I The typical and the exceptional in physics

[Ken G]

On the general topic of realism, there are two ways to state what realism is in physics, one which is perfectly attuned to the goals of science, and the other, the more standard way, which I claim has nothing to do with science at all:
1) standard way: physics is the study of what is real, independent of our physics. Reality thus gives meaning to the notion of doing physics. (How would we ever know that? How does that help us do physics, when we can just do the physics anyway?)
2) workable way: physics is a tool that we use to decide what we will regard as real. Physics thus gives meaning to the notion of reality. (Here we have an operational meaning of real that is accessible and useful.)
Notice how the first is ontological, useless, and untestable, while the second is epistemological, useful, and is all about how we test our concept of reality constantly.

 

[zonde]

On the general topic of realism, there are two ways to state what realism is in physics, one which is perfectly attuned to the goals of science, and the other, the more standard way, which I claim has nothing to do with science at all:
1) standard way: physics is the study of what is real, independent of our physics. Reality thus gives meaning to the notion of doing physics. (How would we ever know that? How does that help us do physics, when we can just do the physics anyway?)
2) workable way: physics is a tool that we use to decide what we will regard as real. Physics thus gives meaning to the notion of reality. (Here we have an operational meaning of real that is accessible and useful.)
Notice how the first is ontological, useless, and untestable, while the second is epistemological, useful, and is all about how we test our concept of reality constantly.

I can propose non scientific test for the first statement: all valid descriptions of reality can be joined in one consistent system.
So can you justify requirement that descriptions should be mutually consistent without claiming that there is reality?

 

[Ken G]

I can propose non scientific test for the first statement: all valid descriptions of reality can be joined in one consistent system.

You have proposed a test, but you have not supplied evidence that the test is ever passed. Isn't that a problem-- a test that is not passed?

So can you justify requirement that descriptions should be mutually consistent without claiming that there is reality?

Yes, I take the epistemological approach of simply asserting that I seek mutually consistent descriptions. Notice how easily I handle the failure to achieve the goal, it is simply a goal whether I achieve it or not!

 

[zonde]

You have proposed a test, but you have not supplied evidence that the test is ever passed. Isn't that a problem-- a test that is not passed?

Pilot wave theory consistently unifies particle and wave descriptions.

Yes, I take the epistemological approach of simply asserting that I seek mutually consistent descriptions. Notice how easily I handle the failure to achieve the goal, it is simply a goal whether I achieve it or not!

You haven't provided justification for that assertion. And the ease with which you give up the goal I see as a drawback of your approach.

 

[ddd123]

So how we use information is helped by imagining that the information is "about" something, I've no problem with that. But it only matters to how we process and use that information, what we think the information is "about" is not, itself, information, it's more like a kind of crutch that supports our information processing without adding anything to that information. It's how we think, so it's just more epistemology, disguised as what the epistemology is "about."

If we want to get this philosophical, we might as well do it right. You are confusing information and epistemology: what is beyond information (which you assert, for science, is everything) is semantics (in the actual sense, not the usual ironic figure of speech). Science needs information and semantics, at a minimum. That is, the meaning of language, and its understanding. If you flatten language to the abstract material of information you lose its meaning so you lose language itself (think about Searle's Chinese room): at best, there are philosophical theories that do away with meaning by positing that language is exhausted by its grammar, or structure (so in that case it's not the information that is fundamental but its structure, at best). But I think those are a little too outlandish.

I think I understand, you are trying to generate a working ontology.

No, I was trying to generate a working epistemology, but again, I'm not sure about it.

 

[stevendaryl]

I'm not sure whether the irony is intentional, or not, but proclamations about what science is and is not is philosophy, rather than science.

I've pointed out before (in a different thread) that some of the greatest advances in physics were not from people trying to get more accurate predictions for a wider range of experiments, but from people trying to understand and address conceptual problems in theories that already existed. Einstein's General Relativity was not motivated by the precession of planetary orbits; it was motivated by Einstein's attempt to reconcile Special Relativity with Newtonian gravity. Dirac's equation of the electron was motivated by his attempt to reconcile quantum mechanics and relativity. Maxwell's equations were an attempt to unite the various empirical laws governing electromagnetism, including Gauss' law, Faraday's law and Ampere's law. Maxwell's biggest original contribution was introducing the "displacement current", and that was motivated by conceptual issues, not by experiment.

It appears to me that the most important advances in physics have always been by people doing what a lot quantum philosophers say shouldn't count as science.

Having said that, I do think that the conceptual issues with quantum mechanics are particularly difficult to make any progress on.

 

[stevendaryl]

I mostly agree with the previous post, and this somewhat less. There is no need to mention physicists, only how they prepare states and what the probable states of the measuring apparata will be.

If the same scenario happened accidentally we would expect the same distribution of outcomes, surely ?

(Or replace the Physicist by an equivalence class ...)

Well, I would certainly be more comfortable with quantum mechanics if it could be formulated without mentioning "preparation" and "measurement". Surely, on a star billions of miles from any humans, nuclear fusion works perfectly fine without anybody preparing anything, and without anybody measuring anything. The standard minimalist interpretation of quantum mechanics would seem to say that it requires a human looking at the star before nuclear fusion in the star has any meaning.

 

[vanhees71]

Well, I would certainly be more comfortable with quantum mechanics if it could be formulated without mentioning "preparation" and "measurement". Surely, on a star billions of miles from any humans, nuclear fusion works perfectly fine without anybody preparing anything, and without anybody measuring anything. The standard minimalist interpretation of quantum mechanics would seem to say that it requires a human looking at the star before nuclear fusion in the star has any meaning.

This is nonsense. In QT nothing, really nothing, depends on whether a human being is looking at something. Nature doesn't care about humans very much.

 

[stevendaryl]

This is nonsense. In QT nothing, really nothing, depends on whether a human being is looking at something. Nature doesn't care about humans very much.

I agree that it is nonsense to believe that physics depends on human observers. But it seems to be a consequence of the "minimalist interpretation" in terms of "preparation" and "measurement". So that's a problem with the minimalist interpretation, in my view.

 

[ddd123]

Don't call it nonsense when you're agreeing with me. I agree that physics doesn't depend on humans. But the formulation of the minimalist interpretation in terms of "preparation procedures" and "measurements" is not appropriate for physics without humans. So the minimalist interpretation is not adequate.

Asher Peres addresses this in his book:

Real life seldom follows the idealized preparation-observation pattern presented throughout this book. Astronomers, for instance, observe spectral lines (i.e., detect photons) which they interpret as due to the presence of atoms or molecules in interstellar space. Obviously, the atoms were there a long time ago in an excited state; they decayed to their ground state, emitting photons which we can now observe, considerably later. These excited atoms were not prepared by us, nor our research assistants. We can only observe them passively. We also observe bigger objects, such as the Moon moving around the Earth, or various planets, without ever having prepared them.
This would cause no conceptual difficulty with quantum theory if the Moon, the planets, the interstellar atoms, etc., had a well defined state ρ. However, I have insisted throughout this book that ρ is not a property of an individual system, but represents the procedure for preparing an ensemble of such systems. How shall we describe situations that have no preparer? […] why should we expect A and B to agree that there is, objectively, a star somewhere in the sky? The reason is that any macroscopic object, such as a star, involves an enormous number of identical subsystems with almost identical properties, in particular identical positions, within the accuracy of our instruments. Thus, a macroscopic object effectively is assembly, which mimics, with a good approximation, a statistical ensemble. Measurements performed on such an assembly have a huge redundancy. In particular, different apparatuses can be used for probing disjoint subassemblies, each one of which is large enough to mimic an infinite ensemble. We can thereby measure, with little dispersion, the expectation values of noncommuting operators.
You must have noted the difference between the present pragmatic approach and the dogmas held in the early chapters of this book. It was then asserted that any operator which can be written by a theorist can also be measured in the laboratory. This fiction was needed in order to establish a formal framework for quantum theory. Now, our goal is different: we want to use a classical language for describing, with a good approximation, macroscopic phenomena.

 

[vanhees71]

Yes, read Peres's book. That's the best to prevent one from getting into these esoterics by trying to provide ontology from science. That's the realm of religion/philosophy, not science!

 

[ddd123]

I'm not sure whether the irony is intentional, or not, but proclamations about what science is and is not is philosophy, rather than science.

Yes it was intentional, but there's not as much irony as you think. When, for example, we talk about falsification, or even more radically, like in this thread, about ontic vs. epistemic interpretations, we are using concepts directly borrowed from philosophy which come with a baggage that isn't necessarily carried along with them properly across the border. For example I see a lot of use of the word "positivist" to refer to instrumentalist interpretations… this is kind of a contradiction from a philosophical point of view, but I don’t want to clutter by explaining why. The point is: to bring all this confusion to an end, either scientists must become also philosophers, or they must find a purely scientific meaning and justification for terms like “ontic”, “epistemic”, “positivist”, “falsification”… Neither of these seem very likely to me so we must accept the confusion and wait for, say, a theory of quantum gravity to solve these issues (that is, get back to science, and put these horrible matters aside for the moment).

 

[stevendaryl]

What I think a sensible type of physical theory looks like is something along the lines of:

In situation such-and-such, such-and-such happens.​

Then we make an experiment out of such a theory by adding stuff along the lines of:

• If you do such-and-such, you can put a system into such-and-such a situation.
• If such and such happens, it will affect our measuring devices in such-and-such a way.

Then the testable claim "If I prepare a system like this, I will get a measurement like that" follows from our theories.

 

[Ken G]

If we want to get this philosophical, we might as well do it right. You are confusing information and epistemology: what is beyond information (which you assert, for science, is everything) is semantics (in the actual sense, not the usual ironic figure of speech). Science needs information and semantics, at a minimum.

I completely agree that science requires semantics, but I don't agree with the association of ontology with semantics. Meaning is modeling, yes, but ontology is not modeling-- ontology is believing the model. That's my whole point here. When the ancient Greeks pictured the universe as geocentric, that was a model. If they could have just said "we choose to model the universe as geocentric for reasons X,Y and Z", that is 100% pure epistemology, and it would have given the world zero problems when Galileo came along and said "actually, a heliocentric model does better with observations A, B and C." That's science. But no, Galileo had to recant his claims, because of ontology, which is never scientific. Ontology is the statement that the universe either really is geocentric, or it really is heliocentric, which turns "geocentric" and "heliocentric" into "geocentrism" and "heliocentrism." But the shift from "ic" to "ism" is a category error in scientific thinking, because geocentric and heliocentric are attributes of models, not attributes of universes. Science models, and judges models-- that means science uses "ics" not "isms." Epistemology, not ontology.

It's the same with determinism, reductionism, localism, even realism. None of those are scientific, because none of them involve the making and testing of models, they involve making untestable claims and falling into the category error of mistaking the attributes of models for the attributes of universes. It's not ontology until you choose to believe it, and the scientist never needs to make any such choice, they only need to test and judge models. Look how much easier that would have made the whole Galileo business!

No, I was trying to generate a working epistemology, but again, I'm not sure about it.

If you are trying to generate an epistemology, then you are agreeing with me-- an epistemology is all about modes of creating expectations toward outcomes, and then testing those expectations. Notice the crucial role of thought in all that. An ontology is a claim about what is, independent of thought-- so it's not at all about expectations and modes of thinking to get some result. I certainly agree that we use pictures to help us process information-- it's not ontology until you claim your picture is what is, when the "ic" of a model, a way of thinking, becomes the "ism" of what is, independent of thought.

 

[stevendaryl]

Yes, read Peres's book. That's the best to prevent one from getting into these esoterics by trying to provide ontology from science. That's the realm of religion/philosophy, not science!

Well, I disagree with you about what science is. But arguing about what science is is philosophy.

 

[Ken G]

Pilot wave theory consistently unifies particle and wave descriptions.

That isn't a test of your claim-- your claim was "all valid descriptions of reality can be joined in one consistent system", not "some aspects of a valid description of reality can be joined in one consistent system." Pilot wave theory does not unify quantum mechanics and gravity, and it's not even clear it unifies quantum mechanics with special relativity, so it does not pass your test there.

You haven't provided justification for that assertion. And the ease with which you give up the goal I see as a drawback of your approach.

My justification is the physics has never, in its entire history, presented a fully unified and self-consistent description of reality. There has always been elements missing from the o ntology, so there has always been the need to pick and choose the theory to suit the situation. If physics has always been a certain way, it is odd to imagine that what it actually does is something different from what it has ever actually done! That doesn't mean giving up the goal, it means recognizing that a goal is like a direction, like walking East, not a destination, like arriving at East. There is no more unscientific attitude than the proclamation "we have arrived at the true ontology of the universe," that's what Cardinal Bellamy said. The scientist is always skeptical, always digging deeper, because science doesn't believe in ontology.

 

[Ken G]

Then the testable claim "If I prepare a system like this, I will get a measurement like that" follows from our theories.

Yes, it seems to me that if we look at the scientific method, nothing could be more clear than that it is an epistemology. It is a system that goes "model, test, repeat." There is never any step in the scientific method that says "now believe that your model is the reality." Not only is that step absent from science, it is anathema to the basic skepticism that is so essential to scientific progress. Steps like that give you the dark ages, and reliance on authority instead of questioning. So I don't see how anyone can think that science invokes ontology, unless one defines ontology as "picture something that the information is about to breathe meaning into the information", which is not what I mean by ontology-- what we picture in our minds is just more epistemology, and allows us to see all interpretations of QT as pictures. Ontology means arguing over which interpretation is the real one, overlooking that even QT will likely be someday replaced. The attributes of models is all the scientist can ever test, and nothing adjudicates those tests other than more observations-- more information. That's epistemology, and it makes the debate go away-- the Bohmian is someone who says "I like to picture a pilot wave when I apply QT", a Copenhagen follower says "I like to picture a quasi-mystical collapse when I do it." When one strips away the need to be saying something about what actually is, and instead can look at it as a kind of mindset, no problems ensue. We are not trying to figure out reality, which by its nature needs to be unique, we are trying to figure out a good way to think about reality, which by its nature does not need to be unique.

 

[ddd123]

I completely agree that science requires semantics, but I don't agree with the association of ontology with semantics.

I never said that.

If you are trying to generate an epistemology, then you are agreeing with me

I was trying to generate an ontology-free epistemology, and explained why it "felt" somewhat unnatural (but then again, block universe has just that same problem so I guess it's not that big of a deal).

 

[vanhees71]

Well, I disagree with you about what science is. But arguing about what science is is philosophy.

This whole thread is not about science but philosophy!

 

[ddd123]

Anyway, to get the bundle back together, @Ken G:

1) why is it, then, that the moment we could picture a fully consistent stable backstory of what happens in between preparation and measurement, we would all instantly embrace it? We could just keep with the computational map from preparation and measurement. So there must be something to that idea.
2) by dropping that idea, you are also dropping attempts to make it work, and that eventually may be a limitation. We could end up overlooking something just because we didn't follow our physical intuition, as stevendaryl said: all you say is formally true, but science doesn't always progress by rigid schemes, it has schizophrenic shifting in between conservation and revolution, out-of-the-blue leaps of intuition and so on.

This is along the lines of Feyerabend's critique of Popper, if you're interested, that he tried to put constraints on science that there's no evidence are needed.

 

[Ken G]

This came up because it seemed to me that the thermal interpretation was based in the epistemology of the scientific method, where one makes observations on systems and tracks standard deviations and how the standard deviations can be beaten down by making correlations with other information that is available. I'm fine with all that, I think the thermal approach is very solid thinking about how we process information as we do quantum mechanics. I was objecting to the effort to turn it into an ontology, in this case by claiming that what we perceive as a standard deviation between different outcomes in an ensemble of preparations represents measurement error. It seems to make collapse go away by saying it's all a kind of mistake, so needn't be worried about, but it doesn't explain where the mistake comes from. But that doesn't make the thermal interpretation any worse than any other, because they all misstep equally, in my view, when they are regarded as ontologies-- statements of how reality is independently of our preferences of thought. If we just accept that all we test is QT itself, and the interpretations are going to be nonunique ways that different minds will prefer to generate pictures, none of which produces a "reality" because that would be a category error, then all these problems go away. The problems go away if we simply distinguish theories, which the scientist tests, from pictures, which the scientist chooses by preference, from some independent reality, which never appears in the scientific method at all.

 

[Ken G]

1) why is it, then, that the moment we could picture a fully consistent stable backstory of what happens in between preparation and measurement, we would all instantly embrace it?

It depends on what you mean by "embrace." One can embrace a model because it tests out well. One can even embrace a model knowing it doesn't work in all situations, we do that all the time. But one does not need to "embrace" in the sense of an ontology! Here's an example of what I mean. I often invoke the concept of a force of gravity in doing calculations, even though I am personally quite skeptical that there is any such thing as a force of gravity in some independent reality (indeed, I'm skeptical that the concept of an independent reality is a coherent notion in the first place, but that's not my point here). So how can I do that? Why am I not crippled in solving equations that deal with gravity as a force if I don't believe that gravity really is a force? It's because ontology isn't anything important in science, we don't use a belief in what is real-- we invoke pictures, or don't invoke them, like putting on gloves if it's cold outside or not if it isn't.

We could just keep with the computational map from preparation and measurement. So there must be something to that idea.

I am not disputing that part of the scientific epistemology is to invoke pictures, the "backstory" to which you speak. This helps us give meaning to the information, the semantics. But that's still not a claim on what is, it's a claim on how we like to think. The problem is when we get lazy and fail to notice any more that these are modes of thinking, not statements of what is. For example, when I say that a hydrogen atom is composed of a proton with an electron orbiting it, I have invoked a picture, a backstory. I could just solve the equations, but no one ever really "shuts up and calculates", we just don't, we need to have our pictures to organize our thinking. But what we really mean is, "the equations I am solving can be motivated if we picture the hydrogen atom as if it were a proton being orbited by an electron", The hypothetical is what turns ontology into epistemology, and the desire to test the hypothetical is what makes it science. We never need to think we are testing if a hydrogen atom really is a proton orbited by an electron, because we don't get to know that anyway, we are testing the theory that invokes that picture, we are testing that way of thinking. That's all we ever do. That's why I say the goal of science is not to understand reality, the goal of science is to improve our ways of thinking, to get better more accurate and more unified results. That's just our goal-- we don't need any other reason for it, but since our goal is to make sense, we can separate the objective aspects of doing tests from the subjective aspects of choosing pictures we prefer. When it isn't ontology, there just isn't any problem with that separation.

2) by dropping that idea, you are also dropping attempts to make it work, and that eventually may be a limitation. We could end up overlooking something just because we didn't follow our physical intuition, as stevendaryl said: all you say is formally true, but science doesn't always progress by rigid schemes, it has schizophrenic shifting in between conservation and revolution, out-of-the-blue leaps of intuition and so on.

I completely agree with all of that, what I reject is that ontology played any role in any of it. It was all an eloquent description of how scientific epistemology works. We seek a consistent backstory because that's how we think, that is the goal we choose for our efforts, not because we expect the backstory to be the reality. Recall the infamous words of Lord Kelvin in 1900: "There is nothing new to be discovered in physics now. All that remains is more and more precise measurement." The question to ask is not why does the scientist seek a consistent backstory, the question is, having found such a thing, why is it true that any scientist worth their salt should immediately set out to find problems with it?

This is along the lines of Feyerabend's critique of Popper, if you're interested, that he tried to put constraints on science that there's no evidence are needed.

I'd have to see Feyerabend's argument, because I thought Popper did a nice job of showing why those constraints are indeed needed. Science all too easily turns into ontology, at which point it becomes too close to dogma.

 

[atyy]

Pilot wave theory does not unify quantum mechanics and gravity, and it's not even clear it unifies quantum mechanics with special relativity, so it does not pass your test there.

The standard model of particle physics does not unify quantum theory with relativity, because of the Landau pole.

 

[atyy]

This is nonsense. In QT nothing, really nothing, depends on whether a human being is looking at something. Nature doesn't care about humans very much.

But if you take the viewpoint that the wave function is epistemic, then quantum theory does depend on epistemic agents (a fancy way generalization of "human beings").

 

[atyy]

Yes, read Peres's book. That's the best to prevent one from getting into these esoterics by trying to provide ontology from science. That's the realm of religion/philosophy, not science!

In fact Peres, like Ballentine, does talk about ontology in a misleading way, eg. p426 of the Peres book:

"It is this reduction of our resolving power which allows the emergence of an objective reality, even if only a fuzzy one."

At this point Peres is introducing collapse, but instead of stating his assumptions clearly, he obscures it with a sleight of hand.

 

[zonde]

My justification is the physics has never, in its entire history, presented a fully unified and self-consistent description of reality.

This was not my point. My point was more modest, that when two theories are mutually inconsistent at least one of them is wrong.

 

[vanhees71]

But if you take the viewpoint that the wave function is epistemic, then quantum theory does depend on epistemic agents (a fancy way generalization of "human beings").

Quantum theory, as all of physics, is epistemic. It's about what we can observe in nature and a quantitative description of what we observe.

In fact Peres, like Ballentine, does talk about ontology in a misleading way, eg. p426 of the Peres book:

"It is this reduction of our resolving power which allows the emergence of an objective reality, even if only a fuzzy one."

At this point Peres is introducing collapse, but instead of stating his assumptions clearly, he obscures it with a sleight of hand.

Well, it's your personal opinion that the scientific core of quantum theory, as opposed to possible metaphysical or religious implications, is misleading. That's the freedom of personal belief. The claim that Peres introduces collapse, is however a distortion of his entire book ;-).

 

[secur]

... arguing about what science is is philosophy.

That's right. In fact arguing about what anything "is", is philosophy.

We are not trying to figure out reality ...

Actually I am trying to figure out reality - but not qua scientist. Indeed the only reason I'm interested in science - these days, since I no longer need a job - is to help me understand what reality "really is". That's the job of philosophy, which uses scientific facts as raw material.

An overall problem with much of this discussion is that philosophy is often considered useless, even bad. The statement "it's philosophy" is taken to mean "it's just philosophy" or "mere philosophy", not worthy of an intelligent person. People are trapped by this incorrect syllogism:

Premises:

1) Only scientific reasoning is valid.
2) But I want to know what reality is.

Conclusion: Therefore the ontological question - what is real - must be scientific, or I can't address it.

Actually both science and philosophy are valid. Ontology is fine, even though it's not science. IMHO.

I confess I can't quite get my head around the viewpoint that the wavefunction is merely descriptive of our 'state of knowledge' ...

There's no problem if one also says that science's idea of a rock (for instance) is "merely descriptive of our 'state of knowledge'". Paradox occurs only when you say a rock is ontologically real but a wavefunction (or at least, some aspect of QM) isn't. Something real can't be composed of a bunch of things that aren't real! The paradox is equally avoided if you suppose they're both real, but qua scientist it's best to have no opinion on the issue, which is the province of philosophy.

 

[atyy]

Quantum theory, as all of physics, is epistemic. It's about what we can observe in nature and a quantitative description of what we observe.

But if quantum theory is about "what we can observe", then quantum theory does depend on whether a human being is looking at something or not.

Well, it's your personal opinion that the scientific core of quantum theory, as opposed to possible metaphysical or religious implications, is misleading. That's the freedom of personal belief. The claim that Peres introduces collapse, is however a distortion of his entire book ;-).

Unfortunately you are wrong about the scientific core of quantum theory. Quantum theory is not about ontology, but it is about "what we can observe".

 

[vanhees71]

Now everything gets mixed up. Of course quantum theory is there because humans have discovered it as a description of what's observed but not more, particularly it doesn't provide ontology (neither does classical physics). The observables are defined operationally by real-world measurement processes. Whether even a classical abstraction as the electromagnetic field has an ontic meaning, is completely outside of science. It is a mathematical description that describes successfully what we observe in the corresponding realm of nature, including phenomena from electrostatic and magnetic forces to light (electromagnetic waves). Whether the field is ontic or not, cannot be clarified in any way but is a matter of personal world view. What can be stated scientifically is that it's a mathematical picture which fits to all observations.