Completeness of Quantum Mechanics

[mn4j]

If Quantum Mechanics is a complete theory, how come it can not predict the individual events of Double-slit or Stern-Gerlach experiments, which are well known experimental facts.
Can any theory be considered complete if it can not predict things that experimenters can measure, such as the time order of 'spin up' and 'spin down' in a Stern-Gerlach experiment or the slow build up of interference patterns by individual "clicks" in a double-slit experiment?

 

[meopemuk]

If Quantum Mechanics is a complete theory, how come it can not predict the individual events of Double-slit or Stern-Gerlach experiments, which are well known experimental facts.
Can any theory be considered complete if it can not predict things that experimenters can measure, such as the time order of 'spin up' and 'spin down' in a Stern-Gerlach experiment or the slow build up of interference patterns by individual "clicks" in a double-slit experiment?

You are right that QM cannot predict these individual events (QM can only calculate probabilities in large ensembles). In this sense QM can be considered an "incomplete theory". My guess is that these events are completely random and unpredictable. It may be true that no theory will be able to describe them ever.

 

[ZapperZ]

If Quantum Mechanics is a complete theory, how come it can not predict the individual events of Double-slit or Stern-Gerlach experiments, which are well known experimental facts.
Can any theory be considered complete if it can not predict things that experimenters can measure, such as the time order of 'spin up' and 'spin down' in a Stern-Gerlach experiment or the slow build up of interference patterns by individual "clicks" in a double-slit experiment?

Forget about QM. Can YOU set up the experiment and you yourself make the experimental prediction on such a thing, independent of any theory? If you can't or could find any means to make that prediction, then QM gives a valid description of the phenomena. To claim that due to some a priori requirement causes QM to not be complete is weaker than QM itself, because such a claim lacks any experimental backing.

Zz.

 

[Phrak]

What does 'complete' mean?

 

[DrChinese]

What does 'complete' mean?

I think that is the issue we find ourselves with. It makes no sense to criticize a theory for not explaining (predicting) everything. No one ever said we know all there is to know about the nature of physical reality.

The issue of completeness was raised with the 1935 EPR paper, "Can a Quantum Mechanical Description of Physical Reality be Considered Complete?". They laid out a specific argument that strongly implied that the predictions of QM, if "complete", did not seem reasonable. So this is where the "completeness" criticism really got started. The EPR argument was clever, but failed upon further examination (Bell, 1964).

The upshot is that the Heisenberg Uncertainty Principle remains as a fundamental limit on our ability to see into the quantum world, and may in fact represent an accurate description of how observation shapes reality.

 

[Phrak]

I think that is the issue we find ourselves with. It makes no sense to criticize a theory for not explaining (predicting) everything. No one ever said we know all there is to know about the nature of physical reality.

The issue of completeness was raised with the 1935 EPR paper, "Can a Quantum Mechanical Description of Physical Reality be Considered Complete?". They laid out a specific argument that strongly implied that the predictions of QM, if "complete", did not seem reasonable. So this is where the "completeness" criticism really got started. The EPR argument was clever, but failed upon further examination (Bell, 1964).

Yes. I don't know what EPR meant be 'complete', either. So here it is.

http://prola.aps.org/abstract/PR/v47/i10/p777_1"

"In a complete theory there is an element corresponding to each element of reality. A sufficient condition for the reality of a physical quantity is the possibility of predicting it with certainly, without disturbing the system."

 

[DrChinese]

Yes. I don't know what EPR meant be 'complete', either. So here it is.

http://prola.aps.org/abstract/PR/v47/i10/p777_1"

"In a complete theory there is an element corresponding to each element of reality. A sufficient condition for the reality of a physical quantity is the possibility of predicting it with certainly, without disturbing the system."

Exactly. Bell showed that this assumption - that you could predict WITHOUT disturbing the system - was incompatible with the predictions of QM. And of course, it also turns out that QM is supported experimentally. So that made EPR's logic flawed.

 

[Phrak]

Exactly. Bell showed that this assumption - that you could predict WITHOUT disturbing the system - was incompatible with the predictions of QM. And of course, it also turns out that QM is supported experimentally. So that made EPR's logic flawed.

I have a very nice 3D graphical model of the Bell Inequality that greatly helps in making intuitive sense of it. But I don't have any decent drawing software to do 3D rendering. Do you know of any for one-off use?

 

[Demystifier]

If Quantum Mechanics is a complete theory, how come it can not predict the individual events of Double-slit or Stern-Gerlach experiments, which are well known experimental facts.
Can any theory be considered complete if it can not predict things that experimenters can measure, such as the time order of 'spin up' and 'spin down' in a Stern-Gerlach experiment or the slow build up of interference patterns by individual "clicks" in a double-slit experiment?

Just to understand what do you mean by "complete".
In your view, can classical mechanics be considered complete, if this theory cannot predict the initial conditions.

 

[mn4j]

Forget about QM. Can YOU set up the experiment and you yourself make the experimental prediction on such a thing, independent of any theory? If you can't or could find any means to make that prediction, then QM gives a valid description of the phenomena. To claim that due to some a priori requirement causes QM to not be complete is weaker than QM itself, because such a claim lacks any experimental backing.
Zz.

Experimentalists measure. Theoreticians explain and predict. It is not the domain of the experimentalists to predict things. Yet if there is a phenomenon which a theory can not explain, there must be "something-else" which when added to the theory will make it more complete. Accepting the incompleteness is just the first step on the long path to finding this "something-else".

Just to understand what do you mean by "complete".
In your view, can classical mechanics be considered complete, if this theory cannot predict the initial conditions.

I do not know if Classical mechanics is complete or not. I do know that there are phenomena that are not currently explainable in the context of Classical mechanics, which is not to say no one will ever be able to explain them using Classical mechanics. However, the topic question was about QM. Is it accepted that there are experimental observations QM can not explain or as is often mentioned in QM discussions, the very act of asking how individual "clicks" are observed to build up an interference pattern over time, is a stupid question.

 

[cam875]

I don't think there is a complete theory yet, is there ? I mean why would all the great theoretical physicists strive for the ultimate theory of everything if classical mechanics, relativity or QM was correct. The reason is because there not complete. This is why new theories emerge such as string theory, to explain and try to merge things that don't work until we get to the bottom of it all, if we ever do.

 

[gendou2]

I don't think there is a complete theory yet, is there? I mean why would all the great theoretical physicists strive for the ultimate theory of everything...

The term "complete theory" is not the same as "theory of everything".
The word complete in this context means that the theory logically complete, not that it necessarily describes all natural phenomena.

The argument I have heard against quantum mechanics being complete is that it relies on a classical "observer" lying outside the studied system.

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

 

[DrChinese]

Accepting the incompleteness is just the first step on the long path to finding this "something-else".

You are speaking as if physicists have accepted the completeness of QM, turned off the lights and gone home. Nothing could be further from the truth.

The usage of the word "complete" in this context does not mean what you are implying. EPR used it a specific way, to attempt to show that there was a self-contradiction in QM. There wasn't such a contradiction after all, but that does not mean that QM - as it is now - is a final theory. I assume better will come along; in fact, things advance daily.

 

[meopemuk]

... as is often mentioned in QM discussions, the very act of asking how individual "clicks" are observed to build up an interference pattern over time, is a stupid question.

No this question is not stupid. Individual clicks are observable facts, and physical theory is supposed to say something about them. Unfortunately, QM does not even attempt to say anything about individual clicks. QM can calculate probabilities very accurately, but "individual clicks" are simply beyond this theory. I see only two possibilities to resolve this conundrum: 1) some kind of "hidden variable" deterministic theory, which would explain quantum randomness; 2) the randomness of "individual clicks" is an inherent property of nature, which simply cannot be explained. My guess is that #2 is correct.

 

[turbo]

You are speaking as if physicists have accepted the completeness of QM, turned off the lights and gone home. Nothing could be further from the truth.

The usage of the word "complete" in this context does not mean what you are implying. EPR used it a specific way, to attempt to show that there was a self-contradiction in QM. There wasn't such a contradiction after all, but that does not mean that QM - as it is now - is a final theory. I assume better will come along; in fact, things advance daily.

In Penrose's presentations, he speculates how classical and quantum theories might have to be modified in order to allow for some sort of reconciliation. It's an interesting mental exercise to figure where the conflicts lie and where the capitulations (or re-statements of the problems) might be most fruitfully explored and exploited. My gut instinct is that GR will have to take some pretty huge hits, and that quantum mechanics will come out relatively unscathed. Einstein feared as much in his 1924 essay "On the Ether".

 

[Demystifier]

I see only two possibilities to resolve this conundrum: 1) some kind of "hidden variable" deterministic theory, which would explain quantum randomness; 2) the randomness of "individual clicks" is an inherent property of nature, which simply cannot be explained. My guess is that #2 is correct.

Even if #2 is correct, a complete version of QM may still need hidden variables, though not deterministic ones. The role of hidden variables is not only to explain causes of individual events, but even more importantly, to specify the objective (even if random) properties of physical systems without regard if they are measured or not.

 

[ZapperZ]

Experimentalists measure. Theoreticians explain and predict. It is not the domain of the experimentalists to predict things. Yet if there is a phenomenon which a theory can not explain, there must be "something-else" which when added to the theory will make it more complete. Accepting the incompleteness is just the first step on the long path to finding this "something-else".

But this is exactly my point! The measurement can be completely independent of any kind of theoretical description of the system. If I do a which-way experiment, I do NOT depend on QM's description to figure out if I have photons coincidence, or only one click at a time! I can design the experiment and give it to someone completely ignorant of QM do the experiment.

So my question is, if you do this experiment, what do you think you will get?

All this claim about QM being incomplete is very tiring. This is as IF there is anything at all out there that is "complete". We continue to test QM and to push its boundaries, so it is not as if we are sitting down on our fat rear ends being content with what we already have. So there are people who continue to test and challenge it. But these must be done via valid means. The most convincing way to show that something may not work all the time is via experimental evidence, and not simply via personal preferences or a matter of tastes! Saying "Oh, QM must be incomplete because there must be something wrong if it can't predict for certain which way the photon must go through in a 2-slit experiment" is arguing based on personal tastes! Someone else, like me, have no trouble with it because as an experimentalist, it is what it is when this is what you get empirically! Without any experimental evidence, all arguments are nothing more than a discussion about favorite colors.

Zz.

 

[meopemuk]

... to specify the objective (even if random) properties of physical systems without regard if they are measured or not.

I don't think this is a role of science. If a property is not measured, then it is not objective. In order to be "objective" a property must be "measurable", at least in principle. Otherwise, we will be arguing endlessly about this property without any chance to resolve our argument by the only available objective tool - the experiment.

Sure, I can accept that "Moon is still there, even when nobody is looking". But this statement does not belong to physics. It belongs to philosophy (or religion). A correct physical statement should be verifiable by experiment. The above statement is not verifiable, bacause it involves the clause "when nobody is looking".

 

[DrChinese]

Sure, I can accept that "Moon is still there, even when nobody is looking". But this statement does not belong to physics. It belongs to philosophy (or religion). A correct physical statement should be verifiable by experiment. The above statement is not verifiable, bacause it involves the clause "when nobody is looking".

Sorry, but you have not been following closely enough. The evidence appears strongly as if the moon is NOT there when no one is looking. Or maybe it just dashed to the other side of the universe in the meantime.

 

[Demystifier]

The evidence appears strongly as if the moon is NOT there when no one is looking.

Only if you assume locality. Otherwise, not so strongly.

 

[Demystifier]

I don't think this is a role of science. If a property is not measured, then it is not objective. In order to be "objective" a property must be "measurable", at least in principle.

The role of science is defined by scientists themselves. If a significant number of scientists thinks that the question of objective reality is scientific, then, by definition, the question of objective reality is scientific too. Such scientists certainly do exist, but the question is whether their number is "significant".

Anyway, personally I think that "objective reality" is scientifically meaningful in the following sense. You may construct a THEORY that postulates the existence of objective reality. If such a theory has a better predictive or explanatory power on measured phenomena than some other theory that does not postulate the existence of objective reality, then it is justified to think that objective reality exists. This is what I try to do with the Bohmian interpretation, to show that such a formulation of QM has a better predictive or explanatory power on measured phenomena than standard QM. I have several results suggesting that it is indeed the case if one considers phenomena BEYOND those described by non-relativistic QM.

 

[ThomasT]

... the randomness of "individual clicks" is an inherent property of nature ...

Randomness refers to predictability, which only has meaning wrt instrumental behavior. The fact that ensemble and aggregated quantum experimental phenomena are rather predictable seems to me to suggest that Nature is deterministic. However, it doesn't necessarily follow that a pseudo-realistic, causal, metaphysical (hidden variable) description of the deep reality of quantum phenomena would produce more accurate predictions of individual instrumental events. (And if it doesn't, then quantum theory without the metaphysical embellishment is just as complete and just as 'descriptive'.) The randomness of individual quantum events is evidence that our knowledege of Nature is incomplete. The principles of the orthodox interpretation of quantum theory lead to the conclusion that our knowledge of Nature MUST (now and forever) be incomplete.

The role of hidden variables is not only to explain causes of individual events, but even more importantly, to specify the objective (even if random) properties of physical systems without regard if they are measured or not.

The "objective properties of physical systems" refers to instrumental behavior. Quantum theory already includes everything which can be taken into account regarding the objective properties of quantum scale physical systems. Doesn't it? If so, then it's as complete as any theory of quantum scale phenomena can be.

The evidence appears strongly as if the moon is NOT there when no one is looking.

Unless you're speaking metaphorically, I should think that the evidence (direct observations of the Moon, and trajectory calculations based on gravitational theory ) suggests that it's occupying a certain spatial volume relative to the Earth even when no one is looking at it.

The role of science is defined by scientists themselves. If a significant number of scientists thinks that the question of objective reality is scientific, then, by definition, the question of objective reality is scientific too. Such scientists certainly do exist, but the question is whether their number is "significant".

I assume that you're using the term 'objective reality' as a synonym for a dynamic reality underlying and precipitating chains of instrumental events which presumably amplify underlying phenomena.

This is a legitimate use of the term I suppose, however, confusion with it's other meanings can be avoided by referring to an 'underlying reality' as an 'underlying reality'.

Another use of the term 'objective' is to refer to phenomena which are amenable to our direct sensory perception (ie., amenable to public verification), such as instrumental data. This is the meaning that I normally assume is intended by 'objective reality' when used in connection with the physical sciences. Physical science is differentiated from metaphysics in that statements and disputes about reality are evaluated and ultimately decided by appealing to objective reality, ie., experimental results.

Anyway, personally I think that "objective reality" is scientifically meaningful in the following sense. You may construct a THEORY that postulates the existence of objective reality.

Ok, you can construct a theory with some components that are intended as qualitative descriptors of underlying phenomena.

If such a theory has a better predictive or explanatory power on measured phenomena than some other theory that does not postulate the existence of objective reality, then it is justified to think that objective reality exists.

If such a theory demonstrates better predictive power, then we would have to seriously consider its metaphysical components as corresponding in some significant way to an underlying reality. But if it doesn't have better predictive power, then the metaphysical embellishments are just that and nothing more.

This is what I try to do with the Bohmian interpretation, to show that such a formulation of QM has a better predictive or explanatory power on measured phenomena than standard QM. I have several results suggesting that it is indeed the case if one considers phenomena BEYOND those described by non-relativistic QM.[ /QUOTE]What does the Bohmian formulation predict more accurately than standard quantum mechanics? What does the Bohmian formulation predict that standard quantum mechanics doesn't?

 

[The Dagda]

But this is exactly my point! The measurement can be completely independent of any kind of theoretical description of the system. If I do a which-way experiment, I do NOT depend on QM's description to figure out if I have photons coincidence, or only one click at a time! I can design the experiment and give it to someone completely ignorant of QM do the experiment.

So my question is, if you do this experiment, what do you think you will get?

All this claim about QM being incomplete is very tiring. This is as IF there is anything at all out there that is "complete". We continue to test QM and to push its boundaries, so it is not as if we are sitting down on our fat rear ends being content with what we already have. So there are people who continue to test and challenge it. But these must be done via valid means. The most convincing way to show that something may not work all the time is via experimental evidence, and not simply via personal preferences or a matter of tastes! Saying "Oh, QM must be incomplete because there must be something wrong if it can't predict for certain which way the photon must go through in a 2-slit experiment" is arguing based on personal tastes! Someone else, like me, have no trouble with it because as an experimentalist, it is what it is when this is what you get empirically! Without any experimental evidence, all arguments are nothing more than a discussion about favorite colors.

Zz.

Well said. The argument from ignorance is also the modus operandi of the fundamentalist and the fringe scientist and the downright crackpot. Who would want to work in a science that was complete in absolute terms anyway?

 

[The Dagda]

The role of science is defined by scientists themselves. If a significant number of scientists thinks that the question of objective reality is scientific, then, by definition, the question of objective reality is scientific too. Such scientists certainly do exist, but the question is whether their number is "significant".

Anyway, personally I think that "objective reality" is scientifically meaningful in the following sense. You may construct a THEORY that postulates the existence of objective reality. If such a theory has a better predictive or explanatory power on measured phenomena than some other theory that does not postulate the existence of objective reality, then it is justified to think that objective reality exists. This is what I try to do with the Bohmian interpretation, to show that such a formulation of QM has a better predictive or explanatory power on measured phenomena than standard QM. I have several results suggesting that it is indeed the case if one considers phenomena BEYOND those described by non-relativistic QM.

Ideas shmideas, shut up and calculate.

 

[jms5631]

I'm sorry if someone has already said something similar, I haven't had time to read the entire thread. It seems to me that dissatisfaction with quantum mechanics always stems from personal philosophical predilections and intuitive discomfort rather then demonstrable insufficiencies. The fact is, QM is the most rigorously tested framework ever produced by man, and continues to pass every test in its current form. Evidence for its shortcomings will come only when it fails to predict an observed phenomena, and not before. That is not to say the quantum formalism will never change, who knows what a coherent theory of quantum gravity will bring, but for the time being, we should accept the framework as certainly being empirically reliable, and cease arguing for its incompleteness without evidence.

 

[meopemuk]

Evidence for its [QM] shortcomings will come only when it fails to predict an observed phenomena, and not before.

In this thread we discussed one obvious "shortcoming" of QM: this theory cannot predict actual physical events, but only their probabilities. For example, there is no way to predict which point on the screen will be hit by the next electron in a double-slit experiment. QM can only say that some points on the screen will be hit more frequently than others. There is no way to predict the exact time sequence of clicks in a Geiger counter. The best QM can do is to calculate the average frequency of the clicks.

Does this mean that QM is an incomplete theory? My answer is "yes". Does this mean that a more "complete" theory is possible? My answer is "no".