Anti-realist Interpretations of QM

[Paul Colby]

A high speed camera could show us the number to be picked in advance.

You just redefined the measurement apparatus. The camera shows the pick. Big Deal.

Now, what you say about quantum measurement is certainly true. That wasn't what I was commenting on.

 

[Lynch101]

The notion of "hidden variables" has several different meanings:
1) Any variables additional to the wave function.
2) Any variables that have values even when they are not measured.
3) Variables that determine the measurable outcomes, so that the outcomes are actually deterministic, rather than random.

So what notion of "hidden variables" do you have in mind?

I would be inclined towards 1) and 3).

Am I correct in saying that, according to the instrumentalist and anti-realist positions, the wave function doesn't describe the properties of the system prior to measurement? In this sense then, the properties of the system prior to measurement - if such properties exist - would be considered hidden variables according to 1) above?

I had thought that hidden variables necessarily implied an underlying deterministic Universe, but would stochastic interpretations that treat the wave function realistically also involve hidden variables?

 

[Lynch101]

Interpretations in which the observer is a participant in creating (objective) reality (i.e. the results of a measurement chosen by the subjective observer) could be called "anti-realist". Another label often applied is "subjective realism" although of course every person has their own meanings for these terms which may or may not be considered consensus.

What role does the wave function play in such interpretations? Am I wrong in thinking that it treats the wave function instrumentally also i.e. as a predictive tool?

Are such interpretations idealisitic in the sense that they say that there is no mind-independent reality?

 

[Lynch101]

Not an expert on the schools of thought or whatever this thread is about so I don't know. However, the electromagnetic field (or better our abstraction of it) does have a state or an ensemble of states prior to measurement. We can make statistical statements about said measurements. The electromagnetic field exists and can be discussed prior to measurement. So I don't really get the anti-realists position at all.

I struggle with the anti-realist position myself because [based on my possibly incorrect interpretation of it] I think there are some insurmountable hurdles to it. The question of how something without any properties can interact with a measuring device in the first instance, for example. Also, in what sense can something without properties even be said to be a part of the Universe?

As I say, that is based on my current interpretation of it but I have been known to be inaccurate in my understanding

 

[Lynch101]

The anti-realist says that a specific outcome only materializes once a final choice of measurement basis occurs. Therefore the outcome is observer dependent. And it is meaningless to assert that there was a well-defined value to the measured property without that.

Would that be closer to the instrumentalist position?

My understanding is that the anti-realist position has to go further than simply saying it is meaningless to assert that there was a well-defined value to the measured property, prior to the final choice of measurement basis. I think it has to deny that there is any property whatsoever prior to the measurement. To allow that there is some unquantified property of the system prior to measurement would, I was thinking, be a realistic, hidden variables interpretation?

 

[Paul Colby]

The question of how something without any properties can interact with a measuring device in the first instance, for example.

In my opinion one can't have such a discussion without being careful about the "something" one is actually discussing. The electromagnetic field has properties, photons being one. Knowing the state of the field one can make statistical statements about photon detections. However, a photon really isn't "a thing" in the same sense as the moon or a beachball.

 

[Lynch101]

1. ...The realist says they have well-defined values at all times, regardless of observer.

Is it possible to take a somewhat "weaker" realist position and remain agnostic on whether properties have well defined values at all times and simply say that there are properties?

 

[DrChinese]

I struggle with the anti-realist position myself ...

Just be aware that the alternative is one (such as Bohmian Mechanics) which contain/require FTL mechanisms. Nothing wrong with that at all, perfectly acceptable. But there is a good reason there are anti-realistic interpretations: a) they naturally explain the EPR paradox; b) locality may be retained in all respects, a la Bell.

 

[Lynch101]

In my opinion one can't have such a discussion without being careful about the "something" one is actually discussing. The electromagnetic field has properties, photons being one. Knowing the state of the field one can make statistical statements about photon detections. However, a photon really isn't "a thing" in the same sense as the moon or a beachball.

Sorry, I meant in the broadest possible sense. It might be better to ask how can anything without properties can interact with a measurement device?

 

[DrChinese]

Is it possible to take a somewhat "weaker" realist position and remain agnostic on whether properties have well defined values at all times and simply say that there are properties?

I'm sure. In fact several posters earlier mentioned that you might define "realism" as you see it. That may not really help much though. Once you address Bell, you see how difficult that is.

Many papers have been written attempting to provide alternative definitions of realism and locality such that Bell does not apply. That hasn't worked out so well. Quick, name the guy who got a Nobel for getting around Bell!

Those who go with the "weaker" definitions end up on a lonely road. And one that provides no more useful description or interpretation. You may as well end up with "shut up and calculate" and take no particular interpretation addressing Bell.

 

[Lynch101]

Just be aware that the alternative is one (such as Bohmian Mechanics) which contain/require FTL mechanisms. Nothing wrong with that at all, perfectly acceptable. But there is a good reason there are anti-realistic interpretations: a) they naturally explain the EPR paradox; b) locality may be retained in all respects, a la Bell.

I'm wondering if we are necessarily using the same interpretation of "anti-realism" here. I'm taking it to be the position that the system in question has absolutely no properties whatsoever, prior to measurement.

Are there other interpretations of non-locality that don't involve FTL mechanisms? I thought someone mentioned some alternatives in another thread - I'll see if I can find it.

 

[DrChinese]

However, a photon really isn't "a thing" in the same sense as the moon or a beachball.

It's a quantum object. And it really doesn't matter which quantum object we are referring to. The electron's non-commuting properties are just as simultaneously "real" or "not-real" as those of any photon.

 

[Paul Colby]

It might be better to ask how can anything without properties can interact with a measurement device?

Well, I don't think such things exist in exactly the way you are expressing it. One can show experimentally that a spin-1/2 system has no spin-value prior to measurement. That's not to say the spin-1/2 system doesn't exist or have spin values one can measure.

 

[Lynch101]

I'm sure. In fact several posters earlier mentioned that you might define "realism" as you see it. That may not really help much though. Once you address Bell, you see how difficult that is.

Many papers have been written attempting to provide alternative definitions of realism and locality such that Bell does not apply. That hasn't worked out so well. Quick, name the guy who got a Nobel for getting around Bell!

Those who go with the "weaker" definitions end up on a lonely road. And one that provides no more useful description or interpretation. You may as well end up with "shut up and calculate" and take no particular interpretation addressing Bell.

I think I have a bit more study to do before even being able to solve a question on Bell, not to mind anything else I'm mindful that I have mentioned in my previous post that we might be talking at cross purposes with regard to the term "anti-realism" and you may have replied to it before I post this, but just for clarity, taking the "weaker" position is more just for the purpose of juxtaposing it with the "anti-realist" position, as I understand it.

My understanding of the "anti-realist" position is that it states that the quantum system has no properties whatsoever prior to measurement. This to me would seem to be more problematic than an FTL mechanism which cannot be used for signalling.

 

[DrChinese]

I'm wondering if we are necessarily using the same interpretation of "anti-realism" here. I'm taking it to be the position that the system in question has absolutely no properties whatsoever, prior to measurement.

An entangled electron, for example, lacks well defined values for its entangled properties. You could say those properties are in a superposition of states. Whether it has "absolutely no properties" is just by definition at that point.

 

[Lynch101]

Well, I don't think such things exist in exactly the way you are expressing it. One can show experimentally that a spin-1/2 system has no spin-value prior to measurement. That's not to say the spin-1/2 system doesn't exist or have spin values one can measure.

This might be where my understanding is lacking. My understanding of the "anti-realist" position is that it says it is meaningless to talk about the properties of the system prior to measurement because it does not have any properties prior to measurement. I was inclined to think that allowing for properties of any kind would be to allow for hidden variables and necessitate a realistic interpretation.

 

[DrChinese]

My understanding of the "anti-realist" position is that it states that the quantum system has no properties whatsoever prior to measurement. This to me would seem to be more problematic than an FTL mechanism which cannot be used for signalling.

A perfectly reasonable position.

@Demystifier : You can thank me later for driving another one over to your side of the road.

 

[Paul Colby]

This might be where my understanding is lacking. My understanding of the "anti-realist" position is that it says it is meaningless to talk about the properties of the system prior to measurement because it does not have any properties prior to measurement. I was inclined to think that allowing for properties of any kind would be to allow for hidden variables and necessitate a realistic interpretation.

Perhaps I'm not using the same definition of property. A particle (AKA a quantum excitation of a quantum field) has a property called spin. What's being discussed is the specific value of spin prior to measurement not being determined.

 

[Lynch101]

An entangled electron, for example, lacks well defined values for its entangled properties. You could say those properties are in a superposition of states. Whether it has "absolutely no properties" is just by definition at that point.

Apologies, I'm not clear on what you mean by the sentence I emboldened above, as in how do you mean it is by definition?

In the anti-realist position does the wave function represent an ontic state i.e. does it correspond to the physical reality of the entangled photons, or does it just give the probabilities for measurement outcomes relating to the entangled photons? Or is there some other aspect of the mathematical formalism that governs that?

I'm not sure if I've used the relevant terms accurately enough there. I'm not sure if this will help, but I'm wondering if there is the mathematical formalism describing the superposition is ontic in a similar - but obviously very different - sense to how the wave function is ontic in Bohmian Mechanics?

 

[Lynch101]

Perhaps I'm not using the same definition of property. A particle (AKA a quantum excitation of a quantum field) has a property called spin. What's being discussed is the specific value of spin prior to measurement not being determined.

I do think we have slightly different ideas in mind. When we talk about things like spin, position, and momentum, we are talking about very specific properties. In this sense we might talk about whether or not those properties have specific or definite values prior to measurement.

I'm talking in the more general sense of any property whatsoever. To my understanding the "anti-realist" position says that the quantum system has no properties whatsoever, in the general sense. My interpretation is that the alternative to this is to say that the system does have some [unquantified] properties. My understanding is that to say the system has unquantified properties would be to say that it has hidden variables and is realistic.

 

[Paul Colby]

I'm talking in the more general sense of any property whatsoever. To my understanding the "anti-realist" position says that the quantum system has no properties whatsoever, in the general sense. My interpretation is that the alternative to this is to say that the system does have some [unquantified] properties. My understanding is that to say the system has unquantified properties would be to say that it has hidden variables and is realistic.

What anti-realist believe or say is beyond me. You're best listening to people more versed in what they actually say. They sound unhinged but that's only my opinion based on the snippets in this thread.

So, the axioms of QM actually provide a means of being clear about "properties." I read property as an observable. An observable is an operator which has a spectrum, a set of allowed values that may be observed. This spectrum is unchanging and I would call it a property.

Now, upon measurement only one of the spectrum set is obtained as a result. This result is the property-value.

 

[DrChinese]

1. Apologies, I'm not clear on what you mean by the sentence I emboldened above, as in how do you mean it is by definition?

2. I'm not sure if I've used the relevant terms accurately enough...

1. At some point, everyone reverts to their own language to describe things. And similarly, it is impossible to say if your words are close in meaning to mine. If you believe the following, that a particle in a superposition of states lacks a well-defined value independent of a measurement basis chosen by an observer, then I would say you have adopted an "anti-realist" position. I would call it "non-realistic" rather than "anti-realist" but I can't see any difference.

"An entangled electron, for example, lacks well defined values for its entangled properties. You could say those properties are in a superposition of states."2. This is why there is an Interpretations sub forum. No one is...

 

[DrChinese]

What anti-realist believe or say is beyond me.

Bell's Theorem says: No physical theory of local Hidden Variables can ever reproduce all of the predictions of Quantum Mechanics.

Quantum Mechanics requires us to abandon at least one of "locality" and "realism" (hidden variables which may be unknowable). The anti-realist position would be that realism is abandoned, presumably so you can retain locality.

There are a number of interpretations that reject realism. In the time symmetric group, there is a mechanism of "handshake" (however that works) between the quantum object being observed and the observer. Therefore there is observer dependence and you have a subjective reality - retaining locality.

 

[Lynch101]

So, the axioms of QM actually provide a means of being clear about "properties." I read property as an observable. An observable is an operator which has a spectrum, a set of allowed values that may be observed. This spectrum is unchanging and I would call it a property.

Now, upon measurement only one of the spectrum set is obtained as a result. This result is the property-value.

Definining "properties" as "observables" seems to come back, somewhat, to the point about what we can say about the system prior to measurement.

My difficulty in defining properties as observables, but that might just be due to the fact that I don't have a good working knowledge of the various terms, so I can't articulate my point using the more precise terms. "Beable" might be the more suitable term - I just haven't used it enough yet. From @Demystifier's paper in his signature:

Beable: That word was coined by John Bell. It means the same as ontology: stuff which is there irrespective of observation. The concept of a beable is central to Bohmians, but not to instrumentalists.

Another difficulty I have with defining "properties" as "observables" is that "beables" are distinguished from observables but we might still talk about the properties of beables or perhaps say nothing more than the statemement, "beables have properties". Is there also a question of whether or not the observable can be said to be an inherent property of the system?

If we talk about the properties of a system prior to measurement, it also distinguishes a property from an observed property.It's in this broad sense of the term property that I tend to talk, to try to clarify my interpretation of "anti-realism". To try and use the term "beable": the "anti-realist" position seems to say that there are no beables prior to measurement.

In terms that seem more intuitive to me, I would say that the anti-realist position says that the system has no properties whatsoever prior to measurement. It is this which leads me to question how anything without properties can interact with a measurement device in the first place?

 

[Lynch101]

1. At some point, everyone reverts to their own language to describe things.

Thank you DrChinese, this is reassuring (if not comforting ).

And similarly, it is impossible to say if your words are close in meaning to mine. If you believe the following, that a particle in a superposition of states lacks a well-defined value independent of a measurement basis chosen by an observer, then I would say you have adopted an "anti-realist" position. I would call it "non-realistic" rather than "anti-realist" but I can't see any difference.

I would be inclined to distinguish between "non-realist" and "anti-realist", where non-realism and instrumentalism are effectively one and the same. My understanding is that it says that a particle in a superposition of states lacks a well-defined value independent of a measurement basis chosen by an observer. It does not make any ontological inferences from the mathematical formalism.

"Anti-realism" on the other hand goes further, it does make ontological inferences from the mathematical formalism. It seems to say that a particle in a superposition of states lacks a well-defined value independent of a measurement basis chosen by an observer, because a particle in a superposition does not have any properties prior to measurement.

"An entangled electron, for example, lacks well defined values for its entangled properties. You could say those properties are in a superposition of states."

This might be where my limited understanding of the mathematics is hindering me. My understanding is that, according to the "anti-realist" position, the mathematical formalism which describes the superposition only gives the probabilities for the measurement outcome of the entangled particles, meaning that it doesn't describe the ontic state of the particles prior to measurement.

If the entangled state defined by the mathematics is an ontic state that would make it a realistic interpretation, wouldn't it? Whereas, if the superposition described by the mathematical formalism is just a tool for predicting the outcome of experiments, then it doesn't describe the system prior to measurement. The instrumentalist/"non-realist" position stops there, but the "anti-realist" position would go further, as above.

That would be my understanding of it.

2. This is why there is an Interpretations sub forum. No one is...

 

[Paul Colby]

Quantum Mechanics requires us to abandon at least one of "locality" and "realism" (hidden variables which may be unknowable). The anti-realist position would be that realism is abandoned, presumably so you can retain locality.

So, would it be accurate to say;

realism = systems have property values prior to measurement?

Since the realism view is born out of a classical macroscopic intuition we evolved with, I say that it's unsupported by the measurements and should be abandoned. There is no reason to expect it to hold for purely quantum objects.

 

[Paul Colby]

In terms that seem more intuitive to me, I would say that the anti-realist position says that the system has no properties whatsoever prior to measurement. It is this which leads me to question how anything without properties can interact with a measurement device in the first place?

I can only repeat what I've said. I believe you are conflating having a value with having a possible set of observable values. Things interact because they do and we have the formalism and data to prove it.

 

[Morbert]

In terms that seem more intuitive to me, I would say that the anti-realist position says that the system has no properties whatsoever prior to measurement. It is this which leads me to question how anything without properties can interact with a measurement device in the first place?

Anti-realism re/ QM pertains to the anti-realism of properties considered by QM (i.e. those represented by regions of Hilbert space), as opposed to all properties entirely.

 

[Demystifier]

Am I correct in saying that, according to the instrumentalist and anti-realist positions, the wave function doesn't describe the properties of the system prior to measurement?

Yes.

In this sense then, the properties of the system prior to measurement - if such properties exist - would be considered hidden variables according to 1) above?

Yes.

I had thought that hidden variables necessarily implied an underlying deterministic Universe, but would stochastic interpretations that treat the wave function realistically also involve hidden variables?

Yes, in the sense 1) and 2).

 

[Demystifier]

It is this which leads me to question how anything without properties can interact with a measurement device in the first place?

Exactly!