Wednesday, September 06, 2017

Interpretations of Quantum Mechanics: The Cat’s View

Something else I made for the book but later removed. (Click to enlarge.)


45 comments:

Patrice Ayme' said...

Why did you remove it? Sounds pretty fair to me.

Sabine Hossenfelder said...

Why did I remove it? To begin with it was pretty pointless. Also, style doesn't fit with the rest of the book. It's not a funny book really. Then I removed the whole explanation in the text about consistent histories because I found it to be unnecessary and more confusing than enlightening, and somehow I felt with only 5 items the list wouldn't really be a list.

I think in the end I just got fed up with it and that was that.

Matthew Rapaport said...

I know how busy you must be by how much the time between your posts increases. You are in a new reference frame! :) I hope everything is going well for you!

Andrew Wells said...
This comment has been removed by the author.
Fizeg said...

Except the last thing is not really a proper description of the Copenhagen interpretation, rather of the "Shut up and calculate". The proper one would be: I can only describe the result of what happened between me and the cat with "The cat is dead" and let's close this topic, please.

Uncle Al said...

Is reality determinant (GR), probabilistic (QM), or compliant given noise and decoherence?

https://www.uni-due.de/~hp0198/pubs/prl7.pdf
...DOI:10.1103/PhysRevLett.103.023202

Hund's Paradox tests the cat. We can manipulate single molecules within a tuned cavity in extreme vacuum. Replace D2S2, chiral by a rotation barrier, with a lean and robust polycycle lacking any facile racemization pathway. Resolved chiral 2-norbornanone, twistane, and especially D_3-trishomocubane derivatives end hand-waving.

Alexander McLin said...

I noticed the transactional interpretation is missing from your list. While I have no strong preferences for any one of interpretations proposed to date, I'm now curious to whether you had any favored opinions about any one on the list?

Matthew Rapaport said...

I know you are not asking me, but my own favorite ("Transactional Interpretation" Kastner/Cramer) is not even mentioned. I have asked Dr. H about this one before but I believe she is unfamiliar with it.

Sabine Hossenfelder said...

Alexander,

There are like 30 or so more interpretations that aren't on the list. I just picked the most widely known ones. As to the transactional interpretation, it's too vague for my comfort. I have no idea how to actually calculate anything with it. The papers that I read on it basically didn't contain a mathematical formulation, it was just words.

Sabine Hossenfelder said...

Matthew,

Yes, you have asked about this before, so have some other people. I am familiar with it, I just think it's bla-bla. The reason it's not on the list is that nobody uses it. I don't know if one can even use it for anything, other than asking me about it that is.

Louis Tagliaferro said...

I never thought it worthwhile to give too much thought to any of the QM interpretations; however as a non-cat lover your Copenhagen looks pretty good to me!

Tam Hunt said...

What do you think of decoherence interpretations? I've been reading a bit about them in relation to process philosophy and it sounds like a promising avenue for resolving many issues like the measurement problem.

ppnl said...

I have never understood what QBism is trying to say. Of course opening the box updates our knowledge of the cat. That must be the least controversial claim ever. But nothing follows from it.

I prefer "shut up and calculate" along with an appreciation for decoherence. Nothing else seems to be needed and it would take surprising experimental results to move me beyond this.

Matthew Rapaport said...

Thank you for the reply Dr. H. not quite understanding tho. How do any of the "interpretations" (speculations about the underlying ontology) help you calculate? For example Everett's "many worlds interpretation" is very popular in the physics community. How does it contribute to calculations?

Sabine Hossenfelder said...

Matthew,

Many Worlds has pretty much the same axioms as Copenhagen except that then you do decoherence and call that a splitting. You can do calculations with that just fine. Now tell me how I'm supposed to calculate a measurement out come in the transactional interpretation.

Sabine Hossenfelder said...

ppnl,

QBism, in my perspective, is the common denominator. It's what everybody agrees on is correct. All the other interpretations add something to it. I'm fine with QBism, but also not sure what one learns from it.

Sabine Hossenfelder said...

Tam,

Decoherence isn't an interpretation it's a fact. And it doesn't explain the measurement problem.

Paul Hayes said...

QBism is not a common denominator from the perspective of interpretation. The common denominator in (most) interpretations is the mathematical formalism but interpretation is not part of that. There is a fundamental dichotomy in 'interpretation space' between the psi-epistemic and the psi-ontic interpretations:

The results reviewed in this paper aim to show that the quantum state must be ontic (a state of reality) rather than epistemic (a state of knowledge). What does this mean and why is it important? The word “ontology” derives from the Greek word for “being” and refers to the branch of metaphysics that concerns the character of things that exist. In the present context, an ontic state refers to something that objectively exists in the world, independently of any observer or agent. In other words, ontic states are the things that would still exist if all intelligent beings were suddenly wiped out from the universe. On the other hand, “epistemology” is the branch of philosophy that studies of the nature and scope of knowledge. An epistemic state is therefore a description of what an observer currently knows about a physical system. It is something that exists in the mind of the observer rather than in the external physical world

QBism is a psi-epistemic, neo-Copenhagen interpretation. The common denominator among these is that they are free of the spookiness [e.g. nonlocality] and pseudo-problems [e.g. the measurement problem] which plague the psi-ontic interpretations.

Another way to look at it is that the psi-epistemic, neo-Copenhagen interpretations are the ones that have, at least tacitly, recognised that QM is just probability theory applied to mechanics, rather than a bizarre new kind of mechanics.

Calin Pop said...

Sabine,
What's the difference, in your view, between QBism and the Copenhagen interpretation?
Thanks

Sabine Hossenfelder said...

Paul,

Nobody denies that the wave-function carries information and that, if you make a measurement, you update your information and so on. What psi-ontic interpretations do is they *add* something to that, in that the wave-function is in one way or the other more than just a carrier of knowledge, etc, whereas psi-epistemic interpretations basically say "that's it" or maybe just "I don't want to talk about this any longer."

Sabine Hossenfelder said...

Calin,

Roughly speaking QBism is more 'personal' in that it's relevant whose knowledge you are talking about, different observers have different knowledge and everyone tells their own story about what's happening. Besides this it's explicitly Bayesian which I don't think is a formulation that was originally brought into the Copenhagen interpretation (though it could have been).

Theophanes Raptis said...

http://www.urbandictionary.com/define.php?term=life%20sucks%20and%20then%20you%20die

https://www.youtube.com/watch?v=lGXT_icWZok

opamanfred said...

I agree with Matthew. Interpretations are interpretations. They should tell you something about the (supposed) ontological nature of the various objects and operations of the theory (wave function, measurement, etc). To calculate actual quantities, you've got the mathematical formalism, or rather formalisms because there are several of them. And they all give the same result, no matter what your favorite interpretation is. Actually, you could even do away with any interpretation and just "shut up and calculate". That's the truly minimal option, the common denominator of QM.

And by the way, "Spontaneous collapse" should not be in the list at all. It is not an interpretation, it is a different theory, a modification of QM. All the other options do not change one iota to the actual results of QM, whereas "Spontaneous collapse" does.

Sabine Hossenfelder said...

I agree that interpretations is a misnomer, but it's common to list collapse models among them. I have extensively complained elsewhere that we need a phenomenology of quantum foundations rather than talk about interpretations.

Having said that, these interpretations are not just interpretations of the same set of axioms, because if that was so we could all go home and be done with it. They actually claim there are different starting points to arrive at the same axioms. Pilot-wave is the best example. Same theory, differently formulated. Same with QBism, many worlds, histories, etc. The transactional interpretation otoh, I can't see how it reproduces the axioms that we know to work to begin with.

Paul Hayes said...

Okay, I see where you're coming from but I think it's a bit too general/vague: nobody would deny that psi "carries information" but nobody would deny that a Maxwell wave "carries information" either. And looking at it that way works well enough for pilot waves but not so well for many worlds which, famously, starts out with a "follow the maths" psi-ontic interpretation and struggles to recover - add - the Born rule.

The [Bayesian] subjectivity in 'the' CI isn't explicit and wasn't/isn't recognised by everyone, no, but it's there:

The EPR experiment pin-points the need for subjectivity in quantum probability; the same need in classical probability has been known and used since Bayes.

https://web.archive.org/web/20151117174141/http://www.mth.kcl.ac.uk/~streater/EPR.html

Matthew Rapaport said...

Thanks Dr. H. but I'm afraid I cannot help relate the ontological interpretation to calculation. I can only read the "popular books" on these interpretations but I have read a number of them. What struck me about "TI" is that the *ontology* she proposes accounts in a natural way for super-position, entanglement, action at distance, etc. All the others seem to struggle here. [Apparent] paradoxes remain paradoxes, but they are not paradoxes in the TI ontology but come out "natural"..

ppnl said...

Paul Hays,

I think the difference between ontic and epistemic is mostly a useless word game when it comes to quantum mechanics. On the limit knowledge exists as a state of matter and thus is ontological. Quantum mechanics combine the observer and the experiment in a way that makes a pigs breakfast of the difference between ontic and epistemic. The fact that we are conscious beings confuse the fact that knowledge is in the end an ontological state that exists as a thing in some sense.

If a human opens a box they will know if the cat is alive or dead.

If an unconscious robot opens the box then it will "know" in some sense if the cat is alive or dead as it will store the information in its memory. This will cause the cat to decohere. A human may not know if the cat is alive or dead but a decohered cat is indistinguishable from a cat that is in a particular state.

If a rock hits the box causing the box to pop open then the rock will interact with the cat thus decohering it. Again the physical state of the cat is stored as a physical state of the rock.

The physics is the same in all three cases but our consciousness creates the illusion that we have a special roll in the process. The quantum measurement paradox is an illusion created by our existence as conscious beings.

But then what is consciousness?



John MacDonald said...

I believe you made the right decision when you removed the cat's list.

Stuart said...

Niels Bohr never so the Schrodinger cat as a paradox since an observation made by the detector alone was sufficient to collapse the wave function and therefore kill the cat. I concur with his point of view.

Unknown said...

No interpretation of quantum mechanics is needed: Moshe Shapiro showed (J. Phys. A: Math. Theor. 41 (2008) 175303 (9pp) that quantum mechanics simply follows from canonical invariance. What is mysterious is that observations have the character of numbers, and that simple symmetries are present (translation through time, rotation, translation through space).

Arun said...

Whenever Schrodinger observed his cat as alive and dead, he also did and did not note down his observation.

Paul Hayes said...

ppnl,

I almost agree with that except that I don't know what consciousness is but don't think it matters anyway, and to see that there is a useful difference between ontic and epistemic I would just take your observation -

"On the limit knowledge exists as a state of matter and thus is ontological."

- a little further, to include the (quantum) model of the cat in the box together with the rules of QM that the human or robot - but probably not the rock - might be using. So from either a [psi-]epistemic perspective or that ontic meta-epistemic perspective it's a (quantum) description of the cat that 'decoheres' rather than the cat itself.

JimV said...

"But then what is consciousness?"

It is the robot's operating system (handling external inputs by referring them to underlying, unmonitored subroutines, and transmitting the results as external actions) when the robot does not know it is a robot.

(At last, a question here I can answer - at least, to my own satisfaction.)

akidbelle said...

Hi Sabine,

how can you make a box that isolates the cat completely? Think of neutrinos brushing each of its atoms permanently. Think also of its gravitation pull toward everything and the corresponding work.

Thanks,
J.

Sabine Hossenfelder said...

akidbelle,

Don't know what you mean. I'm not in the business of making boxes, other than drawing them around equations. Best,

B.

ppnl said...

akidbelle,

You can't. It is a practical impossibility. But that's ok because it is only a thought experiment. Nobody is ever going to put a cat in superposition.

But you can replace the cat with a large atom, protein, buckyball... up until you reach the limits of experimental practicality.

ppnl said...

Stuart,

You mean the detector inside the box? Why can't it be in superposition? We could for example imagine a detector as small as a molecule and things that size certainly can be in superposition. As you increase the size of the detector there is no reason that it cannot be in superposition. There are practical problems but this is a thought experiment.

Uncle Al said...

@Unknown "simple symmetries are present (translation through time, rotation, translation through space)"

Symmetries via homogeneity are simple. Rotation via isotropy is ugly when absent S_n symmetries, that being chirality. Our chiral universe is modeled by elegantly mirror-symmetric maths. "Berry curvature is an anti-symmetric second-rank tensor..." targets specific problems.

Non-classical gravitations are rigorously derived and empirically wrong. Repair the problem. Observe chiral anisotropic hadron challenges to “isotropic” spacetime. Theory says they cannot succeed! Do they succeed? (The cat is chiral – L-amino acids and D-sugars.)

akidbelle said...

Sabine, ppnl, thanks.

I mean that one must be very cautious with though experiment. Nature may be protected against full isolation; the main premise of the problem may be wrong.

As far as I know, the only place to isolate the cat is within a black hole. Then good luck to measuring its state...

Best,
J.

Paul Hayes said...

Sabine, have you lost my response to ppnl?

Unknown said...

A newspaper presents this
http://www.telegraph.co.uk/technology/google/google-doodle/10237347/Schrodingers-Cat-explained.html

Rob van Son (Not a physicist, just an amateur) said...

@aikibelle
"Nature may be protected against full isolation; the main premise of the problem may be wrong."

That is not how I understood the problem.

The aim of the box is to prevent the objects inside to exchange information (entanglement, photons, or particles) with the outside. So, for the duration of the experiment, there should be no EM or particle interactions that carry information about the inside state to the outside. Neutrinos or gravity most likely would not carry off information about the state of the cat. Photons would do carry information. Sound would too.

The practical problems of shielding a box carrying a cat in free fall in vacuum from all incoming and outgoing radiation are daunting. But I do not see a law of nature that prohibids it. Or am I missing something?


Phillip Helbig said...

"There are like 30 or so more interpretations that aren't on the list."

As Penrose said, there can be more interpretations than people who work in the field, since some can be in a mixed state, holding more than one interpretation at once.

Seriously, what do you think of Rovelli's relational interpretation? (Maybe Carlo himself will chime in here!)

ppnl said...

Paul Hayes,

Why not the rock?!?

Decoherence is decoherence is decoherence and all of it is indistinguishable from absolute wave collapse. It makes no difference to the physics if a human or a rock opens the box. Otherwise you could devise an experiment that could tell you what opened the box. In the absence of any such experiment people, robots and rocks are functionally equivalent. It does not matter what model any of these are using as the experiment will go the same regardless.

And yes my point is that consciousness seems to be irrelevant to the process. But what consciousness does do is deeply confuse us when we try to analyze what is happening. I only ask what consciousness was to highlight the deeply confusing nature of consciousness. It was not clear that consciousness was irrelevant to the founders of QM and many get confused by it still today. What is consciousness that it could so confuse us? An unconscious robot would seem to have an advantage over us.

ppnl said...

Rob,

Exactly so. The point of the cat is that there is no such protection built into QM to prevent superposition of cats. If you want such protection it seems you will have to add it as an extra. Cats are so hard to isolate anyway that for all practical purposes no protection is needed.