Is the self-consistency principle a scientific or philosophical idea?

[Spathi]

TL;DR Summary

I have heard that the Novikoc's self-consistency principle is a consequence of the stationary-action principle in quantum mechanics. Does this mean that it is scientific?

As I understand it, the Novikov's principle is either philosophical or scientific. I heard that it is a consequence of the stationary-action principle (principle of least action) in quantum mechanics. Does this mean that Novikov's principle is scientific?

I also heard the following: in quantum mechanics, instead of the stationary-action principle, there is a functional integral that diverges for closed timelike curves. Is this so, and what does this mean?

My question can be formulated differently - does modern physics allow time travel (if it does, then obviously Novikov's principle is a scientific hypothesis). I do not understand the article by S. Lloyd “Closed Timelike Curves via Postselection: Theory and Experimental Test of Consistency”. In this article, the authors describe an experiment where they tried to cause a time paradox and a random factor counteracted this. It seems to me that if Novikov's principle worked in its original understanding, the universe would not have allowed Lloyd to conduct this experiment, in the same way as the universe does not allow a time traveler to kill his grandfather (as it is described in popular sci-fi).

 

[phinds]

does modern physics allow time travel

Only in the forward direction and only at one second per second unless you want to get into time dilation, but that requires enormously expensive rockets and stuff.

 

[Demystifier]

Self-consistency is a logical principle. Logic underlies both philosophy and science. In this sense, the self-consistency principle is more fundamental than any purely philosophic or purely scientific principle. In my opinion, the Novikov self-consistency principle is almost a tautology, i.e. it makes no sense to question the validity of it. Instead, what has to be questioned in the presence of CTC's is the assumption that humans have free will, or more mathematically, that in principle any initial condition is physically possible.

 

[PeroK]

Self-consistency is a logical principle. Logic underlies both philosophy and science. In this sense, the self-consistency principle is more fundamental than any purely philosophic or purely scientific principle. In my opinion, the Novikov self-consistency principle is almost a tautology, i.e. it makes no sense to question the validity of it. Instead, what has to be questioned in the presence of CTC's is the assumption that humans have free will, or more mathematically, that in principle any initial condition is physically possible.

It's not so much "free will" - I don't believe in that as a thing in itself. It's about the complex biological systems that can reason about the universe. It's difficult to imagine that this ability to respond to stimuli would somehow be constrained in ways that are almost superdeterministic.

For example, suppose I am close to developing the technology to transport myself back one day (through a CTC). I can have my plan all ready in advance. The CTC machine is built and:

1) If I appear from tomorrow, then I am not going to get into the CTC "time" machine.

2) If I don't appear from tomorrow, then I am going to get into my CTC "time" machine the next day.

It's difficult to see what would force me to change my plans. Perhaps, indeed, in respect of getting into the CTC machine I and my future copy will become like automatons and be unable to execute my disruptive plan. I will somehow be compelled to do everything as "remembered" by my future self. In other respects, I can stick to my plans.

It seems to me equally plausible that the self-consistent principle is invalid. That I will be able to stick to my disruptive plans and that if the laws of physics allow a CTC of this sort, then they must allow some sort of MWI-like branching.

I can't see there is any way to disentangle this unless we can do an experiment and see what happens.

It's possible that nature finds a way to enforce consistency and in the above scenario my future self and I can have a laugh about the things that we are allowed to do and the things that we know we can't do, even if we can write them down as a plan and think about them.

 

[PeroK]

A further thought about this is that it's related to the idea of a machine that could predict precisely what a person is going to do. This is plausible as long as the person does not see the prediction. But, if they see the prediction, then a self-referencial loop is set up that could break the system. Either:

1) The machine works and the person is strangely compelled to do what it says. It would be like they had been hypnotised. In particular, they would be conscious of their inability to carry out any preconceived plan not to do what the machine says.

2) Such a machine is impossible for sufficently complex systems, such as human beings.

A CTC in the scenario I described could be used to construct such a situation, with the future self acting like a prediction machine. Again, it could all be pre-planned that the future self carries a written record of the activities of the day before. To maintain consistency, both would need to be effectively hypnotised into following the instructions.

Again, this can't simply be a case that they don't know what they are doing. Intelligence and the ability to conceive of a plan to break the system must lead to a realisable compulsion to do things. I'm not saying this is impossible, but I'm saying that we would be able to arrange the circumstances where we are conscious of the compulsion.

In a way, it might not feel so different from the daily compulsions we feel to do some things and not others. That said, it seems to me that it would feel an extremely powerful impulse to do something - that it would be physically impossible to do otherwise. And, of course, we would know in advance what we are bound to do.

Finally, of course, the original self would be compelled to get into the CTC machine and disappear; and the future self would equally be compelled not to! You would remain with a memory of the strange day spent with your future self, of getting into the CTC machine, then the day repeated with your past self - two memories of the day, one from each perspective!

 

[Demystifier]

@PeroK Stop thinking of the $t$ variable in equations as "time"! Instead, think of it as a spatial variable, or just as an abstract mathematical variable without any physical interpretation. Then, suddenly, all your arguments above stop to make any sense and self-consistency becomes paradox-free, doesn't it?

Another way to say the same thing is that one has to take the block-universe point of view. See also my https://fqxi.org/community/forum/topic/259

 

[Demystifier]

1) The machine works and the person is strangely compelled to do what it says. It would be like they had been hypnotised. In particular, they would be conscious of their inability to carry out any preconceived plan not to do what the machine says.

Of course. Even without CTC's, I don't see how can humans be any different from machines, if we assume that both are governed by the laws of physics. Either humans are machines or the laws of physics are not universally valid. See also my https://arxiv.org/abs/1006.0338

 

[Demystifier]

It's difficult to see what would force me to change my plans.

Replace "me" and "my" with "machine" and "machine's", and everything should become crystal clear.

 

[PeroK]

Replace "me" and "my" with "machine" and "machine's", and everything should become crystal clear.

I don't think you can satisfactorily think about the universe without acknowledging that human intelligence creates the ability of the system to analyse itself. And, in a similar way to how Goedel got mathematics to talk about itself, this creates a nontrivial complexity.

I've no argument with the question of free will, but I would argue against the inability of humans to self analyse.

Otherwise, I believe you are into superdeterminism and everything, including your arvix papers, are just meaningless random gibberish.

We've had this difference of opinion before regarding mathematics, which I don’t believe is an illusion created by random quantum phenomena. The ability of humans to produce abstract mathematical models of the universe - and abstract mathematics generally - is one strong piece of evidence that intelligence is a complicating factor in all this.

The one thing I am certain of is that it is not valid to dismiss consciousness and the potential for the universe to self analyse as irrelevant.

 

[PeroK]

Immediately you have a paradox that if you believe you are no different from an unconscious, unintelligent machine, then there is no basis on which your philosophy of physics has any merit! It's just random noise.

You deny the very thing that is needed to give what you think and write any meaning.

 

[Spathi]

It seems to me equally plausible that the self-consistent principle is invalid. That I will be able to stick to my disruptive plans and that if the laws of physics allow a CTC of this sort, then they must allow some sort of MWI-like branching.

Maybe different types of time travel are possible: a travel to another universe, which does not cause a paradox, and a travel to your own universe, for which the principle of self-consistency works?

I can't see there is any way to disentangle this unless we can do an experiment and see what happens.

But it seems that S. Lloyd has conducted such an experiment:

https://arxiv.org/pdf/1005.2219.pdf

Here's a quote from that article:

FIG. 4: Probability that time travel succeeds and the probes are found in the same state (red circles) or in opposite states (blue diamonds). When the quantum gun “misfires”, the polarization qubit is not flipped and the probe qubits are found in either the 00 or 11 state. As the accuracy θ of the quantum gun increases from 0 to π, the probability that the teleportation succeeds decreases. When the quantum gun “kills” the photon (flips the polarization qubit), the probes record opposite values (01 or 10). The probability that the probe qubits are found in either the 10 or 01 state is 0.01 ± 0.04, indicating that the photons never succeed in traveling back in time and killing their former selves. Solid curves correspond to theoretical predictions.

I do not understand how this experiment was carried out, and why the universe did not prevent Lloyd from conducting it. If we expand this reasoning, it is not clear why we exist at all, if the universe protects itself from paradoxes and, accordingly, from the emergence of civilizations that can cause a paradox.

 

[Demystifier]

@PeroK you missed a part of my point. If you are right that humans are not machines in a way you suggest, I'm fine with this, but then physics as we know it is incomplete. Hence, when a part of physics clashes with such understanding of humans, then it's very likely that this part of physics is wrong. In this case, it seems reasonable to assume that the part of physics that predicts CTC's is wrong. Which resolves the paradox.

 

[PeroK]

@PeroK you missed a part of my point. If you are right that humans are not machines in a way you suggest, I'm fine with this, but then physics as we know it is incomplete. Hence, when a part of physics clashes with such understanding of humans, then it's very likely that this part of physics is wrong. In this case, it seems reasonable to assume that the part of physics that predicts CTC's is wrong. Which resolves the paradox.

We are biological machines but we are intelligent machines. I don't see it as a question of physics being incomplete. There is nothing magical or unphysical about intelligence. It arises through the laws of physics. But, once evolved it can analyse the universe.

As an example, take the difference between a tree and a car. You might argue they have both evolved through the laws of physics. But, I see a fundamental difference between them. The modern motor car has evolved because intelligent life evolved and, unlike nature, applied intelligent design. The car's components were chosen not randomly but with planning (I.e. the foreknowledge of how they would operate).

It's this foreknowledge of what we expect to happen that distinguishes intelligent analysis of the universe from lower level quantum mechanical processes.

I don't believe there is anything missing in the evolution of intelligence. It's that as the complexity of a system grows it may develop characteristics that are not inherent in the fundamental processes.

This is why we can see a potential paradox in CTC. On the one hand we are governed by fundamental processes, which in principle we cannot overrule. But, on the other hand we can plan to subvert those processes. And plan in advance to do something that the laws of physics suggest we are not allowed to do. At the very least, we are conscious of what's happening.

 

[Demystifier]

This is why we can see a potential paradox in CTC. On the one hand we are governed by fundamental processes, which in principle we cannot overrule. But, on the other hand we can plan to subvert those processes.

Fine, but then we know that those two principles are not equally fundamental. The first, that we cannot overrule, is fundamental and hence always valid. The second, that we can plan to subvert, is not fundamental so does not necessarily need to be valid. The second principle is emergent, meaning that it can only emerge when the conditions needed for its emergence are fulfilled. If, along CTC's, those conditions are not fulfilled, then in such conditions the second principle does not emerge. Paradox resolved.

Here is an analogy. Usually a bird can plan to move in all 3 dimensions. But if you put the bird in a planar cage, then she can only move in 2 dimensions and cannot plan to move in the 3rd dimension. As simple as that. If the bird lived in the planar cage for the whole life, she would not even feel that she misses something by not being able to move in the 3rd dimension. It will never occur to her mind that she could decide to move in the 3rd dimension, just as it never occurs to your mind that you could decide to move in the 4th spatial dimension.

 

[PeroK]

I'm different from a bird - although some corvids have high intelligence - I can learn mathematics. I can conceive not only four spatial dimensions, but understand something abstract like $n$ dimensions where $n$ is arbitrary.

An atom cannot do that, nor a cell, but if atoms are arranged in a sufficiently complex system, then the whole system transcends the abilities of the components.

 

[Demystifier]

I'm different from a bird - although some corvids have high intelligence - I can learn mathematics. I can conceive not only four spatial dimensions, but understand something abstract like $n$ dimensions where $n$ is arbitrary.

An atom cannot do that, nor a cell, but if atoms are arranged in a sufficiently complex system, then the whole system transcends the abilities of the components.

So what? Does it contradict anything I said in #14?

 

[PeroK]

So what? Does it contradict anything I said in #14?

Perhaps not.

 

[PeterDonis]

It seems to me that if Novikov's principle worked in its original understanding, the universe would not have allowed Lloyd to conduct this experiment

I don't see why you would conclude that. The self-consistency principle does not say that experimenters can't try to violate self-consistency. It just says that, if they try, they will always fail.

 

[PeterDonis]

does modern physics allow time travel

It depends on what you mean by "allow".

In (non-quantum) General Relativity, there are definitely known solutions of the Einstein Field Equations that contain closed timelike curves. Most physicists believe that these solutions are not physically reasonable and would never be realized in our actual universe. But the solutions are there and are mathematically consistent. Any such solution would also satisfy the Novikov self-consistency principle since GR is not a quantum theory and so at any given event in spacetime, only one thing happens, even if that event is on a closed timelike curve and an observer can "pass" that event multiple times.

In non-relativistic quantum mechanics, you can formulate scenarios that include time travel and appear to violate the Novikov self-consistency principle, but such formulations always involve hand-waving, since you can just declare by fiat that the quantum states and Hamiltonians take whatever form they need to to realize your scenario, without considering whether such quantum states could actually be prepared or whether such Hamiltonians could actually be realized by any experimental device.

In relativistic quantum mechanics, i.e., quantum field theory, as it is usually formulated, the background spacetime is Minkowski spacetime, which does not contain closed timelike curves, so questions about time travel cannot even be formulated. In order to do so, one would need to formulate a quantum field theory on a background spacetime that did contain closed timelike curves, such as Godel spacetime. I'm not aware of any actual attempts to do this in the literature.

 

[PeterDonis]

I heard that it is a consequence of the stationary-action principle (principle of least action) in quantum mechanics.

I also heard the following: in quantum mechanics, instead of the stationary-action principle, there is a functional integral that diverges for closed timelike curves.

"I heard" is not a valid reference. Can you give actual references to where you found these claims?

 

[Spathi]

"I heard" is not a valid reference. Can you give actual references to where you found these claims?

I talk on many forums, including science forums. This thesis was told to me on a programmers forum by a member who seems to know QM well, especially the multiuniverse interpretation

 

[PeterDonis]

In this article, the authors describe an experiment where they tried to cause a time paradox

No, that's not what they were doing. They were using quantum teleportation, not time travel; any claims about what the results of their experiment say about time travel are based on the assumption that Lloyd's claimed analogy in the paper between quantum teleportation and time travel is correct.

 

[PeterDonis]

I talk on many forums, including science forums. This thesis was told to me on a programmers forum by a member who seems to know QM well, especially the multiuniverse interpretation

That's still not a valid reference. You should be looking at peer-reviewed papers; there are plenty in this field. (Also, there are a lot of people out there on the Internet who might seem to you to know QM well, but who actually don't.)

However, even leaving that aside, if you are going to reference something someone else said on some other forum, the very least you can do is provide a link to what they said. How do we know you are describing what they said correctly?

 

[Jarvis323]

This is why we can see a potential paradox in CTC. On the one hand we are governed by fundamental processes, which in principle we cannot overrule. But, on the other hand we can plan to subvert those processes. And plan in advance to do something that the laws of physics suggest we are not allowed to do. At the very least, we are conscious of what's happening.

I think you might be overestimating the importance of intelligence. You can program a simple machine to subvert a process with a simple if statement. In nature, many such simple systems are bound to just exist without us having to first create them. That they are designed to subvert a process isn't relevant, only that they subvert the process.

 

[PeroK]

You can program a simple machine to subvert a process with a simple if statement.

You need intelligence to build and program such a machine. And, you need intelligence to conceive of what the laws of nature are in the first place; let alone plan to subvert them!

 

[Jarvis323]

You need intelligence to build and program such a machine.

This is doubtful. First, why does the key to subverting a process with access to a CTC need to be complex? And how complex? And why do you think self reference a la Godel is important? Do you think humans can do things computers can't?

Where non-trivial self analysis factors in as a necessary condition for some class of subversion isn't clear. But, even if you need high complexity, or some form of computation/self analysis, it is not a fact that we can say that the universe doesn't perform non-trivial computation outside of human control. Note there is a low bar to Turing completeness.

 

[PeroK]

Do you think humans can do things computers can't?

Yes, invent computers in the first place. Humans can evolve from nature directly; computers cannot. They need intelligent design.

Anyway, I'm not sure what your argument is except argument for argument's sake.

 

[PeterDonis]

If you are right that humans are not machines in a way you suggest, I'm fine with this, but then physics as we know it is incomplete.

There is no need to claim that humans cannot be machines if we are intelligent and conscious. It's perfectly possible that we are intelligent, conscious machines.

 

[Spathi]

However, even leaving that aside, if you are going to reference something someone else said on some other forum, the very least you can do is provide a link to what they said. How do we know you are describing what they said correctly?

I am Russian, this was a Russian forum. Some talks about this integral were when I asked a similar quesion on physics.stackexchange.com:

https://physics.stackexchange.com/q...e-a-scientific-or-a-philosophical-idea/671478

 

[berkeman]

Thread closed for Moderation...

 

[PeterDonis]

After moderator review, the thread will remain closed.