Arguments Against Superdeterminism

[octelcogopod]

The only problem I can see that could argue against superdeterminism is consciousness.
If we were talking about anything else in the world I would have no problem believing that they were all things controlled by the quantum forces.
I do however see a gap between quarks and electrons, and qualia.
So far there has been no way to bridge the mental with the physical, and if you think about it, it becomes increasingly harder to do so.
This is not a question about free will (although it's related) but rather how the subjective conscious states can emerge from physical matter and energy.
You can measure and scrutinize the brain all you want, but never are you able to capture the actual subjective experience.
And the only way to do so seems to be to translate neuron relations into known subjective states, if we were able to know every possible neuron configuration. But even in that scenario the magical property of the consciousness is gone.

Consciousness is not just a property of the brain, it is a property of the senses, and the environment around those senses perceive.
But if superdeterminism was true, we should be able to pick up this experience directly in the brain, because the brain would have to be the carrier of all such information.
There's also the free will question of course. If superdeterminism is true, how can we make a choice? No matter on what level the choice is, I can choose to pick up the apple or not, so that would mean deterministic events would control all my emotions and thoughts, but how could I be aware then? Am I really aware?

 

[ueit]

You are arguing both sides of the same argument! Either a) electroweak theory is accurate, and there is no room for SD in it currently; or b) SD is completely outside our current Physics.

I pointed out that a) contradicts your hypothesis. So clearly SD is outside of what we know. That makes it 100% as speculative as the existence of God, so where is the science in any of this?

1. Please substantiate your first statement (a). Where exactly did you provide evidence for your assertion that electroweak theory is incompatible with SD ?

 

[DrChinese]

1. Please substantiate your first statement (a). Where exactly did you provide evidence for your assertion that electroweak theory is incompatible with SD ?

Read any of the Weinberg/Salaam work and follow-on work, and you will see that there is no mention of a superdeterministic mechanism. Yet there would need to be for there to be a violation of Bell Inequalities when the angle setting is determined by radioactive decay (i.e. randomly as far as we know). The mechanism you propose (which is not really a specific proposal at all) requires that either the electromagnetic properties are "right" for emission, which requires that the source knows the detector settings in advance. But they won't be selected until AFTER the photon pair is in flight.

I really must point out that anyone can postulate a non-falsifiable hypothesis (which is completely useless in all respects) regarding any existing physical theory. Hey, maybe oranges are really tiny solar systems but they simply act like edible fruit due to superdeterminism!

Why don't you own up to the true purpose of your question, which I believe is to find a back-handed way to keep local realism in play? If you were serious about superdeterminism per se, you would start by looking for evidence IN FAVOR of SD (of course there is none currently) rather than asking for evidence AGAINST (as you have done).

 

[Doc Al]

The way a SD theory might explain entanglement has nothing to do with a specific configuration at the Big-Bang. The correlations are not a result of fine-tuning of the original conditions but a result of source-detector interaction. The mechanism is as follows:

1. The detector is described by some particle configuration (it doesn't matter what, no conspirational fine-tuning required).
2. Each particle has an associated classical-like local field. This field is not one of the known fields (EM, weak, etc.) but those fields are assumed to arise as an effective description of the assumed fundamental field. This field has also infinite range so that it can communicate the particle configuration anywhere, at light speed.
3. A particle is only emitted when the detectors' field has a certain, "favorable", value, corresponding to a certain detector configuration.
4. The properties of the emitted particle is a function of the above field.

The result of the above proposed mechanism, both the properties of the emitted particle (spin for example), and the future detector configuration depend on the detector's configuration in the past, therefore they cannot be assumed to be statistical independent, hence the possible violation of Bell's inequalities by a local theory.

(1) I see no justification in your model for your statement that future detector positions depend on past detector positions. I don't see anything in your model that prevents detector positions from being chosen "randomly", determined perhaps by the polarization of some cosmic microwave background photon.

(2) If I understand your model properly, it satisfies Bell's locality conditions and thus cannot agree with experiment in all situations. Even if the "field" from the detector communicates with and influences the emission of the particles, the actual detector positions can be randomly chosen at the very last instant (as in double delayed choice experiments) just before the already-emitted particles reach the detectors.

 

[ThomasT]

The way a SD theory might explain entanglement has nothing to do with a specific configuration at the Big-Bang. The correlations are not a result of fine-tuning of the original conditions but a result of source-detector interaction.

The correlation is between the angular difference of the crossed polarizers and the rate of joint detection. The detectors placed after the polarizers are always both set the same. The polarizer settings are varied. So, I assume that by detector you mean polarizer.

Keeping that in mind:

The mechanism is as follows:

1. The detector is described by some particle configuration (it doesn't matter what, no conspirational fine-tuning required).
2. Each particle has an associated classical-like local field. This field is not one of the known fields (EM, weak, etc.) but those fields are assumed to arise as an effective description of the assumed fundamental field. This field has also infinite range so that it can communicate the particle configuration anywhere, at light speed.
3. A particle is only emitted when the detectors' field has a certain, "favorable", value, corresponding to a certain detector configuration.

What about when the polarizer settings are varied while the particles are in flight?

4. The properties of the emitted particle is a function of the above field.

If this were true, then wouldn't it be possible to predict the sequences (and not just the rates) of both individual and joint detections?

The result of the above proposed mechanism, both the properties of the emitted particle (spin for example), and the future detector configuration depend on the detector's configuration in the past, therefore they cannot be assumed to be statistical independent ...

They're dependent due to the pairing process, which is based on the assumption that the polarizers are analyzing the same (or a related) property wrt any given pair of detection attributes. The problem is that it's unknown where the relationship between the entangled particles is produced. The standard assumption is that it's happening via the emission process, and that everything is happening according to the principle of locality via transmissions less than or equal to c. But the precise qualitative charactaristics of the emitted disturbances is unknown. So, standard qm can't give a realist, or explicitly local, account. However, from optics, and the conservation laws, and the known statistical dependencies, etc., standard qm gives an accurate statistical account of the joint state in a nonfactorable form -- which doesn't rule out the possibilities of locality, or nonlocality, or ftl transmissions. And, locality and light speed limit remain the defacto standard assumptions.

... hence the possible violation of Bell's inequalities by a local theory.

I think that the Aspect experiment with time-varying analyzers, as well as the demonstrated independence between polarizer orientation and individual detection in every Bell experiment, make your proposal unacceptable.

 

[ueit]

No, that's a rather narrow and misleading view. Life is an emergent property of quarks and electrons. There is very Big difference between a dead person and an alive person, though they are made up of the same quarks and electrons.

The "big" difference boils down to a difference in the particle configuration, nothing else. Do you have evidence for the existence of something else?

Without free-will, everything is an illusion.

I don't see how this follows.

Why didn't you say consciousness instead of brain? Brain is not consciousness, though they are obviously related.

Consciousness is an emergent, macroscopic property of the particle configuration of the brain. It has no relevance at fundamental level.

This explanation, along with all explanations of anything are completely senseless in a superdeterministic universe.

How is a fundamentally probabilistic universe different in this aspect?

True, but if you posit that everything is a consequence of a pre-determined configuration of exteremely low entropy at the Big Bang, it raises more questions than it answers.

Those questions being...?

How about "Free will(and Life) is an emergent property in our universe, because there are infinite number of universes".

I do not see how this answers what I've asked you:

"Give me a good account of how "our personal subjective experience of reality" appears in a universe that is not superedetrministic."?

How does the number of universes change anything?

I think your idea is radical(that's generelly a good thing when applied to reality), but it degrades science, understanding, knowledge and logic and as such is absurd. If free will is an illusion, how is your theory not another illusion/delusion?

Free will is an assumption we make, not an empirical observation. The fact that this particular assumption is wrong does not imply that our direct observations are delusional.

 

[ueit]

Read any of the Weinberg/Salaam work and follow-on work, and you will see that there is no mention of a superdeterministic mechanism. Yet there would need to be for there to be a violation of Bell Inequalities when the angle setting is determined by radioactive decay (i.e. randomly as far as we know). The mechanism you propose (which is not really a specific proposal at all) requires that either the electromagnetic properties are "right" for emission, which requires that the source knows the detector settings in advance. But they won't be selected until AFTER the photon pair is in flight.

Look, tell me which of the following assumptions you find incompatible with the papers you refer to:

1. Each particle (electron, quark, neutrino, etc.) has a well defined trajectory.
2. Each particle is accompanied by an infinite range, local field.
3. The trajectory of each particle is determined by the structure of this field.

I really must point out that anyone can postulate a non-falsifiable hypothesis (which is completely useless in all respects) regarding any existing physical theory. Hey, maybe oranges are really tiny solar systems but they simply act like edible fruit due to superdeterminism!

SD refers to a class of possible theories (like the class of non-local theories like BM or GRW), it is not a unique theory. Therefore it is too soon to say if those theories are falsifiable or not. I see no a-priori reason to assert that they are not. I see no relevance of your "orange" analogy to SD. SD does not claim that we are deluded. Our observations are correct. SD only imposes some additional constraints on what experimental results are possible to be observed.

Why don't you own up to the true purpose of your question, which I believe is to find a back-handed way to keep local realism in play? If you were serious about superdeterminism per se, you would start by looking for evidence IN FAVOR of SD (of course there is none currently) rather than asking for evidence AGAINST (as you have done).

SD is usually dismissed because of various reasons ('t Hooft is an exception). For now I want to evaluate how strong the arguments against the idea are. Also, I do find EPR experiments as pointing towards SD because the other explanations are very close to a belief in god as you have mentioned (non-locality, unfalsifiable many worlds, etc). While SD may seem counter-intuitive it doesn't contradict (at least in an obvious way) no well established scientific result.

 

[WaveJumper]

The "big" difference boils down to a difference in the particle configuration, nothing else. Do you have evidence for the existence of something else?

I think you have misunderstood my statement. I said that a particluar particle configuration does not mean a living entity. This same configuration can also mean a dead entity(person, animal, etc.). It's not only the configuration that is at play, in fact it's the configuration that causes the emergence of a totally New phenomenon - that of Life. It's still a mistery what it really is that causes a particluar dumb quantum particles configuration to "come alive". Your guess that this emergent phenomenon is a deterministic process is a speculative unfalsifiable guess.

I don't see how this follows.

If every event in the universe is pre-determined by an awful long chain of reactions(incl. your free will), how can we ever know anything for sure? There may be no universe at all, how would we tell if a deterministic process isn't causing us hallucinations of an objectively existent universe? We can't say anything with certainty about anything. We can only say - "A deterministic process is causing/forcing us to believe there is A, B or C".

Consciousness is an emergent, macroscopic property of the particle configuration of the brain. It has no relevance at fundamental level.

What do you mean by "fundamental level"? There are good arguments to believe that at the most fundamental level, something and nothing are one and the same and that all known concepts from our experience are squeezed into non-existence.

How is a fundamentally probabilistic universe different in this aspect?

True randomness(whatever that is) is a pre-requisite for probabalistic genetic occurences. I don't think our observations point to there being a particlular chain of events in the past that lead to particular genetic mutations.

Those questions being...?

If you are putting forward the Simulation Argument, do say so. I find it rather thought provoking and i think it makes much more sense than a "bare" superdeterministic universe that has no first cause. If you had said from the onset that that's what you believe, i think your worldview would have met even more recognition(but maybe that's just me).

I do not see how this answers what I've asked you:

"Give me a good account of how "our personal subjective experience of reality" appears in a universe that is not superedetrministic."?

True Randomness in the quantum vacuum. Wait infinity. Infinity is a long time. Vacuum fluctuations come and go, constantly creating particle-antiparticle pairs. Occasionally whole atoms are created. emember, you have infinity on your side. After 10^5678 "years" - there goes BOOM, a giant quantum fluctuation gives birth to a whole universe like ours(Bolzmann brains). This of course pre-supposes the existence of a spacelike medium. It sounds somewhat abstract, but what we perceive as reality is already abstract enough.

How does the number of universes change anything?

It's the existence of true randomness that counts. Infinite universes, or infinite random quantum fluctuations(some of which create whole atoms and are theorized to be able to create even universes) all rely on randomness as a pre-requsite.

Free will is an assumption we make, not an empirical observation.

How can we make the assumption that we have free-will if we don't have the free-will to make assumptions? This is circular reasoning, not much different than the liar's paradox - The liar said - "I am a liar".

Generally speaking, it sounds like you have the virtual reality argument in mind. If it is so, i have nothing against it, as i see reasons to believe that this might be plausible.

The fact that this particular assumption is wrong does not imply that our direct observations are delusional.

How are we not delusional? We believe we are making our own choices, we believe we live our own lives in our own ways, we believe we are not "pre-programmed" cause-effect robots. How is this not a delusion in a Superdeterministic universe?


I am surprised you didn't invoke the block view of the universe(as per GR) and its all-at-once existence of past, present and future. You could make a strong case on determinism there.

 

[ThomasT]

ueit, I'm still not sure what the distinguishing characteristics of a superdeterministic theory are.

Anyway, here's a good article by Ghirardi for those who haven't read it:

http://plato.stanford.edu/entries/qm-collapse/

 

[SW VandeCarr]

If arguments against SD are wanted, it seems one the most obvious is that it negates the concepts of entropy and information. SD would imply the entropy of any system anywhere at any time is zero since there is never any objective uncertainty as to a system's state or evolution. Then we are dealing with subjective uncertainty only. However subjective uncertainty would also be predetermined, as our "subjective" state is also a function of a system with zero entropy. (Note: I'm using 'subjective' and 'objective' as if there were a fundamental difference. However,I agree with others here that our notion of 'objectivity' is related to issues of consistency and clarity of descriptions).

EDIT: What happens to "the arrow of time" if thermodynamic entropy is always zero?

 

[ueit]

The correlation is between the angular difference of the crossed polarizers and the rate of joint detection. The detectors placed after the polarizers are always both set the same. The polarizer settings are varied. So, I assume that by detector you mean polarizer.

By "detector" I mean the device or group of devices that measure the spin. It might be polarizer + photon detector, or a Stern-Gerlach device or something else. What is important is that the "detector" also includes whatever is used to change its orientation (be it an electric engine, a human, a monkey pressing a button, etc.). Everything that has a contribution to the decision regarding the measurement axis is included in the generic name of "detector".

What about when the polarizer settings are varied while the particles are in flight?

As I have said, "a particle is only emitted when the detectors' field has a certain, "favorable", value, corresponding to a certain detector configuration." Because the evolution of the detector is deterministic, its future orientation is "fixed". The "change" while the particle is in flight is nothing but the detector's deterministic evolution which is "known" by the particle since emission. In other words, you cannot "fool" the particle. The particle "knows" what will happen because it knows the value of the field in the past + deterministic evolution law.

If this were true, then wouldn't it be possible to predict the sequences (and not just the rates) of both individual and joint detections?

Sure, but only if you know the exact value of the field at particle's place.

I think that the Aspect experiment with time-varying analyzers, as well as the demonstrated independence between polarizer orientation and individual detection in every Bell experiment, make your proposal unacceptable.

It doesn't, see above.

 

[ueit]

ueit, I'm still not sure what the distinguishing characteristics of a superdeterministic theory are.

See my answer above.

Anyway, here's a good article by Ghirardi for those who haven't read it:

http://plato.stanford.edu/entries/qm-collapse/

Thanks!

 

[Doc Al]

By "detector" I mean the device or group of devices that measure the spin. It might be polarizer + photon detector, or a Stern-Gerlach device or something else. What is important is that the "detector" also includes whatever is used to change its orientation (be it an electric engine, a human, a monkey pressing a button, etc.). Everything that has a contribution to the decision regarding the measurement axis is included in the generic name of "detector".

Including the polarization of the cosmic microwave background photon that was used to "choose" the detector setting, right?

As I have said, "a particle is only emitted when the detectors' field has a certain, "favorable", value, corresponding to a certain detector configuration." Because the evolution of the detector is deterministic, its future orientation is "fixed". The "change" while the particle is in flight is nothing but the detector's deterministic evolution which is "known" by the particle since emission. In other words, you cannot "fool" the particle. The particle "knows" what will happen because it knows the value of the field in the past + deterministic evolution law.

So basically the emitter knows the state of the entire universe and thus certainly can predict the detector settings at any future time and choose to emit or not emit particles accordingly. Good one! Of course the detector is no dummy--it also knows the state of the entire universe and can predict the behavior of the emitter and act accordingly. (I assume the behavior of the emitter is just as deterministic as is that of the detector.)

All I see is some vague handwaving that somehow everything works out in the end. Where is the physics?

 

[ueit]

Including the polarization of the cosmic microwave background photon that was used to "choose" the detector setting, right?

Right.

So basically the emitter knows the state of the entire universe and thus certainly can predict the detector settings at any future time and choose to emit or not emit particles accordingly.

Indeed.

Good one! Of course the detector is no dummy--it also knows the state of the entire universe and can predict the behavior of the emitter and act accordingly. (I assume the behavior of the emitter is just as deterministic as is that of the detector.)

Sure, but as we are not interested in the particles emitted from detectors I've let them out.

All I see is some vague handwaving that somehow everything works out in the end. Where is the physics?

Can you show that there is something inconsistent in the above assumptions? Can you show mathematically that such a behavior is not possible? If you can, it is great. Rejecting a possible mechanism is good science. But maybe, such a mechanism works and leads to testable predictions (for example one may find out that only a small class of fields lead to predictions that are consistent with QM).

 

[Doc Al]

Can you show that there is something inconsistent in the above assumptions? Can you show mathematically that such a behavior is not possible? If you can, it is great. Rejecting a possible mechanism is good science. But maybe, such a mechanism works and leads to testable predictions (for example one may find out that only a small class of fields lead to predictions that are consistent with QM).

I stated up front that it's "possible" (meaning: not immediately self-contradictory), as I think Bell did as well. So what? It's also "possible" that you (and all of PF) are just a figment of my imagination.

You have not provided or described any mechanism. What experiment would you propose to falsify your proposed "mechanism"? To get anywhere, you need a specific physical mechanism.

 

[ueit]

I stated up front that it's "possible" (meaning: not immediately self-contradictory), as I think Bell did as well. So what? It's also "possible" that you (and all of PF) are just a figment of my imagination.

You have not provided or described any mechanism. What experiment would you propose to falsify your proposed "mechanism"? To get anywhere, you need a specific physical mechanism.

In order to falsify the mechanism one should propose a clear mathematical structure. I am not able to propose it. But if you want a "cheap" example of a SD theory just take Bohm's interpretation, as it is, and replace in the equation the instantaneous, "present" particle distribution with a past distribution so that locality is observed. The "present" distribution is then "predicted" by the particle from the past one.

It would be more interesting to find a formulation that is not such obviously ad-hoc, but for the time being I was only interested if there are well formulated arguments against SD.

 

[ueit]

If arguments against SD are wanted, it seems one the most obvious is that it negates the concepts of entropy and information. SD would imply the entropy of any system anywhere at any time is zero since there is never any objective uncertainty as to a system's state or evolution. Then we are dealing with subjective uncertainty only. However subjective uncertainty would also be predetermined, as our "subjective" state is also a function of a system with zero entropy. (Note: I'm using 'subjective' and 'objective' as if there were a fundamental difference. However,I agree with others here that our notion of 'objectivity' is related to issues of consistency and clarity of descriptions).

EDIT: What happens to "the arrow of time" if thermodynamic entropy is always zero?

Are you saying that a classical gas, composed of molecules with strict deterministic behavior would not obey the laws of thermodynamics?

 

[SW VandeCarr]

Are you saying that a classical gas, composed of molecules with strict deterministic behavior would not obey the laws of thermodynamics?

I'm making a reductio ad absurdum argument. If SD is true, then all events occur with probability one. If you plug p=1 into the Shannon equation (which differs from the Boltzmann entropy equation only by the choice of the constant) you get zero entropy. A block universe under SD (as I understand SD) is completely defined at all space-time points. Therefore its entropy would be zero everywhere all the time. The universe exists in just one possible state.

http://en.wikipedia.org/wiki/Boltzmann_entropy

EDIT:http://en.wikipedia.org/wiki/Entropy_(statistical_thermodynamics)

I'm posting these links, not because I don't think most people on this thread know what I'm talking about, but because some might not, and I just want to be specific regarding what I'm talking about.

 

[Count Iblis]

I'm making a reductio ad absurdum argument. If SD is true, then all events occur with probability one. If you plug p=1 into the Shannon equation (which differs from the Boltzmann entropy equation only by the choice of the constant) you get zero entropy. A block universe under SD (as I understand SD) is completely defined at all space-time points. Therefore its entropy would be zero everywhere all the time. The universe exists in just one possible state.

http://en.wikipedia.org/wiki/Boltzmann_entropy

This is the fine grained entropy, not the coarse grained entropy used in thermodynamics. The fine grained entropy is always zero, even in "ordinary physics".

 

[SW VandeCarr]

This is the fine grained entropy, not the coarse grained entropy used in thermodynamics. The fine grained entropy is always zero, even in "ordinary physics".

Just a short response for now. What kind of entropy does the equation I cited ($$S_{B}$$) describe? If it's the "fine grained" entropy and it's always zero, what good is it?

 

[ThomasT]

By "detector" I mean the device or group of devices that measure the spin. It might be polarizer + photon detector, or a Stern-Gerlach device or something else. What is important is that the "detector" also includes whatever is used to change its orientation (be it an electric engine, a human, a monkey pressing a button, etc.). Everything that has a contribution to the decision regarding the measurement axis is included in the generic name of "detector".

Using the standard referents for emitter, polarizer, and detector, in a simple optical Bell test setup involving emitter, 2 polarizers, and 2 detectors it's pretty easy to demonstrate that the polarizer settings aren't determined by the detector settings, or by the emitter, or by anything else in the design protocol except "whatever is used to change" the polarizer settings.

It's been demonstrated that the method that's used to change the polarizer settings, and whether it's a randomized process or not, isn't important wrt joint detection rate. What is important is the settings that are associated with the detection attributes via the pairing process -- not how the settings themselves were generated.

As I have said, "a particle is only emitted when the detectors' field has a certain, "favorable", value, corresponding to a certain detector configuration." Because the evolution of the detector is deterministic, its future orientation is "fixed". The "change" while the particle is in flight is nothing but the detector's deterministic evolution which is "known" by the particle since emission. In other words, you cannot "fool" the particle. The particle "knows" what will happen because it knows the value of the field in the past + deterministic evolution law.

It's already well established that detector orientations don't trigger emissions -- and changing the settings while the emissions are in flight has no observable effect on the correlations. If you want to say that these in-flight changes are having some (hidden) effect, then either there are some sort of nonlocal hidden variable(s) involved, or, as you suggest, there's some sort of heretofor unknown, and undetectable, local field that's determining the correlations. Your suggestion seems as contrived as the nonlocal models -- as well as somewhat incoherent wrt what's already known (ie., wrt working models). Anyway, flesh out the details of it, submit it to the appropriate forum, and maybe somebody who knows more than I do will agree with your approach.

I still don't know what the distinguishing characteristics of a superdeterministic theory are. Can you give a general definition of superdeterminism that differentiates it from determinism? If not, then you're OP is just asking for (conclusive or definitive) arguments against the assumption of determinism. There aren't any. So, the assumption that the deep reality of Nature is deterministic remains the defacto standard assumption underlying all physical science. It isn't mentioned simply because it doesn't have to be. It's generally taken for granted -- not dismissed.

 

[SW VandeCarr]

I still don't know what the distinguishing characteristics of a superdeterministic theory are. Can you give a general definition of superdeterminism that differentiates it from determinism? If not, then you're OP is just asking for (conclusive or definitive) arguments against the assumption of determinism. There aren't any. So, the assumption that the deep reality of Nature is deterministic remains the defacto standard assumption underlying all physical science. It isn't mentioned simply because it doesn't have to be. It's generally taken for granted -- not dismissed.

I don't know how SD is defined either. Perhaps you can address my concerns stated in post 53. I'm basing my argument on Boltzmann entropy.

 

[ThomasT]

I don't know how SD is defined either. Perhaps you can address my concerns stated in post 53. I'm basing my argument on Boltzmann entropy.

Since we're not sure exactly what ueit means by superdeterminism, let's assume for the moment that it's just an emphatic form of the term determinism (equivalent to saying that Nature is absolutely deterministic or really really deterministic, which is equivalent to the standard meaning of determinism).

If arguments against SD are wanted, it seems one the most obvious is that it negates the concepts of entropy and information.

Determinism is an assumption about the underlying nature of the reality that we observe. It's a given as far as physical science's search for fundamental, as well as emergent, dynamical laws.

It isn't obvious to me, and I don't understandl, how the assumption of determinism "negates the concepts of entropy and information".

As I've mentioned, afaik, there are no (and I don't think there can be any) definitive or conclusive arguments against determinism. Entropy might be irrelevant for this, and our information seems to support the assumption of determinism.

Maybe I just don't understand your argument. So, if you could lay it out, step by step, that might help.

 

[ueit]

I don't know how SD is defined either. Perhaps you can address my concerns stated in post 53. I'm basing my argument on Boltzmann entropy.

As ThomasT says, there is no difference between superdeterminism and determinism. However, classical deterministic theories can work with free, external parameters because either there is no long-range field (billiard balls) either the field vanishes with distance (classical EM and gravitational fields). The distinctive characteristic of SD (as I propose it) is that such a separation observer-observed is not possible because the interaction between distant particles never becomes negligible. It is a quantitative, not qualitative difference. The closest analogous would be a system (galaxy) of black holes in general relativity. In order to model the trajectory of one BH you need to know the distribution of all other BH-s.

Regarding your argument, AFAIK a reversible system has a constant entropy. If our universe is SD then it has a constant entropy. However we cannot measure the entropy of the universe, only of a part of it. But this part is not reversible because the interactions with the environment are not taken into account, therefore the entropy may increase.

 

[SW VandeCarr]

Maybe I just don't understand your argument. So, if you could lay it out, step by step, that might help.

The basic argument is fairly straightforward. Entropy, both in the Shannon IT context and in the Gibbs thermodynamic context, is a logarithmic function of probabilities The Gibbs entropy equation:

$$S=-k\sum_{i}p_{i}lnp_{i}$$

The probabilities are based on the number of states in which a system can exist. For example a sequence of ten fair coin tosses has 1024 possible states, each state having an equal probability of 1/1024. In thermodynamics, we're talking about the macrostate of a system as a composition of N microstates. The model of the evolution of a thermodynamic system is a Markov process involving the transition probabilities from one state to another.

Given this background, if you assume SD, I'm arguing that there is just one state in which a system can exist at any point in time. The whole notion of a probability becomes irrelevant as an objective concept. Entropy is a function of the number of possible states in which a system can exist. If there is only one state, entropy is equal to zero. Moreover, since the evolution of a system is strictly determined, there is no Markov process.

In the coin toss example, the ten toss sequence is predetermined. There is only one possible outcome before the coin is actually tossed. At best we an only talk about subjective entropy: that is, our uncertainty as to the outcome. This is what SD says. Now, if you take this one step further and ask what subjective uncertainty actually is, it's a state of a system, the system being our brain and the state of knowledge represented in the brain. That's my reductio ad absurdum. We only imagine there are multiple possible outcomes. In reality (under SD) there is only one possible outcome and probabilities have no objective meaning. How can we talk about the entropy of the universe increasing when there is no objective entropy?

 

[SW VandeCarr]

As ThomasT says, there is no difference between superdeterminism and determinism.

Regarding your argument, AFAIK a reversible system has a constant entropy. If our universe is SD then it has a constant entropy. However we cannot measure the entropy of the universe, only of a part of it. But this part is not reversible because the interactions with the environment are not taken into account, therefore the entropy may increase.

I saw your post after I posted my last post quoting ThomasT. If you concede that the entropy of the universe is constant under SD, what does that say about the Second Law? As far as local environments, I'm not sure your argument rescues the Second Law. Local entropy may increase, decrease, or remain constant with the background of a constant entropy universe. In any case, entropy is imaginary if there are no objective probabilities. IMHO, you can have SD or the Second Law, but not both.

 

[Count Iblis]

Just a short response for now. What kind of entropy does the equation I cited ($$S_{B}$$) describe? If it's the "fine grained" entropy and it's always zero, what good is it?

That formula is completely general and can describe any kind of enetropy. To get the entropy we use in practice, you always have to use a coarse graining procedure to define the probabilities.

If you have a given number number of molecules in a given volume and you would exactly specify the energy of the system, then there is only one quantum state the system vcan be in. So, what you do is you explicitely specify some small energy uncertainty delta E and then count the number of microstates that are within that small energy nterval. Then, the fundamental assumption being that all these microstates are equally likely, yields the Boltzmann formula:

S = k Log(Omega)

where Omega is the number of microstates.


Check out http://en.wikipedia.org/wiki/Fundamental_thermodynamic_relation#Derivation_from_first_principles"

dS = dQ/T

from S = k Log(Omega)


to see the importance of specifying the delta E.

So, in the end we arrive at:

dS = dE/T + X dx/T

where X is the generalized force and x and external parameter. And for a reversible change we have dE + X dx = dQ

 

[WaveJumper]

In order to falsify the mechanism one should propose a clear mathematical structure. I am not able to propose it. But if you want a "cheap" example of a SD theory just take Bohm's interpretation, as it is, and replace in the equation the instantaneous, "present" particle distribution with a past distribution so that locality is observed. The "present" distribution is then "predicted" by the particle from the past one.

It would be more interesting to find a formulation that is not such obviously ad-hoc, but for the time being I was only interested if there are well formulated arguments against SD.

Lets push your model a bit deeper and see if it survives. We know from QCD, that more than 99% of all mass of matter is concentrated in the nuclei and the mass of quarks only adds up to a few percent. The rest of the mass(>96%) comes from virtual gluons that randomly pop into existence and disappear again from the quantum vacuum. The Higgs field is theorized to make up the other few percent and give mass to quarks and electrons through virtual Higgs bosons, and it is thought to derive its energy from the quantum vacuum too. So it appears we are very close to having evidence that all of physical matter in space-time emerges from timeless and spaceless Planck scale through random quantum fluctuations. This may resolve the biggest of all questions - "Why is there something instead of nothing?" through adjustments to both how we view "something" and how we view "nothing". But one could wonder - are quantum fluctuations really random if they have the ability to create such immense volumes of information at our macro scale(the whole structure of reality as we see it)? Only infinity and the notion that given infinity, everything that can occur will occur in the quantum vacuum can provide a somewhat coherent explanation. What's your opinion?

 

[SW VandeCarr]

That formula is completely general and can describe any kind of enetropy. To get the entropy we use in practice, you always have to use a coarse graining procedure to define the probabilities.

Thanks for clarifying that. Given that the Gibbs formula is good for both fine and coarse grained entropy, it would seem that SD would restrict the Second Law to specific experiments at most, but that the Second Law is not a universal principle. Therefore, with SD, we cannot explain the arrow of time in terms of the Second Law, nor can we justify entropy as an objective concept. (see my previous posts.) SD may even mean that we have to give up the idea of randomness at all scales, even random quantum fluctuations at Planck scales.

 

[ThomasT]

SW VandeCarr, it seems to me that probabilistic descriptions and the concepts of entropy and information are independent from the assumption of determinism.

It's assumed that the underlying evolution of any physical system is deterministic. This assumption is objective in the sense that, and insofar as, it's inferrable from the development of dynamical laws and principles that correspond with the objective data.

However, the assumption of determinism doesn't inform us about the details of the underlying (real) state or evolution of any physical system. Even though it's assumed that any physical system can be in only one (real) state at any given time, those (real) states are generally unknown and the behavior of observable objects is increasingly upredictable as the dependence of the behavior on unknown factors increases. So, probabilistic descriptions are often necessary, and their use doesn't contradict the underlying assumption of determinism.

You wrote:

We only imagine there are multiple possible outcomes. In reality (under SD) there is only one possible outcome and probabilities have no objective meaning.

If the underlying states and dynamics were known, then probabilistic descriptions would be obviated. But they aren't.

However, this doesn't mean that probabilistic descriptions aren't objective. It isn't our imaginations that tell us that the tossed coin is going to come up either heads or tails.

How can we talk about the entropy of the universe increasing when there is no objective entropy?

I don't know quite how to think about the entropy of the universe. Is there a definitive statement about the entropy of the universe? I've seen several different values given, none of which are 0.

In any case, entropy is connected with the dissipation of energy and the arrow of time -- the archetypal example of which is the radiative arrow. Drop a pebble into a smooth pool of water. The evolution of the wavefront is deterministic, isn't it?

I don't think your argument is why ueit's proposal should be rejected. There are other reasons, not the least of which is the notion of fields whose strength doesn't diminish with distance.

 

[ThomasT]

Therefore, with SD, we cannot explain the arrow of time in terms of the Second Law, nor can we justify entropy as an objective concept. (see my previous posts.) SD may even mean that we have to give up the idea of randomness at all scales, even random quantum fluctuations at Planck scales.

The Second Law doesn't explain the arrow of time. It's just a generalization of it. Since observations are so far in agreement with it, it's kept.

Entropy, in its many forms, is very much an objective concept insofar as it depends on measurements.

The assumption of determinism isn't at odds with randomness. Randomness refers to unpredictability. We use words like random and spontaneous when we can't specify causal antecedents. This doesn't mean that there aren't any.

 

[SW VandeCarr]

However, this doesn't mean that probabilistic descriptions aren't objective. It isn't our imaginations that tell us that the tossed coin is going to come up either heads or tails.

I don't know quite how to think about the entropy of the universe. Is there a definitive statement about the entropy of the universe? I've seen several different values given, none of which are 0.

The entropy of the universe, whatever it might be, is definitely not 0. ueit has already agreed that SD implies a constant entropy for the universe and if the universe is in just one possible spacetime state (block universe), all events occur with a real probability of one, which yields zero when plugged into the Gibbs equation for entropy. This is an argument against SD.

Regarding probabilities, if the uncertainty is due only to a lack of complete information, the probabilities are not objective. They would be objective only if we assume that nature is fundamentally probabilistic and true randomness actually exists.

 

[ThomasT]

ueit has already agreed that SD implies a constant entropy for the universe ...

He's also agreed that SD is synonymous with standard determinism. There's a better name for what he's proposing, which I'll suggest to him. Anyway, determinism doesn't imply a constant entropy.

... and if the universe is in just one possible spacetime state (block universe), all events occur with a real probability of one, which yields zero when plugged into the Gibbs equation for entropy. This is an argument against SD.

I don't think the block universe model should be taken literally, as a realistic representation. The universe is assumed to be in one possible, transitory, spatial configuration at any given time wrt evolutionary (deterministic), presentist models.

Saying that all events occur with a real probability of one is meaningless. Probabilities are applicable before, not after, the facts of observation.

Regarding probabilities, if the uncertainty is due only to a lack of complete information, the probabilities are not objective.

If we had complete information we wouldn't need probabilities. What is non objective about the observation that a tossed coin will come up either heads or tails?

They would be objective only if we assumed that nature was fundamentally probabilistic and true randomness actually existed.

No. Probabilities are objective when they're based on observable possibilities. Randomness refers to our observations, not the deep reality of Nature. True randomness does exist. There are lots of things that we really can't predict.

Why would we assume that Nature is fundamentally probabilistic when there are so many reasons to believe that it isn't?

 

[SW VandeCarr]

Why would we assume that Nature is fundamentally probabilistic when there are so many reasons to believe that it isn't?

I'm not assuming anything. I don't know. I'm saying if...then. Given determinism, the future is as well determined as the past. We just don't know for certain what it will be. Therefore, it's our uncertainty that probabilities quantify.

 

[ThomasT]

I'm not assuming anything. I don't know. I'm saying if...then. Given determinism, the future is as well determined as the past. We just don't know for certain what it will be. Therefore, it's our uncertainty that probabilities quantify.

Yes, it's our uncertanties that probabilities quantfy. We assume an underlying determinism (for many good reasons). But we don't know the details of that underlying determinism. Hence, the need for probabilistic descriptiions.

ueit asks for arguments against determinism. Afaik, there aren't any -- at least no definitive ones. And, as far as I can tell from this thread you haven't given any.

But, ueit's proposed explanation for Bell-type correlations is problematic for reasons that I've given.

So, where are we?