arrow of time evidenced by our objective records of the evolution of our world is just an untestable fiction that we've, collectively, created.

What's the problem with that view?

 

[Ken G]

I would say the data doesn't run anyway, it's just a bunch of numbers. We set up the experiment based on a certain way of thinking about things. But in the "final exam question" I gave in my example, we can set up an experiment to test that. It's a bizarre experiment, but here's how it works: we just send out a bunch of particles, and record their trajectories (it's a classical experiment here), and then we look at their final location and velocity. We take our theory to predict where they were 5 seconds in the past, and then we check our stored data stream to test our prediction. That's a perfect example of "data running backward," and such an experiment can be used to support a story in which the "final" location of the particle compelled certain things to happen "earlier" in the data stream, things we were unaware of until we checked that data stream.

[qoute] The given correlation is: If a muon is observed, then an electron and two neutrinos will be observed.

Actually, all you can say is that there is a certain probability that they will be observed along some trajectory. You don't know when the muon will decay. Similarly, all my story can say is that the convergent paths have a probability of creating a muon. The same probabilities will feature in both those stories, it's the same matrix element used in a different way.

So my point is still that your story is fine as a setting up a causal relationship, however, I think you will be hard pressed to make predictions for new phenomena.

It still predicts, it just predicts things that you would view as being in the past. That's not its fault, it's yours!

 

[Ken G]

can speak of no physical principles or dynamical laws governing the evolution of our universe), you seem to be presenting the view that we have no apprehension of a reality 'out there' functioning independent of the way we process sensory data -- that, for example, the arrow of time evidenced by our objective records of the evolution of our world is just an untestable fiction that we've, collectively, created.

What's the problem with that view?

There aren't any problems with it that are in nature, they come from us and how we think. It's not so much that I know our normal thinking must be wrong, it's that we can't know it's right, because we can trace its origin as coming from us. I'm not saying there's anything wrong with telling stories about reality that can be traced back to us, we do that starting as little children and it's a healthy brain development. I'm just locating where it happens.

 

[ThomasT]

There aren't any problems with it that are in nature, they come from us and how we think. It's not so much that I know our normal thinking must be wrong, it's that we can't know it's right, because we can trace its origin as coming from us. I'm not saying there's anything wrong with telling stories about reality that can be traced back to us, we do that starting as little children and it's a healthy brain development. I'm just locating where it happens.

Ok, so you're saying that causality and the arrow of time can be traced to our sensory apprehension of the world. I agree. Is there a problem with that? After all, our objectified sensory apprehension of the world (eg., instrumental behavior) is the basis for evaluating theories. It's all we've got. Isn't it?

The idea of time reversal is a fiction that, because it contradicts our objective observations and reasonable inferences from those observations, isn't taken too seriously by mainstream physics. Would you agree with that?

Regarding Demystifier's concern, I had said that:

... I think that Demystifier's main intent is to say that the nonlocality of Bohmian qm is essentially the same as, and indeed is grounded in, the nonlocality of standard qm, so that this is not a reason to dismiss Bohmian qm.

Would you agree with that?

 

[Ken G]

Ok, so you're saying that causality and the arrow of time can be traced to our sensory apprehension of the world. I agree. Is there a problem with that?

Only for those who wish to hold that reality is just the way we perceive it, that we are not adding an interpretive layer to everything we perceive. If the perception is the reality, then the sign of the arrow has to be there whether we are around to perceive it or not, but if the sign of the arrow comes from us, then part of what we think we know about reality is more like something we know about ourselves. I claim it's important to recognize that sometimes when we think we are looking out the window at the world around us, someone has played a trick on us and put a mirror there instead.

After all, our objectified sensory apprehension of the world (eg., instrumental behavior) is the basis for evaluating theories. It's all we've got. Isn't it?

Absolutely, but even if our eyes are all we've got in my analogy, we can still try to recognize the difference between a window and a mirror. I think we can track the disconnect between reversible laws and irreversible experience, and it isn't just another law, or it's a different kind of law.

The idea of time reversal is a fiction that, because it contradicts our objective observations and reasonable inferences from those observations, isn't taken too seriously by mainstream physics. Would you agree with that?

I'm not sure what you mean by "the idea of time reversal." My claim is that if a baby was trained to be a scientist by giving them only movies to watch, no personal experience, and if every movie they watched was run backward, that child could still develop (in principle-- in practice this would of course be a disaster because of the absence of interaction) perfectly usable scientific principles that would serve perfectly well to understand and predict any movie they saw from that point on-- so long as it was shown backward! If so, then backward time forms a perfectly equivalent scientific medium to the one we recognize as the "normal" way, and indeed all the elementary interactions come out just the same, so it's essentially the same science. But the interpretation of "cause and effect", the stories that get told around those words, would be very different. It wouldn't be that causes and effects are inverted, it would be that causes follow effects and are compelled by them, more so than caused by them.

Regarding Demystifier's concern, I had said that:

Would you agree with that?

I agree with the statement about Bohmian mechanics, but I would have expected Demystifier's conclusion from that to have been "so that's why Bohmian mechanics isn't delivering any goods", or words to that effect.

 

[jfy4]

Actually, all you can say is that there is a certain probability that they will be observed along some trajectory. You don't know when the muon will decay.

Fair enough.

Similarly, all my story can say is that the convergent paths have a probability of creating a muon.

But wait, take this case again with the "reverse cause and effect" story. Now as an experimenter, I will always detect the muon first. Now the experimenter certainly does not have any convergent paths at his/her disposal to account for the muon (as of yet). Would the prediction (from this story, from the correlated data) say that "the reason for this muon is that there possibly will be (for the experimenter) an electron and two neutrinos (detected, possibly soon) that prompted this muon to be there?" This seems circular, and contrived... While I'd allow this interpretation, I'd much rather a prediction, from a different story (from the same correlated data) that says, "This muon will decay, and the experimenter can (with a particular probability) expect to detect an electron and two neutrinos." Would that be a fair interpretation of the prediction I can expect from the "reverse cause and effect" story?

 

[Ken G]

But wait, take this case again with the "reverse cause and effect" story. Now as an experimenter, I will always detect the muon first.

Not necessarily, it depends on how the experiment is designed. Let's say you have a bubble chamber that keeps records that are not inspected. You start scanning the records for evidence of an electron and two neutrinos on a diverging trajectory, and when you find it, you then start looking for evidence, in the records, that there was a muon there. Then, as the experimenter, you will always detect the electron first, not the muon! It all depends on the constraints of how you are setting up the story you are telling, this is the crux of my point.

Now the experimenter certainly does not have any convergent paths at his/her disposal to account for the muon (as of yet). Would the prediction (from this story, from the correlated data) say that "the reason for this muon is that there possibly will be (for the experimenter) an electron and two neutrinos (detected, possibly soon) that prompted this muon to be there?"

No, to tell the story in the reverse way, you have to set up the experiment the way I described. Now, if you really did that, you'd say "well of course the muon was there first, and caused the electron, but they way I examined the records I discovered all that in reverse order." To which I say "how do you know you discovered it in reverse order, it's just the way you decided to tell the story. I'll tell the story just like it happened in my experience of scanning those records-- first there was an electron, and some convergent neutrinos, so this compelled the necessity for a muon to appear at the vertex, and sure enough, there it was." Nothing adjudicates between our stories except the way we experience our daily lives, they are both perfectly good physics because both can be used to test the laws of physics.

While I'd allow this interpretation, I'd much rather a prediction, from a different story (from the same correlated data) that says, "This muon will decay, and the experimenter can (with a particular probability) expect to detect an electron and two neutrinos."

No question, we tell the story the way we do because it makes more sense to us that way. All I'm saying is that "making more sense" is coming from us, from how we process information in our daily lives, the actual physics of that muon doesn't care which way we tell the story. Ergo, the sign of the arrow of time comes from us, it is a logical convention we have built up from the way we experience things, but the physics in the absence of us does not care about that convention.

 

[jfy4]

Not necessarily, it depends on how the experiment is designed. Let's say you have a bubble chamber that keeps records that are not inspected. You start scanning the records for evidence of an electron and two neutrinos on a diverging trajectory, and when you find it, you then start looking for evidence, in the records, that there was a muon there. Then, as the experimenter, you will always detect the electron first, not the muon! It all depends on the constraints of how you are setting up the story you are telling, this is the crux of my point.

I know we are beating this poor muon to death, but, with bubble chambers, aren't the particles passed into the chamber, meaning at least some of there properties are known and the particle is maneuvered into the chamber, doesn't this presuppose the existence of the muon, before the cause again. In which case isn't this once again 'the prediction (from this story, from the correlated data) says that "the reason for this muon is that there possibly will be (for the experimenter) the paths of an electron and two neutrinos that prompted this muon to be there'?"

EDIT: But I suppose you could set up a bubble chamber experiment where you had no prior interaction with the particles... curious...

EDIT2: Wait, yes, your interpretation is allowed for particular circumstances, but that still does not reenforce anything about the predictability of a theory where supposedly the cause occurs after the effect. My question to you still is: How, given initial data, are you going to make predictions about new experiments when the initial data is an effect of a cause which has not taken place yet...? That is, will every prediction's story be: "The initial conditions of this experiment are simply an effect of a cause that will be discovered soon." In which case I still hold to my previous statement from my prior post.

 

[ThomasT]

Only for those who wish to hold that reality is just the way we perceive it, that we are not adding an interpretive layer to everything we perceive. If the perception is the reality, then the sign of the arrow has to be there whether we are around to perceive it or not, but if the sign of the arrow comes from us, then part of what we think we know about reality is more like something we know about ourselves. I claim it's important to recognize that sometimes when we think we are looking out the window at the world around us, someone has played a trick on us and put a mirror there instead.

What reason is there to think that the arrow of time comes from us? For example, the radiative arrow of time. Disturbances are observed to propagate away from points of origin. And this seems to be true no matter what the scale or the medium. Is there some reason to think that, contrary to our observations, our apprehension of the radiative arrow of time doesn't correspond to what's happening in nature?

I think we can track the disconnect between reversible laws and irreversible experience, and it isn't just another law, or it's a different kind of law.

What are you calling time reversible laws? The basic equations of motion? They're just equations of motion. They can track the motion of the evolutions that we actually observe as well as the reverse of those evolutions. They don't give priority to, or establish the reality of, either. Our observations do that. Physics doesn't have a fundamental dynamical law that reflects our observation of the evolution of the universe and the arrow of time. But I think that eventually it will, and this would resolve the pseudo-conflict between so called time reversible equations of motion (which are thusly mischaracterized because they have nothing to do with time) and the observed irreversibility of the evolution of our universe.

I'm not sure what you mean by "the idea of time reversal." My claim is that if a baby was trained to be a scientist by giving them only movies to watch, no personal experience, and if every movie they watched was run backward, that child could still develop (in principle-- in practice this would of course be a disaster because of the absence of interaction) perfectly usable scientific principles that would serve perfectly well to understand and predict any movie they saw from that point on-- so long as it was shown backward! If so, then backward time forms a perfectly equivalent scientific medium to the one we recognize as the "normal" way, and indeed all the elementary interactions come out just the same, so it's essentially the same science. But the interpretation of "cause and effect", the stories that get told around those words, would be very different. It wouldn't be that causes and effects are inverted, it would be that causes follow effects and are compelled by them, more so than caused by them.

Let's say we've got a really fast camera and we take 1000 pictures of the sequence of events of a cup being knocked off a table to it breaking into lots of smaller pieces that come to rest on the floor. Picture #1 depicts the contact of a hand and the cup, and pictures #2 through #1000 depict the cup falling to the floor and breaking into small pieces. Each picture is unique, and each picture is more like the pictures closer to it in number in the index or sequence than the pictures that are more distant from it in the sequence. This is what is meant by the arrow of time.

A time reversed evolution would simply reverse the numbers, the indexing, of the pictures, with picture #1000 being picture #1 and so on.

The thing is, time reversal is just an idea. It's never ever been observed to happen. To say that it's even just possible is an entirely ungrounded assertion.

Physics is about why/how nature evolves the way it's observed to evolve. In answering that question it also answers questions about why nature doesn't evolve the way it isn't observed to evolve. Backward time isn't a scientific medium because so far it doesn't exist, and there are some good reasons to believe that can't and never will.

I agree with the statement about Bohmian mechanics, but I would have expected Demystifier's conclusion from that to have been "so that's why Bohmian mechanics isn't delivering any goods", or words to that effect.

I think it's time for Demystifier to weigh back in, resolve some things, and then the thread can be closed.

 

[Ken G]

I know we are beating this poor muon to death, but, with bubble chambers, aren't the particles passed into the chamber, meaning at least some of there properties are known and the particle is maneuvered into the chamber, doesn't this presuppose the existence of the muon, before the cause again.

Again, it all depends on how the experiment is set up. Saying that the sign of the arrow of time stems from how we set up experiments does not refute my claim that it stems from how the experiment is set up. If it's coming from us, then any story that sounds like the things we do is going to sound like it has a sign to the arrow of time.

In which case isn't this once again 'the prediction (from this story, from the correlated data) says that "the reason for this muon is that there possibly will be (for the experimenter) the paths of an electron and two neutrinos that prompted this muon to be there'?"

That's not quite what I have in mind, you are telling the story with the same sign of the arrow of time as we normally use, but trying to invert cause and effect. That won't work, the story itself has to be told with a reversed sign of time to be an example of the time convention at work. The story I'm talking about sounds like "the reason this muon came into being is it was compelled by the presence of converging neutrinos and electron." The time arrow is reversed there, and what we think of as the cause is compelled by the effect.

EDIT: But I suppose you could set up a bubble chamber experiment where you had no prior interaction with the particles... curious...

Yes, you could stumble onto a bubble chamber set up by someone else that you had no knowledge about, all you get is the record of the tracks, which you are, for some reason, forced to inspect in reverse time order. You are the child watching a moving in reverse.

EDIT2: Wait, yes, your interpretation is allowed for particular circumstances, but that still does not reenforce anything about the predictability of a theory where supposedly the cause occurs after the effect. My question to you still is: How, given initial data, are you going to make predictions about new experiments when the initial data is an effect of a cause which has not taken place yet...?

That the cause has not taken place yet is purely a matter of convention, stemming from how we see the world. How would you predict an outcome if you didn't know the initial data? It's the same question.

 

[jfy4]

That the cause has not taken place yet is purely a matter of convention, stemming from how we see the world. How would you predict an outcome if you didn't know the initial data? It's the same question.

Once again, fair enough, but until we have AI, or bubble chambers start working on theoretical physics, humans have to begin with initial data. I really do think I see where you are coming from here, except I can't seem to imagine a scenario where one begins with initial data, but is not allow to use it to make predictions about what will happen (since it is not responsible for causing anything that will be observed), but is somehow able to make predictions with data that doesn't exist yet about what one is observing currently...

 

[Ken G]

Let me clarify some things because I don't think I'm quite getting across. It all goes back to the OP, I did not intend a hijack-- the OP asks if there's anything in the causation concept that is fundamentally different from a certain flavor of correlation. I'm only taking the next step of asking what is the flavor of correlation that leads to a concept of causation, and I'm noticing that this flavor is a kind of back-story that involves in a fundamental way the arrow of time, but it does not involve in a fundamental way a sign to the arrow of time, because changing the sign of the arrow of time just creates a mapping to the back story that is still perfectly good physics.

A lot of what I'm saying can be easily misheard as something I am not saying. One thing I am not saying is that since the elementary interactions are time-reversible, there isn't anything strange that happens in a movie shown backward. That's been done to death, it's thermodynamics. I'm also not saying that running time backward turns causes into effects, because what we mean by a cause-to-effect relationship is an increasing entropy relationship, and that does not reverse when we reverse the sign of time. What I am talking about reversing is nothing other than the direction that time "actually goes", and the back-story mapping between causes and effects, to show that we really cannot support any such fundamental notion beyond how we happen to think about time.

So I'm not asking what if everything we call a cause is actually an effect, I'm saying that if we think causes must temporally precede effects, we cannot actually justify that claim by anything that nature is doing independently of us. Effects could temporally precede causes, and then every effect compels its own cause. If we see shards of an explosion converging together in the movie shown backward, then the existence of those shards on that convergent path compels that a bomb is going to blow up there. The shards are still the effect and the bomb still the cause, but the effect comes first and compels the cause. This is all perfectly good physics, and indeed is precisely the physics that a child who only ever saw movies shown backward would develop (if their brains could develop normally, which seems unlikely in such a stultified environment). In principle, that child could get all the same laws that we have, all the same cross sections and matrix elements, all the same equations that we are solving-- expressly because all those laws are time reversible. They would just have a different back story: specifically, effects compel the causes that follow after them.

Ergo, we simply don't know the sign of the arrow of time, we cannot say that time "really flows forward", this is just a convention of how we think about and experience the universe. It is yet one more way in which our science is our interaction with our environment, not something the environment is doing independently of our science.

 

[jfy4]

Let me clarify some things because I don't think I'm quite getting across. It all goes back to the OP, I did not intend a hijack-- the OP asks if there's anything in the causation concept that is fundamentally different from a certain flavor of correlation. I'm only taking the next step of asking what is the flavor of correlation that leads to a concept of causation, and I'm noticing that this flavor is a kind of back-story that involves in a fundamental way the arrow of time, but it does not involve in a fundamental way a sign to the arrow of time, because changing the sign of the arrow of time just creates a mapping to the back story that is still perfectly good physics.

A lot of what I'm saying can be easily misheard as something I am not saying. One thing I am not saying is that since the elementary interactions are time-reversible, there isn't anything strange that happens in a movie shown backward. That's been done to death, it's thermodynamics. I'm also not saying that running time backward turns causes into effects, because what we mean by a cause-to-effect relationship is an increasing entropy relationship, and that does not reverse when we reverse the sign of time. What I am talking about reversing is nothing other than the direction that time "actually goes", and the back-story mapping between causes and effects, to show that we really cannot support any such fundamental notion beyond how we happen to think about time.

So I'm not asking what if everything we call a cause is actually an effect, I'm saying that if we think causes must temporally precede effects, we cannot actually justify that claim by anything that nature is doing independently of us. Effects could temporally precede causes, and then every effect compels its own cause. If we see shards of an explosion converging together in the movie shown backward, then the existence of those shards on that convergent path compels that a bomb is going to blow up there. The shards are still the effect and the bomb still the cause, but the effect comes first and compels the cause. This is all perfectly good physics, and indeed is precisely the physics that a child who only ever saw movies shown backward would develop (if their brains could develop normally, which seems unlikely in such a stultified environment). In principle, that child could get all the same laws that we have, all the same cross sections and matrix elements, all the same equations that we are solving-- expressly because all those laws are time reversible. They would just have a different back story: specifically, effects compel the causes that follow after them.

Ergo, we simply don't know the sign of the arrow of time, we cannot say that time "really flows forward", this is just a convention of how we think about and experience the universe. It is yet one more way in which our science is our interaction with our environment, not something the environment is doing independently of our science.

As of now, I have no problem with anything you wrote up above, (and yes I'm sorry we got off topic from the original OP) what I thought we were discussing was how, utilizing that story and implementing it into the world in which we do experience will cause problems for people who want to predict what will happen. I think the problem comes about precisely because none of the available data now, can be considered to cause what will happen, and hence if any attempt is going to be made to predict anything and be consistent with the "backwards story" you will need to know the future, only to "predict" the present, at least so it seems to me.

With these two things said, maybe a PM apology is in order to Demystifier and we should start a new thread about prediction in a "reverse-causality" story?

 

[Ken G]

As of now, I have no problem with anything you wrote up above,

Excellent, because that means we agree that a sign to the arrow of time is not actually a physical imperative, it is a kind of architectural detail involving how we think. That is all I'm trying to say.

what I thought we were discussing was how, utilizing that story and implementing it into the world in which we do experience will cause problems for people who want to predict what will happen.

Yes, that's tracking where in our intelligent processing the difficulty emerges, i.e., where the sign comes from.

I think the problem comes about precisely because none of the available data now, can be considered to cause what will happen, and hence if any attempt is going to be made to predict anything and be consistent with the "backwards story" you will need to know the future, only to "predict" the present, at least so it seems to me.

Yes, if time is running backward, then needing the future to predict the present is quite natural, but it just doesn't gibe very well with the fact that our brains store memories of the past, not knowledge of the future. This would seem to be a crucial element of where the sign of the arrow of time comes from, and it goes away if there is not a memory-analyzing intelligence in the story.

With these two things said, maybe a PM apology is in order to Demystifier and we should start a new thread about prediction in a "reverse-causality" story?

Yes, you're probably right about the apology, a hijack is what eventually occurred but I would say that understanding the nature of causation is a key element to tracking its connection to correlation, so it wasn't a non sequitur. Do you think we have more ground to cover, or are we basically in agreement at this point?

 

[Demystifier]

But I must say, there is SOME sense in which "standard" QM is more local than Bohmian QM. But this has nothing to do with correlation vs. causation. Instead, it has to do with the positivistic philosophy according to which only observed phenomenon is a phenomenon.

To explain this, consider Alice measuring the particle on the left and Bob measuring the particle on the right, and let the wave function be |up>|down>+|down>|up>.

Let Alice find the left particle in the state up at time t. The crucial question is the following: Does it mean that the right particle is in the state down at t?

If it does, then it is a correlation-nonlocality, which is not any "weaker" than nonlocality involved in Bohmian QM.

But "standard" QM, or at least one version of it, denies that it means that the right particle is in the state down at t. Even if Bob measures it and finds that it is, it is of no relevance to Alice because at time t she couldn't possibly know what Bob has measured at t. Of course, if she trusts the THEORY called QM, she could CALCULATE that the state measured by Bob should be down. But calculation is not a measurement. For Alice, the state of the right particle at time t is not observed at time t. She can observe it only later, after a time t + Delta t, where Delta t is time needed for a signal to come from Bob to Alice. So for her, the right particle is in the state down only later, which saves locality.

In this way, one can save even correlation-locality. But the price is very big. Not only that unobserved phenomenons are not physical, but even phenomenons observed by someone else are not physical. This is a logically consistent way of thinking, but is that really what physics should be about?

Now I have further developed this idea, which resulted in a LOCAL hidden-variable model compatible with QM:
http://xxx.lanl.gov/abs/1112.2034

 

[bohm2]

Now I have further developed this idea, which resulted in a LOCAL hidden-variable model compatible with QM:
http://xxx.lanl.gov/abs/1112.2034

Am I the only one who noticed some similarity between this local and solipsistic hidden-variable model by Demystifier and Leibniz's concept of "monads"? Maybe I'm misinterpreting it but here are some quotes regarding Leibniz's monads:

Thus, in creating monads, God endows them with their own internal laws or program, and it is by virtue of these internal laws alone that their subsequent perceptions evolve in harmony with those of other monads...Strictly speaking, since all of a monads's states are explicable solely by reference to its own nature, there need not even be other monads (solipsism). Leibniz, then, introduces his doctrine of pre-established harmony to guarantee inter-substantial correspondence. Thus,

...pre-established harmony explains why, despite the fact that a substance's behaviour is causally accounted for solely by reference to its own nature, there nevertheless are inter-substantial regularities and correspondences which are not matters of fortuitous chance.

Given Leibniz's requirement that all the states and perceptions of any individual substance are internally predetermined or pre-programmed into that substance (so that a substance's states in no way depend on the nature and states of other created substances), Leibniz is then, driven to invoke the doctrine of pre-established harmony so that one substance's perceptions will evolve in harmony (i.e. 'mirror') with those of another. More generally, this harmony or 'isomorphism' will necessarily hold between the infinite set of all substances (so that each monadic history, in some sense, 'reflects' every other). Thus, as Rescher points out,

...the only interaction between monads arises in the reciprocal 'perception' built into their mutual accord by pre-established harmony. The only thing monads can 'do' in relation to one another is to perceive, and to agree (more or less) in their successive states.

Thus, each substance (monad) is like a 'spiritual automaton' programmed in such a way that its states and perceptions "represent the universe in a very exact way, though with relative degrees of distinctness"; like a kind of private picture gallery representing ('mirroring') from its own unique perspective all the other galleries. While no two substances are allowed to interact with each other (as extrinsic relational properties are not possible), their internal states nonetheless 'mirror' one another so 'harmoniously' (by way of pre-established harmony), that it seems as if they really are interacting (i.e. 'phenomenal' interaction). Since, Leibniz also holds that "ultimate reality" can be completely characterized by the infinite set of monads and their corresponding perceptions and appetitions ('mirroring' each other to infinity), then the whole question of an "independent objective reality" (i.e. what exactly do the monads mirror?), becomes superfluous, since, strictly speaking, there is nothing "outside" of the "mirroring monads" to mirror. Leibniz, however, held,

...that for there to be a perceived objective reality it is sufficient simply that there be agreement and correspondences between subjective states of different substances. It is not necessary that those states causally result from any other than the substance whose states they are.

Finally, since in Leibniz's ontology, these mathematical-like, isomorphic relationships, preclude the possibility for inter-substantial relations, the only kind of relational properties allowed (at least, at the "groundfloor" metaphysic of simple substances), are necessarily, intrinsic; consequently, all relational properties of any individual must be ultimately, reducible to non-relational predicates or properties of that individual.

 

[Ken G]

I think I understand Leibnitz's ideas about monads better now, thank you. In addition to connections with hidden variables ideas, there seems also to be a connection between "pre-established harmonies" and the idea of "cosmic censorship" or other systems-level interpretations. It sounds to me like Leibnitz was not happy with purely imagining the actions of all the monads, he felt they wouldn't hang together without organizational principles.

 

[bohm2]

Now I have further developed this idea, which resulted in a LOCAL hidden-variable model compatible with QM:
http://xxx.lanl.gov/abs/1112.2034

In that paper, I didn't understand this part:

Of course, Alice can hear that Bob tells her what his observations are. That information about Bob available to Alice can be correlated with other Alice’s observations, in agreement with predictions of QM. But the point is that an Alice’s observation is correlated with another Alice’s observation, not with another Bob’s observation. What Alice observes that Bob tells her that he observes is not necessarily what Bob really observes, and not because Bob is lying, but because Bob as observed by Alice is not real. There is a real Bob, but this is not the one observed by Alice.

Maybe I'm misunderstanding but given this solipsic/monadic model, I still don't understand how it's possible that we humans seem to have inter-subject correspondance/agreement?

Yet, there is one crucial difference. In the many-world interpretation, there is a copy of each observer in any of the branches. In our solipsistic interpretation, for each observer there is only one copy living in only one of the branches.

So if I'm understanding you:

1. In MWI there is a different copy of you in each branch.
2. In Bohmian model, there is only 1 observer/particle that goes in one branch but then the problem has always been how to explain the so-called "empty" branches:

For every branch of the wavefunction containing the actual particle trajectories, there are countless other branches corresponding to every other potential ‘world’ which would have been realized had the particle positions been different. The effects of decoherence soon disable the influence of other branches on the particle trajectories, leaving much of the wavefunction redundant. Nonetheless these redundant branches are an essential element of BM...This criticism of BM has led several authors to argue that BM is little more than a version of the many-worlds interpretation in which the particle trajectories are a way to select one particular world...

3. In this solipsistic/monadic/Leibnizian version there is 1 observer in each of the branches but they each don't know they are in different branches because they have no way of really comparing with each other or am I misunderstanding it?

 

[Ken G]

What is the scientific difference between two different observers in different branches, and two copies of the same observer in different branches? I don't think the concept is coherent. Communication is crucial for the whole concept of observation, whether it be communication between parts of our brain, or communication between scientists to establish the concept of objectivity. If physics needs a concept of an observer, and a concept of what observers agree on, then we cannot take that concept of an observer outside the domain for which it was developed, and not expect some problems. So I don't think the question is whether or not the observers in the different branches are the same observer, copies of the same observer, or different observers, I think the question is what is science and what is interpretation of what is science.

 

[Demystifier]

So if I'm understanding you:

1. In MWI there is a different copy of you in each branch.
2. In Bohmian model, there is only 1 observer/particle that goes in one branch but then the problem has always been how to explain the so-called "empty" branches:

3. In this solipsistic/monadic/Leibnizian version there is 1 observer in each of the branches but they each don't know they are in different branches because they have no way of really comparing with each other or am I misunderstanding it?

You have understood 1. correctly.

You have understood 2. incorrectly. In particular, there is no problem of empty branches in Bohmian mechanics. What some view as a problem in Bohmian mechanics is nonlocality, which a solipsistic approach removes.

You have understood 3. almost correctly. The incorrect part in 3. is "there is 1 observer in each of the branches". For example, if there are two observers (A and B) and two branches (1 and 2), then there are 4 different possibilities:
1) A in branch 1, B in branch 2.
2) A in branch 2, B in branch 1.
3) Both A and B in branch 1.
4) Both A and B in branch 2.

More precisely, the possibilities 1)-4) are possibilities for the local approach (Sec. 4.2).
For the nonlocal approach (Sec. 4.1), as well as in the Bohmian case, only 3) and 4) are allowed.

Is it clearer now?

 

[Ken G]

But here's the $64,000 question: what determines which branch the two observers, A and B, are in?

 

[Demystifier]

But here's the $64,000 question: what determines which branch the two observers, A and B, are in?

That's easy: the initial particle positions.
Can I get my $64,000 now?

 

[bohm2]

You have understood 2. incorrectly. In particular, there is no problem of empty branches in Bohmian mechanics. What some view as a problem in Bohmian mechanics is nonlocality, which a solipsistic approach removes...Is it clearer now?

Thanks. This issue of "empty waves" is still confusing me because even some pro-Bohmian physicists (see quote above from Daniel J. Bedingham) have raised some issues on this topic. Even in Bohm's book that I read, he didn't give a very good explanation. The just of it, is the following criticism:

An obvious strategy for defeating the above argument in the Bohmian case is to claim that wavefunction-stuff is just not the kind of stuff from which objects like cats could be made, even in principle. One might even claim that the wavefunction is not any kind of “stuff” at all, but is merely a mathematical device for calculating the motions of the Bohmian particles. If either of these claims could be substantiated, then one would have a principled reason to deny that empty branches could contain cats, either dead or alive, or any other measurement outcomes for that matter.

Against this strategy, however, Deutsch writes of the empty branches (or “unoccupied grooves”) that “it is no good saying that they are merely a theoretical construct and do not exist physically, for they continually jostle both each other and the ‘occupied’ groove, affecting its trajectory” . Since empty branches interact with each other and with the occupied branch, and empty branches are nothing but aspects of the wavefunction, the wavefunction must be real a physical entity and not just a mathematical construct.

Empty Waves in Bohmian Quantum Mechanics
http://philsci-archive.pitt.edu/2899/

Valentini discussing Bohmian "empty waves" responds:

Furthermore, in realistic models of the classical limit, one does not obtain localised pieces of an ontological pilot wave following alternative macroscopic trajectories: from a de Broglie-Bohm viewpoint, alternative trajectories are merely mathematical and not ontological.

De Broglie-Bohm Pilot-Wave Theory: Many Worlds in Denial?
http://www.tcm.phy.cam.ac.uk/~mdt26/local_papers/valentini_2008_denial.pdf

But I thought the wave function even in Valentini's de-Broglian interpretation was not just mathematical but represents a new type of non-local "causal" agent as suggested in this video by him?

Valentini video- from Perimeter Institute The nature of the wave function in deBroglie’s pilot wave theory
http://streamer.perimeterinstitute.ca/Flash/3f521d41-f0a9-4e47-a8c7-e1fd3a4c63c8/viewer.html

 

[bohm2]

You have understood 2. incorrectly. In particular, there is no problem of empty branches in Bohmian mechanics.

Here's Bohm's explanation of the empty branches (it begins on p. 335):

While it is clear that what is automatically called the measurement process has thus been given an overall causal and objective description, one may nevertheless ask what is the meaning of all the “empty” wave packets (i.e., those not containing particles). Those still satisfy Schrödinger’s equation, but are nevertheless permanently inactive, in the sense that they never manifest themselves in the movements of the particles at all. Such packets seem to be floating, almost like wraiths in a strange multi-dimensional world. One can see, for example, in the “many worlds” interpretation of Everett, the problem is dealt with in a certain way, as each packet would correspond to a different universe, with its different measuring instrument (along with its different human observer as well). What do all these “empty” packets signify in the causal interpretation?...

As long as the measuring apparatus interacts reversibly with the classical system, channels that are inactive with regard to the particles are still potentially active. But as soon as the irreversible interaction with the registration device takes place, the channels not containing particles are permanently inactive. In the usual language, we would say that the information has been “lost”, but as with the diffusion of ink particles, it has merely ceased to be capable of acting in the manifest domain...

Before the decision is made, each of these possibilities constitutes a kind of information. This may be displayed virtually in imagination as the sort of activities that would follow if we decided on one of these possibilities. Immediately after we make such a decision, there is still the possibility of altering it. However, as we engage in more and more activities that are consequent on this decision, we will find it harder and harder to change it. For we are increasingly caught up in its irreversible consequences and sooner or later we would have to say that the decision can no longer be altered. Until that moment, the information in the other possibilities was still potentially active, but from that point on such information is permanently inactive. The analogy to the quantum situation is clear for the information in the unoccupied wave packet becomes more and more inactive as more and more irreversible processes are set in train by the channel that is actually active.

An ontological basis for the quantum theory
http://www.tcm.phy.cam.ac.uk/~mdt26/local_papers/bohm_hiley_kaloyerou_1986.pdf

I still don't understand what Bohm means here?

And here is Brown's criticism of Valentini's position:

According to Valentini, the unoccupied component is merely “simulating” the approximately classical motion of the atom. Valentini further claims in section 6 that the treatment of the analogous, and more pressing, case of a superposition of non-overlapping packets representing distinct macroscopic arrangements is conceptually just the same. But in both cases, this notion of simulation is hard to reconcile with the plausible claim in that, even in pilot-wave theory taken on its own terms, the intrinsic properties of quantum systems such as mass (both inertial and gravitational), charge and magnetic moment pertain to (at least) the pilot-wave. If in the second case the macroscopic systems involve contain human observers, and the superposition is defined relative to the appropriate decoherence basis, it is hard to see why phenomenologically the unoccupied component does not have the same status as it does in the Everett picture.

Comment on Valentini, “De Broglie-Bohm Pilot-Wave Theory: Many Worlds in Denial?”
http://www.tcm.phy.cam.ac.uk/~mdt26/local_papers/brown_on_valentini.pdf

 

[Ken G]

That's easy: the initial particle positions.
Can I get my $64,000 now?

No, because an initial particle position should not determine something about the observer, it's backward. That's the "empty branch" problem in spades.

 

[Demystifier]

because an initial particle position should not determine something about the observer,

I don't understand why do you say this. In the solipsistic HV interpretation the observer IS a bunch of particles, so particle positions do determine the observer.

 

[Ken G]

I don't understand why do you say this. In the solipsistic HV interpretation the observer IS a bunch of particles, so particle positions do determine the observer.

But that is not a closed stance, because you have the particle positions determining the observer, yet it is quite demonstrably true in physics that the observer determines the particle positions. Your tiger is chasing its own tail. Whether or not that is a serious objection depends critically on what goals you have for the picture you are espousing.

This is how I look at it: the sole reason for invoking classical trajectories in an interpretation of any theory of dynamics is that we understand them well. This is a valid reason for Bohmian interpretations-- the goal of an interpretation is to bring a physics theory into the realm of what we can understand. So we have a valid question-- can we find a way to think about the theory of relativistically consistent quantum mechanics using a picture of classical trajectories underpinning our understanding? That is a valid goal to establish, and that is what you have attempted to do, possible successfully-- I have not delved deeply enough nor are enough of an expert to say one way or the other. I certainly see no obvious flaw in the endeavor. My main point is that we must distinguish three very different things:

1) physics theories: these make predictions by borrowing from some formal mathematical structure, but take no stance on the meaning or lessons of the predictions, only their practical usefulness and demonstrable success.

2) interpretations of physics theories: these assist in the application of the theory, and can give us a sense of meaning behind the theory, but are not unique and are chosen based on pedagogical and philosophical considerations, not testable predictions.

3) claims on reality: it is natural for any physicist who adopts some form of realism to want to take successful theories and use them to make claims on how reality works. This can inform our search for new theories also, and generally borrows from particular interpretations of existing theories. But these are not the same as interpretations of theories, because they go beyond the theory and make assertions about reality, possibly in the form of lessons we have learned about reality (not about our theories). Some reject this step altogether, but most physicists wish to take their understanding of the theories and graduate them to an understanding of some aspect of reality.

Thus, I'm saying that anyone who would endeavor to explore the terrain of the requirements for relativistic QM and how it can be made consistent with classical trajectory concepts needs to decide which of those three areas they are shooting for, or what kind of combination of all three. It's important to distinguish claims made in the various areas, because they are judged by different criteria. I'm not completely clear the claims you are making about the solipsistic HV interpretation, as to which of those areas are being invoked, but I do think that certain criticisms of it might fall victim to misjudging the appropriate sector, thereby applying inappropriate criteria. Clarifying the sectors thus clarifies the appropriate criteria, and can effectively deflect much of the criticism.

 

[Demystifier]

yet it is quite demonstrably true in physics that the observer determines the particle positions.

It is true in experimental physics but not necessarily in theoretical physics, depending on how exactly one interprets the role of observer in theoretical physics. In particular, if one assumes that the known theoretical laws of physics (either deterministic or probabilistic) describe also the observer, then it is not true that observer determines the particle positions. Instead, particle positions are determined by the laws of physics and the corresponding initial conditions.

 

[Ken G]

It is true in experimental physics but not necessarily in theoretical physics, depending on how exactly one interprets the role of observer in theoretical physics.

Yes, I suppose I am exposing myself as an empiricist. But all the same, it is clear that what we choose to regard as "laws" we do so by virtue of agreement with experiment, not logical necessity (the ancient Greeks were a little unclear on that, but Galileo straightened them out). Thus, it is inescapable that our ability to interact with nature, and perceive and interpret that interaction, must enter into anything we can call physics.

What's more, In particular, if one assumes that the known theoretical laws of physics (either deterministic or probabilistic) describe also the observer, then it is not true that observer determines the particle positions. Instead, particle positions are determined by the laws of physics and the corresponding initial conditions.

But that is the logical morass I was referring to. Of course you may assume whatever you like, but physics is not about making assumptions, it is about verifying them. How are you going to verify that the particle positions are determined by "the laws of physics" (whatever that is) without first using an observer to establish that? And how is that observer going to establish that the particle positions, thus established by the observer, are what determines the observer? If we stick to what we can demonstrate, we cannot hold that particle positions determine observers.

 

[bohm2]

If we stick to what we can demonstrate, we cannot hold that particle positions determine observers.

But it seems reasonable to assume that the observers are made of that same kind of stuff as other stuff. Which is the reason some have difficulty with the orthodox view. As Maudlin writes:

The reason that this problem does not come up in practice is because the ‘standard’ interpretation is a legacy of the Copenhagen view, and the Copenhagen view does not postulate wavefunction monism. Copenhagenism insisted on the necessity of having a classical description somewhere, the description of the ‘measurement situation’: the infamous Copenhagen ‘cut’ was exactly between a quantum realm and a classical realm. And the classical description would, of course, be in terms of local beables, so there is no problem applying a spacetime transformation to it. Within this sort of a dualistic picture the problem of spacetime transformations of the wavefunction can be approached. The problem, of course, is that this sort of dualistic ontology is impossible to take seriously: no one ever thought that there were really two different sorts of physical systems, the classical and the quantum, that somehow interact. If that were the view, then the ‘cut’ would be a matter of physical fact: somewhere the classical and quantum bits of ontology would actually meet. Furthermore, it is evident that the ‘classical objects’, measuring apparatus and so on, are composed out of ‘quantum stuff’ (electrons, protons, and so on), so this cannot really be a dualistic ontology. In the confused morass of Copenhagenism, the observation that the ‘cut’ could, For All Practical Purposes, be moved about at will within a large range was taken to show that the cut itself corresponded not to a physical fact but to a convention, or something like that. But if the theory can be formulated without a cut at all, let it be so formulated. Having removed the cut and put everything in the quantum ontology, one would evidently remove all the local beables, and all the problems we have been discussing would return.