The Measurement Problem

[Sherlock]

Having read lots of different ways of describing
the Measurement Problem, I must say that I still
don't understand exactly what it is.

That is, I think I understand what the stuff
I've read is saying, but there doesn't seem to
me to be much of a problem.

Apparently, I'm missing something. So, if
anybody at PhysicsForums does understand why
the Measurement Problem is important, then an
explanation that I can understand would be
appreciated.

[vanesch]

Having read lots of different ways of describing
the Measurement Problem, I must say that I still
don't understand exactly what it is.


The measurement problem is essentially: how and when do wavefunctions give probablities of outcomes ?

Now, you might think you know the answer to that one: if you measure a quantity a of a system in state |psi>, then with it is associated an hermitean operator A, and then the probability to find result a_k (an eigenvalue of A) is given by <psi|P_k|psi> where P_k is the projector on the eigenspace of operator A, with eigenvalue a_k.
The state of the system after this measurement, after having obtained value a_k, is then P_k |psi>, renormalized to unity norm.

There are people that say that there is no problem (the "shut up and calculate" crowd).

But there are a lot of hypotheses that go into the above statement which are somehow contradictory if you dig deeper.
The most obvious problem is the "projection postulate" which gives us the state after the measurement to be P_k |psi>, given by a random choice of P_k. However, all physical interaction is normally described by a hermitean operator called the hamiltonian, giving rise to a unitary evolution operator ; the above "collapse" of the wavefunction is incompatible with such an evolution (the collapse can never be the effect of a unitary operator). So whatever this "measurement" does, it cannot be described by a physical interaction in quantum theory.

This is the most fundamental problem I think: measurements seem to be something else than physical interactions given by electromagnetic, weak or strong interactions (of which we know the hamiltonians) ; and even different from any potential physical interaction that can be described in quantum theory.

Another problem with this projection is that it is bluntly non local and cannot be made lorentz invariant, so it is hard to see how lorentz invariance appears on the macroscopic level, and how it is required for the dynamical description (in QFT for instance), while there is some part of the theory that is bluntly non-local.

Yet another problem you have is the association of an operator describing a measurement (the hermitean operator with its eigenfunctions) and the apparatus that performs the measurement. Indeed, the only thing a measurement apparatus does physically is to *amplify* a certain property of the system to macroscopic levels. For instance, a Stern-Gerlach apparatus amplifies the spin state of the atom in the field direction into macroscopic positions. But why do we consider that the atom can be in superpositions of spin states, but we naturally accept that it must be seen in ONE position ? So somehow we take for granted a certain preferred quantity on the macroscopic level (here, position). This is 'sneaking in' a preferred basis which does not really exist in quantum theory, and without it, there is no way to associate the correct "measurement basis" for the apparatus, and there is no way to find out to what hermitean operator it corresponds.

Finally a problem is that you decide arbitrarily WHAT is your system and what is not (usually you stop short of doing too complex calculations). But a priori your hermitean operator and projection depend on the level of complexity you want to use in order to apply the measurement: for instance, when you measure a light beam, do you only include the light beam, or do you include the solid-state processes in the photocathode ? And do you include the electron transport and so on from dynode to dynode ? ...

All the above problems are problems when you consider that a measurement *really* projects out something during a meausurement (which is then not a physical interaction), but something else, which is not clearly defined.

If however, you consider that a measurement is a physical interaction, described by a unitary operator as all physical interaction is in quantum theory, you encounter other problems. The most obvious one is that you then end up with an observer which is entangled with states of the system, because unitary evolution cannot select one of the states.
The second problem is that even in order to find out what are the alternatives, you have to write out the final state in *a particular basis*.
And finally, the question is how the right probabilities can arise out of such a wave function (the question of the Born rule).

So we have this funny situation that if you "naively" apply the rule for calculating outcomes without thinking too much, everything works out fine and you get perfect agreement with experiment. And the more you think about what exactly is going on, the more problems you encounter :bugeye:

cheers,
Patrick.

[werty]

I all this QM problems just might be one big hint that zen budhism is the correct way to go. In zen the problem is, the more you try not to think the more you think, in QM the problem is the more you learn the less you understand. I heard rumors about some very bright QM researchers that say that they have understood where the collapse of the wavefunction comes from, eg. what collapses all levels of entanglement, however they havent published anything about it, just claim they know. Maybe QM is the path to knowing zen, and zen the path to knowing QM. :uhh:

[Adrian Baker]

I all this QM problems just might be one big hint that zen budhism is the correct way to go....

Or it might just mean that you shouldn't try to find philosophical meaning to Physics data....

[marlon]

The measurement problem is essentially: how and when do wavefunctions give probablities of outcomes ?

Now, you might think you know the answer to that one: if you measure a quantity a of a system in state |psi>, then with it is associated an hermitean operator A, and then the probability to find result a_k (an eigenvalue of A) is given by <psi|P_k|psi> where P_k is the projector on the eigenspace of operator A, with eigenvalue a_k.
The state of the system after this measurement, after having obtained value a_k, is then P_k |psi>, renormalized to unity norm.

There are people that say that there is no problem (the "shut up and calculate" crowd).


I would be a member of that crowd.
But maybe i didn't dig deep enough. QM as a formalism has no problem what so ever in my opinion. There are several interpretations but isn't the outcome all the same. By outcome i mean the real physical values like energy values of a certain system. As long as these different interpretations yield different physical results, then YES there is a problem but to my kowledge this is not the case. Or am i terribly wrong here ?

I never understood the real nature of these "measurement problems", i think.


The most obvious problem is the "projection postulate" which gives us the state after the measurement to be P_k |psi>, given by a random choice of P_k. However, all physical interaction is normally described by a hermitean operator called the hamiltonian, giving rise to a unitary evolution operator ; the above "collapse" of the wavefunction is incompatible with such an evolution (the collapse can never be the effect of a unitary operator). So whatever this "measurement" does, it cannot be described by a physical interaction in quantum theory.


I see your point but i don't understand why that is a problem. A measurement is just an operation right ? So why does it have to be physical interaction. I mean, you just measure the interaction. Or am i not getting your point ?


This is the most fundamental problem I think: measurements seem to be something else than physical interactions given by electromagnetic, weak or strong interactions (of which we know the hamiltonians) ; and even different from any potential physical interaction that can be described in quantum theory.

To me, that just sounds very good because a measurement IS something else. I mean, the measurement is just about detecting what the EM-radiation has "produced"




For instance, a Stern-Gerlach apparatus amplifies the spin state of the atom in the field direction into macroscopic positions. But why do we consider that the atom can be in superpositions of spin states, but we naturally accept that it must be seen in ONE position ?


Well, what else would be the real useful physical outcome ? I mean, we should be able to work with the results of QM, right ? I really don't see the problem. We get one postion because the superposition is broken due to the measurement ? This is a fundamental postualte...So ?

I mean, the results we get are consistent with experiment, what else would you want ?


Finally a problem is that you decide arbitrarily WHAT is your system and what is not (usually you stop short of doing too complex calculations). But a priori your hermitean operator and projection depend on the level of complexity you want to use in order to apply the measurement: for instance, when you measure a light beam, do you only include the light beam, or do you include the solid-state processes in the photocathode ? And do you include the electron transport and so on from dynode to dynode ? ...


Well, for each case, depending on what you want to study and the available data you have, you can apply the QM-formalism in its original nature. So ?



I maybe missing a lot here, but in all honesty, i have been using QM for several years now and i never ever fully understood these socalled measurement problems. I mean, the fact that QM is completely consistent with experiments is the most convincing proof of its accuracy and if it ain't broken, don't fix it...

Vanesch, i hope you can help me out here, because i would really like to understand

regards

marlon

[Elros]

According to the postulate of QM the closed system(isolated from its environment and any measurement process) evolves according to time dependent schrödinger equation.
This equation is deterministic and it is a unitary transformation(conserves the norm)
The mathematical framework of QM desribing interactions and other physical phenomena is based on the unitary and deterministic formalism. However when the collapse postulate enters the picture there arises a "conceptual inconsistency" which is hearth of the measurement problem. The collapse postulate is indeterministic and non-unitary. The system is observed in one its determinate eigenstates and the act of the measurement concerning the choice of the eigenstate is indeterministic. Although the system is in a superposition state prior to measurement, by the reduction of the state vector it collapses to one its eigenstates( say |a>) but this demands renormalization so that the probability to find the system |a> after the second measurement is 1. This means that the first measurement is not unitary.

In conclusion, on one side we have deterministic and unitary schrödinger equation until the measurement and on the other side we have non-unitary and indetermisinistic collapse postulate when the measurement enters the picture. There is an discontiunity between these two conceptually inconsistent views.

[Juan R.]

I would be a member of that crowd.
But maybe i didn't dig deep enough. QM as a formalism has no problem what so ever in my opinion. There are several interpretations but isn't the outcome all the same. By outcome i mean the real physical values like energy values of a certain system. As long as these different interpretations yield different physical results, then YES there is a problem but to my kowledge this is not the case. Or am i terribly wrong here ?

I never understood the real nature of these "measurement problems", i think.


I am sorry to say this marlon but both Elros and vanesch are right.

For example, people with no idea of chemistry (just basic notions) as Hawking, Weinberg, Gell-Mann or Witten have claimed that all of chemistry is already known in many popular media, books, interviews, etc. In one of his last works, Hawking ruminates about his very wrong interpretations of QM and after claim that all, absolutely all of chemistry is derived from quantum physics. Then he said that wait that chemists agree with him.

That ignorance about real chemistry of real laboratories broadly contrasts with standard chemical thinking:

"It is also essential to recognize at the outset that from the perspective of quantum chemistry there are substantial fundamental problems, which are not well understood, centring round the imperative to introduce the classical concept of molecular structure into the formalism. These difficulties do not arise in atomic physics [or in particle physics or string M-theory I add] where the use of atomic eigenstates and the photon Fock space to provide the reference states for a perturbation theory treatment is entirely straightforward; in general, however, eigenstates of the full molecular Hamiltonian do not describe chemical species which we understand in terms of isomerism and functional groups, and the Born-Oppenheimer approximation does not solve this problem"

What is the solution?

It is "straightforward". A generalization of QM, using a new very advanced mathematical formalism is the basis for the rigorous description of molecules and other systems that are do not adequately described by pure QM.

In fact, a simple argument in quantum chemistry shows that QM cannot be correct, since that contradices well-known chemical data about isomers.

A generalized theory that, of course, reduce exactly to QM in "atomic physics" situations is also the basis for the rigorous description of classical systems as the famous cat.

Once i said on physicalforums that canonical theory was ultra-advanced. People without arguments did joke of this and called to me squizophrenic.

Take the last Brushels School's RHS formulation for quantum-LPS. It is lot of times more advanced than the easy standard quantum formalism used, for example in QFT. In fact, Prigogine theory is based in a new branch of mathematics still being developed. Many specialists (including mathematicians) opine that it is exccesive in some mathematical requirements.

Canonical theory is still more advanced and permits us a rigorous analisys of some mathematical aspcets cannot be rigorously addresed with the new math of RHSs or Gelfand triplets.

In fact, the collapse of Hilbert space structure is not related to the divergences of the perturbative expansion of vectors asociated with Poincare resonances. In fact, the appeal to new vectors in L-space where the novel spectral decomposition of observables is achieved is not consistent after all.

Another related problem is the "hidden" link between both topological duals.

Of course, Hawking, Weinberg, Gell-Mann, or Witten have no idea of this.

Of course, as said in the official launching letter (www.canonicalscience.com) even some renombred chemical thinkers as Pople have misguided the point. This is not so strange by several motives:

- Pople is a old man. This signifies that he is well-versed in the old initial QM, but his knowledge of most modern approaches is rather discutible. When i have 70 years, I will not know many new theories and approaches sure, but young people will can offer me new perspectives :biggrin:.

- Pople was mathematician. His approach to problems is from the mathematical side (with an eminent mathematical mind) and often he does not understand completely the most chemical aspects. Somewhat like my mathematical research cannot be good enough for a pure mathematician. They take my work with care and correct my possible errors :biggrin: .

- Pople has centred his research precisely in the obtaining of "energies" once the BO approach is invoked, ad hoc, in the molecular Schrödinger equation. :biggrin:

- One would remember that his Nobel Prize was for work in computational models, newer for conceptual or theoretical issues of a more fundamental face. He newer explained the problem of molecular structure. In fact, in his famous program Gaussian, one enters first the Z-matrix from 19th century chemical theory and then the algorithm computes energies, spin densities, Abelian subgroups, orbitals, hexadecapoles moments, and all that from "Hphy = Ephy" :biggrin:

[vanesch]

I would be a member of that crowd.
But maybe i didn't dig deep enough. QM as a formalism has no problem what so ever in my opinion. There are several interpretations but isn't the outcome all the same. By outcome i mean the real physical values like energy values of a certain system. As long as these different interpretations yield different physical results, then YES there is a problem but to my kowledge this is not the case. Or am i terribly wrong here ?


The different interpretations arise exactly because of the existance of a problem. Now, as I pointed out before, if you "naively" use quantum theory, for simple enough systems, and clear enough built measurement apparatus, then you get out the right results.
But there is clearly something wrong, from a fundamental viewpoint, in the "standard textbook" formulation of quantum theory (even though for all practical purposes, this works, in the case of simple systems, and even complex systems, if you apply the right tricks at the right moment).

There are two ways of viewing quantum theory: or it is a fundamental theory that describes all of physics (including classical physics) ; or it is a theory with a limited domain, that somehow interfaces with classical physics. The last view is the Copenhagen view: namely that macroscopic objects follow classical physics. Mind you: Copenhagen doesn't say that for macroscopic objects, classical physics is _a good approximation_ to the "true" quantum theory. No, Copenhagen says that on macroscopic scales, quantum theory is not applicable, and you have to use *classical* physics as a fundamental theory. On the other hand, microscopic physics is ruled by quantum physics, and there must be somewhere an interface between both, where classical physics takes over from quantum physics. It is this take-over that is ruled by the projection postulate. The problem with this is: WHERE is this transition ? What physical processes, and what objects, belong still to the quantum realm, and what objects are ruled by classical theory ? Or are both limiting forms of a more general theory ? The transition is usually called the "Heisenberg cut". But you admit that the way things are formulated do not give you any hint of what exactly constitutes the transition from quantum theory to classical theory. Take photodetection for instance. One can say that "detecting the photon" gives you the transition from quantum to classical descriptions. But when you analyse a photodetector, you can describe the photo-electric process by a hamiltonian. And once the electron is emitted, you can describe its motion through vacuum also as a free particle in QM. So when, suddenly, is QM NOT valid anymore and does classical physics take over ? An extra problem here is that whatever more general theory (with limiting cases quantum theory for microscopic systems, and classical physics for macroscopic systems) is behind this (and will hence clearly define the boundary), it will have to handle the non-locality of EPR situations.

The other view, namely that quantum theory is a fundamental theory ruling ALL of physics, is the many-worlds view (in one form or another). It tells you that ALL physical processes are described by quantum theory. So all physical interactions are described by hamiltonians, and unitary evolution. I already illustrated several times here, that as long as no satisfying "more general" theory is available, this view is my favorite view, BTW. Most people seem to agree with the statement that "quantum theory is a universally valid theory", but deny the weirdness of many worlds. Well, you CANNOT ESCAPE many worlds if you consider quantum theory universally valid !!

But even there, things are very bizarre. Indeed, as Elros said, unitary quantum theory just makes wavefunctions evolve deterministically. There is no natural appearance of any stochastic nature. So, strange as it may seem, CLASSICAL PHYSICS DOES NOT FOLLOW from unitary quantum theory. You need the Born rule somehow, to introduce the probabilistic nature, but it is inherently NOT PRESENT in unitary quantum theory. And if EVERYTHING evolves unitarily, your body included, then you end up entangled with all the stuff around you. For instance, the superposition principle is valid all the way up, and when you look at a voltmeter, and that voltmeter measured a quantum property (say, 2.5 V if an electron was spin up, and 0 V if an electron is spin down), then your body sees the two states of the voltmeter, in two terms of the "universal wavefunction". So, how the hell do you only see one state ? And why does the world appear classical to us ?

As I already tried to argue here a few times, with my (probably disliked) "consciousness" thing, there is a way out, but admit at least that the issue is not that simple ! And certainly not as simple as textbooks seem to indicate (although I think they are right to do so: you first need to learn the machinery before being able to discuss these issues).


I never understood the real nature of these "measurement problems", i think.


Well, I hope you now see a bit more the issue (even if you can consider the issue "solved" if you accept a) that there is a more general theory of which classical and quantum theory are limiting cases (only, we don't have that theory ! or b) that you accept many worlds views with consciousness issues).


I see your point but i don't understand why that is a problem. A measurement is just an operation right ? So why does it have to be physical interaction. I mean, you just measure the interaction. Or am i not getting your point ?


Ah, that's a theorist at work :biggrin: ! I'm talking about the interaction between the measurement apparatus and your system under study. Why cannot this be also a physical interaction ?


To me, that just sounds very good because a measurement IS something else. I mean, the measurement is just about detecting what the EM-radiation has "produced"


Tell me, what did it "produce" ? An electron ? And can this process be described (in principle) by a hamiltonian, or not ?



Well, what else would be the real useful physical outcome ? I mean, we should be able to work with the results of QM, right ? I really don't see the problem. We get one postion because the superposition is broken due to the measurement ? This is a fundamental postualte...So ?


No, this is not a fundamental postulate ! Consider a Stern-gerlach setup, and a beam of atoms going in (like in the first chapter of Sakurai). Now, if you decide to "look at it" the position of the atoms after a machine is a "measurement", but if you feed them into the next one, it is a "quantum state that evolves". Don't you see the rather contradictory aspects, according to what you tend to pay attention to ?


I mean, the results we get are consistent with experiment, what else would you want ?


A consistent theory ! You don't want a working theory in which concepts change according to your personal preferences. In one case, the interaction of EM radiation with a metal is the "system under study" and evolves according to a hamiltonian, unitarily, because you happen to be a solid-state theorist that writes papers about the photo-electric effect ; in another situation, exactly that same interaction is now a "measurement" because you're a quantum optician who looks at EM beams. But if it is a measurement, leading to a collapse of the wavefunction, then the hamiltonian that the solid-state theorist uses, doesn't exist for the quantum optician (otherwise a unitary evolution would be at its origin).

Now, quantum theory seems to be very forgiving. I think that that is the main contribution of decoherence theory. It seems that if we sneak in classical concepts late enough, then everything happens as if it didn't matter where. But in one way or another, we seem to need to sneak in classical concepts, of which, I repeat, quantum theory has no natural explanation.

I'd say that all these issues are complicated enough to deserve the title of "problem".

cheers,
Patrick.

[vanesch]

"It is also essential to recognize at the outset that from the perspective of quantum chemistry there are substantial fundamental problems, which are not well understood, centring round the imperative to introduce the classical concept of molecular structure into the formalism. These difficulties do not arise in atomic physics [or in particle physics or string M-theory I add] where the use of atomic eigenstates and the photon Fock space to provide the reference states for a perturbation theory treatment is entirely straightforward; in general, however, eigenstates of the full molecular Hamiltonian do not describe chemical species which we understand in terms of isomerism and functional groups, and the Born-Oppenheimer approximation does not solve this problem"


The point you raise is an important one. For people (marlon :-) not aware of this issue, it goes like follows: How do you explain the robustness of isomeres (molecules who are each-other's mirror image). Let's call the state of the "left-handed" molecule" |L> and "the righthanded molecule" |R>.
It is obvious that nor L, nor R, are eigenstates of the molecular hamiltonian (at least if you take into account only QED ! If you introduce the weak interaction, then things are different!), but that |L> + |R> and |L> - |R> are such eigenstates, with different energies. So how come that you can make molecules in, say, the |L> state, which are pharmaceuticals, and which are kept for days in the chemist's cupboard, while if you eat the |R> version, you get sick ?

This is indeed a computational issue in quantum chemistry. However, I do not agree with what Juan R writes, in that this cannot be solved. Probably these |L> and |R> states are stabilised due to interaction with the environment (for instance EM interaction with the EM field). Indeed, chiral molecules are optically active, which means that they couple to the EM field. So probably if you take into account this interaction, this introduces an "effective superselection rule", which forbids the longlivety of "|L> + |R>" states.

cheers,
Patrick.

[ZapperZ]

The point you raise is an important one. For people (marlon :-) not aware of this issue, it goes like follows: How do you explain the robustness of isomeres (molecules who are each-other's mirror image). Let's call the state of the "left-handed" molecule" |L> and "the righthanded molecule" |R>.
It is obvious that nor L, nor R, are eigenstates of the molecular hamiltonian (at least if you take into account only QED ! If you introduce the weak interaction, then things are different!), but that |L> + |R> and |L> - |R> are such eigenstates, with different energies. So how come that you can make molecules in, say, the |L> state, which are pharmaceuticals, and which are kept for days in the chemist's cupboard, while if you eat the |R> version, you get sick ?

This is indeed a computational issue in quantum chemistry. However, I do not agree with what Juan R writes, in that this cannot be solved. Probably these |L> and |R> states are stabilised due to interaction with the environment (for instance EM interaction with the EM field). Indeed, chiral molecules are optically active, which means that they couple to the EM field. So probably if you take into account this interaction, this introduces an "effective superselection rule", which forbids the longlivety of "|L> + |R>" states.

cheers,
Patrick.

Now correct me if I'm wrong. I essentially agree with what you said. I especially disagree with Juan R in saying such isomers "violates" or cannot be explained by QM. However, quantum chemistry is an extremely complex field in which you usually do not even have even a Hamiltonian to start with. This is often the case when you are dealing with many-body interactions, as in the case with condensed matter. In many cases, what you have to resort to is a different starting point than QFT and it's perturbative approach. Often, the most popular methodology is using the Density Functional Theory (DFT). I believe this is the most common way in tackling many problems in quantum chemistry.

However, this is still quantum mechanics, and not a different beast. So to say that QM fails in such a thing is misleading at best.

Zz.

[marlon]

There are two ways of viewing quantum theory: or it is a fundamental theory that describes all of physics (including classical physics) ; or it is a theory with a limited domain, that somehow interfaces with classical physics. The last view is the Copenhagen view: namely that macroscopic objects follow classical physics. Mind you: Copenhagen doesn't say that for macroscopic objects, classical physics is _a good approximation_ to the "true" quantum theory.No, Copenhagen says that on macroscopic scales, quantum theory is not applicable, and you have to use *classical* physics as a fundamental theory.


ok, up till now, you are reading my mind...i totally agree and that is how i have always looked at QM...so far no problemo...


On the other hand, microscopic physics is ruled by quantum physics, and there must be somewhere an interface between both, where classical physics takes over from quantum physics. It is this take-over that is ruled by the projection postulate. The problem with this is: WHERE is this transition ? What physical processes, and what objects, belong still to the quantum realm, and what objects are ruled by classical theory ? Or are both limiting forms of a more general theory ? The transition is usually called the "Heisenberg cut". But you admit that the way things are formulated do not give you any hint of what exactly constitutes the transition from quantum theory to classical theory.


Yes, i totally agree. However i don't see why that justifies that the Copenhagen theory is incorrect or incomplete. I mean why can't you just apply both QM as well as classical physics onto a phenomenon of which you are not sure to what regime it belongs. I mean, experiment will prove you either right or wrong. This is how i see it, but please feel free to instruct me if i am being to naive here.




Take photodetection for instance. One can say that "detecting the photon" gives you the transition from quantum to classical descriptions.


Why ? I would say this is a QM thing...


An extra problem here is that whatever more general theory (with limiting cases quantum theory for microscopic systems, and classical physics for macroscopic systems) is behind this (and will hence clearly define the boundary), it will have to handle the non-locality of EPR situations.


Well ok but in all honest aren't we assigning to much importance to this EPR paradox. I mean, the discription is correct and are there any EPR-related subjects that clearly demonstrate that QM provides us with the wrong answers when calculating certain physical variables ? I think no, so then i wonder, what is all the fuss about ? Maybe this paradox is due to our misinterpretation of the QM-formalism ???


The other view, namely that quantum theory is a fundamental theory ruling ALL of physics, is the many-worlds view (in one form or another). It tells you that ALL physical processes are described by quantum theory. So all physical interactions are described by hamiltonians, and unitary evolution.

But even there, things are very bizarre. Indeed, as Elros said, unitary quantum theory just makes wavefunctions evolve deterministically. There is no natural appearance of any stochastic nature.


Why not. Maybe i am getting this wrong but let me give you this example : suppose you have a wavefunction that is the superposition of two terms like 1 and 0 (let's work with qubits). Now, when a measurement is done you get either 0 or 1 with a certain probability. Isn't the many worlds view based upon saying that if we get a 1, the 0 must occur in another universe with a certain probability ? I have no problem with this view but i don't see the use of it ? Could you enlighten me here, if i am seeing things falsely here...


So, strange as it may seem, CLASSICAL PHYSICS DOES NOT FOLLOW from unitary quantum theory. You need the Born rule somehow, to introduce the probabilistic nature, but it is inherently NOT PRESENT in unitary quantum theory.


ok, in all honesty, i am lost here...Could you provide me with a specific example or some links to more info ? I would be really greatful...


And if EVERYTHING evolves unitarily, your body included, then you end up entangled with all the stuff around you. For instance, the superposition principle is valid all the way up, and when you look at a voltmeter, and that voltmeter measured a quantum property (say, 2.5 V if an electron was spin up, and 0 V if an electron is spin down), then your body sees the two states of the voltmeter, in two terms of the "universal wavefunction". So, how the hell do you only see one state ? And why does the world appear classical to us ?


ok, i see your point but i keep on having a lot of difficulties with the exact use of this vision. I mean, it seems absurd to me that you 'create' that much problems with something of which you know the outcome...I really think i still am not getting the point here.



Tell me, what did it "produce" ? An electron ? And can this process be described (in principle) by a hamiltonian, or not ?


No, what i meant is that the measurement just observes and reflects the outcome of some interaction, not the interaction itself



No, this is not a fundamental postulate ! Consider a Stern-gerlach setup, and a beam of atoms going in (like in the first chapter of Sakurai). Now, if you decide to "look at it" the position of the atoms after a machine is a "measurement", but if you feed them into the next one, it is a "quantum state that evolves". Don't you see the rather contradictory aspects, according to what you tend to pay attention to ?

I see,

So are you saying that a measurement is also an evolving quantum state ?



regards
marlon

[marlon]

So how come that you can make molecules in, say, the |L> state, which are pharmaceuticals, and which are kept for days in the chemist's cupboard, while if you eat the |R> version, you get sick ?


Err i'd say biochemistry :rofl:

but why do i have the feeling that i again am not getting your point ??? :uhh: :confused: :rofl:

marlon

[vanesch]

ok, up till now, you are reading my mind...i totally agree and that is how i have always looked at QM...so far no problemo...


Well, I already have a problem, in that I find it rather schizofrenic to have two different, incompatible theories to describe nature. There should be one fundamental set of axioms that describe everything, no ? Probably if you can live with the idea that the microscopic world is ruled by different physics than the macroscopic world, then you don't have any problems either with the measurement problem. I exaggerate a bit, but that sounds to me like on mondays, you apply Newtonian physics, and on fridays, you apply general relativity, except during holidays. If that works in practice, why not.

I thought that the more prevailing view was that there was ONE theory, namely quantum theory, that underlies everything. After all, what's the point in trying to incorporate gravity into any quantum theory then ? Just also say that, well, gravity applies in a certain set of situations, and quantum theory in another, no ?

Isn't the aim of physics to find the laws of nature, which apply universally ? And try to find out what are the underlying principles to all of it ? And not chop up situations into different domains, and have different principles, laws etc... apply to each of those different domains ?



Yes, i totally agree. However i don't see why that justifies that the Copenhagen theory is incorrect or incomplete. I mean why can't you just apply both QM as well as classical physics onto a phenomenon of which you are not sure to what regime it belongs. I mean, experiment will prove you either right or wrong. This is how i see it, but please feel free to instruct me if i am being to naive here.


Well, I find that 1) a very pragmatic view :-) and 2) a strange view for a theoretical physicist. That would mean that you finally have no predictive power *in principle*. Because you don't know what principles apply, in either situation. So you just have a box with different theories, and apply each one of them, until one fits the experiment ?


Well ok but in all honest aren't we assigning to much importance to this EPR paradox. I mean, the discription is correct and are there any EPR-related subjects that clearly demonstrate that QM provides us with the wrong answers when calculating certain physical variables ? I think no, so then i wonder, what is all the fuss about ? Maybe this paradox is due to our misinterpretation of the QM-formalism ???


There is no "paradox" within quantum theory in EPR. The difficulty comes exactly if you consider a transition to classical physics, because then you clearly have non-local effects on your hand, and there is no limit on their spacelike separation. Now, locality is considered a very important principle, as well in quantum theory as in classical physics. Now (as I tried to outline several times) within an MWI view, EPR situations are compatible with locality. But not if an objective transition to a classical theory is supposed to happen. Now, again, I can understand that there is no need for any explanation if you can already accept the fact that the rules of physics change according to the situation to which you apply them.


Why not. Maybe i am getting this wrong but let me give you this example : suppose you have a wavefunction that is the superposition of two terms like 1 and 0 (let's work with qubits). Now, when a measurement is done you get either 0 or 1 with a certain probability. Isn't the many worlds view based upon saying that if we get a 1, the 0 must occur in another universe with a certain probability ? I have no problem with this view but i don't see the use of it ? Could you enlighten me here, if i am seeing things falsely here...


There is absolutely no use of a "many worlds" view in a probabilistic setting. The point of a MWI view is not just to say that, if A happens with probability P, then not(A) happens somewhere else with probability 1-P, even though that is often how it is represented. The point is that if you already know for what reason A happens with probability P, then we can stop there. MWI finds its origin in that we do not know of a reason for why A happens with a probability P if we take quantum theory as a universally valid theory. But if you do not consider quantum theory as universally valid, but just only applicable to the microscopic world, then MWI is completely superfluous.

However, I'd like to ask something. Maybe I'm wrong, but I thought that it is somehow impossible to combine in a completely consistent way, classical theories with quantum theories ; this brings us back to the discussions between Einstein and Bohr. The point is that objects which are ruled by classical mechanics should not be subject to the Heisenberg uncertainty principle (indeed, if they are classical, they are described by a point in phase space). Now, if you have such objects, I thought that it was always possible to design a (thought) experiment that violates the HUP for microscopic systems. Like, imagine that I have a classical mirror. Then, in principle I can know its position and momentum better than permitted by the HUP. If I now use such 2 mirrors to do an interference experiment with photons, I can in principle know which way a photon took by looking at the momentum transfer (recoil) onto one of the mirrors, while keeping the mirrors sufficiently in place to allow for an interference pattern to build up. If the mirrors satisfy the HUP, then this is impossible because in order to be able to measure the delta-P of the recoil, I have to admit a delta-X which destroys all interference (the mirror is not in a fixed position). But if the mirror is classical, I don't see why it should satisfy the HUP...

cheers,
Patrick.

[vanesch]

Err i'd say biochemistry :rofl:
but why do i have the feeling that i again am not getting your point ???


Well, the point is of course (I thought that was clear) that by symmetry, |L> + |R> and |L>-|R> are eigenstates of the molecular hamiltonian, with slightly different energies. So if you start out with an |L> state and you apply time evolution with that hamiltonian, the state would oscillate between |L> and |R> with a period which is given by the (tiny) energy difference. So how come then that the stuff remains in an |L> state for days ?

It is a bit puzzling, but probably not more so than why a hydrogen atom in an excited state doesn't remain there. The point is that the environment interactions count and that you cannot just work with the molecular hamiltonian as such. But it is an extremely interesting situation ! Small molecules tend to be rather in eigenstates of the molecular hamiltonian, and larger molecules take on a "more classical" configuration, like chiral states, where we "know where the atoms are".

cheers,
Patrick.

[vanesch]

Ha, that's funny, I seem to be about on the same line as d'Espagnat.

Here's some reading: quant-ph/0402121 ; I'll try to find more references...

Of course there is a good overview in quant-ph/0312059 as I already mentionned before.

cheers,
Patrick.

[marlon]

Ha, that's funny, I seem to be about on the same line as d'Espagnat.

Here's some reading: quant-ph/0402121 ; I'll try to find more references...

Of course there is a good overview in quant-ph/0312059 as I already mentionned before.

cheers,
Patrick.

i will certainly browse these through...thanks you very much...i have to go now, for my DFT-lessons

marlon :)

[marlon]

Well, I already have a problem, in that I find it rather schizofrenic to have two different, incompatible theories to describe nature. There should be one fundamental set of axioms that describe everything, no ? Probably if you can live with the idea that the microscopic world is ruled by different physics than the macroscopic world, then you don't have any problems either with the measurement problem. I exaggerate a bit, but that sounds to me like on mondays, you apply Newtonian physics, and on fridays, you apply general relativity, except during holidays. If that works in practice, why not.


mm, i see your point and i have to admit there is some truth in it. However is still keep on having difficulties with 'the way out' of these problems. I mean, i don't see how the different interpretations can solve this one. Besides, the HUP is not violated by classical objects, it does not apply to them. Why ? Well, because For classical objects, the deBroglie wavelength (let's call this L) is much much smaller then the actual dimension of the object. For an electron, this wavelength is much bigger then the actual size of the electron.
Let's call the actual size of the object T.


Let's look at the problem with the error in momentum ∆p first. Since the error in position ∆x is directly related to L/2, and L is vanishingly small for a typical macroscopic entity, it follows from the original Heisenberg’s uncertainty principle that the measurement error in its momentum, ∆p, is extremely large; which is in conflict with reality.

Ocay, but this is resolved by recognizing that the measurement error in momentum, as implied by the Heisenberg’s uncertainty principle, is caused by the inability (of a classical tool) to determine the exact location of the quantum entity within its de Broglie’s wavelength.

This problem is caused by the fact that L >>T which is particular to quantum entities and does not apply to classical entities, where L< T
So the measurement error in momentum does not occur for a classical object because it's size T is bigger then the deBroglie wavelength L...In this case, we can always localize the object because the length-scale we use to measure (the deBroglie wavelength) is smaller then the actual object.

For an electron, this is a bit like saying that you need a distance scale of (let's say) of one metre in order to determin and 'see' the actual electron because it's deBroglie wavelength (which defines the entity : electron) is one meter. But the actual electron is much smaller in dimension, so you cannot determin it's position and momentum exactly. Beware that the distance scales i gave are just to exemplify and they are ofcourse much smaller in QM.


Why can't we use such a way of reasoning in order to determin whether a phenomenon is classical or QM ? aybe this is wrong, but my point is hat we should look for these kinds of alogritms in order to verify the physical nature, you see ? I really don't see the use of these interpretations...




There is no "paradox" within quantum theory in EPR. The difficulty comes exactly if you consider a transition to classical physics,


yes and that is what bothers me. Einstein said for something to be an element of reality you should be able to determin it at all times and its value should be fixed and you should be able to get it with absolute certainty, right ? Well, i say this is wrong because QM-phenomena are probabilistic in nature and the HUP is inherent to how mother nature operates. I mean, who is Einstein to say stuff like : well, this is an eement of reality for this and that reason...I mean this is just his opinion purely based on philisophycal reasoning...QM works, point final, that is how i think about it. You cannot just use classical reasoning in QM. If you did that, well then there wouldn't be any stable atoms, ya see ?






The point is that if you already know for what reason A happens with probability P, then we can stop there. MWI finds its origin in that we do not know of a reason for why A happens with a probability P if we take quantum theory as a universally valid theory.


correct me if i am wrong but isn't that the same as asking why gravity exists ? Physics does not answer that question in my opinion. It only describes nature, it does not tell us why nature works, the way she does...


The point is that objects which are ruled by classical mechanics should not be subject to the Heisenberg uncertainty principle (indeed, if they are classical, they are described by a point in phase space). Now, if you have such objects, I thought that it was always possible to design a (thought) experiment that violates the HUP for microscopic systems.


sound very strange to me
Again, why compare two things that are totally different in nature ?


Like, imagine that I have a classical mirror. Then, in principle I can know its position and momentum better than permitted by the HUP. If I now use such 2 mirrors to do an interference experiment with photons, I can in principle know which way a photon took by looking at the momentum transfer (recoil) onto one of the mirrors, while keeping the mirrors sufficiently in place to allow for an interference pattern to build up. If the mirrors satisfy the HUP, then this is impossible because in order to be able to measure the delta-P of the recoil, I have to admit a delta-X which destroys all interference (the mirror is not in a fixed position). But if the mirror is classical, I don't see why it should satisfy the HUP...


ofcourse, the mirror is classical, it does not obey the HUP because you are in another physical regime. In the classical regime, the HUP does not exist...and don't mix this with the analogy of the HUP with diffraction of waves...

marlon

[vanesch]

Here's some reading ; I did some bibliographical selection ; didn't read (or am able to read!) everything myself, but I think it contains interesting pointers for those interested in the matter... I tried to leave out the most cranky texts ; most have been published somewhere.

quant-ph/0403184 Authors: Rachael M. McDermott, Ian H. Redmount

Some of the most enduring questions in physics--including the quantum measurement problem and the quantization of gravity--involve the interaction of a quantum system with a classical environment. Two linearly coupled harmonic oscillators provide a simple, exactly soluble model for exploring such interaction. Even the ground state of a pair of identical oscillators exhibits effects on the quantum nature of one oscillator, e.g., a diminution of position uncertainty, and an increase in momentum uncertainty and uncertainty product, from their unperturbed values. Interaction between quantum and classical oscillators is simulated by constructing a quantum state with one oscillator initially in its ground state, the other in a coherent or Glauber state. The subsequent wave function for this state is calculated exactly, both for identical and distinct oscillators. The reduced probability distribution for the quantum oscillator, and its position and momentum expectation values and uncertainties, are obtained from this wave function. The oscillator acquires an oscillation amplitude corresponding to a beating between the normal modes of the system; the behavior of the position and momentum uncertainties can become quite complicated. For oscillators with equal unperturbed frequencies, i.e., at resonance, the uncertainties exhibit a time-dependent quantum squeezing which can be extreme.

quant-ph/0504199 Authors: Jean Schneider

I show that the quantum measurement problem can be understood if the measurement is seen as a ``speech act'' in the sense of modern language theory. The reduction of the state vector is in this perspective an intersubjectice -- or better a-subjective -- symbolic process. I then give some perspectives on applications to the ``Mind-Body problem''.


gr-qc/0503046 Authors: Abel Camacho

The possible description of the vacuum of quantum gravity through the so called kappa--Poincare group is analyzed considering some of the consequences of this symmetry in the path integral formulation of nonrelativistic quantum theory. This study is carried out with two cases, firstly, a free particle, and finally, the situation of a particle immersed in a homogeneous gravitational field. It will be shown that the kappa--Poincare group implies the loss of some of the basic properties associated to Feynman's path integral. For instance, loss of the group characteristic related to the time dependence of the evolution operator, or the breakdown of the composition law for amplitudes of events occurring successively in time. Additionally some similarities between the present idea and the so called restricted path integral formalism will be underlined. These analogies advocate the claim that if the kappa--Poincare group contains some of the physical information of the quantum gravity vacuum, then this vacuum could entail decoherence. This last result will also allow us to consider the possibility of analyzing the continuous measurement problem of quantum theory from a group--theoretical point of view, but now taking into account the kappa--Poincare symmetries.


quant-ph/0412144 Authors: Pradip Kumar Chatterjee

In order to resolve the measurement problem of Quantum Mechanics, non-unitary time evolution has been derived from the unitarity of standard quantum formalism. New wave functions of free and non-free quantum systems follow from Schroedinger equation after inserting an ansatz. Quantum systems show up as probability waves before measurement. A pure entangled state of a composite system evolves non-unitarily, only to disentangle itself into a definite state after reduction at the measurement point. A classical space-time point is created momentarily in this event. Unitarity is restored at that point. The non-Hermitian observables defined in the domain of rigged Hilbert space transform into Hermitian ones at the measurement point. The problem of preferred basis is resolved by the requirement of specifying the position of measurement point. Two theorems prove that time is a non-Hermitian operator, thus placing space and time on an equal footing. Bound states are found to need discrete space-time, which supports its use in loop quantum gravity. Non-unitarity in the theory helps buttress the no-boundary proposal; and uncertainty relation makes a leeway to singularity-free Quantum Cosmology. Quantum Mechanics also accommodates complex and negative probabilities.

quant-ph/0403094 Authors: Harvey R. Brown, David Wallace

The quantum theory of de Broglie and Bohm solves the measurement problem, but the hypothetical corpuscles play no role in the argument. The solution finds a more natural home in the Everett interpretation.

quant-ph/0402121 Authors: Bernard d'Espagnat

It is generally agreed that decoherence theory is, if not a complete answer, at least a great step forward towards a solution of the quantum measurement problem. It is shown here however that in the cases in which a sentient being is explicitly assumed to take cognizance of the outcome the reasons we have for judging this way are not totally consistent, so that the question has to be considered anew. It is pointed out that the way the Broglie-Bohm model solves the riddle suggests a possible clue, consisting in assuming that even very simple systems may have some sort of a proto-consciousness, but that their ``internal states of consciousness'' are not predictive. It is, next, easily shown that if we imagine the systems get larger, in virtue of decoherence their internal states of consciousness progressively gain in predictive value. So that, for macro-systems, they may be identified (in practice) with the predictive states of consciousness on which we ground our observational predictions. The possibilities of carrying over this idea to standard quantum mechanics are then investigated. Conditions of conceptual consistency are considered and found rather strict, and, finally, two solutions emerge, differing conceptually very much from one another but in both of which the, possibly non-predictive, generalized internal states of consciousness play a crucial role.

quant-ph/0402044 Authors: S.Mayburov

It's argued that Information-Theoretical restrictions for the systems selfdescription are important for Quantum Measurement Problem. They are described by O information system restricted states R formalism by Breuer and can be obtained also in Algebraic QM considering Segal Algebra of O observables. From Segal theorem it's shown that R describes the random measurement outcomes in the individual events.

quant-ph/0307044 Authors: G. B. Lesovik, A. V. Lebedev, G. Blatter

Although quantum mechanics is a mature theory, fundamental problems discussed during its time of foundation have remained with us to this day. These problems are centered on the problematic relation between the quantum and classical worlds. The most famous element is the measurement problem, i.e., the measurement of a quantum system by a classical apparatus, and the concomitant phenomena of wave packet reduction, the appearance of probability, and the problems related to Schr\"odinger cat states. A fundamental question in this context is whether quantum mechanics can bootstrap itself to the classical world: is quantum mechanics self-consistent, such that the measurement process can be understood within quantum mechanics itself, or does this process require additional elements from the realm outside of traditional quantum mechanics? Here, we point to a problematic aspect in the traditional Schr\"odinger cat argument which can be overcome through its extension with a proper macroscopic preparation device; the deliberate creation of a cat state and its identification then turns into a non-trivial problem requiring the determination of the evolution of a quantum system entangled with a macroscopic reservoir. We describe a new type of wave-function correlator testing for the appearance of Schr\"odinger cat states and discuss its implications for theories deriving the wave function collapse from a unitary evolution.

quant-ph/0302160 Authors: R. Srikanth

Is the dynamical evolution of physical systems objectively a manifestation of information processing by the universe? We find that an affirmative answer has important consequences for the measurement problem. In particular, we calculate the amount of quantum information processing involved in the evolution of physical systems, assuming a finite degree of fine-graining of Hilbert space. This assumption is shown to imply that there is a finite capacity to sustain the immense entanglement that measurement entails. When this capacity is overwhelmed, the system's unitary evolution becomes computationally unstable and the system suffers an information transition (`collapse'). Classical behaviour arises from the rapid cycles of unitary evolution and information transitions.
Thus, the fine-graining of Hilbert space determines the location of the `Heisenberg cut', the mesoscopic threshold separating the microscopic, quantum system from the macroscopic, classical environment. The model can be viewed as a probablistic complement to decoherence, that completes the measurement process by turning decohered improper mixtures of states into proper mixtures. It is shown to provide a natural resolution to the measurement problem and the basis problem.

quant-ph/0112095 Author: Stephen L. Adler

We discuss why, contrary to claims recently made by P. W. Anderson, decoherence has not solved the quantum measurement problem.

quant-ph/0105126 Authors: Diederik Aerts

In the hidden measurement formalism that we develop in Brussels we explain the quantum structure as due to the presence of two effects, (a) a real change of state of the system under influence of the measurement and, (b) a lack of knowledge about a deeper deterministic reality of the measurement process. We show that the presence of these two effects leads to the major part of the quantum mechanical structure of a theory where the measurements contain the two mentioned effects. We present a quantum machine, where we can illustrate in a simple way how the quantum structure arises as a consequence of the two effects. We introduce a parameter 'epsilon' that measures the amount of the lack of knowledge on the measurement process, and by varying this parameter, we describe a continuous evolution from a quantum structure (maximal lack of knowledge) to a classical structure (zero lack of knowledge). We show that for intermediate values of epsilon we find a new type of structure that is neither quantum nor classical. We analyze the quantum paradoxes and show that they can be divided into two groups: (1) The group (measurement problem and Schrodingers cat paradox) where the paradoxical aspects arise mainly from the application of standard quantum theory as a general theory (e.g. also describing the measurement apparatus). This type of paradox disappears in the hidden measurement formalism. (2) A second group collecting the paradoxes connected to the effect of non-locality (the Einstein-Podolsky-Rosen paradox and the violation of Bell inequalities). We show that these paradoxes are internally resolved because the effect of non-locality turns out to be a fundamental property of the hidden measurement formalism itself.

quant-ph/0006104 Authors: S.Mayburov

The quantum measurement problem considered for measuring system (MS) consist of measured state S (particle), detector D and information processing device (observer) O. It's shown that O states selfreference structure results in principal nonobservability of MS interference terms which discriminate pure and mixed S states. Such observables restriction permit to construct for MS states subjective representation (SR) which describes probabilistic evolution for measurement events observed by $O$ and his subjective information about S values. SR is dual and nonequivalent to MS Hilbert space $H$ for external observer $O'$. Due to it SR evolution is compatible with Schrodinger linear MS evolution observed by $O'$. It's argued that SR evolution corresponds to S state collapse for individual events observed by $O$.

gr-qc/9808033 Author: Jeeva Anandan

The quantum measurement problem and various unsuccessful attempts to resolve it are reviewed. A suggestion by Diosi and Penrose for the half life of the quantum superposition of two Newtonian gravitational fields is generalized to an arbitrary quantum superposition of relativistic, but weak, gravitational fields. The nature of the ``collapse'' process of the wave function is examined.

quant-ph/9802020 Authors: Carlo Rovelli

Without addressing the measurement problem (i.e. what causes the wave function to ``collapse'', or to ``branch'', or a history to become realized, or a property to actualize), I discuss the problem of the timing of the quantum measurement: assuming that in an appropriate sense a measurement happens, when precisely does it happen? This question can be posed within most interpretations of quantum mechanics. By introducing the operator M, which measures whether or not the quantum measurement has happened, I suggest that, contrary to what is often claimed, quantum mechanics does provide a precise answer to this question, although a somewhat surprising one.

quant-ph/9609002 Authors: Carlo Rovelli

I suggest that the common unease with taking quantum mechanics as a fundamental description of nature (the "measurement problem") could derive from the use of an incorrect notion, as the unease with the Lorentz transformations before Einstein derived from the notion of observer-independent time. I suggest that this incorrect notion is the notion of observer-independent state of a system (or observer-independent values of physical quantities). I reformulate the problem of the "interpretation of quantum mechanics" as the problem of deriving the formalism from a few simple physical postulates. I consider a reformulation of quantum mechanics in terms of information theory. All systems are assumed to be equivalent, there is no observer-observed distinction, and the theory describes only the information that systems have about each other; nevertheless, the theory is complete.

quant-ph/9506020 Authors: H. D. Zeh

Introduction to the theory of decoherence. Contents: 1. The phenomenon of decoherence: superpositions, superselection rules, decoherence by "measurements". 2. Observables as a derivable concept. 3. The measurement problem. 4. Density matrix, coarse graining, and "events". 5. Conclusions.




cheers,
Patrick.

[marlon]

Wow Vanesch, impressive thread...do i need to read all of this :) ?

just kiddin, thanks for the info, really appreciate it

regards

marlon

[vanesch]

Let's look at the problem with the error in momentum ∆p first. Since the error in position ∆x is directly related to L/2, and L is vanishingly small for a typical macroscopic entity, it follows from the original Heisenberg’s uncertainty principle that the measurement error in its momentum, ∆p, is extremely large; which is in conflict with reality.


Eh, no, the Heisenberg relation dx.dp >= hbar/2 works with ABSOLUTE x and p, not with relative errors. So the momentum error dp for a mirror, or for an electron, is the same, given a certain dx (assuming for a moment that the HUP is valid for a mirror, which I think it is, BTW). So if you know that your mirror is within a position dx << L/2, then the dp must be larger than a certain momentum, but that momentum corresponds still to extremely low speeds for a macroscopic object. Only, it is relatively big as compared to the momentum that will be transmitted by the photon banging on it, which makes it impossible to find out whether the photon bounced off the mirror or not.
However, if the HUP is NOT valid for the mirror, then I can, in one way or another, make sure that it is in a rather well known position x, and has a momentum which is very much below the transmitted momentum by the photon. After the photon has passed by, I can then wait for, say half an hour, and if the mirror did displace appreciably from its original position (assuming it could float freely during that half hour), I know the photon came by.
But given the large scale of the object as compared to a reasonable dx and the large scale of the object's momentum (even at very low speeds) to the associated dp of the HUP, we can say that there exist quantum states (coherent states) for the object which are, in all respects, compatible with what we usually take for granted in classical physics, with usual magnitudes, and with usual errors for position and momentum. But that's no proof that the quantum description is invalid ! So it is not because the RELATIVE errors dx/T and dp / P are very tiny that the HUP is not valid for the mirror.


So the measurement error in momentum does not occur for a classical object because it's size T is bigger then the deBroglie wavelength L...In this case, we can always localize the object because the length-scale we use to measure (the deBroglie wavelength) is smaller then the actual object.


No, this only means that we can make wavepackets (coherent states) which are on a relative scale, well-localized, and with a relative accuracy of the momentum (meaning that the blurry spot in phase space is small with respect to the immense volume of phase space a macroscopic object has ; but the absolute volume of the blurry spot always is of order hbar^3).


You cannot just use classical reasoning in QM. If you did that, well then there wouldn't be any stable atoms, ya see ?


You don't have to convince me :-) But apparently at some point you have to use classical reasoning and that's where difficulties arise (also called "the measurement problem")


correct me if i am wrong but isn't that the same as asking why gravity exists ? Physics does not answer that question in my opinion. It only describes nature, it does not tell us why nature works, the way she does...


Well, that would be nice, if physics could do that. But the situation is less evident. It is as if you say that Newtonian gravity works in our solar system but not for, say the solar system around another star X (if ever it had a solar system). There is not really a criterion for mass (a crystal lattice can have a quantum mechanical behaviour: phonons, while a single electron can have a "classical" behaviour when looking at its track in a particle tracker), there is not really a criterion for distance (EPR results have been established over 50 miles if I'm not mistaking). So what's the criterion for applying classical physics, and when do we use quantum theory ? And what happens with the border cases ?


ofcourse, the mirror is classical, it does not obey the HUP because you are in another physical regime. In the classical regime, the HUP does not exist...


How small does a mirror have to be for the HUP to apply to it ?

cheers,
Patrick.

[ZapperZ]

Here's some reading ; I did some bibliographical selection ; didn't read (or am able to read!) everything myself, but I think it contains interesting pointers for those interested in the matter... I tried to leave out the most cranky texts ; most have been published somewhere.

You probably were citing stuff that can be readily available online. Still, a very conpicuous absence in your list is the two major articles by Leggett, who HAS made significant impact on this issue. While one needs a subscription to get to both of these articles, I don't think it would be complete to make a list such as yours and not even cite these:

1. A.J. Leggett, J. Phys. Cond. Matt., v.14, p.415 (2002).
2. A.J. Leggett, Science v.307, p.871 (2005).

I think is summary in Ref. [2] clearly sums up the "Quantum Measurement Problem" as we know of today and the leading experiments that indicate where it appears to be heading.

Zz.

[vanesch]

You probably were citing stuff that can be readily available online.

Thanks for the extra references. I didn't mean in any way to be "complete" ! I just browsed through the arxiv, picking out what seemed interesting.

cheers,
Patrick.

[Juan R.]

Multiple answer

The quote that I posted is a usual thinking in quantum chemistry and molecular physics. The author was not referring to optical isomers, was talking about isomers in general.

The basic problem is that a molecule cannot be represented from pure QM.

Your comments about optical isomerism do not solve the problem. The problem is not find a |L> or |R> state and attempt to explain why is stable or not. The problem is that one needs first introduce the correct molecular state into the quantum description. That molecular state is derived from outside of QM and, morever, violates QM.

This is the problem with QM. By this reason, the author said that BO approximation does not solve the problem. It is not a problem about stability of L or R isomers, the most basic question is not why there is not a L + R state, the question is if there is a L (or R) stable state, that state can be splited into several states according to molecular geometries: L1, L2, L3, etc. Why may one select a priori one of those “substates” and ignore the rest violating the superposition principle?

Answer, because molecules are not pure quantum systems, are a mixture of quantum and classical systems. The pure quantum part (electrons) verify very well standard QM, the nuclear framework is a semiclassical object and does not follow the nuclear Schrödinger equation.

Note the emphasis of the expertise on that

“centring round the imperative to introduce the classical concept of molecular structure into the formalism.”

It is not about optical isomers in biological systems, where I think that has been already solved. Stability of an isomer over the other is due to a contact-like interaction between quarks and electrons. Stability and the arising of biological isomers appears to be well explained in electroweak quantum chemistry.

Therefore, the real question about molecular structure is not a computational issue inside QM. It is that molecular structure is only explained when abandoned the very basis of quantum mechanics, precisely when one abandon carefully certain quantum superposition that are newer observed. Not only one may abandon QM, still is need the appeal to pure chemistry (e.g. the chemical theory of functional groups say which is the correct molecule).

In the same form that quantum superposition principle does not work for cats, it does not work, in general, for molecules. In fact, the full application of QM to molecules give you wrong answer in several molecular properties, it is only when one eliminate certain quantum interferences that one obtain the correct answer. QM fails because cannot explain the classical word.

The problem is basically that of the “Copenhagen” (Bohr) interpretation. There are two worlds: quantum and classical, and one cannot explain the other. The great interest of molecular systems is that are “mixed” systems between the famous classical cat and quantum electrons. There one can study the famous Bohr frontier between both worlds.

DFT is basic in computational quantum chemistry, specially in the work of “big” molecules, but research in quantum chemistry (specially theoretical one) is mainly guided by other issues: CI, MP, CC, etc.

Quantum chemistry is not applied QM. QC could be described like QM(only part of them) + chemical theory.

In fact, I remark again that if one uses pure QM in quantum chemistry and introduce a pure wavefunction for a molecule one obtains the wrong answer for molecular properties, in general.

Now, my views. All attempt to derive classical word from QM is condemned to failure, simply because both description are incompatible. QM is based in superposition of quantum states. Classical physics is based in non-superposition.

Still, there are people that continue in the traditional way of “deriving” (nobody achieved it) non-superposition from superposition. This is obviously impossible, and the true basis of the more than 100 years failure of the program.

That is the reason that all approaches from old multiple worlds to recent decoherence have failed for solving the problem.

The most successful approach is recognize that both (classical and quantum) are two different theories and then formulate a theory that derive both like special cases. That theory is already formulated in my approach. The superposition principle arises in determined situations (e.g. in an atom) but is not applicable in others (e.g. in a cat). All satisfactory formulas used in the laboratories of the world:

- pure quantum. E.g. Schrödinguer
- pure classical. E.g. Newton equation
- semiquantum (or semiclassical) equations. E.g. Caldeira-Legget,
- etc.

arise like special case from my theory.

I think that sophisticated Brussels-School (Prigogine) theory (a “simplification” from my theory) has highlighted several important key questions that can convince you why QM does not work always.

In QM one has vectors of kind

|phi> = c1 |1> + c2 |2>

with “geometry” |c1|^2 + |c2|^2 = 1

but in molecules, cats, etc. that vector does not offer the correct outcomes of averages for observables O. One is obligated to compute the averages <O> with new vectors.

|rho>> = b1 |1>> + b2 |2>>

with “geometry” |b1| + |b2| = 1

this vector is not a special case of above vector is a new vector. Prigogine develops his theory in a RH space, and the new vectors |j>> are a generalization of Gamow states (QM works with Dirac states |j> in usual H space).

My theory is more general still (work in L and S spaces) and explain why for an atom, the first vector work very well, but for cats, work better the second vector.



The uncertainty principle can be incompatible with classical physics. The basis of the error is easily seen in the elementary particle in a box model. Let me take the next uncertainty relation

deltap · deltax = h

still see the effect of taking h --> 0 in Heisemberg formula. Most people indicate that

deltap · deltax = 0

implies that one can measure position and impulse perfectly for a classical object. That is wrong. deltap vanish but deltax is of the order of the side of the box in that model. In a box of 2 metres, the uncertainty for a classical object from QM in this model is of the order of 2 metres. Incompatible with experience!!! a Ball into a room is perfectly localized!!!

Yes, the model is highly idealized, but one can see where is the problem with incompatibility with experience. The problem is not the use of an irrealistic model. The problem is that it is the product that appears in Heisember relation instead of each uncertainty by separate. Therefore localization of x always implies macroscopic delocalization of p and viceverse independently of the specific model used. Textbooks simply ignore that.

From QM one cannot obtain none of classical equations (Newton equation is not obtained from QM). Moreover, application of QM to classical physics offers wrong answers. Bohr known that, Einstein known that, most of modern physicists prefer ignore it.

We may search for a new theory.

[ZapperZ]

Multiple answer

The quote that I posted is a usual thinking in quantum chemistry and molecular physics. The author was not referring to optical isomers, was talking about isomers in general.

The basic problem is that a molecule cannot be represented from pure QM.

You are implying that ALL molecules cannot be represented from "pure QM", whatever that is? And H2 molecule that is clearly described by QM having the bonding-antibonding bonds cannot be described by QM? Not only that, if this is true, then QM would have even MORE of a trouble describing solids in general since the many-body effects are even more apparent in such a system. Yet, we have a well-established solid state physics.

Again, the basic FACT that DFT and other methodologies have been quite successful, at the very least, with coming up with a reasonable ground state of very complex chemical and solid state system shows that one simply cannot say that QM cannot represent such systems. One only needs to look at how successful the calculation for the band structure of solids to be convinced of this.

Your comments about optical isomerism do not solve the problem. The problem is not find a |L> or |R> state and attempt to explain why is stable or not. The problem is that one needs first introduce the correct molecular state into the quantum description. That molecular state is derived from outside of QM and, morever, violates QM.
.
.
.

In the same form that quantum superposition principle does not work for cats, it does not work, in general, for molecules. In fact, the full application of QM to molecules give you wrong answer in several molecular properties, it is only when one eliminate certain quantum interferences that one obtain the correct answer. QM fails because cannot explain the classical word.

.
.
.

Now, my views. All attempt to derive classical word from QM is condemned to failure, simply because both description are incompatible. QM is based in superposition of quantum states. Classical physics is based in non-superposition.

.
.
.
My theory is more general still (work in L and S spaces) and explain why for an atom, the first vector work very well, but for cats, work better the second vector.

Please give peer-reviewed citations where your "theory" has been published, including citations of all so-called violations of QM in "molecules".

Zz.

[Juan R.]

You are implying that ALL molecules cannot be represented from "pure QM", whatever that is? And H2 molecule that is clearly described by QM having the bonding-antibonding bonds cannot be described by QM? Not only that, if this is true, then QM would have even MORE of a trouble describing solids in general since the many-body effects are even more apparent in such a system. Yet, we have a well-established solid state physics.

Again, the basic FACT that DFT and other methodologies have been quite successful, at the very least, with coming up with a reasonable ground state of very complex chemical and solid state system shows that one simply cannot say that QM cannot represent such systems. One only needs to look at how successful the calculation for the band structure of solids to be convinced of this.



Please give peer-reviewed citations where your "theory" has been published, including citations of all so-called violations of QM in "molecules".

Zz.



I am talking in a general form. Most of molecules cannot be studied from QM. Very small and linear gas molecules like H2 are rather well described, in general, from QM.

When I said QM i am refering to pure QM. I am not talking about computational schemes used in solid state physics or quantum chemistry where pure QM is not used.

It is my impresion that you have little idea of the "application" of QM formalism to those fields. They work only in part, the total application of QM to solid state chemistry fails completely, the same is valid for the physics.

Please, if QM works perfectly in the description of molecules or solids in your view, could you highlight here what procedure do you follow for computing, for instance, bonding state for the methyl acetylene molecule? Take for example the omnipresent Gaussian 98.

I known "a bit" the comercially available computational models (as it is clear in the page 8 of the official launching letter www.canonicalscience.com) and also very advanced models. DFT relies in the same basic premises that WFT, both are good electronic structure methods. But i am talking about consider a molecule or a solid like a quantum system. That is newer considered in usual QM applications that you are citing.

Your appeal to the successful calculation for the band structure of solids is precisely a proof that I am saiyng is correct. It is obvious that I explained why Schrödinger equation work for the derivation of those bands. I had experience in computational issues (I know algorithms of Gaussian 98, GAMESS, etc.). In fact, I neglected a position in the university like a computational assistant for the quantum dep. group precisely for developing this new theory.

Your "arguments" on my view are rather irrelevant and precisely are a proof that I am saying. Please read with more care my previous post.




Please give peer-reviewed citations where your "theory" has been published, including citations of all so-called violations of QM in "molecules".

Zz.


Since that you solicite peer-reviewed citations on my ""theory"", posibly on a irrelevant attempt to convince you that I am saying cannot be true.

I prefer ""reply"" you with two paragraphs extracted from an elementary educative work published in the Journal of Chemical Education 77 in the year 2000.

"Woolley and others have claimed that a purely quantum mechanical description involving the raw molecular Hamiltonian without use of the Born–Oppenheimer approximation does not require the attribution of any structure to molecules."

"Most chemists react with complete incredulity to the view that structure is nothing but a metaphor, pointing out the seemingly overwhelming evidence for structure that comes from spectroscopic and other structural studies. They suggest that if a deep quantum mechanical analysis reveals molecular structure to be a mathematical artifact, then the fault must lie with present-day quantum mechanics and not with the deeply entrenched chemical notion of structure."

Precisely, my theory (perfectly verified in multiple experiments up at second order in Lv and first order in time perturbation expansion) demonstrate that fault... of QM on a sound mathematical basis. The idea that QM does not work in molecules, solids, liquids, cats, etc. is well-known by many specialists. If you (like Weinberg, Gell-Mann, Hawking, etc.) newer heard before, it is not my problem.

[Juan R.]

I said

"I am talking in a general form. Most of molecules cannot be studied from QM. Very small and linear gas molecules like H2 are rather well described, in general, from QM."

I was thinking from a theoretical point of view. Note that also pure QM is not used when one compute spectrosciopies frecuency for the H2 with a Gaussian 98. The program does not use pure QM, in fact, the solution that one can read in the output file violates QM. True?

If you are an expert in computational algorihtms I assume that you would know why.

A "pista" for you: What is the Z-matrix for the H2?

[ZapperZ]

I am talking in a general form. Most of molecules cannot be studied from QM. Very small and linear gas molecules like H2 are rather well described, in general, from QM.

When I said QM i am refering to pure QM. I am not talking about computational schemes used in solid state physics or quantum chemistry where pure QM is not used.

I'm sorry, but what in the world is "pure QM"? Are you saying that in DFT, when they really start with the Hamiltonian, but instead, focused on the charge density, that is no longer "pure QM"? Or are you also insisting that if I write the many-body hamiltonian using the single-particle Green's function, that is no longer "pure QM"? Who made this distinction? Since WHEN does this ever matter, considering that QM itself has several different formulations?

It is my impresion that you have little idea of the "application" of QM formalism to those fields. They work only in part, the total application of QM to solid state chemistry fails completely, the same is valid for the physics.

Please, if QM works perfectly in the description of molecules or solids in your view, could you highlight here what procedure do you follow for computing, for instance, bonding state for the methyl acetylene molecule? Take for example the omnipresent Gaussian 98.

This is all MOOT since obviously, you're making a distinction between "pure QM" (whatever that is) and how things are "computed". It appears that you have a fixation that "pure QM" must be those wavefunction superpositions described by the naive description of what QM is. This is a VERY jaundice and primitive view of QM and ignores that fact that ANY theory evolves as one applies it to more and more situations. That's like saying that just because certain problems are too difficult to be solve using standard Newton's force equation, that classical mechanics is wrong even if I can solve it using the Lagrangian method, as IF they are not one of the same.

I want to know who here studied 2nd quantization and Green's function/perturbation approach, and DFT, but without first having the need to learn "basic" QM.

Zz.

[Juan R.]

If you know like works DFT, like works WFT, like the Gaussian 98 computes the bond state for a molecule, like a solid state physicist or an inorganic chemist compute charge density for the Si(111), etc. then you will know exactly that is pure QM.

Please read with care, and after re-read again my previous posts.


I think that you are posting here without idea on the field. Since my traumatic experience in my previous post in gravitation in this same forum, where people with no idea of the field but lots of garbage posted completely wrong posts. I said that I would not waste my time with irrelevant posts. My free time is limited and cannot reply people who has not studied the fields minimally.

I repeat for you the previous paragraph now in bold:

"Most chemists react with complete incredulity to the view that structure is nothing but a metaphor, pointing out the seemingly overwhelming evidence for structure that comes from spectroscopic and other structural studies. They suggest that if a deep quantum mechanical analysis reveals molecular structure to be a mathematical artifact, then the fault must lie with present-day quantum mechanics and not with the deeply entrenched chemical notion of structure."

Note that was extracted from an elementary article in an educative journal!!!! I was not quoting the last fad in the more advanced research journal. If you have no knowledge of those basic points, I recommend you to begin your study before post irrelevant/wrong comments.

This is the usual procedure for a scientist...


Still let me offer you again the crucial part of the educative paper paragraph:

"They suggest that if a deep quantum mechanical analysis reveals molecular structure to be a mathematical artifact, then the fault must lie with present-day quantum mechanics and not with the deeply entrenched chemical notion of structure."

"present-day quantum mechanics" sound like "pure QM". It refers to current available QM. It is not talking about any specific (e.g. Hamiltonian) formulation of QM, it is saying that old (19th century) chemical theory is compatible with experimental data, still current (today) QM is not, because cannot explain molecular structure. This is solved with my approach and explain why "deep quantum mechanics" or "pure QM" or simply QM is not used in that "sucess" that you consider, and why in solid state physics or molecular chemistry one does not use pure QM or Qm is short, just a part of QM... precisely the part that work.

If you has studied a bit of computational methods you would replied to me several post ago what is the computational procedure used in solid state physics and quantum chemistry and that part of QM is used in the computerized algorithms and that part is not used because violate experimental results.

"Most chemists react with complete incredulity to the view that structure is nothing but a metaphor, pointing out the seemingly overwhelming evidence for structure that comes from spectroscopic and other structural studies..."

[ZapperZ]

If you know like works DFT, like works WFT, like the Gaussian 98 computes the bond state for a molecule, like a solid state physicist or an inorganic chemist compute charge density for the Si(111), etc. then you will know exactly that is pure QM.

Please read with care, and after re-read again my previous posts.


I think that you are posting here without idea on the field. Since my traumatic experience in my previous post in gravitation in this same forum, where people with no idea of the field but lots of garbage posted completely wrong posts. I said that I would not waste my time with irrelevant posts. My free time is limited and cannot reply people who has not studied the fields minimally.

I repeat for you the previous paragraph now in bold:

"Most chemists react with complete incredulity to the view that structure is nothing but a metaphor, pointing out the seemingly overwhelming evidence for structure that comes from spectroscopic and other structural studies. They suggest that if a deep quantum mechanical analysis reveals molecular structure to be a mathematical artifact, then the fault must lie with present-day quantum mechanics and not with the deeply entrenched chemical notion of structure."

Note that was extracted from an elementary article in an educative journal!!!! I was not quoting the last fad in the more advanced research journal. If you have no knowledge of those basic points, I recommend you to begin your study before post irrelevant/wrong comments.

And you keep failing to cite even the NAME of this "educative journal". I have asked you for a citation on where "your theory" has been published. And you have provided none.

"present-day quantum mechanics" sound like "pure QM". It refers to current available QM. It is not talking about any specific (e.g. Hamiltonian) formulation of QM, it is saying that old (19th century) chemical theory is compatible with experimental data, still current (today) QM is not, because cannot explain molecular structure. This is solved with my approach and explain why "deep quantum mechanics" or "pure QM" or simply QM is not used in that "sucess" that you consider, and why in solid state physics or molecular chemistry one does not use pure QM or Qm is short, just a part of QM... precisely the part that work.

Excuse me? What is "deep QM" or "pure QM"? Are these something you made up?

Where exactly in solid state physics that "one does not use pure QM, just a part of QM... precisely the part that work"? Please open Ashcroft and Mermin (since we are in the mood to educate people here) and show me where we abandon a part of QM. You are making generalized statement without bothering to offer concrete proofs.

You also have not defined what you mean by "pure QM". As far as I have seen, this exists only within the confine of your imagination. Unless you are willing to be explicit in defining what it is, then you are using a lot of wiggle room to randomly define and catch whatever suits your fancy. If you think DFT, for example, is STILL "pure QM", then the fact that it CAN describe a whole zoo of quantum chemistry phenomena is enough of an evidence that it works!

If you has studied a bit of computational methods you would replied to me several post ago what is the computational procedure used in solid state physics and quantum chemistry and that part of QM is used in the computerized algorithms and that part is not used because violate experimental results.

I did a quantum monte carlo routine for Du Pont to simulate and calculate the activation energy on the surface of a catalyst used in the petroleum industry. I used PLENTY of QM in there, thankyouverymuch. And the most important thing - IT WORKS! People in such industries have no patience for "philosophy" or empty claims of so-and-so. If you don't deliver something that WILL work, you're done!

Again, let me repeat. If you think you have evidence that QM is wrong, or simply has no ability to explain a certain phenomenon, I would welcome you to get it published in Nature or Science, since such an observation would be of a paradigm-shaking caliber. Write down explicitly why the isomer example you brought up is in violation of QM, or something it simply beyond what QM can handle and publish it, and I'll show you the theory of high-Tc superconductivity! Claiming that you have a theory that does all these fancy things is a dime a dozen - one can find hundreds of those in Crank Dot Net. Either you clearly cite where your "theory" has been published, or this thread will be shoved into the la-la land of Theory Development.

Zz.

[marlon]

Eh, no, the Heisenberg relation dx.dp >= hbar/2 works with ABSOLUTE x and p, not with relative errors. So the momentum error dp for a mirror, or for an electron, is the same, given a certain dx (assuming for a moment that the HUP is valid for a mirror, which I think it is, BTW).


yes, indeed you are right, i see where i went wrong...



So if you know that your mirror is within a position dx << L/2, then the dp must be larger than a certain momentum, but that momentum corresponds still to extremely low speeds for a macroscopic object.


Why ??? dp has to exceed a certain momentum but how can you say that this momentum corresponds to very slow speeds for the macroscopic object ???Because the mass is very big ?


Only, it is relatively big as compared to the momentum that will be transmitted by the photon banging on it, which makes it impossible to find out whether the photon bounced off the mirror or not.

why ??? Sorry, but i really don't follow you on this one...Could you please clarify. I think i am misinterpreting what you wanna say, so before i start make claims based upon false assumtions i politely ask you please could you elaborate on these last two paragraphs. Perhaps a specific example (perhaps with the deBroglie wavelength) ?


However, if the HUP is NOT valid for the mirror, then I can, in one way or another, make sure that it is in a rather well known position x, and has a momentum which is very much below the transmitted momentum by the photon. After the photon has passed by, I can then wait for, say half an hour, and if the mirror did displace appreciably from its original position (assuming it could float freely during that half hour), I know the photon came by.


True story...yes


But given the large scale of the object as compared to a reasonable dx and the large scale of the object's momentum (even at very low speeds) to the associated dp of the HUP, we can say that there exist quantum states (coherent states) for the object which are, in all respects, compatible with what we usually take for granted in classical physics, with usual magnitudes, and with usual errors for position and momentum. But that's no proof that the quantum description is invalid ! So it is not because the RELATIVE errors dx/T and dp / P are very tiny that the HUP is not valid for the mirror.


This i normally should get only it is essential that i understand your first two paragraphs...which i do not


How small does a mirror have to be for the HUP to apply to it ?

cheers,
Patrick.
I was gonna bring the deBroglie wavelength but that cannot be true...ok, I DON'T KNOW... :cry:

marlon

ps : Vanesch, thanks for instructing me on this...i have never paid any attention to this stuff but it is getting more interesting by the post. For my phd i only use QM in a very pragmatic way...same goes for what i did in college...i never gave the actual formalism any deep consideration, as you very well can see :blushing:

[vanesch]

Why ??? dp has to exceed a certain momentum but how can you say that this momentum corresponds to very slow speeds for the macroscopic object ???Because the mass is very big ?


The point I wanted to make was only this: for a macroscopic object, if the HUP applies to it (meaning that it is somehow a quantum object, with a wavefunction) or not is hard to say, because the HUP works with *absolute* errors dx and dp, and if you reduce these to relative errors on the x and p of a macroscopic object this leads usually to very very tiny relative errors, giving us the impression that we very accurately know x and p. So with, or without HUP applying to the macroscopic object, we have the impression that we know accurately x and p (because our impression of accuracy is based on relative errors). So you could somehow say that then HUP doesn't "matter" for a macroscopic object. I wanted to indicate that it still makes a difference, because if the HUP doesn't apply to a macroscopic object, even though it is going to be very difficult, you could in principle know its position and momentum BETTER than the HUP ; and then you can use this knowledge to violate the HUP for microscopic objects, making quantum theory inconsistent for microscopic objects.

Don't look too much behind it: the point I tried to make was very simple.

cheers,
Patrick.

[Juan R.]

Zapperz

"And you keep failing to cite even the NAME of this "educative journal"."

If you don't read my previous post before reply, let me close this communication with you, it is obvious that your interest here is simply to do an attempt to waste my time with stupid questions or solicity to me eight times the same thing. As an example I said the journal, the number, the year...


If you consider that in the computation of molecular or solid properties one is using QM, it is not my problem.

A recomendation for you.

Write a paper claiming that a molecule or a solid can be perfectly described with QM. In fact, you could demonstrate that I am wrong :-), but the most interesting is that true chemical and physical expertises in the field will aknowledge to you for showing that they were wrong in the claim of that QM does not explain molecular structure (or solids). After you could write a paper and send, for instance, to the International Journal of Quantum Chemistry. I think that you could solve the most difficult open problem in quantum chemistry and receive the next Nobel Prize.

Congratulations!!!


If you don't know that is done in DFT: What is computed, How, or When. Your words are irrelevant for me.

I newer studied cosmology. I would newer post a topic claiming that Hoyle is wrong when he critizes the big bang. I newer would post a comment saying "Hey Hoyle you are a crackpot, you are wrong, all cosmologists agree in that big bang is correct, take this book you can see that big bang si correct, you would have in your mind a wrong conception of the topic, please could you cite some of your peer-review work? If you claim that big ban is wrong, please write a paper for Nature since that would be a sensation"...

I first would study cosmology, after study Hoyle's work and after offer my own interpretation (either correct or wrong) of if he i is milseading or no.

You would sincerely, update your knowledge. Even some philosophers of science know that i am talking and the science and thechnologies that are being developed around those new ideas (yes some engineers are already working in the field)!!

"I did a quantum monte carlo routine for Du Pont to simulate and calculate the activation energy on the surface of a catalyst used in the petroleum industry. I used PLENTY of QM in there, thankyouverymuch. And the most important thing - IT WORKS! People in such industries have no patience for "philosophy" or empty claims of so-and-so. If you don't deliver something that WILL work, you're done!"

I was not talking about philosophy, i was talking about real science and thecnology. If you compute the wavefunction for a fullerene in gas phase and your data fit perfectly with recent experimental data. Good for you. I know the answer. QM fails. QM also fails in simulation of catalysis. Sure, I have studied the topic at one rather high level. Once an expertise in molecular dynamics, computing activation energies for several molecular processes using CC, MP2-4, and CI wrote a four pages reply to one of my highly thechnical works. I show that they were not using QM in their computations of activation energies, their description of electronic structure, etc., just a mixture of quantum formulas more ad hoc formulas violating QM. He wrote his "arguments" of why QM worked perfectly and my paper was wrong. After of some correspondence and one detailed thecnical reply by my part, he wrote in his report that main criticism to my paper that he could do
"was that i listed few literature for the broad study that i had done."

His thecnical and mathematical arguments on favor of pure QM vanish...


If you cannot distinguish between QM and that you did, that part of QM you used in QMC, that part of QM you used in the computation of the energy of the solid and that parts of QM are sistematically leaved out of the models (because if one introduces in the computation one obtains the WRONG answers) i am atonished. Specially seing your "arrogance" in the posts.

Other people can distinguish. You would be not attack to me (us) by that...


That is all by my part!! Thanks ad good luck for you.

[ZapperZ]

Zapperz

"And you keep failing to cite even the NAME of this "educative journal"."

If you don't read my previous post before reply, let me close this communication with you, it is obvious that your interest here is simply to do an attempt to waste my time with stupid questions or solicity to me eight times the same thing. As an example I said the journal, the number, the year...

I went back and double check, and unless I missed something, the ONLY reference you cite is this:

"I prefer ""reply"" you with two paragraphs extracted from an elementary educative work published in the Journal of Chemical Education 77 in the year 2000."

That is like using Am. J. of Phys, or even Eur. J. of Phys. article to show that a whole branch of physics is WRONG! Yet, the rest of the quantum chemistry community publishing in their own peer-reviewed journal continues to use "pure QM"? So what's wrong with this picture?

And you STILL have not defined what you meant by "pure QM". If there is a pure QM, then there must also be an "impure QM", or else why make such a meaningless distinction. So what is an example of an "impure QM"?

As I read the rest of your postings, you seem to be confusing between the degree of complexity, and the INHERENT inability of QM to describe a system. Would you throw out classical mechanics just because it cannot produce a complete solution to the most generalized 3-body problem? And forget about 4, 5, 6, etc bodies. The study of N-body problem deals with just this, finding as many closed solutions as possible within a certain configuration, but NOT the most generalized case because it is just plain daunting to do such a thing.

Now go look back and your "quantum chemistry" case. No quantum chemists that I've talked to (and BTW, a huge number of people working in quantum chemistry are physicists!) hold an opinion anywhere even close to what you are spewing here. Forget about "molecular" chemistry. Anything beyond He atom is complicated enough for textbook QM to solve for the energy spectrum. We have to resort to approximation, simplifications, special cases, etc. Add another atom to form a molecule and it gets worse. Well, what about adding a gazillion atoms/molecules and let them interact with each other? This is what we do in condensed matter physics!

I still want to know where, exactly, in solid state physics that we dump "pure QM". I'll make your "guessing field" even wider... Where exactly in condensed matter/many-body physics, such as that defined in G.D. Mahan's book, where we abandoned "pure QM" had to resort to other things simply because QM failed?

Zz.

[vanesch]

Some references:

* phys Rev A vol 33, p 2245 + references herein for the history of this "conflict" and a recent work

* phys rev lett vol 88 p 123001

cond-mat/0211217

and the discussion about the problem raised by Juan.

(just some ad hoc picked up references...)

cheers,
Patrick.

[ZapperZ]

Some references:

* phys Rev A vol 33, p 2245 + references herein for the history of this "conflict" and a recent work

* phys rev lett vol 88 p 123001

cond-mat/0211217

and the discussion about the problem raised by Juan.

(just some ad hoc picked up references...)

cheers,
Patrick.

I'm able to comment on the PRL article since I have seen it briefly before, and when I went to look at it, I remember this clearly because this was one had a Phil Anderson's citation refering to his "localized state" (see Ref. 10 - this is THE same Phil Anderson in the Anderson-Lauglin-Pines axis of emergent phenomena). I have to read the other two to make any intelligent comment.

But from what I have understood, this has nothing to do with "pure QM" being wrong. It is exactly the illustration of what I have said regarding the degree of complexities involved when more than 2 bodies with large degree of freedom are involved. The fact that they DID start off their model using Anderson's localized state due to the dipole-dipole interactons, it would be highly dubious to claim that "pure QM" doesn't work here. And then they invoke tunneling phenomenon AND mean-field approximations (both of which I know of very well).

So I don't see how this supports what he has been trying to say.

Zz.

[vanesch]

So I don't see how this supports what he has been trying to say.


It wasn't supposed to be ! But when you read the (rather old) phys rev A article, apparently the chemical community at one time was divided into people thinking QM didn't work somehow, and others ; I have to say I was aware that there were some difficulties to obtain explicit structure from the molecular hamiltonian alone (because of naive symmetry considerations) ; I didn't think part of the chemical community would go and think that QM was not correct, and apparently at a certain point in time that seemed to be the case.

The other references indeed try to indicate that when you include some environmental interaction, localization appears, also from within QM.

So no, it takes some work, but QM can explain localization of the nuclear backbone ; at least in simple molecules. But that shows the way for larger molecules.

cheers,
Patrick.

[ZapperZ]

It wasn't supposed to be ! But when you read the (rather old) phys rev A article, apparently the chemical community at one time was divided into people thinking QM didn't work somehow, and others ; I have to say I was aware that there were some difficulties to obtain explicit structure from the molecular hamiltonian alone (because of naive symmetry considerations) ; I didn't think part of the chemical community would go and think that QM was not correct, and apparently at a certain point in time that seemed to be the case.

The other references indeed try to indicate that when you include some environmental interaction, localization appears, also from within QM.

So no, it takes some work, but QM can explain localization of the nuclear backbone ; at least in simple molecules. But that shows the way for larger molecules.

cheers,
Patrick.

Ah, OK. I should read that PRA paper when I have the chance (this week is really bad since I'm in the middle of a run and I go off for my vacation on Sat).

What you just described is not surprising to me. I can give you a parallel scenario in condensed matter. Band structure calculations, which has been extremely successful at describing everything from metals to semiconductors to insulators, failed and failed MISERABLY when we try to describe materials such as chromium, TiN, and even high-Tc superconductors (HTS). We're not talking about something even close here - it predicts that undoped compound of HTS should be a metal when in reality, it's a damn insulator (ceramic)! So it isn't even in the same ballpark!

It is only when one looks at the nature of the complexity of the material, and all the stronger-than-usual electron-electron interactions (similar to the dipole-dipole interactions) that one sees that the original premise and assumption of band theory calculations just doesn't work. But it is the model that is at fault, not "pure QM", because now we know that the description for those family of material (Mott insulators) make full use of QM and QFT.

Zz.

[Juan R.]

ZapperZ

Yes ZapperZ you are very smart. You are right. i am completely wrong. QM work correctly. It is fantastic. Please, ommit reply to me.

vanesch

Yes vanesch. There are two currents in quantum chemistry and molecular physics:

One claims that QM is totally inconsistent with chemistry and experimental data.

Other claims that QM is complete and may explain molecular structure.

The interesting of the debate is that none simpathizer of the second view is providing significant advance in the research beyond the "may explain", none consistent explanation of experimental data, and often they remain silent when you present the result of solid scientific and mathematical research against their preconceived ideas and "models".

In last 30 years the general status of quantum chemistry has passed from "QM explains molecules (or solids)". To the quotes that i posted some days ago of that actually "nobody" know how solve the obvious contradiction webteen QM and molecular structure and many claim for a faul of QM.

Woolley and others have claimed that a purely quantum mechanical description involving the raw molecular Hamiltonian without use of the Born–Oppenheimer approximation does not require the attribution of any structure to molecules.

Most chemists react with complete incredulity to the view that structure is nothing but a metaphor, pointing out the seemingly overwhelming evidence for structure that comes from spectroscopic and other structural studies. They suggest that if a deep quantum mechanical analysis reveals molecular structure to be a mathematical artifact, then the fault must lie with present-day quantum mechanics and not with the deeply entrenched chemical notion of structure.

It is also essential to recognize at the outset that from the perspective of quantum chemistry there are substantial fundamental problems, which are not well understood, centring round the imperative to introduce the classical concept of molecular structure into the formalism. These difficulties do not arise in atomic physics [or in particle physics or string M-theory I add] where the use of atomic eigenstates and the photon Fock space to provide the reference states for a perturbation theory treatment is entirely straightforward; in general, however, eigenstates of the full molecular Hamiltonian do not describe chemical species which we understand in terms of isomerism and functional groups, and the Born-Oppenheimer approximation does not solve this problem"

Note: a solid is simply a giant molecule. Therefore the same problem arises in solid state physics or chemistry.

Some people claim for solving it in the framework of new advanced theories. I am one of those, but ZapperZ (who has not read literature) ignore it. The idea of that QM fails for explaining molecules is so extended between chemical comunity that even I can cite an article published in an educative journal!! Imagine that is being done in research journals by several groups in the world. Articles that ZapperZ, simply, has not read still he before critique the ideas of others with stupid posts on stupid question like 3 bodies effects (?), DFT densities, and lot of garbage (?) with no link with I am really saying.

Anyone working seriously in molecular structure problems know the recent attempts to obtain molecular structure from ad hoc modifications of QM inspired in decoherent effects, quantum gravity spacetime foam contributions, modifications of the mathematical space where the spectral decomposition is done, complex scaling, etc.

1º) Those models are not rigorous. Nobody can convince to me (and others people that has a strict meaning of the word "rigor") that Zurek, Omnes, etc. work in decoherence stuff is mathematically rigorous. Yes, the works were published in peer-review literature. But continue to be wrong, so wrong like that famous physical review paper claiming for a violation of thermodynamics.