Is QM Inherently Non-local in EPR and Bell Discussions?

[vanesch]

You know what I hate the most about this kind of arguments, is that you always leave something unexplained (something weird, magical has to be there). The next step you have to take is to explain conciousness by a physical theory which uses consciousness as a fixed, postulated, concept.

I think that all this just indicates that we have not yet a full understanding of physics, which is - I would think - a totally trivial statement.
However, it is not by tossing out all we know that you get a better understanding of course.

So, actually, you are not solving anything, you are just pushing a perverse scheme a step further. I would like to know from you where your consciousness was in the beginning of the universe, since clearly something must have reduced the state there (the universe is entirely classical...).

Aren't you being a bit axiomatic here ? What says that the universe is entirely classical ? If that were true it wouldn't be necessary to use quantum theory of course. You can just as well state that the universe is entirely Newtonian or Aristotelian. It is not by stating this that things have to be this way. I'm just presenting a view of quantum theory, incomplete as it may be, that shows you that the problems that make you toss everything out of the window can be seen in a different light.
Of course Newtonian physics and Coulomb electrostatics are much nicer and better understood. They give less rise to interpretational problems... but then they don't correspond to observations in certain circumstances.

So we have a formalism that works (= makes correct predictions FAPP). You've presented us with a riddle as a gedanken experiment which uses non-existing interactions to provide for "extended measurements", and when you look at it through MWI glasses, you see simply more clearly that your "measurement interaction" cannot be compatible with known, local, unitary laws.

Moreover, your consciousness does not solve many problems : I do not see for example how you would get out the second law of thermodynamics (this is much nastier at the quantum level than the classical one).

Because I will always experience a branch with a (relatively) high Hilbert norm, and in those branches, that law is respected, no ?

If you like Penrose in that respect, then you must realize that the scheme he has for quantum gravity is not covariant ...
I also think that gravity is playing an important part in quantum mechanics, but then CLASSICAL gravity not some undefined dream as QUANTUM gravity.

The combination of gravity and quantum theory is still an open question, and it is silly to claim a priori what view will prevail. I can just as well claim that neither general covariance, neither the superposition principle will survive and that we will be in for something totally new. But all that is speculation, and one speculation is as good as the next. That is still no reason to toss out our actual knowledge and CERTAINLY no good reason to go back 90 years.

 

[vanesch]

(measurement does not happen at a spacetime point, it does need a spacetime region, an apparatus registers only when there is peak which goes over a certain threshold).

Measurement does not happen AT ALL.

If you want to interpret your measurement results, you have to trace out the degrees of freedom of the field under observation, the resulting dynamics of the traced out density matrix is not unitary although the total dynamics (field + observer) is. Similarly the dynamics of the field under observation is not unitary when you trace out the degrees of freedom of the detector.

You don't have to consider this approximated dynamics as anything real, do you. It is just a shortcut in a calculation.

Moreover, there is no theorem which says that unitarity implies causality and vice versa (otherwise the wightman axioms would be abundant).

Of course unitarity is not sufficient. You also need local dynamics ! Interactions that only act locally. That's what goes wrong with your extended measurement: you cannot build that using known interactions.

At least we know that electroweak and strong interactions satisfy these criteria...

 

[Careful]

That is still no reason to toss out our actual knowledge and CERTAINLY no good reason to go back 90 years.

I do not go back 90 years in time; there have been many sensible people in the last 90 years which have, although in unfortunate circomstances, conducted good research outside the mainstream. The shortage in your approach is not just a lack of respect for common sense but the absence of a good axiomatic system of physical principles (so all this is just patchwork). The sad thing about this whole story is that most people do not even research classical theories to the bone. There is an overwhelming number of staments done by quantum theorists concerning the presumed fact that only quantum mechanics explains some cherished experiments, most of these are plainly wrong! What I try to tell to people is that CLASSICAL gravity has surprising implications on the microscale most physicists are not even aware of, which come very close to quantum phenomena. It seems therefore logical that people explore this beautiful/rational theory to the end. As I said QM is for the moment a good effective scheme but certainly not a physical theory and filled with contradictions (I would like to see how your non-local consciousness state solves the measurement/superluminal signalling problem). Classical chaotic phenomena are not understood and progress in physics is not going to be made without a good axiomatic system.

 

[Careful]

Measurement does not happen AT ALL.
You don't have to consider this approximated dynamics as anything real, do you. It is just a shortcut in a calculation.
Of course unitarity is not sufficient. You also need local dynamics ! Interactions that only act locally. That's what goes wrong with your extended measurement: you cannot build that using known interactions.
At least we know that electroweak and strong interactions satisfy these criteria...

Look, you fall over words now. The question is how are you extracting a classical number which you note down on your sheet of paper from a quantal field. You need some non local averaging procedure for that; I call that measurement (how you implement it is your own business, but you should do it in a realistic way). This has nothing to do with strong or weak interactions, this has to do with when we put a cross and when not and that is clearly a real process. In my opinion it makes not even sense to take any quantum theory and speak about one measurement in a temporal sense (since quantum theory is about predicting results of a series of measurements). So far we have been speaking about reduction of density matrices in QFT, but this is not what we should actually do since these computations are really about non temporal experiments. But anyway, most people do not seem to bother about these *small details*.

I think this approximated dynamics is very very real since that is the only way you can put crosses. Moreover, it was not clear at all in the beginning you did not want to implement state reduction which we all know not to be local. The fact that I used a non local observable or not did not matter in that respect. So now, you still have to invent your scheme in which I can measure non local (and even local!) obervables without violating causality. Good luck!

Moreover, if you claim that general covariance does not survive, then (a) you have a hard job in explaining why GR is sooo good (as successful as you dear QM) and (b) why don't we go back to Newtonian days all together (go back 350 years back in time).

Moreover, most quantum gravitists expect QG only apply at the big bang and deep into black holes, the rest is entirely classical (apart from the cosmological constant perhaps). About the second law of thermo: my knowledge is that it is most of the time respected although not always (Poincare recurrency times seem to have more severe consequences in QM than in classical thermo).

 

[vanesch]

Look, you fall over words now. The question is how are you extracting a classical number which you note down on your sheet of paper from a quantal field. You need some non local averaging procedure for that; I call that measurement (how you implement it is your own business, but you should do it in a realistic way).

What I am trying to say is that this cross is put there locally (locally, on the time scale of writing down a cross, being several milliseconds, so what we call "local" here is a spacetime blob which extends over several milliseconds/lightmilliseconds in all directions) and that all "result of measurement" that resulted in me writing down that cross or not, if it came from a region a lightyear across (spacelike) needs to be totally in my past lightcone. So the averaging you are talking about (over the region a lightyear across) can be 'non-local' but will then take about half a year before it reaches me and I can decide to "put down a cross" or not.
This cross is there then "really" only for me ; I don't know how others experience this. So it isn't sure that "a real classical number has been extracted" in any way but my own conscious observation. In my conscious awareness of the world, this looks then now as a "real classical number", but it could just as well be that the paper is in a superposition, one state with, and another state without a cross on it, and I'm only consciously aware of the paper with a cross, while someone else may only be aware of the same paper without a cross.

This has nothing to do with strong or weak interactions, this has to do with when we put a cross and when not and that is clearly a real process.

As I try to point out, it isn't so clear that this is a "real process". It gives me maybe only the awareness of some reality, but that's just my experience, and maybe not someone else's. (I'm sure that this makes you jump up and down your chair ) But that's exactly MWI...

In my opinion it makes not even sense to take any quantum theory and speak about one measurement in a temporal sense (since quantum theory is about predicting results of a series of measurements).

I don't see why you say that. This is only the epistemological view of QM: a technique for calculating statistical predictions. But then it becomes very hard to implement physical principles into the theory. Even "locality" doesn't mean anything, because of course the numbers printed on a sheet are "local to the sheet".

Moreover, it was not clear at all in the beginning you did not want to implement state reduction which we all know not to be local. The fact that I used a non local observable or not did not matter in that respect. So now, you still have to invent your scheme in which I can measure non local (and even local!) obervables without violating causality. Good luck!

I think there is no difficulty there (but I told you that it will involve too much calculation on a too involved example to do it explicitly here). I take observers to be "pointlike" (at least on the scale of their conscious experience, which must be of the order of milliseconds or so), and all interaction, including "measurements" to be described by the standard unitary evolution dictated by the electroweak and strong interaction lagrangians or whatever, knowing that these have Green's functions which vanish outside of the lightcone. As such it should be clear that the state of entanglement of the local observer cannot be influenced by what happens outside of its past lightcone (as no unitary interaction will be able to propagate to it, using the Green's functions), and the state of entanglement of the local observer is exactly what describes the local observer's experience.

Moreover, if you claim that general covariance does not survive, then (a) you have a hard job in explaining why GR is sooo good (as successful as you dear QM) and (b) why don't we go back to Newtonian days all together (go back 350 years back in time).

I didn't say that general covariance will not survive ! I say it is an OPEN QUESTION.

Moreover, most quantum gravitists expect QG only apply at the big bang and deep into black holes, the rest is entirely classical (apart from the cosmological constant perhaps).

Black holes are BIG THINGS compared to people ! When you see that Hawking considers superpositions of spacetimes over billions of years (the time for gas to contract into a star, then a black hole, and then have the black hole evaporate in interference with the gas finally not contracting into a star) then having Alice in two states for a couple of years doesn't seem so extravagant !

About the second law of thermo: my knowledge is that it is most of the time respected although not always (Poincare recurrency times seem to have more severe consequences in QM than in classical thermo).

I am not so very fluent in these sophisticated applications of statistical mechanics.

 

[Careful]

QUOTE]
Just a few small notes for now :
(a) Black holes are not necessarily big things at all by any standard, they can be as small as the want to and there are VERY good reasons to think of elementary particles as black holes or similar gravitational configurations.
(b) My claim is that you HAVE to break covariance if you do stick to QM as it stands (if you are interested we can have a deeper chat about that)
(c) I am not going to answer on your consciousness crackpot stuff for now, just saw the lord of the rings and it indeed gives me the creeps but I am too tired to jump out of my chair :zzz:
(d) I am curious how you will put your cross at the orgin of the cosmic microwave background though.

I am off for the weekend so you can plunge yourself in your highly personal though universally connected consciousness state.

 

[Hurkyl]

So you seem to claim that performing the measurement at a, or not, when the B measurement is performed, changes the outcomes of C ?
Let us take an initial state |psi> which is u|a+> + v|a->, |a+> and |a-> being the two eigenstates of A (and also of C, since they commute).

As far as I know, there should be four eigenstates:

|a+, c+>
|a+, c->
|a-, c+>
|a-, c->

(Actually, there should be another parameter denoting the stuff that A and C don't care about)

 

[Hans de Vries]

(a) Black holes are not necessarily big things at all by any standard, they can be as small as the want to and there are VERY good reasons to think of elementary particles as black holes or similar gravitational configurations.

It's well known that the Schwartzschild radius only becomes equal to the
Compton Radius at Planck's scale. It's the very definition of Planck's scale!
This is 10¹⁹ times the energy scale of our common elementary particles.

That's how remote this proposal is from the generally accepted laws of physics...Regards, Hans

 

[vanesch]

As far as I know, there should be four eigenstates:
|a+, c+>
|a+, c->
|a-, c+>
|a-, c->
(Actually, there should be another parameter denoting the stuff that A and C don't care about)

Yes, you're right, I only treated a specific example where A and C were equal, and with a 2-state system (in a 2-dim space, I even think that you don't have a choice but to take their eigenspaces equal). I wasn't doing things in all generality. I probably should, and even do so with the BL and B

 

[vanesch]

Just a few small notes for now :
(a) Black holes are not necessarily big things at all by any standard, they can be as small as the want to and there are VERY good reasons to think of elementary particles as black holes or similar gravitational configurations.
(b) My claim is that you HAVE to break covariance if you do stick to QM as it stands (if you are interested we can have a deeper chat about that)
(c) I am not going to answer on your consciousness crackpot stuff for now, just saw the lord of the rings and it indeed gives me the creeps but I am too tired to jump out of my chair :zzz:
(d) I am curious how you will put your cross at the orgin of the cosmic microwave background though.
I am off for the weekend so you can plunge yourself in your highly personal though universally connected consciousness state.

Concerning your claim (b), I think you are right but I'm not expert enough in quantum gravity to understand this issue entirely. I think you're crashing into a wide open door if you claim that there is a conflict between QM and GR, and that it is an OPEN QUESTION on how to solve this. Your wishful dreams of particles being black holes are just as speculative and unfounded as any other random claim - who was the one requesting a theorem here ?
Concerning the "consciousness crackpot stuff", my point of view is this one: it is - in my opinion - the view that fits best with the formalism of QM as we have it today. I'm surely not entirely happy with it myself, but I stick to it as long as I have to stick to QM. Do I really think that this is how the world works ? My answer is simply that I don't know, and I think that anybody who claims he knows is deluding himself. As we don't have a final physical theory yet, we cannot say (and in fact, I don't know if we'll have such a theory one day - how can we know ?) My personal preference (but what does that mean?) would be that I somehow hope that this is NOT how the world really works. The only thing that this view allows me is to have a clearer understanding on how to apply the QM formalism - that's why I found your example very instructive.
On the other hand, even though I agree that it sounds crazy, it is not *that* crazy, you only have to get used to it. The idea that "observation" is something linked to a conscious experience is not so much far fetched after all ! In your beloved GR, one could call it just as well a crackpot idea that *time* is something observer-related. That you could go traveling and come back younger than your kids. Or that sitting on a high building just does the same (ok, the effect is tiny :-) That's also an idea to get used to.
At the end of the day, it doesn't matter what story we tell around a theory. What matters is the formalism and the general principles from which it is derived, and how well this formalism can explain experimental results. There, QM is for the moment unbeatable. And the day when there will be deviations it will be extremely interesting and instructive, but that isn't the case yet. So you're simply stuck with it for the moment.

 

[Careful]

It's well known that the Schwartzschild radius only becomes equal to the
Compton Radius at Planck's scale. It's the very definition of Planck's scale!
This is 10¹⁹ times the energy scale of our common elementary particles.

That's how remote this proposal is from the generally accepted laws of physics...


Regards, Hans

But that is not the issue: I did nowhere claim that the the Schwartzschild radius has to be of the order of the comption scale (of say an electron). What I alluded to however is, say, that indirectly (through the Einstein Maxwell equations) there are important gravitationally induced electromagnetic phenomena on the compton scale (of say the electron), see Rosquist 2004 (on gr-qc I believe and references therein). This does not involve a Kerr black hole solution (with an event horizon) but a spin dominated Kerr solution which gives a naked singularity (the electron is by far spin dominated if you believe in electron spin at least). You do NOT have to go to Planck scale energies to get interesting phenomena out (that is a common misunderstanding). This is a possible mechanism which might allow you to put forward a realistic continuum electron model and explain its structural stability, something particle physicists cannot even dream of.

cheers,

Careful

 

[vanesch]

But that is not the issue: I did nowhere claim that the the Schwartzschild radius has to be of the order of the comption scale (of say an electron). What I alluded to however is, say, that indirectly (through the Einstein Maxwell equations) there are important gravitationally induced electromagnetic phenomena on the compton scale (of say the electron), see Rosquist 2004 (on gr-qc I believe and references therein).

Hey, this post is an unexpected proof of MW ! You, Careful, went on a weekend, in your conscious experience:

I am off for the weekend so you can plunge yourself in your highly personal though universally connected consciousness state.

Nevertheless, in MY conscious experience, you came back and you posted again on PF

 

[vanesch]

see Rosquist 2004 (on gr-qc I believe and references therein).

I suppose you mean the paper:
gr-qc/0412064

It surely is thought-provoking !

 

[Careful]

What matters is the formalism and the general principles from which it is derived, and how well this formalism can explain experimental results. There, QM is for the moment unbeatable. And the day when there will be deviations it will be extremely interesting and instructive, but that isn't the case yet. So you're simply stuck with it for the moment.

On the other hand, even though I agree that it sounds crazy, it is not *that* crazy, you only have to get used to it. The idea that "observation" is something linked to a conscious experience is not so much far fetched after all ! In your beloved GR, one could call it just as well a crackpot idea that *time* is something observer-related. That you could go traveling and come back younger than your kids. .

I knew you wanted to go back 350 years back in time time is not observer related, time is the same for all inertial observers in minkowski (that is a common misunderstanding of the twin paradox). Time changes however when you accelerate and deviate from the geodesic path between spacetime points A and B, that is what you need to do in order get back home and such effects have been measured already.

Moreover, I wanted still to make a few comments on your ``measurement´´ procedure:
(a) I do not understand why you want to keep observables at all since they were introduced in the first place to make observation and you propose something which not related at all to this.
(b) in your ``reasoning´´ concerning the spacetime consciousness blob (did I understand that well ? ) you make a common mistake of introducing a global lorentz frame (since you speak about big spacelike distances), in which your consciousness must operate (so your observers are global at all and not local which is what I meant with your universal consciousness).
(c) It remains crystal clear that any copenhagen scheme is still in trouble when they use non local observables (in the standard way, involving QCD and all that)
(d) my claim for strong gravitationally induced gravitational effects at the compton scale is far from empty (see my reply to de Vries, also Carter and Wheeler have made similar observations at the end of the sixties even. ) and is very well supported indeed (for further references : see Cooperstock et al.)
I am one of the very few people around who don't go to the Planck scale at all in order to find interesting gravitational effects and to explain ``quantal´´ phenomena.

The rest is too crazy to answer, I am not stuck with QM at all, as I said many of its predictions have classical answers. There are a few challenges left true, but I one would succeed in solving these then one has ``quantum gravity´´ for free.

Cheers,

Careful

 

[Careful]

Hey, this post is an unexpected proof of MW ! You, Careful, went on a weekend, in your conscious experience:
Nevertheless, in MY conscious experience, you came back and you posted again on PF

I know, I cheated a bit
Now I have to go, otherwise my wife kills me... :!)

Cheers,

careful

 

[vanesch]

I knew you wanted to go back 350 years back in time time is not observer related, time is the same for all inertial observers in minkowski (that is a common misunderstanding of the twin paradox). Time changes however when you accelerate and deviate from the geodesic path between spacetime points A and B, that is what you need to do in order get back home and such effects have been measured already.

Just to avoid all misconceptions: I *know* that time is observer-dependent! But when you first hear it, being brought up in a Newtonian picture, you could have as a first reaction that this is a "crackpot idea". I wanted to draw the parallel that if you grew up with a classical relativistic picture, the idea that people could be in two places at once, though only observing one of them, can sound like a "crackpot idea" too. Nevertheless, that is the *fundamental idea* behind quantum theory: the superposition principle: if you can be here, and you can be there, then you can also be in both places at once.
The only small problem we have with this otherwise beautiful idea is that, well, we don't observe that (ahum... ). We do seem to observe the indirect consequences of it, however. So you need then to say that you will only be consciously aware of one of the states.

(a) I do not understand why you want to keep observables at all since they were introduced in the first place to make observation and you propose something which not related at all to this.

If by observables, you mean those famous hermitean operators with eigenstates in which you're supposed to flip ? They are only a useful mathematical summary of the very complicated unitary interaction - in fact environmental decoherence theory is the justification for that approach. A hermitean operator is nothing else but a "bag of orthogonal eigenspaces".
The "bag of orthogonal eigenspaces" is then nothing else but the family of disjoint, environmentally stable final states of the specific unitary evolution operator that describes the action of the measurement apparatus, calculated back to the states of the system to which it is to be applied (eg, the states of the system which will remain entangled with those environmentally stable final states of the overall system+measurement+environment).

(b) in your ``reasoning´´ concerning the spacetime consciousness blob (did I understand that well ? ) you make a common mistake of introducing a global lorentz frame (since you speak about big spacelike distances), in which your consciousness must operate (so your observers are global at all and not local which is what I meant with your universal consciousness).

? I do make the assumption that the piece of spacetime around the "conscious event" is about Minkowskian, but if it weren't, my body wouldn't be there ! I only wanted to say that at the end of the day, when you consciously look at your results, this takes a certain time, and occupies a certain space of course (the time of becoming consciously aware and the size of your brain, for instance) and that we should consider this blob as being "one event" and not going to subdivide this in smaller pieces of spacetime (the left side of my brain, or the right side, the beginning of my becoming aware, or the end...). This is just as in relativity books, where you consider the "explosion of a fire cracker" to be an event ; you shouldn't then nitpick over the length of the firecracker or the duration of the explosion.
I only wanted to point out that this "event" (this blob, if you want) must have the "experienced measurement interactions" entirely in its past lightcone. So the "measurement" is only complete at that point (and even a bit later). If you insist on using projection, you should only apply it when the measurement is complete, meaning, at that time (using of course a foliation of spacetime - the very reason I don't want to consider this projection because I don't want such a foliation).

(c) It remains crystal clear that any copenhagen scheme is still in trouble when they use non local observables (in the standard way, involving QCD and all that)

I think it is an abuse. You can DEFINE non-local hermitean operators, after all, why not. But it is an abuse to call it a measurement because you will not be able to construct a measurement apparatus which will involve a unitary evolution (due to the electrons and protons of its constituent parts) and couple in such a way to the environment that it will RESULT in an environmentally stable set of states and that can be traced back to the non-local hermitean operator, BEFORE a spacetime event is reached where the entire non-local region is in its past lightcone.

(d) my claim for strong gravitationally induced gravitational effects at the compton scale is far from empty (see my reply to de Vries, also Carter and Wheeler have made similar observations at the end of the sixties even. ) and is very well supported indeed (for further references : see Cooperstock et al.)
I am one of the very few people around who don't go to the Planck scale at all in order to find interesting gravitational effects and to explain ``quantal´´ phenomena.

If there's no error in the paper you cited (I'm not enough of a relativist to check, although I can follow the arguments), it is indeed surprising that GR phenomena appear already on the scale of 10^(-13) cm.
But there's a lot of work to do before you can claim "equivalence" with quantum phenomena.
What is a great success of QFT (dispite its lot of deseases) is, I'd think, the prediction of particles from fields: the very reason why the electon field comes in "lumps of equal electrons".

The rest is too crazy to answer, I am not stuck with QM at all, as I said many of its predictions have classical answers. There are a few challenges left true, but I one would succeed in solving these then one has ``quantum gravity´´ for free.

Of course. But as you say, there are still a few challenges left.

 

[Careful]

you shouldn't then nitpick over the length of the firecracker or the duration of the explosion.
I only wanted to point out that this "event" (this blob, if you want) must have the "experienced measurement interactions" entirely in its past lightcone. .
If there's no error in the paper you cited (I'm not enough of a relativist to check, although I can follow the arguments), it is indeed surprising that GR phenomena appear already on the scale of 10^(-13) cm.
But there's a lot of work to do before you can claim "equivalence" with quantum phenomena.
What is a great success of QFT (dispite its lot of deseases) is, I'd think, the prediction of particles from fields: the very reason why the electon field comes in "lumps of equal electrons".
Of course. But as you say, there are still a few challenges left.

Your blush makes you more attractive Ok, I am knitpicking over the length of the firecracker as you say it and I agree with you that the non local observables are an abuse (already for quite some time but I had too much fun with your consciousness). Now, is the firecracker important or not (is this tiny violation significant)? For all practical purposes (FAPP) of course not, however my evil mind could of course cook up an idealized thought experiment with many fire crackers placed in sequence as to violate causality with some big amount (although that is not possible in practice, but on a sufficiently long timescale it would be). The reason why I was knitpicking here is because it is IMPORTANT to know the details and we have now been talking for quite some time here (with many more details involved) about something which most people take to be as obvious.

My overall feeling in this, is that this FAPP attitude is not going to lead us anywhere (you accused me before of being too axiomatic when I said that your consciousness really does not solve anything and now you had to invoque another argument of knitpicking). Theories have to start from *exact* principles, and investigate the full consequences. The only theory really worhwhile doing this for is GR because of its immense axiomatic beauty and extreme accuracy. The paper of Rosquist is correct. Concerning your argument about what QFT is really good for, I cannot unfortunately disagree more. The particle concept of QFT is worthless. Its big succes however is the accuracy of S matrix predictions, a miracle indeed if you realize what mathematical ``magic´´ is needed for achieving this.

Have a nice weekend (true this time)

 

[vanesch]

My overall feeling in this, is that this FAPP attitude is not going to lead us anywhere (you accused me before of being too axiomatic when I said that your consciousness really does not solve anything and now you had to invoque another argument of knitpicking).

No, you miss the point - it is a pity you do not want, just for the sake of argument, place yourself in the MWI viewpoint. You seem to have the impression that it is all handwaving, but it is not that at all. There are of course very fuzzy concepts such as consciousness, but human perception IS a fuzzy thing in the end! In MWI, there is a very strict law of evolution, it is unitary evolution, period. As such, your body ends up in an entangled state with many other states, and the only extra thing that is claimed, is that you are consciously aware of ONE of these states, according to the Born rule.
The precise interaction doesn't matter - this is not a matter of FAPP ; it is a matter of what the experimenter, at the end of the day, when he studies the output of his computerized experiment, observes. If, at that point, you can say that the experimenter will experience his body state according to the Born rule, then, thanks to decoherence theory, this Born rule trickles down to the exact system his experimental setup has been studying. There's no "FAPP" here, this is entirely exact. The only vague point is what happens exactly when he reads his printout, but if it can be accepted that this results in applying the Born rule to this (exact interaction, using unitarity) superposition of body states in one way or another, we're home.
You can even trace this back 15 billion years if you want (except for the fact that we don't have quantum gravity), and have the experimenter choose from all possible states that occurred since the Big Bang - and nevertheless, he'll pick out a state (corresponding to the Born rule) which makes you apply the Born rule all the way back.
So the Earth did, and didn't form ; the sun did and didn't form... but we happen to have choosen a branch where it did form.

Theories have to start from *exact* principles, and investigate the full consequences. The only theory really worhwhile doing this for is GR because of its immense axiomatic beauty and extreme accuracy.

Newtonian mechanics is also in this case: a nice theoretical framework of immense axiomatic beauty and extreme accuracy. The dirty physics however, would also like to have predictive power that fits with experiment.

Concerning your argument about what QFT is really good for, I cannot unfortunately disagree more. The particle concept of QFT is worthless.

In free field theory, I'd say that the derivation of the particle concept is rather straight forward ! The problem comes in with the interactions.

Its big succes however is the accuracy of S matrix predictions, a miracle indeed if you realize what mathematical ``magic´´ is needed for achieving this.

I will not deny the mathematical difficulties of QFT. But there are A LOT of successes (like the recent hadron mass predictions using lattice QCD), which should make you wonder how it could be that a totally misguided idea leads to so much correct results.

 

[Careful]

The problem comes in with the interactions.
I will not deny the mathematical difficulties of QFT. But there are A LOT of successes (like the recent hadron mass predictions using lattice QCD), which should make you wonder how it could be that a totally misguided idea leads to so much correct results.

I will be short about this
(a) you have the comments of Penrose which are entirely justified (concerning the clever mixing)
(b) you have a preferred basis problem (which coarse graining do you apply?)
(c) you cannot explain the perception state (I already made that comment)
and so on and so on. It is true that in a *well defined QFT* involving *local* projection operators on a universal perception state, you can save locality but then try to give me an example of a local perception projection operator whose outcome corresponds to the measurement of a non local observable such as in our region B. The only way, in my opinion to solve schroedingers cat, is to make quantum theory non linear (another reason why I am keen on the self field approach) just as all realistic processes in nature are. About QFT ,who cares about the free field ?? (and even then you have to be careful). Look, as mentioned before, I think that QFT despite its contemporary uglyness, is a worthwhile theory, a bit in the sense that thermodynamics is. It does not give a realistic explanation of the processes in the microword, but it gives very good statistical predictions concerning outcomes of scattering experiments (as well as masses of gauge particles) just as thermodynamics can serve extremely well for engineering. I have nothing against QFT in that respect, I just think it cannot serve as a basis ingredient for a theory of quantum gravity.

There are other approaches in QM which do in my opinion a much better job in trying to solve the cat problem: this is penroses OR theory, and a theory which is called the Brussels Vienna interpretation of QM (very abstract for now still) but these two suffer still from incompatibility with special relativity as far as I know. A fully classical alternative (such as I try) is another logical way out.

By the way, do not dare to compare Newtonian mechanics to GR.

 

[vanesch]

(b) you have a preferred basis problem (which coarse graining do you apply?)

This could be solved "by postulate". Within the MWIers I'm kind of heretic in that I'm convinced that you cannot *derive* the Born rule logically from unitary quantum theory, so that you need to ADD extra postulates which describe conscious observation ; one of them being the Born rule, and why not, another postulating the basis in which it occurs. This comes in fact very close to von Neumann, except that the act of observation is not something that has physical consequences (as does the projection postulate), but only affects conscious observation. 'True' MWI-ers spend a lot of effort trying to derive that somehow from unitary QM, but I think that that is fundamentally impossible.

(c) you cannot explain the perception state (I already made that comment)
and so on and so on.

That is just added by postulate: you have the physical world, entirely gouverned by unitary QM, and you have the "mental world" which couples to the physical world through some extra postulates, which "sample" the wavefunction in a certain way so that this corresponds to our habitual perception.

It is true that in a *well defined QFT* involving *local* projection operators on a universal perception state, you can save locality but then try to give me an example of a local perception projection operator whose outcome corresponds to the measurement of a non local observable such as in our region B.

Well, build me an apparatus that does the measurement, and I'll give you the local perception operator! It will be nothing else but the unitary evolution operator associated with the physics of the apparatus.

The only way, in my opinion to solve schroedingers cat, is to make quantum theory non linear (another reason why I am keen on the self field approach) just as all realistic processes in nature are.

It is A way, but not the ONLY way. I would think that if somehow gravity could introduce, in a "natural" way, this non-linearity, that would be a good idea. Just adding non-linearity for the sake of obtaining a projection is a fudge factor. And we have to accept the possibility that the superposition principle IS fundamental and strict. If that's the case, I don't see any way out except the MWI style. And, you'll have to admit it, for the moment there's no experimental indication of a deviation from the superposition principle (except maybe for the trivial fact that we don't directly experience it )

Look, as mentioned before, I think that QFT despite its contemporary uglyness, is a worthwhile theory, a bit in the sense that thermodynamics is. It does not give a realistic explanation of the processes in the microword, but it gives very good statistical predictions concerning outcomes of scattering experiments (as well as masses of gauge particles) just as thermodynamics can serve extremely well for engineering. I have nothing against QFT in that respect, I just think it cannot serve as a basis ingredient for a theory of quantum gravity.
There are other approaches in QM which do in my opinion a much better job in trying to solve the cat problem: this is penroses OR theory, and a theory which is called the Brussels Vienna interpretation of QM (very abstract for now still) but these two suffer still from incompatibility with special relativity as far as I know. A fully classical alternative (such as I try) is another logical way out.

If by your comments, you want to state that there are still unsolved problems in physics, and especially with QFT, and that QM's interpretation is far from clear, and that it is very well possible that it will need modifications in the future, I couldn't agree more. However, as you also point out, many of the suggested ways out are in their infancy on the formal level, in that they introduce a fundamentally different formalism for which none has yet, in its whole, shown the same experimental accuracy and efficiency as quantum theory as it stands today. I think that the danger of following these paths is that one is blinded by the consistency of the formalism one is develloping, and forgets that it needs to work also in the lab - a bit the result of some nostalgia to times when physics was clearer. Of course, one cannot work on 10 approaches at the same time, and in the end it is just gut-feeling which guides one to choose an approach - as this is a very personal matter, it is hard to discuss.
My viewpoint is that any approach that will have a fundamental impossibility to explain entanglement in the way QM does, is a very dangerous bet, because too many close hits have confirmed QM when it uses entanglement, in very different circumstances.

By the way, do not dare to compare Newtonian mechanics to GR.

I don't see why. I would even say that Newtonian mechanics of fundamental elastic spheres, with Newtonian gravity and Coulomb electrostatics, is a very very clear, axiomatically well-defined and interpretationally totally clear theory. It doesn't fit certain experiments, but so what ? It is axiomatically even better defined than GR, so if you want to hide into a clear, axiomatic world view, why not go for that one ?
In the words of Weinberg: "I don't see what's wrong with a Newtonian universe with fundamental spheres. It's simply not ours."

 

[Careful]

This could be solved "by postulate". Within the MWIers I'm kind of heretic in that I'm convinced that you cannot *derive* the Born rule logically from unitary quantum theory, so that you need to ADD extra postulates which describe conscious observation ; one of them being the Born rule, and why not, another postulating the basis in which it occurs. This comes in fact very close to von Neumann, except that the act of observation is not something that has physical consequences (as does the projection postulate), but only affects conscious observation. 'True' MWI-ers spend a lot of effort trying to derive that somehow from unitary QM, but I think that that is fundamentally impossible.
."

I agree that it is impossible to do this (I don't see a real difference here with the environmental decoherence game). However, I by far do not agree with you that you could just take this as an extra postulate. Then, you are not really explaining anything - on the contrary - you are simply putting the possible outcomes of your pre-dediced experiments in by hand. Moreover, as I pointed out to you several times, such attitude is worthless if you speak about the entire universe. There, reduction happens without any experimentator being around for deciding which setup one shoud do, moreover in a sum over histories framework you would even have tremendous difficulties in defining the experiment itself when doing quantum gravity. As said before, you are really taking a perverse game one step further.

.
Well, build me an apparatus that does the measurement, and I'll give you the local perception operator! It will be nothing else but the unitary evolution operator associated with the physics of the apparatus.
It is A way, but not the ONLY way. I would think that if somehow gravity could introduce, in a "natural" way, this non-linearity, that would be a good idea. Just adding non-linearity for the sake of obtaining a projection is a fudge factor. And we have to accept the possibility that the superposition principle IS fundamental and strict. If that's the case, I don't see any way out except the MWI style. And, you'll have to admit it, for the moment there's no experimental indication of a deviation from the superposition principle (except maybe for the trivial fact that we don't directly experience it )

I do not have to build you this apparatus, you do! I pointed out to you that non local observables, which are widely used in QFT pose a potential measurement problem. You answered, that if I would take into account a perception field and state, and limit myself to local projection operators on these ``mental states´´ (you can even choose here to do the reduction or not) then I can still save locality. I agreed modulo the technical worries of a well defined QFT and physical worries as to the reality and dynamics behind the mental field and its coupling to physics fields. Now, I asked you to make this scheme concrete: try to link ``observables´´ of the real field to the singular projection operators on mental states, singular because they live on a set of measure zero. This is not my problem, and even if you could solve it, which is possible in principe, then still it is not a meaninful scheme in my opinion unless you go to a OR type of scheme. But then you would have to join you fellow MWI compadres.

Of course, this non linearity is to come from gravity. Moreover, there is no reason for keeping the superpostion principle at all (as you said, we do not see it and I thought quantum physicists are all strong positivists, so logically abandonning this should be of no worry). As said before, it is this very principle which is responsible for the measurement problem.

If by your comments, you want to state that there are still unsolved problems in physics, and especially with QFT, and that QM's interpretation is far from clear, and that it is very well possible that it will need modifications in the future, I couldn't agree more. However, as you also point out, many of the suggested ways out are in their infancy on the formal level, in that they introduce a fundamentally different formalism for which none has yet, in its whole, shown the same experimental accuracy and efficiency as quantum theory as it stands today. I think that the danger of following these paths is that one is blinded by the consistency of the formalism one is develloping, and forgets that it needs to work also in the lab - a bit the result of some nostalgia to times when physics was clearer. Of course, one cannot work on 10 approaches at the same time, and in the end it is just gut-feeling which guides one to choose an approach - as this is a very personal matter, it is hard to discuss.

Indeed it is hard to discuss, let's just respect each others approaches and argue on points of consistency.

My viewpoint is that any approach that will have a fundamental impossibility to explain entanglement in the way QM does, is a very dangerous bet, because too many close hits have confirmed QM when it uses entanglement, in very different circumstances.
I don't see why. I would even say that Newtonian mechanics of fundamental elastic spheres, with Newtonian gravity and Coulomb electrostatics, is a very very clear, axiomatically well-defined and interpretationally totally clear theory. It doesn't fit certain experiments, but so what ? It is axiomatically even better defined than GR, so if you want to hide into a clear, axiomatic world view, why not go for that one ?
In the words of Weinberg: "I don't see what's wrong with a Newtonian universe with fundamental spheres. It's simply not ours

[/QUOTE]

Concerning your entanglement impossiblity, I would say that this bet is as dangerous for you as it is for me. There is no distinction as yet and if the experiments would keep on turning out inconclusive (or in favor of realism) then from the esthetical point of view, the local realist attitude is certainly preferred. QM can only vindicate when the experiments are fully successful and even then there are dirty, realist ways out. Concerning your Newtonian comments, you really don't seem to appreciate the full beauty of GR: locality and space time coordinate invariance are the most powerful and simplifying principles in nature.

 

[vanesch]

Then, you are not really explaining anything - on the contrary - you are simply putting the possible outcomes of your pre-dediced experiments in by hand.

That's the unfortunate fate of anything that is introduced by postulate...

Moreover, as I pointed out to you several times, such attitude is worthless if you speak about the entire universe. There, reduction happens without any experimentator being around for deciding which setup one shoud do

Or it doesn't happen ! Why does reduction have to happen ? If there's no observer, there's no need to have the wavefunction reduce to one of the "observable" states, as there aren't any. This is one of the big advantages of considering the unitary evolution "all the way": you can have - in principle - a wavefunction of the universe - which poses indeed a problem if you need state reduction, because what observer is going to do so.
The nice thing about the MW view (also, probably more appropriately called relative state interpretation) is that you simply split the universe in "observer" + "rest of the universe" and that you consider that the observable states are those that are product states of the two subsystems.

, moreover in a sum over histories framework you would even have tremendous difficulties in defining the experiment itself when doing quantum gravity.

All this of course in a universe without gravity, because we can't yet do this...

Moreover, there is no reason for keeping the superpostion principle at all (as you said, we do not see it and I thought quantum physicists are all strong positivists, so logically abandonning this should be of no worry). As said before, it is this very principle which is responsible for the measurement problem.

Unfortunately it is also the principle at the basis of the successful quantum formalism !

Concerning your entanglement impossiblity, I would say that this bet is as dangerous for you as it is for me. There is no distinction as yet and if the experiments would keep on turning out inconclusive (or in favor of realism) then from the esthetical point of view, the local realist attitude is certainly preferred.

That's where we differ fundamentally in opinion: the tests are NOT inconclusive. The tests follow EXACTLY what is expected by quantum theory, including the description of the apparatus. There is not one of these situations where you simply present the description of the experimental situation to just any quantum physicist, and where he doesn't turn up, after some straightforward calculation, with exactly those numbers that are also found during the experiment. In doing so, he did use quantum entanglement.
Ignoring the indicative value of these experiments is what I call "a dangerous bet".
Let us take a (ridiculous) example: imagine that you have a world view where, for some or other fundamental philosophical reason, it is impossible to have gravitational attraction between a mass like the sun, and a planet like Jupiter, at the earth-sun distance. Imagine you're living in the 17th century and there's a crazy Brit, called Newton, which has a theory with gravity in 1/r^2, which goes against your world view.
Now, this theory works of course on the sun-mercury, venus ... distance, but of course it is "experimentally" impossible to put Jupiter on the Earth orbit. So you say, see, this Newton guy's theory doesn't PROVE that the Sun-Jupiter interaction, if it were at the sun-earth distance, would be there. All measurements done today confirm my statement and are inconclusive about a potential gravitational interaction of a Sun-Jupiter system at a Sun-earth distance. That's 50 years now that people have been trying to confirm, without success, that Jupiter, placed at the Earth orbit, would follow Newton's laws.
Wouldn't you find such a claim totally ridiculous, in that if the Newtonian scheme is *confirmed* by experiment for all the cases of the planets in the solar system, that there is very very little room for a view where it WOULDN'T work in exactly that situation which gives you a conceptual problem ?

Concerning your Newtonian comments, you really don't seem to appreciate the full beauty of GR: locality and space time coordinate invariance are the most powerful and simplifying principles in nature.

Oh, but I do ! I only wanted to indicate that it is not sufficient to have a beautiful, powerful, simplifying principle. It needs to fit experiment also. And I DO find Newtonian physics more "beautiful, powerful an simplifying" than GR or QM. It is much more intuitive and clear... only it doesn't work in all cases, so that's it.
I think that QFT and GR will, one day, go the same way.

 

[Careful]

Or it doesn't happen ! Why does reduction have to happen ? If there's no observer, there's no need to have the wavefunction reduce to one of the "observable" states, as there aren't any. This is one of the big advantages of considering the unitary evolution "all the way": you can have - in principle - a wavefunction of the universe - which poses indeed a problem if you need state reduction, because what observer is going to do so.
The nice thing about the MW view (also, probably more appropriately called relative state interpretation) is that you simply split the universe in "observer" + "rest of the universe" and that you consider that the observable states are those that are product states of the two subsystems.
All this of course in a universe without gravity, because we can't yet do this... .

Come on, you know as well as I do that this position leads to ridiculous situations where the moon is not there unless we consciously percieve a photon scattered by it (Penrose mocks with this for a good reason). Again, you do NOT explain what macroscopic is, this is my main comment and you put it away handwavingly.

Look, the nice thing about GR is that it tells us that everything is dynamical; so your split observer + rest is very very ugly from that point of view.

That's where we differ fundamentally in opinion: the tests are NOT inconclusive. The tests follow EXACTLY what is expected by quantum theory, including the description of the apparatus. There is not one of these situations where you simply present the description of the experimental situation to just any quantum physicist, and where he doesn't turn up, after some straightforward calculation, with exactly those numbers that are also found during the experiment. In doing so, he did use quantum entanglement.
Ignoring the indicative value of these experiments is what I call "a dangerous bet".
Let us take a (ridiculous) example: QUOTE]

Your example is truly ridiculous. Look, you see the issue too one sided. For me, it is ok that you take Schroedinger as a good description together with its linearity and I acknowledge that this theory has predictive successes. BUT at the same time YOU do not want to see that your cheap tricks are putting one and the same problem on more and more religious and fuzzy grounds. What I try to tell you all along is that this travesty is the result of a HISTORICAL process, not of any *absolute* scientific reason. It is in my opinion much more natural, if you start from classical Hamiltonian dynamics to derive the non linear self field equations than the N particle Schroedinger equation (and in this derivation you do not even need to speak about a product state, neither do you need second quantization to take into account a non stationary radiation field). What I want to say here is that the compelling need to fit with experiment would have gotten anothor face if people would have taken that road and there would not even be spoken about entanglement at all. There is in my view no need to go over to configuration space techniques and this comment has been around since the first days of QM (unfortunately, people wanted to push the program too fast). The ONLY thing you say is that QM can be fitted, by taking into account realistic measurement setups, to the experimental data, but in the same way can local realism do that. The only difference is that QM is more advanced but this in an issue of MONEY and policy and not of intrinsic scientific value. Both positions are logically consistent and that is why your example is ridiculous. I would even dare to say that QM at the macroscopic level is more underdevelopped as classical theories are on the microscopic level (but you have a fictitious mechanism which you do not wish to explain to dispose of that comment). Moreover quantum gravity is not a problem in my reasoning, it is however a terrible (and unsolvable one in my opinon) in your line of thought.

Cheers,

careful

 

[vanesch]

Come on, you know as well as I do that this position leads to ridiculous situations where the moon is not there unless we consciously percieve a photon scattered by it (Penrose mocks with this for a good reason).

What is fundamentally wrong with that idea ? Weird, true ; but fundamentally wrong ? Not really. I don't say that things *have* to be that way, but why *can't* they be that way ? In fact, if you take MWI a bit further, in most of the universal wavefunction, the moon and the sun aren't even there ! But we just happen to observe that part of it where they are. I don't see why this position is untenable or ridiculous. If a scientific theory can explain one's perception, isn't that all one can require of it ?

I agree that things would certainly be more intuitively confortable if we didn't have to go into these considerations, in the same way as things would be more comfortable if we could have a universal time. But if a successful formalism forces us in some way to take up these considerations, what's wrong with that ?

Look, the nice thing about GR is that it tells us that everything is dynamical; so your split observer + rest is very very ugly from that point of view.

You still seem to see an observer as something absolute. But it is not. A stone could be an observer - a conscious one. You'll never find out (that's a well-known philosophical problem). There is not more a fundamental problem in considering "observer + rest" this way, than to consider a world line of an observer in GR, and the way this observer sees the universe. In exactly the same way, a "quantum observer" will see what happens along its "quantum world line", this time including its entanglement with whatever it is interacting locally with and making his "Born rule choices" in tracing out its world line.
This is not so very much stranger than an observer falling into a black hole observing (just before getting crushed) the entire universe's future (and hence being fried by all the radiation he gets at once).

It is in my opinion much more natural, if you start from classical Hamiltonian dynamics to derive the non linear self field equations than the N particle Schroedinger equation (and in this derivation you do not even need to speak about a product state, neither do you need second quantization to take into account a non stationary radiation field).

It may sound natural, but this doesn't work ! Many *solved* problems in QM have not much hope of being correctly handled that way ; I gave you the examples of configuration interaction in quantum chemistry, but there are miriads of examples, in solid state physics and condensed matter physics.

The ONLY thing you say is that QM can be fitted, by taking into account realistic measurement setups, to the experimental data, but in the same way can local realism do that.

That's the point: it does NOT have to be "fitted". The photoelectric effect in the photomultipliers for instance can be quantum-mechanically described. The workfunction of the metal can be described quantum-mechanically. There are no specific "fudge factors" that apply in the case of these EPR experiments. The phenomena leading to the detection process are well-described. That doesn't mean of course that no empirical calibration is used, but NOT MORE than for any other experimental technique. No "new concepts" have to be plugged in the theory to have the natural descriptive machinery of the process, of the beam splitters, of the detectors etc...
All LR models have to propose new concepts made up for the purpose, and often involve "unknown" workings of the experimental material (such as the detectors). For instance, Santos' SED, as far as I understands it, posits that EM radiation with energy h omega/2 is "present" in every mode, but that photodetectors have calibrated that away in order to give out 0 in what we think is "dark" but is in fact the background radiation. But if you now apply *thermodynamics* to this, you'd find that a bottle of black ink would soon start to boil if it truly absorbed all this radiation ! The remark is then that SED is only meant to describe *optical* phenomena, and that it can mimick EPR results in low-efficiency detectors that way. Ok, but you can't have such an EM theory that works for optics, but not for thermodynamics !

The only difference is that QM is more advanced but this in an issue of MONEY and policy and not of intrinsic scientific value. Both positions are logically consistent and that is why your example is ridiculous. I would even dare to say that QM at the macroscopic level is more underdevelopped as classical theories are on the microscopic level (but you have a fictitious mechanism which you do not wish to explain to dispose of that comment).

Except that apart from *conceptual* problems (which I will not deny - although they may be less severe than you seem to imply), we have good formulas which work FAPP ! It serves no purpose to have a clear conceptual framework when you don't have working formulas !

Moreover quantum gravity is not a problem in my reasoning, it is however a terrible (and unsolvable one in my opinon) in your line of thought.

Ok, but quantum chemistry (amongst other things - like a lot of stuff in solid state physics) is a problem in YOUR reasoning. You have some vague hope that this can be solved, but it is on just as fuzzy grounds if not more, than the quantum gravity problems from the QM side. Only, on the quantum chemistry side, there's a lot of experimental data, while on the quantum gravity side, there's not much for the moment on the experimental side. So you should first get the data we already have, right, before tackling what we don't even have, don't you think ?
As I said before, the solution space in QM is much bigger than any classical field problem - so if you succeed in obtaining a *correct* way of doing, in a classical way, quantum chemistry, this will be computationally MUCH more efficient.

 

[DrChinese]

In case it wasn't clear, what I described is not an allegory - the game could be played by real prisoners and captors, and presuming the prisonors can carry concealed entangled particles and stern gerlach appartuses(!) their probability of being released goes up to 85%. And no, it doesn't allow for superluminal communication between the two prisoners, but it certainly would seem to require superluminal communication between the particles in order to achieve.

I know it's not an allegory, I'm familiar with the game and some of its variations. Depending on the exact version of the game (questions, rules, number of players, etc.), you play it different ways. But where is the magic? The quantum players get a "tool" the others players don't get. But you must always bring together the results before you can see anything special happening.

It is pretty cool that there seems to be FTL "something" even though there is no (apparent) way to exploit this "something".

 

[Careful]

You still seem to see an observer as something absolute. But it is not. A stone could be an observer - a conscious one. You'll never find out (that's a well-known philosophical problem). There is not more a fundamental problem in considering "observer + rest" this way, than to consider a world line of an observer in GR, and the way this observer sees the universe. In exactly the same way, a "quantum observer" will see what happens along its "quantum world line", this time including its entanglement with whatever it is interacting locally with and making his "Born rule choices" in tracing out its world line.
This is not so very much stranger than an observer falling into a black hole observing (just before getting crushed) the entire universe's future (and hence being fried by all the radiation he gets at once).

Come on: conscious stones, we will come to highly intelligent electrons later on ! (that is one way to get realist theories which give quantum outcomes) Let me give some further objections to your already highly unrealistic point of view. I understand very well what you mean by your local observation operators, what you don't seem to understand is that you will need to create another bunch of superselection rules in order to figure out what the *localized* ``ground state´´ it is, your conscious mind is into, as well as the dynamical rules of the mind field operators (you are nowhere near doing this as far as I know) *** however for the localized vacuum consciousness state you might construct an adapted coordinate system in a tube around the classical worline and use the exponential map on the orthogonal space of the worldline; this could allow you to play the same Fourier analysis trick for functions vanishing outside this tube. However, I do not see how you could avoid the localized mind state from diffusing (which is one particular aspect of the cat problem) outside the tube so my guess is that you end up with a global state after all. Moreover, it is (a) a notoriously difficult problem to select the vacuum state for the universe and in any semiclassical approach (it is done wrt a preferred global congruence of observers in the classical solution) (b) your conscious mind state has to be localized, so it will be very difficult for you to couple it to the traditional QFT particle vacuum states since these are after all globally defined (that is one reason why I said the notion of particle in QFT is worth nothing) as well as to divise dynamics which keeps it local (the cat you do not want to tame). Now in quantum gravity, this becomes even more hopeless... moreover in quantum gravity, one of the motivations is that we figure out how a singularity smears out inside a black hole (your observers cannot even come there).

Short about the rest: I think the point SED is trying to make is that there is a background field which is in *equilibrium* with the environment, moreover, I am sure they introduce cutoffs in the modes too (no infinite energies). They do not explain the equilibrium from microphysics (which is for now still a weak point in my view), it is put as a constraint on the stochastic dynamics. I admit what I try to point out are HOPES. As we both know, the self field approach does not deviate much (the examples you gave showed a deviation of about two percent) from ``real´´ QM and that should be seen as hopeful. So, I have a clear guideline (indeed quantum chemistry) and a small percentage to bridge, and no, there has been no real attempt based in gravitation and classical EM to do this as far as I am aware of. I have surprised you a few times already, hope to do it in a more decisive way later on
It was nice talking to you.

Cheers,

Careful

PS: concering your computational efficiency, I disagree again. I will have to work with quite complicated matter models and highly non linear dynamics so that spoils it a bit.

 

[vanesch]

your conscious mind is into, as well as the dynamical rules of the mind field operators

That's the error often made: the "mind" is not a physical degree of freedom. There is no such thing as a "mind operator". Minds are simply associated with whatever physical structure we consider to be an observer ; or better, with the possible states these physical structures can be in. They are, if you want, an "asterix" on ONE term in the state of the universe in which one writes the state of the universe in a Schmidt decomposition (or a coarse-grained such decomposition) into "this physical structure" x "all the rest". In order to do so, you might need a foliation of spacetime ; why not take the one that corresponds to an observer frame that corresponds to the state of motion of the "previous" state of that physical structure, with an asterix ? (of course different states of the same physical structure can be in different motion states, and hence correspond to different world lines, but let's say that we only care about the one with the asterix). If a new interaction (unitary) entangles this "asterix state" with another state of something else, then the "asterix" goes now, following the Born rule, to one of these terms. This is then experienced by that mind as a "measurement" and is only aware of the state with the asterix.

I fail to see where the fundamental difficulty comes from (as long as we do not do quantum gravity).

PS: concering your computational efficiency, I disagree again. I will have to work with quite complicated matter models and highly non linear dynamics so that spoils it a bit.

I don't know if you have looked into computational quantum chemistry, but the "full" problem is almost intractable, hence the many smart selections of those few states that "ought to contribute". On something like a molecule that interests me, CF_4, you have 6 + 4 x 9 = 42 electrons, which means that (naively, ok, there are symmetries to be exploited) you have to find a set of solution functions of 126 real variables to find, as a result of a partial differential equation of second order in these 126 variables. True, the differential equation is linear, so what ? If I'd take, say, as an approximation, 50 hydrogen orbitals as a set of basis functions, then this gives me a priori a linear system of 50^42 variables to solve (again, modulo symmetries).
If I take a non-linear classical problem with an "electron field", I don't care how many of them are around; if I cut space up into, say, 100.000 cells in each dimension, I "only" have to solve a system of 10^15 variables to solve (iteratively, true, because the system is non-linear). That's difficult, but still easier.

cheers,
Patrick.

 

[Hurkyl]

The only way, in my opinion to solve schroedingers cat, is to make quantum theory non linear (another reason why I am keen on the self field approach) just as all realistic processes in nature are.

Out of curiousity, what do you find wrong with the following "solution"?

Suppose you had an experiment to detect whether the cat is "both" dead and alive. (according to whatever meaning you wanted to ascribe to this)

Then, since any state can be decomposed into two states, the first of which is internally consistent with the cat being 100% alive (and thus your experiment gives you a "no" answer, and the other internally consistent with the cat being 100% dead (and thus your experiment gives you a "no" answer), we must conclude that your experiment must always give a "no" answer.

In other words, there can be no observable oddities from the Schrödinger cat scenario.

 

[member 11137]

This discussion appears to be “semantic” dependant. If you accept a non professional in your discussion, first nobody seems to have the same definition for what the word “local” means. Secondly, and if you consider that physical phenomenon at quantum scale are concerning waves and correlated functions extended through space time (a little bit like at the surface of the see) I would intuitively say that: 1°) a wave alone is everywhere in space and in time. Such that two different observers placed at two different points in the universe could “see” the same and unique wave at the same time (“same time” being dependant on the practical tools that these two observers own to exchange and correlate informations concerning the wave). But we neglect here the fact that they need at least one other wave to transfer these informations concerning the first one; 2°) the concept of locality certainly is depending on the scale to which phenomenon are observed. This is for a part explaining why it is difficult to adopt a common point of view on what is local or not. For a phenomenon with the Planck’s length occurring inside a proton (e.g.), or between the proton and an electron, adopting a kind of classical point of view it is not sure that the proton or that the electron appears to be local… On the other side, the waves (functions) associated with this proton and this electron certainly extend until the region where this quantum phenomenon exists. 3°) For me and because of this, the concept of waves makes that nothing can be totally local as long as it stays inside the light- or the information-cone of something else. 4°) This vision explains the next one: space time itself is a fluctuating phenomenon. 5°) Concerning my vision: if a subset of these fluctuations is corresponding to what we call an identified particle, nothing is theoretically forbidding that a part of the particles are dual one or able to become dual, like two twin sisters, like two correlated balls of billiard. Can this vision help to understand the EPR experiment?

 

[member 11137]

To complete the above post I could recommand a recent article by Aharonov where (if I understand correctly) he is defending the idea that any event stays at the boarder between the past and the future; based on this idea he gives a re-definition of the probabilities, a.s.a...

 

[Sherlock]

Out of curiousity, what do you find wrong with the following "solution"?
Suppose you had an experiment to detect whether the cat is "both" dead and alive. (according to whatever meaning you wanted to ascribe to this)
Then, since any state can be decomposed into two states, the first of which is internally consistent with the cat being 100% alive (and thus your experiment gives you a "no" answer, and the other internally consistent with the cat being 100% dead (and thus your experiment gives you a "no" answer), we must conclude that your experiment must always give a "no" answer.
In other words, there can be no observable oddities from the Schrödinger cat scenario.

It makes sense to me. I've learned to think of the S-cat thing, and associated qm interpretational issues, as a non-problem. Sort of like the ancient Greek argument that arrows can't reach their targets. (A lot of Greeks died believing that.) :-)

Anyway, I've reached a conclusion about the question of this thread that I think is ok. Quantum theory is not inherently non-local -- and interpreting the theory to be local or non-local is a matter of how one chooses to understand the *bases* for the formalism (not what sort of non-physical stuff one can generate from its use, or abuse).

 

[Careful]

I don't know if you have looked into computational quantum chemistry, but the "full" problem is almost intractable, hence the many smart selections of those few states that "ought to contribute". On something like a molecule that interests me, CF_4, you have 6 + 4 x 9 = 42 electrons, which means that (naively, ok, there are symmetries to be exploited) you have to find a set of solution functions of 126 real variables to find, as a result of a partial differential equation of second order in these 126 variables. True, the differential equation is linear, so what ? If I'd take, say, as an approximation, 50 hydrogen orbitals as a set of basis functions, then this gives me a priori a linear system of 50^42 variables to solve (again, modulo symmetries).
If I take a non-linear classical problem with an "electron field", I don't care how many of them are around; if I cut space up into, say, 100.000 cells in each dimension, I "only" have to solve a system of 10^15 variables to solve (iteratively, true, because the system is non-linear). That's difficult, but still easier.
cheers,
Patrick.

EVERYTHING is PHYSICAL ! Concerning your computability argument: I said I would have to consider realistic electron models and perhaps another local interaction field (a spin field such as in einstein cartan theory perhaps, but I would not like to do that). So I will have to be a lot more creative than what you are hinting to and I do not a priori know what number of degrees of freedom I will end up with. And indeed, hopefully less than the number you hint at; but non linearities always bring subtle difficulties :-)

Cheers,

careful

 

[Careful]

Out of curiousity, what do you find wrong with the following "solution"?
Suppose you had an experiment to detect whether the cat is "both" dead and alive. (according to whatever meaning you wanted to ascribe to this)
Then, since any state can be decomposed into two states, the first of which is internally consistent with the cat being 100% alive (and thus your experiment gives you a "no" answer, and the other internally consistent with the cat being 100% dead (and thus your experiment gives you a "no" answer), we must conclude that your experiment must always give a "no" answer.
In other words, there can be no observable oddities from the Schrödinger cat scenario.

Huh ?? If you allow for a macroscopic state in which the cat is both alive and dead (an experiment which could decide upon that) then the alive and dead states do not form a complete basis.

 

[Careful]

It makes sense to me. I've learned to think of the S-cat thing, and associated qm interpretational issues, as a non-problem. Sort of like the ancient Greek argument that arrows can't reach their targets. (A lot of Greeks died believing that.) :-)
Anyway, I've reached a conclusion about the question of this thread that I think is ok. Quantum theory is not inherently non-local -- and interpreting the theory to be local or non-local is a matter of how one chooses to understand the *bases* for the formalism (not what sort of non-physical stuff one can generate from its use, or abuse).

How how, the S cat problem attracks still a lot of research (at eminent research centres) and it is not philosophy at all. The rest of your conclusion might be true but (a) we do not know wether realistic interacting QFT's satisfy the Wightman axioms (b) you have to really solve the cat in my opinion.

 

[Hurkyl]

Huh ?? If you allow for a macroscopic state in which the cat is both alive and dead (an experiment which could decide upon that) then the alive and dead states do not form a complete basis.

Let's start with this: why do you think the Schrödinger cat scenario is a problem?

 

[Careful]

Let's start with this: why do you think the Schrödinger cat scenario is a problem?

Very simple: you have to show a couple of things (a) how come do we only observe dead or alive cats (while QM allows for both at the same time), how can we dynamically derive from QM what macroscopic states are allowed in nature (b) why, and from what scale on, do objects satisfy the classical laws of nature and behave as classical non fuzzy objects (take care here: the argument that the mass simply gets large is not sufficient here) (c) why is Bell locality such a good principle from a scale of 10^{-8} meters in our universe while QM would predict entanglement as the generic rule. These topics are the centre of very active research and are not solved by any means if you wish to stick to unitary mechanics. Usually people try to dispose of them by using handwaving (often incorrect (!)) quantum statistical arguments. You undoubtedly saw that Vanesh had to cook up a non dynamical parallel mental world (which he cannot explain) coupled to the physical word in order escape from this. Moreover, most MWI's think differently about this and really want to pursue the points I mentioned.

 

[Hurkyl]

(a) how come do we only observe dead or alive cats (while QM allows for both at the same time)

This is exactly what my post was addressing.

QM allows both at the same time only in the sense that it allows a superposition of states in some of which the cat is unambiguously alive, and the rest in which the cat is unambiguously dead.

Therefore, it is impossible for any experiment to distinguish between a state in which the cat is unambiguously dead or alive, and a state in which we have both "at the same time". In other words, we only observe dead or alive cats.

 

[Careful]

This is exactly what my post was addressing.
QM allows both at the same time only in the sense that it allows a superposition of states in some of which the cat is unambiguously alive, and the rest in which the cat is unambiguously dead.
Therefore, it is impossible for any experiment to distinguish between a state in which the cat is unambiguously dead or alive, and a state in which we have both "at the same time". In other words, we only observe dead or alive cats.

Oh yeh, what in QM tells me that I do not have projection operators on such superpositions (which are clearly non commuting with the pure dead and alive projection state operators). By the way all these states can clearly be distinghuised from a statistical point of view (for example vanesh would be able to do this, since his consciousness does not intervene and has to produce the correct statistical results). First, perhaps tell me what measurement procedure you use, reduction/consciousness or environmental decoherence). You should also comment on the other issues which were subject of many papers of Zurek and others...

 

[Hurkyl]

Careful: none of this is answering my question to you. You have made the claim:

The only way, in my opinion to solve schroedingers cat, is to make quantum theory non linear (another reason why I am keen on the self field approach) just as all realistic processes in nature are.

which, of course, begs the question "In your opinion, why must other approaches fail?"

My inability to prove that an alternate approach completely addresses the question is not a proof that other approaches fail. It is not even evidence that the approach is lacking, since the fact I am not a quantum physicist is more than sufficient to explain my inability!

Oh yeh, what in QM tells me that I do not have projection operators on such superpositions (which are clearly non commuting with the pure dead and alive projection state operators).

I never said you didn't. But, I don't see how such projection operators would allow one to experimentally distinguish between a cat that's alive or dead and a cat that's alive and dead.

Maybe I should explain why I mean more formally? If M is an operator that satisfies:

M(any-living-cat-state) = 0
M(any-dead-cat-state) = 0

Then we must have that M is the zero operator.

In particular, this means there cannot be an experiment that gives one result when applied to a cat that's dead or alive, and some different result when applied to a cat that's both dead and alive.


This answers the question of why we don't perceive some chimeric dead-and-alive hybrid cat: any perception we make must be consistent with a cat that's either dead or alive.


Being able to tell statistically is something entirely different: sure, if we had many different copies of a dead-and-alive superposition, we could observe them and see that some turn into living cats and some turn into dead cats, and thus we know the underling state was a superposition, but AFAIK, that's not the paraodx -- the paradox is why we don't see a chimeric dead-and-alive hybrid cat.

First, perhaps tell me what measurement procedure you use, reduction/consciousness or environmental decoherence).

I'm not entirely sure I understand this question properly.

At the moment, I internalize the notion of measurement as being essentially synonymous with computing a function in a quantum computer: I take as input the state

phi x |unmeasured>

and produce as output:

phi1 x |1> + phi2 x |2> + ...

where phi = phi1 + phi2 + ...

and have thus performed a measurement.

 

[Sherlock]

How how, the S cat problem attracts still a lot of research (at eminent research centres) and it is not philosophy at all.

You say that the S-cat problem is:
a) How come we only observe dead or alive cats, while QM allows for both at the same time.

Hurkyl has provided one way to understand this. The way I think about it is that the cat is a measuring instrument, like a PMT in optical Bell tests. A dead cat means that a particle was emitted from the radioactive substance during a certain time interval initiated by opening the radioactive substance's enclosure. Until the cat dies, it's alive. We know this by continually monitoring it. Just like a PMT either registers a detection, or it doesn't during a certain interval. Of course there's an interval during which the cat is in the process of dying and the PMT is in the process of registering.

QM does not allow for cats to be both alive and dead, or PMTs to both register and not register during a certain instant of time, in any physical sense. Alive, dead, not register, register -- are simply the possible results of the experiment (ie., the macroscopically allowed states -- which QM does specify). Dead cats and PMT-generated data streams are unambiguous and irreversible.

So, this doesn't seem to be a problem.

You continue with:
b) why, and from what scale on, do objects satisfy the classical laws of nature and behave as classical non fuzzy objects?

This is a harder question. The correspondence line is fuzzy and changes as technological capabilities progress. I suppose that distinguishing the quantum and classical 'realities' will always be somewhat fuzzy. (eg., they can quantum entangle systems consisting of thousands of atoms now). But whether a beaker of cyanide gas is broken or not, or a cat dead or alive, is not fuzzy. Whether a piece of radioactive material has emitted a particle or not is somewhat fuzzy. The macroscopic behavior (measurement result) of the barrier that is used to intercept and detect the particle isn't at all fuzzy.

Why is the particle-emission behavior of the radioactive substance fuzzy (ie., random)? Because the only information that we have about these materials pertaining to particle emissions comes from putting detectors next to them and noting when they register. Of course it's more complicated than that, but nonetheless the info is still pretty spotty.

I see this as essentially an instrumentation and detection problem.

And you conclude with:
c) why is Bell locality such a good principle from a scale of 10^{-8} meters in our universe while QM would predict entanglement as the generic rule?


Bell locality isn't at odds with quantum entanglement.

Bell locality, P(A|a) = P(A|a,B,b), isn't really an arbiter of locality. Bell locality is about the independence, or dependence, of observations and their associated statistics. (Assuming that dependence of spacelike separated measurements implies non-locality is, imo, wrong.) QM says that entangled measurements, A and B, and associated observational parameters, aren't independent. Why? Because the disturbances that eventually produced A and B originally came from the same emitter, or had interacted with each other, or were altered in some way common to both, and are being analyzed and detected by the same sorts of devices.

I can't really critique your program, but I wonder why the people who developed quantum theory didn't go that route?

 

[vanesch]

Out of curiousity, what do you find wrong with the following "solution"?
Suppose you had an experiment to detect whether the cat is "both" dead and alive. (according to whatever meaning you wanted to ascribe to this)
Then, since any state can be decomposed into two states, the first of which is internally consistent with the cat being 100% alive (and thus your experiment gives you a "no" answer, and the other internally consistent with the cat being 100% dead (and thus your experiment gives you a "no" answer), we must conclude that your experiment must always give a "no" answer.
In other words, there can be no observable oddities from the Schrödinger cat scenario.

That doesn't work this way. Imagine that |D> is the "dead cat state" and |L> is the "live cat state" (or a representative of the orthogonal spaces that correspond to that property).
If the cat is in a |L>+|D> state, then the operator that is diagonal in the basis {|L>+|D>,|L>-|D>} (the observable corresponding to your proposed experiment) with eigenvalues "yes" and "no", would give "yes" 100% to |L>+|D> state; while it would give "yes" only in 50% for a |L> or for a |D> state. In other words, this is an interference experiment with cats, notoriously difficult to do in the lab.
The reason why it is difficult is that a live, or a dead cat state quickly entangles with the environment, so that you get:
|envL>|L> + |envD>|D> as an overall state. Applying your operator NOW to this entangled state, will result in 50% yes only, as if we only had a |L> or a |D> state ; this is due to the essential orthogonality of |envL> and |envD>.
So once your cat gets entangled with its environment (the air, the radiation field...) you cannot get this interference out anymore. The reason why it is essentially the |envL> and |envD> states that couple respectively, in this basis, to |L> and |D> is due to the specific form of the interaction between a cat and its environment, which is heavily position-dominated (for instance, coulomb interaction is in 1/r)
This is the essence of the environmental decoherence program, explaining why we don't often see "macroscopic quantum interference effects".

 

[vanesch]

You undoubtedly saw that Vanesh had to cook up a non dynamical parallel mental world (which he cannot explain) coupled to the physical word in order escape from this. Moreover, most MWI's think differently about this and really want to pursue the points I mentioned.

Most MWIers are making in any case an implicit assumption of "what is perceived" although many of them forget to say so! As you correctly point out, unitary QM just ends up entangling your body with different classical views. MWIers then tacitly assume that we only perceive ONE of them, but that, in itself, is A FUNDAMENTAL ASSUMPTION that they sometimes forget to specify. There is a priori nothing wrong with a world in which we are aware of BOTH our body states (the body seeing a live cat, and the body seeing a dead cat).
Most efforts MWIers concentrate on is to do 2 things:
1) solve the "preferred basis problem"
2) derive somehow the Born rule from some "more natural" statistics such as state counting, or "rational decision theory" or whatever.
But my point (my heresy in MWI !) is that, even if you had very plausible mathematical arguments for 1) or 2), YOU STILL HAVE TO POSTULATE that you don't observe the "entire state", and YOU STILL HAVE TO POSTULATE the "more natural statistical rule", which applies in your "more naturally derived preferred basis".
In other words, you CANNOT *derive* classical observations from a strictly unitary QM ; you STILL need to add some extra postulate that links the total wavefunction to the "perceived" one. The only discussion is about HOW NATURAL this can be done.
So my viewpoint is that *if you have anyhow to add a postulate* why not make life easy, and simply postulate that an observer will observe consciously this state according to the preferred basis that corresponds to observation (classical world) and using the Born rule.
This is "deus ex machina" of course, and now all derivations of MWI proponents just make these postulates "sound a bit more natural".
But there's no reason WHY I cannot postulate this directly. I could try to go to more "fundamental" postulates, and derive the preferred (position) basis and the Born rule from there, but I can just as well posit it directly in the mean time.
The point to see is that *in any case* a postulate will have to be added in order to find agreement between unitary QM and observation. This is nothing else but a postulate that DETERMINES WHAT IS CONSCIOUS OBSERVATION. No matter how you turn it. Because my *body* IS in a superposition. My brain *IS* in a superposition. If it only depended upon my body, I would see the cat BOTH alive and dead. I wouldn't perceive a classical world. I would see myself both in the grocery store and posting on PF. And, as I want to point out, there's nothing wrong with that - only that's not what we perceive. So any "derivation" of classical perception from unitary QM MUST postulate something about perception. Many MWIers seem to do this implicitly by using "world counting rules" and trying to establish "naturally-looking" preferred bases for this "world counting". But, even if this succeeds, they forget that the STILL need to postulate that this "world counting" must be done in the first place !
cheers,
Patrick.

 

[vanesch]

I wrote:

Because my *body* IS in a superposition. My brain *IS* in a superposition. If it only depended upon my body, I would see the cat BOTH alive and dead. I wouldn't perceive a classical world. I would see myself both in the grocery store and posting on PF.

but I should make something probably clearer. When I say: I would see the cat BOTH alive and dead, then I just mean that: I would be aware of TWO classically-like worlds. I WOULDN'T see the cat in a SUPERPOSITION of live and dead. That problem is solved already by environmental decoherence. There are 3 levels in this problem:

1) "the cat is in a strange state, a superposition of live and dead"

2) "the cat is dead in one "world" and the cat is alive in the other"

3) "with probability 50%, I see a dead cat, with probability 50%, I see a live cat"

1) was Schroedinger's objection (that's a state that doesn't seem to exist). However, environmental decoherence shows us that 1) decays extremely rapidly into 2). Indeed, (|L> + |D>)|env> quickly goes into (|envL>|L> + |envD>|D>), where we can consider these two terms as "worlds".

What is unsolvable without an extra postulate, IMO, is the 2) -> 3) transition. I should see BOTH the world with the live cat and the world with a dead cat, as in 2). Why I only perceive ONE of them, with a certain PROBABILITY, can never be deduced (IMO) from the unitary formalism.

The measurement Careful was talking about, about testing the (|L> + |D>) state, is difficult to perform, because we would need to UNDO the 1) -> 2) evolution.

 

[Sherlock]

That doesn't work this way. Imagine that |D> is the "dead cat state" and |L> is the "live cat state" (or a representative of the orthogonal spaces that correspond to that property).
If the cat is in a |L>+|D> state, then the operator that is diagonal in the basis {|L>+|D>,|L>-|D>} (the observable corresponding to your proposed experiment) with eigenvalues "yes" and "no", would give "yes" 100% to |L>+|D> state; while it would give "yes" only in 50% for a |L> or for a |D> state. In other words, this is an interference experiment with cats, notoriously difficult to do in the lab.
The reason why it is difficult is that a live, or a dead cat state quickly entangles with the environment, so that you get:
|envL>|L> + |envD>|D> as an overall state. Applying your operator NOW to this entangled state, will result in 50% yes only, as if we only had a |L> or a |D> state ; this is due to the essential orthogonality of |envL> and |envD>.
So once your cat gets entangled with its environment (the air, the radiation field...) you cannot get this interference out anymore. The reason why it is essentially the |envL> and |envD> states that couple respectively, in this basis, to |L> and |D> is due to the specific form of the interaction between a cat and its environment, which is heavily position-dominated (for instance, coulomb interaction is in 1/r)
This is the essence of the environmental decoherence program, explaining why we don't often see "macroscopic quantum interference effects".

I thought Hurkyl's point was that QM doesn't allow the observation of an alive-and-dead state composed of separate alive and dead states. Anyway, there's no need to jump through all these hoops about this. We're talking about one 'detector', and it can't be in both, mutually exclusive, possible detection states at the same time.

QM says that, Cat = 1/(sqrt 2) {psi_alive + psi_dead}, means that we will *not* observe a Cat that is both alive and dead. So, what QM says about this isn't really the problem. It's consistent with experimental results. The cat is always observed to be either alive or dead.

The problem is sometimes stated as -- why don't we ever observe anything corresponding to interference between alive and dead cats? (Why don't don't we ever observe cats which are both alive and dead at the same time? A silly question?)

As you note, every macroscopic object (including cats, alive *or* dead) can be considered as a manifestation of interfering wave systems on an environmental scale.

So maybe the problem can be stated as -- why don't we ever observe anthing corresponding to interference between cats (alive or dead) and the environment.

Well, an *interference effect* is what *any* macroscopic object *is*, isn't it?

We can actually see interference, as it's happening, wrt many wave phenomena in macroscopic media. But, the structure of more fundamental media and direct apprehension of disturbances in those media are not currently part of our knowledge.

There is a problem, to be sure. But it's not essentially a problem with quantum theory itself. Maybe it's not even essentially a theoretical problem any more. The media scale, wave-reality of quantum level phenomena is, except for intermittent probes, invisible to us -- and this is the problem, quantum-level phenomena can't be tracked.

Quantum theory is part of the solution to this problem. Interpretations, or reformulations of QM (or entirely different theoretical approaches) are not addressing the real, physical problem.

 

[Careful]

Careful: none of this is answering my question to you. You have made the claim:
which, of course, begs the question "In your opinion, why must other approaches fail?"
.

Ok, I see vanesh answered that already in his first reply. Although, as I mentioned this coupling to the environment is not quite satisfactory (Vanesh doubted that also a while ago), since you are relying in fact on a kind of quantum statistical argument here (exact computations with very small model environments have been done recently) and the Born rule comes out rather well, but not quite all the time (due to recurrency times as far as I understand) (you should search it up in the book of Bernard d'Espagnat or Ghirardi Rimini and Weber also comment on this in their overview paper, you can find on quant-ph). Let me say already this: if you want to stick to copenhagen or want to buy your way out ``kinematically´´ as Vanesh does, and you are pleased with that. Then, for you there is no problem. But this is FAPP as Bell calls it, I am interested in an explanation of the measurement, not in some trick like R (when does it take place, by what act which apparatus should be seen as classical and so on) or a mental world like in MWI.

Ok Vanesh nicely explained what measurement is about in his third message: it is indeed the AND -> OR transition. For sherlock: yes I abused Bell causality here: it should have been local realism (if you want an exact definition, you may ask me one).
The reason why I am interested in a true explanantion is to be found in quantum gravity. I shall come back to that in another post.

 

[vanesch]

Anyway, there's no need to jump through all these hoops about this. We're talking about one 'detector', and it can't be in both, mutually exclusive, possible detection states at the same time.

Well, the problem is that if you take quantum theory seriously, that's exactly what happens: your detector IS in two "mutually exclusive states" at the same time. That's what unitary evolution dictates, and it is the very founding principle of quantum mechanics.
This is called the superposition principle, and it is exactly the same principle that says that an electron in a hydrogen atom is both above and below the nucleus, and to the left and to the right of it, which are also "classically mutually exclusive states". This is exactly what the wavefunction is supposed to mean: the electron is in the state ABOVE the nucleus, is ALSO to the left of it, is ALSO to the right of it, and is ALSO below it, with the amplitudes given by the value of the wavefunction.
A quantum particle that impinges on a screen with several holes goes through the first hole, and ALSO goes through the second hole, and ALSO goes through the third hole.
And if you take this principle seriously all the way (that's what MWI does) then your particle detector SAW the particle, and DIDN'T see the particle. So on the display of the detector it is written "CLICK" AND it is written also "NO CLICK". And if you look at it, your eyes will BOTH see "click" and "no click". And your brain will BOTH register the information of the fact that your eyes saw "click" and that your eyes DIDN'T see click.
Only... you are only consciously aware of ONE of these possibilities.
*IF* quantum theory as we know it applies to all the particles and interactions in this scheme (the atoms of the detector, of your eyes, of your brain etc...) then there is no escaping this conclusion. This is due to the fact that *ALL* interactions we know (electroweak, strong, except for gravity), are, as far as we know in current quantum theory, described by a UNITARY EVOLUTION OPERATOR.
So what are the ways out of this riddle ?
1) this is indeed what happens, and for some strange (?) reason, we are only aware of one of the states. This is the picture I'm advocating - unless we've good indications of the other possibilities.
2) this unitary evolution is a very good approximation which is in fact, slightly non-linear. this can be a minor modification to QM, or this can be just an indication that QM is a good effective theory for something totally different.
3) we've not yet included gravity. Maybe gravity will NOT be described by a unitary evolution operator.
4) there's maybe another interaction that spoils the strictly unitary evolution
5) somehow the act of observation (what's that ?) is a physical process that acts upon the wavefunction (that's the von Neumann view: but WHAT PHYSICS is this act of observation then ?) and reduces the state of whatever you're "observing".

 

[vanesch]

Let me say already this: if you want to stick to copenhagen or want to buy your way out ``kinematically´´ as Vanesh does, and you are pleased with that. Then, for you there is no problem. But this is FAPP as Bell calls it, I am interested in an explanation of the measurement, not in some trick like R (when does it take place, by what act which apparatus should be seen as classical and so on) or a mental world like in MWI.

Just to make this clear: I'm certainly not "happy" with the mental MWI explanation. I only stick to it because, IMO, it is what fits closest to the *current formalism* of quantum theory - in fact I don't worry, because there's probably still a lot to discover in physics, and this won't be our "last theory".
I think the fundamental difference between Careful's view and mine is that I consider that, no matter how FAPP our current QM is, I think its entanglement predictions are (even FAPP) correct, because so many experiments confirm it indirectly. As such *I know that there is a serious problem* and I don't know how to solve it.
Careful's view seems to deny this prediction of entanglement, so that he can rely on some hopes that a more classical field theory including GR will do the trick (and which can of course never produce ideal entanglement predictions because it will be an LR theory). The argument being that you then don't have to face the serious problems, and that everything is much cleaner and nicer.
I think that no matter how nice a scheme one has, the final judge is the experiment, and I wouldn't bet my money on hoping that the entanglement predictions of QM are wrong (even only FAPP), them having had so many indirect successes ; nevertheless if he thinks there's some hope for his approach, why not.
In the mean time (until one has a better WORKING theory), I still think it is interesting to have a story that fits with the currently WORKING theory, which is QM - and that's all I want to do here.

 

[Careful]

I think that no matter how nice a scheme one has, the final judge is the experiment, and I wouldn't bet my money on hoping that the entanglement predictions of QM are wrong (even only FAPP), them having had so many indirect successes ; nevertheless if he thinks there's some hope for his approach, why not.
In the mean time (until one has a better WORKING theory), I still think it is interesting to have a story that fits with the currently WORKING theory, which is QM - and that's all I want to do here.

Indeed, that is the only point where we disagree and that is why EPR experiments are so damn important. My filosofy is that GR and QM are both equally working theories but that GR is by far the superior THEORY. We have to look for a unifying framework and the misery of QM is, if you want to include GR, the superpostion principle. It is notoriously difficult (an in my view even meaningless) to add up different spacetimes. That is why I would argue that one should try to push the GR scheme as far as possible and this is not by any means easier in the sense that you will have to answer questions about ``how do elementary particles look from the inside´´. Moreover, even if this would turn out not to be possible, and if there really turns out to be a non locality scale in nature, then any local realist attempt will clearly bring insight into the more precise nature of this non locality. This would allow one to introduce non locality in a *controlled* way and study the corrections. This is in a sense what Penrose proposes: his OR scheme is a controlled way to exclude macroscopically too different alternatives (only he does not know how to do it in a covariant way). The point I want to make is that the *pure* quantum strategy is too wild for these purposes. Therefore, it makes sense to start with the purest local strategy and see how far one gets. This will certainly provide insightful results, also for QM. My hope is of course that there is no non locality scale at all.

Cheers,

Careful

 

[Hurkyl]

If the cat is in a |L>+|D> state, then the operator that is diagonal in the basis {|L>+|D>,|L>-|D>} (the observable corresponding to your proposed experiment) with eigenvalues "yes" and "no", would give "yes" 100% to |L>+|D> state; while it would give "yes" only in 50% for a |L> or for a |D> state. In other words, this is an interference experiment with cats, notoriously difficult to do in the lab.

Maybe I'm misunderstanding the point of Schrödinger's cat...

I thought the question is "Why don't we ever see a cat that looks like it's a superposition of states?"

and my answer is

"No single experiment can detect a superposition -- they're only detectable by looking at the statistics of repeating an experiment on identical copies of the state"

(At the moment, I claim that the same is true for microscopic states)

Or to put it differently, a superposition of states looks normal, until you do experiments with repeated copies of it and look at the statistics.

 

[Careful]

Maybe I'm misunderstanding the point of Schrödinger's cat...
I thought the question is "Why don't we ever see a cat that looks like it's a superposition of states?"
and my answer is
"No single experiment can detect a superposition -- they're only detectable by looking at the statistics of repeating an experiment on identical copies of the state"
(At the moment, I claim that the same is true for microscopic states)
Or to put it differently, a superposition of states looks normal, until you do experiments with repeated copies of it and look at the statistics.

That is a FAPP argument, which boils down to the statement : ``QM is good as long as we ask the correct questions (correct = good common sense borrowed from everyday life experience)´´. In the Copenhagen framework, this leads to the introduction of a different (and conflicting) dynamics R from U. Nobody knows what R means physically (but is is damn necessary to make predictions since unitary dynamics by itself is not enough as Vanesh points out). People find this unsatisfying since we would like to have ONE unified dynamics which also tells us why we only observe one world. You say: no single experiment can distinguish between both pure eigenstates and a linear combination of eigenstates. That is true (by definition in the copenhagen framework) but the issue is what an experiment IS PHYSICALLY (this is the measurement problem which is usually explained using Schroedingers cat picture). Let me recall you a thought experiment from Penrose about the weather on a planet. It is well known that weather is very chaotic and therefore quantum effects will in general make sure that after suffiently long time the weather state of the planet is a sum of many different macroscopic distinghuishable weather states (for example storm or no storm in america). At this moment in time, you have to wonder why we only percieve one weather state. This what reduction does for you, or Vanesh consciousness does. You might also hope that the classical instability argument is miraculously solved by quantum theory (but this seems not the case, and you might want to percieve this as some specific form of the cat problem). I think that you simply say that BY definition R does the trick and we have commented already that (a) this is not satisfying (b) what about non local observables (c) usually you implement R by use of a global foliation (which brings along causality problems unless you only wonder about local observables, moreover foliations are troublesome in quantum gravity).

This is an issue with a long history; Bohr and Heisenberg were very quickly aware that one had to introduce classical concepts into quantum mechanics in order to make contact with the real world. People nowadays try to dispose of this undesirable feature since QM, if it were to serve as a fundamental theory, should explain classicality and not use it. Vanesh already agreed with me, that if you want to do this in a satisfying and dynamical way, you will have to give up unitarity (and in my view also linearity). You might want to read in the book of de Broglie ``non linear wave mechanics´´ in order to get a deeper understanding of this issue which was around since 1919 as far as I know. My personal opinion about this, is that QM is an effective devise to give statistical results and that the Schroedinger equation therefore is NOT a one particle equation but a devise to describe outcomes of many experiments. Therefore, one should look for a true (classical in my mind) chaotic dynamics whose statistical mechanics exactly coincides with the Schroedinger predictions. Barut has once put this in a nice way: it is not because that 50% of the population is male and 50% female that any person is half man and half woman. :-)