The Philosophy of Quantum Mechanics

[A. Neumaier]

TL;DR Summary

an interesting book on the foundations of quantum physics

I'd like to point to the book The Philosophy of Quantum Mechanics by C. Friebe et al., Springer 2018. It contains many topics usually underrepresented in foundational discussions of quantum physics, in chapters on many-particle systems and quantum field theory. It also has in its last chapter a nice timeline of the development of quantum physics,
as far as its foundational aspects are concerned.

 

[Jehannum]

Sounds interesting. Shame that the philosophy of physics gets such a frosty reception here.

 

[A. Neumaier]

the philosophy of physics gets such a frosty reception here.

Only for a certain type of contribution and/or by some contributors. If the contribution quality is high, quite a number of readers appreciate this topic.

 

[PeterDonis]

Shame that the philosophy of physics gets such a frosty reception here.

To the extent that "philosophy of physics" means "asking questions that cannot be resolved by experimental tests", it tends to lead to discussions that never get resolved, because in the absence of any way of resolving the question by experiment, it's just everybody stating their own opinions and complaining whenever anyone disagrees with them. Such discussions don't really contribute anything to PF's mission of helping people to understand mainstream science.

To the extent that "philosophy of physics" means "clarifying what physics actually says and how what it says relates to experimental tests", it tends to help people understand mainstream science, which is fine.

The problem with the label "philosophy of physics" is that it's often unclear which of the two cases above is being referred to.

 

[ftr]

To the extent that "philosophy of physics" means "clarifying what physics actually says and how what it says relates to experimental tests", it tends to help people understand mainstream science, which is fine.

I think the problem seem to be even deeper and more complicated where the experiments themselves are very hard to explain because it becomes chicken and egg( i.e. is QM suppose to explain the experiment or the experiment suppose to explain QM, that is not clear). Since it is not clear what QM says so the experiment is hard to interpret. Like blackening the plates in double slit, the energy measurement resolution as already discussed in TI thread and so on. Not to mention the conceptual divide between classical instruments and the micro world.

 

[PeterDonis]

it is not clear what QM says

The math of QM, and the predictions it makes, are perfectly clear and easily compared with experimental results; so far the predictions match the results.

If anything is "not clear" about QM, it therefore does not affect our ability to use QM to make accurate predictions.

 

[ftr]

Well, The thread is about "philosophy" which I interpret as we don't know something about it, And history is bound by many examples that experiments could predict but there was a need to know more as to why, and that is how science progresses. I was just highlighting some of the problems of the relation between the theory and the experiment in QM. Of course your point is well understood, whether that is enough or not that is the question!

 

[vanhees71]

In my opinion, philosophy of science in general and of QT in particular are simply not (or in my opinion should not) be the topic discussed in a physics forum. Of course, as any science and humanities, it has interesting aspects and it may even help in some cases to understand the physics better.

Unfortunately, that's definitely not the case in popular-science treatments of the underlying physics nor in this forum. Though the sciencific part of QT is very clear and a great success (there's no single contradiction between experiment and QT today but to the contrary the QT analysis of all experiments and observabtions leads to amazing agreement between this analysis and the outcome of the real-world experiments and observations), the issue gets often spoiled by starting philosophical debates about apparent problems with the "ontology" of QT.

That's why I think, such debates should be shifted to a subforum to distinguish them clearly from the scientific issues of QT not to confuse people who want to learn the science first, before starting to discuss philosophical issues, which reside in the mud of unsharp definitions and the lack of mathematical analyzability.

The best discussion about these interpretational problems, I've seen so far, is in Weinberg, Lectures on Quantum Theory, and I'm saying this although I don't agree with Weinberg on the conclusion that there are still open interpretational problems concerning the scientific part. From the philosophcial point of view there may be problems, but they are not part of science but of humanities. For me QT is satisfactorily interpreted by the minimal statistical interpretation, and from a scientific point of view, one simply has to accept the Born rule (in its most general form applied to mixed as well as pure states) as one of the basic postulates of QT. As Weinberg convincingly shows there seems not to be a way to derive it from the other postulates and thus it seems to be independent.

It's a bit analogous to the debate about Euclidean and non-Euclidean geometry in the 19th century: The axiom of parallels in Euclidean geometry was under debate, because it didn't seem to be so "self-evident" from intuition as the other axioms. Contrary to the situation with QT, this however has lead to very important and fruitful mathematical developments, namely the discovery of non-Euclidean geometry by proving that the parallel axiom may be substituted by other axioms leading to non-Euclidean geometries that are consistent as Euclidean geometry is consistent.

Maybe one day, such investigations of the Born rule also lead to an even better and more complete theories than QT is today. However, I'm very sure that it won't come from some murky philosophical speculations but from the usual hard theoretical and experimental work of physicists, finally solidified by some clearly observable and quantifiable phenomena in the lab. Just inventing some new "interpretation", tailored such to lead to the same physical predictions as standard minimal interpreted QT, won't lead to any progress, already by the very choice of methodology.

 

[Lord Jestocost]

For me QT is satisfactorily interpreted by the minimal statistical interpretation......

On going from the instrumentalist minimal interpretation to the minimal statistical interpretation, isn’t this the first metaphysical step?

 

[vanhees71]

I've no clue what you mean. The minimal statistical interpretation is the minimal statistical interpretation, and it's of course instrumentalist since physics is instrumentalist by definition.

 

[Lord Jestocost]

Come on! You infuse the operational formalism of quantum theory with an ontological statement: that it is related to ensembles of systems.

 

[vanhees71]

In physicists' language: To verify a probabilistic statement you have to measure on a sufficiently large ensemble to verify the statement with a given statistical significance, i.e., it's verified on an ensemble of equally prepared systems. There's nothing ontological or epistemic, it's simply the math of probabilities.

 

[A. Neumaier]

In physicists' language: To verify a probabilistic statement you have to measure on a sufficiently large ensemble to verify the statement with a given statistical significance, i.e., it's verified on an ensemble of equally prepared systems. There's nothing ontological or epistemic, it's simply the math of probabilities.

No. The math of probabilities is silent about its meaning. Probability calculus can even be applied to study prime numbers, which cannot be prepared but are given once and for all.

What you refer to is frequentist statistics, which is an ontological interpretation - compared to Bayesian statistics, which is epistemic but has no concept of statistical significance.

You are simply blind to your own philosophical commitments, and cannot accept that others have different such commitments.

 

[vanhees71]

Well, so far nobody has ever explained to me, how you empirically check probabilistic statements other than using the frequentist interpretation. It's of course not pure math but application of math to real-world problems. Indeed, I don't care about which "ism" I might use.

 

[Auto-Didact]

A common misunderstanding by physicists about the philosophy of physics as a subject or discipline is that physics as a discipline is not so much about what the experiments themselves say, but how practicing physicists within some time period interpret both theory and experiment within some methodology; this leads to an orthodox or 'textbook view' and one or more (not necessarily concordant) expert views on the matter as described by the literature at some point in time.

From the rest of science - which is nowhere near as explicitly mathematicized as physics and where 'natural history' tends to play a dominant role - it has become clear that within scientific practice there actually is no such thing as an uninterpreted theory; there is always an aspect of interpretation involved. In fact, even the research reports of so called pure observations/experiments tend to be interpretation laden, since some kind of theory decides at some level the methodology and therefore what is reported.

The question is therefore not if experimental data is uninterpreted, but whether or not this interpretation represents a breakaway from tradition for that discipline or not. Such breakaways becomes instantly more obvious once a physical theory is viewed as a specific mathematical model within the family of mathematical models commonly referred to as 'canonical physics'; there is a traditional methodology from mathematics for how to do this.

For any physical theory, insisting on having a purely statistical interpretation of said theory and nothing more is evidently a breakaway from the discipline of physics since its inception, because this is a very definite statement about the mathematical nature of physical theory itself i.e. that physics de facto is at bottom statistics and nothing more. Such a purely statistical interpretation of physics is clearly at odds with the rest of the mathematical characterization of physics, which has always had deep relations to analysis, geometry and so on.

In fact, such a purely statistical characterization of what physical theory can be, i.e. what physics is, is explicitly not a consequence-free statement in the realm of mathematics; namely, if physics de facto is just statistics, then this directly implies that much of analysis, geometry and so on are all also de facto reducible to statistics; this is a very striking conjecture which is either true or false. It clearly seems to be false or better said, if one claims that because within mathematics subjects can be treated statistically in an epistemic sense therefore they are reducible to statistics seems to be at best a blatantly absurd misunderstanding of mathematics.

 

[Lord Jestocost]

In physicists' language: To verify a probabilistic statement you have to measure on a sufficiently large ensemble to verify the statement with a given statistical significance, i.e., it's verified on an ensemble of equally prepared systems. There's nothing ontological or epistemic, it's simply the math of probabilities.

In case you are thinking about experimental results, any interpretation of quantum theory might be labeled “statistical”. However, does a probabilistic theory (which you verify by measurements on an ensemble) have to be per se about ensembles?

“The deeper reason for the circumstance that the wave function cannot correspond to any statistical collective lies in the fact that the concept of the wave function belongs to the potentially possible (to experiments not yet performed), while the concept of the statistical collective belongs to the accomplished (to the results of experiments already carried out).”

V. A. Fock, “ON THE INTERPRETATION OF QUANTUM MECHANICS”, Czech J Phys (1957) 7: 643

 

[microsansfil]

What you refer to is frequentist statistics, which is an ontological interpretation - compared to Bayesian statistics, which is epistemic but has no concept of statistical significance.

These are not only philosophical points of view, but from these interpretations come different methods of inferences.

For example, the debate that took place during the discovery of the Higgs Bosons is about the difference between frequentist and Bayesian inferences: http://tonyohagan.co.uk/academic/pdf/HiggsBoson.pdf

/Patrick

 

[OCR]

.
"I’m not but I wouldn’t expect any such call for a reconsideration of the basic principles would be popular until it has results which make it hard to avoid thinking about."

Lee Smolin
Here's a good Guest Post at Backreaction. . . well, IMO anyway.
.

 

[Auto-Didact]

A common misunderstanding by physicists about the philosophy of physics as a subject or discipline is that physics as a discipline is not so much about what the experiments themselves say, but how practicing physicists within some time period interpret both theory and experiment within some methodology; this leads to an orthodox or 'textbook view' and one or more (not necessarily concordant) expert views on the matter as described by the literature at some point in time.

This was worded rather poorly: I meant to say that most physicists wrongly view physics as a discipline which tells us what experiment says, instead of as a codified interpretation of theory and experiment of practicing physicists within the context of some methodology within some time frame.

 

[haushofer]

This was worded rather poorly: I meant to say that most physicists wrongly view physics as a discipline which tells us what experiment says, instead of as a codified interpretation of theory and experiment of practicing physicists within the context of some methodology within some time frame.

And only then you can distinguish between what you call "physics" and "philosophy of physics".

I've never understood how people make that distinction. Your post explains imo why that distinction is so superficial.

 

[vanhees71]

In case you are thinking about experimental results, any interpretation of quantum theory might be labeled “statistical”. However, does a probabilistic theory (which you verify by measurements on an ensemble) have to be per se about ensembles?

“The deeper reason for the circumstance that the wave function cannot correspond to any statistical collective lies in the fact that the concept of the wave function belongs to the potentially possible (to experiments not yet performed), while the concept of the statistical collective belongs to the accomplished (to the results of experiments already carried out).”

V. A. Fock, “ON THE INTERPRETATION OF QUANTUM MECHANICS”, Czech J Phys (1957) 7: 643

This is my question to all people who claim otherwise. How to verify probabilistic predictions if not through some statistics on some ensemble. This needs not to be something like throwing many dies you can also throw one die several times. Also it can be about the statistics of a large many-body system by coarse-graining, i.e., averaging over global observables (e.g., the center of mass (or momentum in the relativistic case) of a macroscopic body, the total energy and conserved charges of a gas) or local coarse-grained observables (like single-particle densities phase space in Boltzmann transport theory or over spatial fluid cells in hydrodynamics etc. etc.).

Sometimes you meet socalled "Bayesianists" who claim that probabilities have some meaning, but none of those could answer my question, how to test probabilistic predictions in experiment by just looking at one single event. This has nothing to do with quantum theory per se but more generally for all probabilistic statements.

We all know, why Fock had to write his philosophical essays on the foundations of quantum theory. Another example is in the appendix of the famous QM book by Blokhinsev...

 

[vanhees71]

This was worded rather poorly: I meant to say that most physicists wrongly view physics as a discipline which tells us what experiment says, instead of as a codified interpretation of theory and experiment of practicing physicists within the context of some methodology within some time frame.

This is one example for utter nonsense non-scientists think to understand about how science works. It's not some phantasy of some community but it's about objective empirical and theoretical work to learn, how nature works, leading to knowledge about nature which is independent of all kinds of sociological factors like religion or philosophical prejudices.

 

[stevendaryl]

Sometimes you meet socalled "Bayesianists" who claim that probabilities have some meaning, but none of those could answer my question, how to test probabilistic predictions in experiment by just looking at one single event. This has nothing to do with quantum theory per se but more generally for all probabilistic statements.

You can't verify a probabilistic prediction with a single event. But you also can't verify it with a million events. The most you can do is to convince yourself that it's very likely to be true.

 

[Dale]

Temporarily closed for moderation

Edit: this will remain closed. It is too philosophical even for the QM forum.