Ontology is to quantum theory what hardware is to computation theory

[Demystifier]

There is no useful way to manifest it

So do they manifest it in an unuseful way? If yes, what way is it?

 

[PeterDonis]

So do they manifest it in an unuseful way? If yes, what way is it?

Why do they have to "manifest" it?

 

[martinbn]

Your view (correct me if I misrepresent it) is that real stuff exists and is "nonlocal" in the Bell sense, which is not in contradiction with the fact that physics is also "local" in another sense. In that sense you and me agree, and it's not at all obvious what do we really disagree about. With intention to clarify the source of our disagreement, let me ask you one additional question. In your view, does this violation of locality in the Bell sense imply a violation of some sort of Lorentz invariance?

I expressed my opinion in the very first post i wrote.

No, i think it doesnt imply Lorentz invariance violation.

 

[vanhees71]

Yes, in that sense they are complete. But there is also another sense in which they are not complete. You may call it metaphysical, but whatever you call it, many physicists think that this is important and interesting too.

You never say, what you consider incomplete. That's also typical for philosophical discussions. You keep it nebulous enough just to never end debates about it ;-).

 

[DrChinese]

... I've just come to wonder about whether the phenomenon of entanglement is due to how the wavefunctions of quantum information are "objectively defined". In the paper "A system's wave function is uniquely determined by its underlying physical state" it is concluded based on free choice that interpreting the wavefunction as an objective reality is possible regardless of all the probability involved (contrary to a classical or hidden variable state of a quantum property as proven by the Bell tests). The article poses a thought experiment with an info set $\Lambda$ of complete knowledge of the starting setup, consisting of wavefunctions etc.

Then for instance, it would be easy with two photons ~100% entangled in polarization to explain it such that the wavefunction of polarization of both photons is determined ~100% by the same subset of $\Lambda$.

A link to your reference: https://iopscience.iop.org/article/10.1088/1367-2630/aa515c

The paper is quite similar to (and comes to essentially the same conclusion as) the much better known PBR paper (his reference 14): https://arxiv.org/abs/1111.3328

The psi-ontic and psi-epistemic debate is a complicated subject, and certainly worth the study. But I would like to say something about your statement in bold. You cannot have 2 identical copies of a predetermined entangled wavefunction and get the QM statistical predictions. That is what Bell demonstrated, even though on the surface it appears as if you easily could. (Of course, if there is action at a distance (nonlocality) then all bets are off.)

If that point is not clear from your readings, I would suggest starting a thread in the main QM forum to discuss why. QM is contextual, which means it is dependent on context (the settings of the measurement devices) for its statistical predictions. There are no other known inputs for entangled scenarios with respect to the observed outcomes.

 

[Demystifier]

Why do they have to "manifest" it?

You think that they care, so I assume that they manifest it somehow, for otherwise why would you think that?

 

[PeterDonis]

You think that they care, so I assume that they manifest it somehow, for otherwise why would you think that?

To be precise, I did not say I think they care; I just said I don't think you can infer that they don't care, from the fact that they don't want to engage in endless discussions with "realists". They might just have better things to do with their time.

However, I also do not think caring about something requires "manifesting" that care in a way that you perceive as showing they care. Who made you the judge?

 

[Demystifier]

I expressed my opinion in the very first post i wrote.

No, i think it doesnt imply Lorentz invariance violation.

So you think something exists, and it is nonlocal in the Bell sense, but it does not violate Lorentz invariance. This, indeed, is very textbook like. But I am not satisfied with it, it's too vague for my taste. For example, textbooks say that it is Lorentz invariant because you cannot send signals faster than light. But I'm not satisfied with it, because I don't think that Lorentz invariance is only about sending signals. Bohmian mechanics is a counterexample, where the equations violate Lorentz invariance and yet signals still cannot be faster than light.

The problem with the textbook style explanation is that it does not write down equations that describe how existing things change during the measurement (unless they write down the collapse postulate, which, as an explicit equation, violates Lorentz invariance), so it is impossible to understand mathematically why Lorentz invariance is not violated. Most attempts to write explicit equations of that sort (Bohmian mechanics is an example, but not the only one) lead to violation of Lorentz invariance, but textbooks reject such equations because they are not useful in making new measurable predictions. Since such equations attempt to explicitly describe the existing things, and since textbooks reject it as useless, I see that as a kind of rejection of realism. They still say that something exists, but they reject a need for writing down equations that describe it. Their reason for rejection is - because it does not make new measurable predictions. But their claim that "something exists" also does not make new measurable predictions, so why do they not reject this claim as well? If they are not anti-realists, then they are inconsistent, which for me is even worst. When I think of them as anti-realists, that's because I'm trying to save their consistency.

Any thoughts?

 

[Demystifier]

To be precise, I did not say I think they care; I just said I don't think you can infer that they don't care, from the fact that they don't want to engage in endless discussions with "realists". They might just have better things to do with their time.

However, I also do not think caring about something requires "manifesting" that care in a way that you perceive as showing they care. Who made you the judge?

If they care but not manifest it, then their care is a hidden variable. Would you say that it is justified to think that a hidden variable exists?

 

[TonyStewart]

I have heard of but do not know the many interpretations of QM yet none are definitively Ontological.

Yet the ones that try are: the Copenhagen interpretation and Heisenberg's counter-theory, Many-worlds interpretation, Pilot-wave theory, and String Theory.

Eric Weinstein has boldly stated ST has stagnated progress in TP for at least the last 40 years. He fears but has openly challenged Edward Witten to end the poisoned-well on String Theory (which is useful for QFT) and make way for funding more progress in TP like 100 years ago. Ref Eric & Lex

 

[haushofer]

You describe astrology accurately but that's exactly not what's done in physics and theory building in physics, which is based on empirical facts and not some epistemological prejudices. If the latter approach is applied, nothing fruitful comes out (e.g., Einstein's search for a so-called "unified field theory" or in more modern times string theory). It's even worse with philosophy. I've not a single example, where philosohpical reasoning has brought any progress in the natural sciences. There's the incomprehensible ineffectiveness of philosophy in the natural sciences (Weinberg)!

Einstein's conviction that the equivalence principle entailed the true nature of gravity, even though the small empirical deviations from Newton would suggest a small adjustment.

And ironically, your attitude towards ontology is highly philosophical.

 

[Structure seeker]

You never say, what you consider incomplete. That's also typical for philosophical discussions. You keep it nebulous enough just to never end debates about it ;-).

If you say the theories are complete because there are no phenomena observed yet that falsify it, I don't understand how you discard realism of everything in general. Either the theories are about reality, or they are fantasies per definition of what fantasy is. If you discard realism, then I am only interested in your theories from the perspective of a psychiatrist. But if you claim the theories are applicable always and everywhere, that means in real life, and that means the physics it describes must be real.

So my question only is: what is about reality in your own interpretation and what isn't? Then we can discuss clearly, for otherwise I'm only interested if I want to connect to the world in your head.

 

[Structure seeker]

And if you don't know what to say, I think that is what @Demystifier means when he says the theories are incomplete in another sense. They don't fulfill the "get real" requirement.

 

[haushofer]

Physicists who proudly proclaim they can do without philosophy remind me of protestant christians who read the bible Sola Scriptura. What good has tradition ever brought us anyway? The text speaks for itself!

Well, only after you've assumed some very non-Sola Scriptura assumptions you're probably not aware of.

The same goes for physics and philosophy/ontology. It already starts with classical physics. If the math and observations would speak for themselves we wouldn't have this deluge of topics on PF about centrifugal forces, the twin paradox etc. etc. And what to think about anomalies? Lack of theory? Observations? Both? Why is dark matter such a big issue now if observations speak for themselves? Why 50 years of string theory with all those false expectations?

Downplaying the role of philosophy turns physics into mere bookkeeping. Mere bookkeepers are capable persons, but don't ask them to develop new economic theories.

 

[Simple question]

Then he'd be wrong. The program can work on a wooden computer, or on a paper/pencil system. The software algorithm is entirely independent of the hardware on which one might choose to run it.

That’s is incorrect.

After reading all those posts, I think that mistake is common, and is a good starting point to focus back on the OP analogy.

Programs are NOT algorithm. Computer do not compute. The ontology is pretty clear:an initial finite sate, with the turing machine. Can it be made out of in wood ? Yes. Can it run Angry Birds ? No.
I dont mean no in a FAPP sense. Understanding what ontology really means boils down to fixing that misconception.

Of course there will be at least two camps, even in computing, about every Paradigm. Like the Functionalist vs the procedurealist. But the issue is not to claim moral high ground. The issue is to produce and explains results, which I would map to Observation in physics.

Any pure algorithm will fail if you do not cope with those menial ontological basics like memory availability. Angry birds don t exist if at least million of bits changes state 30 times per second.
Why ? Because the complete ontology, the territory, include a human brains and eyes able to react to birds, and this is not, at a fundamental level, equivalent to reading millions of penciled Encyclopedia full off one and zero.

Clear ontologies are better than confused ones. Even if string (in physics not computing) is not a successfully ontology, it is a clear and coherent one. I don t think proponent of that ontology where adamant that those string were ’real’ nor interested of what they would be made of.

But formalizing what the territory could be is useful to choose amongs the maps available

 

[Fra]

I thought the point of the thread was to not polarize, as that leads nowhere, but to propose that the two views are complementary in some fuzzy sense. But it seems hard enough to agree on that.

I am very far from what one traditionally means by a "realist", yet there is this interesting connecting with real and subjective, which bridges the other extreme epistemological empirical stance.

Unless I misinterpreted Demystieifer in earlier recent threads, I think we agreed that noone will ever know with certainty what anything really "is", no matter how much we all want to know. But we (or the agent to be specific) always have an "image" or picture, or "map" of our best abduction of what reality is? (wether the agents have maps that harmonize is a separate question for me)

Wasn't THIS the "ontology", relative to a given "theory", what Demystifier entertained? or am I wrong?

I find it helpful to distinguish between the principles, rather than specifics. In principle, the "ontological view" does not necassarily mean "bohmian mechanics", for me it might as well mean "agent or bayesian mechanics" or soemthing that traditionally is very FAR from the old resistance to QM.

I appreciate your patience. Can you try to explain the following to me? If they don't want to engage in such discussions, but still care about these questions, then how is their care manifested?

Does it count if I care about, not ontology of "ultimalte reality" (as we agreed? we can never know), but about the ontology of the MAPs? This is for me the connection. Many TP, doesn't consider the theory as part of reality, it's just part of human science.

We often say, don't mistake the map for the territory, but the map is all we have, and what we revise. In one extreme one can wonder, does it matter what the territory is? Maybe it's just a mess of interacting maps? Part of the idea is of course, that the other maps are hidden, each player only views it's own map.

For me this neither traditional Realist, not typical physicist. I am symphatetic to parts of both sides.

/Fredrik

 

[Structure seeker]

If the map maps out real objects in real territory, I agree. The problem is some maps explicitly state "this here is not real". They give no contrary statement about anything mapped, so there is no way to argue that the territory should be like this, the map gives the impression it is all fantasy.
Thus the map doesn't map any physical thing in these cases, not because it is wrong but because there is no way to decide where it is right.

 

[vanhees71]

If you say the theories are complete because there are no phenomena observed yet that falsify it, I don't understand how you discard realism of everything in general. Either the theories are about reality, or they are fantasies per definition of what fantasy is. If you discard realism, then I am only interested in your theories from the perspective of a psychiatrist. But if you claim the theories are applicable always and everywhere, that means in real life, and that means the physics it describes must be real.

So my question only is: what is about reality in your own interpretation and what isn't? Then we can discuss clearly, for otherwise I'm only interested if I want to connect to the world in your head.

It's an empirical fact that observables don't have predetermined values before the system is not prepared in a state, where this is the case. That's the clear outcome of all the Bell tests, given that relativistic QFT, describing them in accordance with all observations, is local (i.e., microcausal).

 

[Structure seeker]

The mistake is to conclude from Godel's incompleteness theorems that since exact science cannot be based on logic, logic doesn't matter that much.

The thing every map needs to explain, is the existence of empirical facts, facts that anyone can verify anywhere. If you do not assume realism, an alternative logical explanation is needed in order to convince anyone. Well, due to the idea that it is tolerant, people might accept the way you talk, and due to the authoritative position of exact science, they might even believe it. But if I argue logically (which I am not going to stop) realism is the only current explanation of why empirical facts exist.

 

[Structure seeker]

It's an empirical fact that observables don't have predetermined values before the system is not prepared in a state, where this is the case. That's the clear outcome of all the Bell tests, given that relativistic QFT, describing them in accordance with all observations, is local (i.e., microcausal).

I agree that the classical observable does not exist before measurement, but since the wavefunction nature has real effects on observables, this wavefunction is an empirical fact. Then my above post applies, if it is not real, why on earth does it behave the same always and everywhere?

 

[vanhees71]

Obviously you use another meaning for the word "realim" than I. That's another problem with philosophical debates: They invent new meanings for words that are well defined in the sciences. Realism in the context we discuss it here is the assumption that all observables of a system take always determined values. That's not true within QT, where a specific observable only takes a determined values, when the state is prepared such that this is the case. An example is that the system is prepared in a pure state, that is represented by an eigenstate of the self-adjoint operator that represents the observable. Then this observable has the corresponding eigenvalue of this operator as its determined value.

 

[Structure seeker]

If nothing is real before the observable is measured, how can it appear out of things that supposedly are not real?

 

[vanhees71]

Obviously you have a different notion about what "real" means. You have to define it. Otherwise I don't understand what you mean.

 

[Structure seeker]

By measurement you apply a fourier transform to the wavefunction and voila, it becomes an observable. How can the second be real and the first not real?

My definition of real is that something is validly defined: no contradictions with other real objects and independent from other real objects (like how histories can be consistent in the consistent histories interpretation). Two entangled particles together have 1 wavefunction that I call real (but the quanta of 1 particle are not validly defined, their definition correlates with that of the other particle so they must be merged), and locality is sort of preserved because wavefunction collapse can only occur where the wavefunction has nonzero magnitude.

 

[vanhees71]

Nature behaves as she does, and we can figure out how she does by objective observations of this behavior. The fact that quantum theory describes all such observations unanimously implies that observables only take determined values when the system is prepared in a corresponding state, and it's impossible to prepare a system in any state, where all observables take determined values (Heisenberg uncertainty relation).

Wave functions are not observables. Their only meaning is to provide the probabilities (probability distributions) for the outcome of measurements via Born's rule.

Of course a state is not represented by the "wave function" but by the statistical operator (or in the special case of pure states by rays in Hilbert space).

 

[Fra]

If the map maps out real objects in real territory, I agree.

My point was that, the agents best answer to what the "territory is", is encoded by the map. The map is an "abstraction" encoding all we know. (To ask what is the difference between all what we know about soemthing, and what it is, seems confused. IF this bothers you, then perhaps I am a TP after all)

But my point is that the map is REAL. The physical composition of the agent, constraints the map, just like computer hardware constraints the possible programs it can run. Ie the physical agent(ie matter) is the memory device that encodes it's expectations. And when you consider two interacting systems, that entertain maps of each other - this is where it gets interesting, right? IF you don't allow this! Then we get normal QM. QM as it stands obivously has probelms to consistently incorporate the "measurement device" in the "quantum part".

So the complementarity I hinted was that the map/ontology is evolving, as per certain epistemological inference rules (which themselves are constrained by the state ontology - simlary to that the question we can ask today in physics, are constrained by our current theories! Ie the open questions are formulated, in terms of limits or boundaries of existing theory etc).

If we take this interaction seriously, the scientific knowledge and the scientific process, ontology and epistemology goes hand in hand. So if we ponger a say hypothetical big bang "agent/observer" in this sense, it's set of allowable inferences and thus interactions are limited, because that agent is presumed to be very "simple" (low complexity).

What difference this makes? It can save us from finetunig, as we do not need to consider a almost infinite landscape of potential ontologies, we can constrain this landscape by taking the "allowed interaction" from which is constructed - very seriously. Current typical TP does NOT take this seriously IMO.

the map gives the impression it is all fantasy.

If the "fantasy" is what determines tha angents behaviour (hamiltonian), then this makes a difference.

not because it is wrong but because there is no way to decide where it is right.

This why I think this must be seen in the evolutionary (quasi dynamical) perspective, there are not final answers, the quest itself may be as close as the answer we get. In a process of learning of selfoorganisation, there isn't necessarily a final goal or ultimate truth, it's just a constant evolution. The forward direction can only be defined relative to the present, so the future can keep changing.

/Fredrik

 

[Fra]

Nature behaves as she does, and we can figure out how she does by objective observations of this behavior.

The problem is that from the perspective of a single observer/agent, the objectivity is inaccessible; the only think I can accept is observer democracy. It's the same thing, except transient disagreements is not excluded. And I think these "forbidden transitions" is exactly what we need to explore to undertand how the fundamental forces are related.

I think when we used this observer equivalence as a constraint, we also limit the applicability of the theory, to essentially asymptotic data for small subsystems, which makes perfect sense for normal QM.

But with unification I think this constraints gets us into trouble.

/Fredrik

 

[Demystifier]

Obviously you have a different notion about what "real" means. You have to define it. Otherwise I don't understand what you mean.

The title of this thread is a kind of explanation of what "real" (ontology) means. It's an explanation through an analogy, not a definition, but as I explained many times, some concepts are primitive so cannot be defined.

 

[vanhees71]

"Realistic" has a clear meaning in this context: A theory is called realistic if it assumes that all observables always have determined values. This is obviously not fulfilled by QT and that's why QT is non-realistic (though in the case of relativistic QFT it's by construction local).

 

[Fra]

"Realistic" has a clear meaning in this context: A theory is called realistic if it assumes that all observables always have determined values. This is obviously not fulfilled by QT and that's why QT is non-realistic (though in the case of relativistic QFT it's by construction local).

I couldn't resist thisr analogy using Demystifiers metaphors...

Suppose that the territory really is, a mess of "interacting maps"? Where the "maps" are encoded by the matter systems and agents. Or translated to the terms here mess of "communicating computers", each running different software. Obviously the software runnable on any computer must be ported to run in the hardware. And lets suppose all the computers do, is trying to make inferences about the states of all the other computers, using the available information they received from communicating. Which means that what is going on in this game, is that all the parts are constantly learning as much as possible about each other - thus either getting along and cooperating, or killing/hijacking each other?

A corresponding questions is:

Do we think such a game/model, reproduce quantum phenomena, such as entanglement or quantum interference, if the state of all maps are "predetermined"? By the set of all maps, I mean essentiall the set of all possible states of the computer hardware and their software; as if there was a "god view" that could look at ALL the computers at once (bypassing physical communication/inference)?

Can we guess if such model will be excluded by mechanicsms in Bells theorem?

Would such a model paradigm be satisfactory (presuming it is identified) and esthetically satsifactory to some realists?

/Fredrik