Is Consciousness involved in wave function collapse?

[nickelite]

Hello:

I know it's a rather subjective title. But I am no expert in the subject and I've read a lot of information in the Internet that is contradictory.

I have read that the collapse of the wave function requires interaction with an observer. But is that collapse in any way related to the awareness of the observer or to interference with the measuring device alone?. I understand that all the founders of quantum mechanics, like Bohr and Einstein rejected this idea, right?

I would appreciate, if it’s not too much trouble, any links or references that discuss the subject.

Thanks.

 

[San K]

Hello:

I know it's a rather subjective title. But I am no expert in the subject and I've read a lot of information in the Internet that is contradictory.

I have read that the collapse of the wave function requires interaction with an observer. But is that collapse in any way related to the awareness of the observer or to interference with the measuring device alone?. I understand that all the founders of quantum mechanics, like Bohr and Einstein rejected this idea, right?

I would appreciate, if it’s not too much trouble, any links or references that discuss the subject.

Thanks.

this comes up often.

the answer is No.

experiments have proved that it's the setup that causes the "collapse" ...

the collapse happens even when no human/living-entity is watching...

now one could argue further and say/ask -

how do you know if collapse has happened, even if no human is watching?

there is an answer...but then... there's no end to such a line of questioning...

 

[Ken G]

The basic problem is that physics itself demonstrably comes from the consciousness (nothing unconscious is doing any physics), consciousness is built right into the fabric of the scientific method (and here I make no attempt to distinguish "consciousness" from "awareness" or "intelligence" or "understanding", you can try to parse those words if you want!). So the basic problem is, we have no idea to what extent our own thinking processes are inherent to our physics, including our concepts like collapse of the wavefunction. The best we can do is recognize that the concept of a hypothetical observer, with a hypothetical consciousness if necesary, suffices to understand our physics. So there is not a need for a real observer or consciousness to be present to understand the outcome of some physical event, but that is not quite the same thing as saying there doesn't have to be a consciousness at all. Physics is, above all, a kind of language, spoken by an intelligence, that is conscious. Make of that fact whatever you will.

 

[dx]

In quantum physics, the state of a system is given in terms of the statistics of measurements obtainable on the system. Formally, this statistics is given in terms of the 'wavefunction' |\Psi \rangle. When an irreversible interaction between the atomic objects and the measuring apparatus occurs, the wavefunction which for t < [the time of measurement/interaction] previosly represented the probabilities of various values no longer defines the state of the object. Thus the 'collapse' of the wavefunction is due to the irreversibility inherent in the concept of observation itself.

 

[AndyUrquijo]

The collapse of states or wave functions is traditionally related to 'observations'. However there are several common natural systems where states collapse by themselves. Human consiousness or awareness is not relevant at all. You could just make a measuring device that didn't show the results and the measured system would still collapse.

In other words, you can't obtain (exact) information of a quantum system without collapsing its states. But it doesn't work the other way around. You can collapse the state without obtaining any information at all.

 

[G01]

So there is not a need for a real observer or consciousness to be present to understand the outcome of some physical event, but that is not quite the same thing as saying there doesn't have to be a consciousness at all.

What does this have to do with the actual question at hand? Of course, only a conscious intelligence could imagine a hypothetical thought experiment, but this is entirely beside the point. This is not what the OP is asking about.

The OP is asking whether the presence of an actual conscious observer effects the outcome of a quantum measurement. The answer is that it does not.

I think the reason this question comes up so often is semantic. We use the word "observation", but we should use the word "measurement", or as dx says above, "interaction."

During the measurement, the quantum system must interact with the measurement device. The measurement device is inherently classical, being built from many atoms. So, the effect is such that when the measurement happens, our quantum system interacts with the measuring device. This causes decoherence to set in in our quantum system and the wavefunction collapses. Or, in other words, the interaction with the measuring device "washes out" the the interference terms of the systems density matrix.

The point is that consciousness has nothing to do with measurement. By no means do all measurements have to include a conscious observer. Wavefunction collapse will happen because of an interaction with an external environment(the measuring device). The interaction, and thus, the WF collapse, happens regardless of whether the device is run by a human or is a mindless automated robot.

 

[BruceW]

I'm glad to see that I agree with most of the responses on this thread. I was worried that I'd see a lot of non-mainstream opinions on this :)

 

[Ken G]

What does this have to do with the actual question at hand? Of course, only a conscious intelligence could imagine a hypothetical thought experiment, but this is entirely beside the point. This is not what the OP is asking about.

That is not entirely clear. The OP admits to two very separate interpretations, and even recognizing the existence of those two interpretations is important for understanding the answer to either one. "Is consciousness involved" could mean:
1) "Does a conscious observer need to be present to get a wave function collapse?"
Answer: no.
2) "Is consciousness/awareness/intelligence/perception involved in giving meaning to the entire concept of doing a measurement that collapses a wave function?
Answer: yes.
Of course we could always satisfy ourselves with #1, but there's no need to sell short the full ramifications of the question. A naive interpretation of what physics is causes a lot of problems in understanding quantum mechanics, relativity, and thermodynamics, in regard to what is the role of the physicist/observer. I just put it out there, and everything I said is true. It doesn't need to start a whole thread on the issue, unless the OPer wants to go there. As I said, the question itself is unclear on the point.

The OP is asking whether the presence of an actual conscious observer effects the outcome of a quantum measurement.

Correction, that is your interpretation of what was asked. The actual words included "is that collapse in any way related to the awareness of the observer..." So I answered the "in any way" part, because I think the OPer would like to know. Apparently this pushes a button for you.

I think the reason this question comes up so often is semantic. We use the word "observation", but we should use the word "measurement", or as dx says above, "interaction."

Yes, it is certainly a semantic issue. Unfortunately, a lot of people equate "semantic" with "unimportant", when in fact, "semantic" means "cuts to the heart of how we use language in physics, and how impossible physics would be without language." Of course, that returns us to the role of consciousness/intelligence etc., where you do not wish to go.

During the measurement, the quantum system must interact with the measurement device. The measurement device is inherently classical, being built from many atoms. So, the effect is such that when the measurement happens, our quantum system interacts with the measuring device. This causes decoherence to set in in our quantum system and the wavefunction collapses. Or, in other words, the interaction with the measuring device "washes out" the the interference terms of the systems density matrix.

Exactly. Now you are ready for the next question: why do some devices correspond to "quantum observables"? Does nature ever do observations, or just undergo unitary evolution? This will lead you inevitably to interpretations of QM, like Copenhagen and many worlds. It all opens right up when you ask what role does the physicist play in all this. Of course, you don't have to ask yourself that, but maybe the OPer would like to, and I gathered from the OP that it is indeed their interest.

The point is that consciousness has nothing to do with measurement.

The point is, that statement is patently false. The truth is that measurement, as defined and understood and contemplated and used by the conscious physicist, can happen even if no such physicist is present. However, to give semantic meaning to what a measurement even is, this does indeed require a conscious intelligence (so far as we understand what those words mean), who has been there in a fully analogous situation for us to be able to use any of those words meaningfully. In short, a universe with no intelligent beings is a universe that has no measurements, and no wave functions to collapse. That is not an opinion, it is a fact-- in such a universe there are not the words "wavefunction", there is not the concept "amplitude", because there are no words and no concepts in the first place, stuff just happens, presumably the same as it does now minus any concept of "measurement" or "collapse" of anything. The significance of this fact opens up a lot of what physics really means, but does indeed get a bit philosophical, so if the OPer has any specific questions about it, they should probably frame them in terms of quantum mechanical predictions for this section.

The interaction, and thus, the WF collapse, happens regardless of whether the device is run by a human or is a mindless automated robot.

Yes, that's why I said the intelligence that says "what happened" in the first place does not need to be present every time it happens, but they certainly do need to present somewhere that it happens, to even say what the words mean. That turns out to be important to understanding what quantum mechanics is, and in choosing which interpretation you favor.

 

[G01]

That is not entirely clear. The OP admits to two very separate interpretations, and even recognizing the existence of those two interpretations is important for understanding the answer to either one. "Is consciousness involved" could mean:
1) "Does a conscious observer need to be present to get a wave function collapse?"
Answer: no.
2) "Is consciousness/awareness/intelligence/perception involved in giving meaning to the entire concept of doing a measurement that collapses a wave function?
Answer: yes.

The first question is a question about physics and what a physical theory says about the measurement process, meaning the process described in my previous post whereby a quantum system is put into a classical state via decoherence.

The second question is a philosophical question.

Since the OP posted this in the Quantum physics subforum and not the Philosophy subforum. I assumed (I think rightfully) that the OP was interested in the first question. I think the context of the forum this question was posted in removes the ambiguity you describe.

Of course we could always satisfy ourselves with #1, but there's no need to sell short the full ramifications of the question. A naive interpretation of what physics is causes a lot of problems in understanding quantum mechanics, relativity, and thermodynamics, in regard to what is the role of the physicist/observer. I just put it out there, and everything I said is true. It doesn't need to start a whole thread on the issue, unless the OPer wants to go there. As I said, the question itself is unclear on the point.

Again, this falls into the realm of philosophy of science, and thus, if this is what the OP was concerned about, should be posted in the philosophy forum.

Correction, that is your interpretation of what was asked. The actual words included "is that collapse in any way related to the awareness of the observer..." So I answered the "in any way" part, because I think the OPer would like to know. Apparently this pushes a button for you.

I do not think that was the OP was asking about, but either way, we won't know unless he/she posts again.

Yes, it is certainly a semantic issue. Unfortunately, a lot of people equate "semantic" with "unimportant", when in fact, "semantic" means "cuts to the heart of how we use language in physics, and how impossible physics would be without language." Of course, that returns us to the role of consciousness/intelligence etc., where you do not wish to go.

The "role of consciousness/intelligence," in the context you are using the phrase is a philosophical question, not a scientific one. That is why "I do not wish to go there."

The point is, that statement is patently false. The truth is that measurement, as defined and understood and contemplated and used by the conscious physicist, can happen even if no such physicist is present.

My statement was not false. This is exactly what I was saying above. Measurement, as defined in the context of the quantum mechanical measurement problem, does not require the presence of a conscious observer, or as I said above,

The point is that consciousness has nothing to do with measurement. By no means do all measurements have to include a conscious observer. Wavefunction collapse will happen because of an interaction with an external environment(the measuring device). The interaction, and thus, the WF collapse, happens regardless of whether the device is run by a human or is a mindless automated robot.

assuming the standard definition of "measurement" used in QM.

 

[nickelite]

Thanks to all for your replies. You've been very helpful. I study engineering and am new to Quantum Mechanics. I'm glad I asked because I was starting to get a mystical idea of the double-slit experiment, which thankfully I'm done with. Nevertheless I find Quantum Mechanics exiting and intriguing.

With all the misleading information about Quantum Mechanics in the web I got a little off track.
I deduce then, from your answers that the people that claim to have collapsed the wave function merely by looking at some results are making it up(They say the device was still on but it was when they looked at the results that the interference pattern disappeared).

how do you know if collapse has happened, even if no human is watching?

there is an answer...but then... there's no end to such a line of questioning...

If it's not too much trouble I'd like to know the answer. Just out of curiosity.

Cheers!

 

[G01]

I deduce then, from your answers that the people that claim to have collapsed the wave function merely by looking at some results are making it up(They say the device was still on but it was when they looked at the results that the interference pattern disappeared)

Yes, The wavefunction collapse occurs when the measurement is taken, i.e. whenever you actually interact with the system, not when you look at the results of that measurement.

 

[San K]

Thanks to all for your replies. You've been very helpful. I study engineering and am new to Quantum Mechanics. I'm glad I asked because I was starting to get a mystical idea of the double-slit experiment, which thankfully I'm done with. Nevertheless I find Quantum Mechanics exiting and intriguing.

With all the misleading information about Quantum Mechanics in the web I got a little off track.
I deduce then, from your answers that the people that claim to have collapsed the wave function merely by looking at some results are making it up(They say the device was still on but it was when they looked at the results that the interference pattern disappeared).

If it's not too much trouble I'd like to know the answer. Just out of curiosity.

Cheers!

Sure nickelite. You have actually provided the answer (or said what I was thinking ...) ...and so i quote you...below...;)

(They say the device was still on but it was when they looked at the results that the interference pattern disappeared).

or in my words:

Mr Argumentative: how do you know WF has collapsed, without the observer looking at it
Physicist: we know from the results that WF collapsed when the photon was detected/measured
Mr Argumentative: well maybe the WF collapsed when the observer looked at it and not before (when it was measured)? can you disprove this?

 

[Ken G]

The first question is a question about physics and what a physical theory says about the measurement process, meaning the process described in my previous post whereby a quantum system is put into a classical state via decoherence.

The second question is a philosophical question.

Correct, though the line between the two is not as clearly drawn as some would like to imagine. This is actually important for understanding quantum mechanics-- as almost any new student of quantum mechanics knows, setting up and solving the equations can be learned easily enough in a kind of "trained monkey" mode, but understanding what it is saying is the difficult part. I think the OP is in the latter mindset, hence the question.

Since the OP posted this in the Quantum physics subforum and not the Philosophy subforum.

I think you should really take a look at the questions currently being discussed in the quantum physics subforum. I estimate about half have a significant philosophical component that is quite evident in the OP question, this one included.

I assumed (I think rightfully) that the OP was interested in the first question.

And I will choose to reserve judgement on that until hearing if the OPer has any follow-up questions.

I think the context of the forum this question was posted in removes the ambiguity you describe.

Again, no. Indeed right now there are several threads in this subforum grappling with essentially purely philosophical issues about de Broglie-Bohm interpretations and just what "quantum information" really entails. Issues of determinism as a fundamental truth, or just a modeling assumption, are also being addressed. None of those come under the heading of what the quantum theory says about measurement, because the quantum theory only says one thing about measurements: it predicts them.

I do not think that was the OP was asking about, but either way, we won't know unless he/she posts again.

Agreed. And I also agree that their question was largely relevant to the answer you gave, it just isn't the whole story about the role of consciousness in quantum mechanics, so I gave a pointer to the rest of the tale. It was not my intention to start a thread on the various different interpretations of QM unless the OPer had follow-up questions, but I do see their initial question as an excellent entry point into such a discussion. If the forum thinks all issues of interpretations should go to the philosophy subforum instead of the quantum physics subforum, I would completely agree, actually-- it is many of the other members here who instead do not think that is philosophy.

The "role of consciousness/intelligence," in the context you are using the phrase is a philosophical question, not a scientific one. That is why "I do not wish to go there."

Nor does any scientist really, it is the bugbear in the corner. But sometimes, quantum mechanics gives us no choice, unless we strictly adhere to "shut up and calculate". In my experience, no one ever actually does.

My statement was not false. This is exactly what I was saying above. Measurement, as defined in the context of the quantum mechanical measurement problem, does not require the presence of a conscious observer, or as I said above,

Except to be there when the "quantum mechanical measurement problem" gets defined in the first place, which is my point (and is the point of Bohr's school of thought also, I might add). If quantum mechanics measurements teach us anything, it is that the "physicist is just a fly on the wall that we can ignore" attitude just doesn't cut it. That's the issue I would like to raise to the awareness of the OPer, so they don't walk away imagining that naive realism continues to work even in quantum mechanics.

 

[nickelite]

Just to clarify:

I was asking about the more technical question, regarding whether awareness from the results is required in order to collapse the wave-function. Although I enjoyed the more philosophical discussion saying that consciousness is needed for there to exist physics at all.

It seems to me that the hype that has been created around the double slit experiment has made a diffuse line between the scientific question and the philosophical question.

It has led many to assume that it is the consciousness, PER SE, what collapses the wave function, and makes an electron appear in a determinate position. But from what I have learned in this discussion, it's not the observer but the act of measuring, it is the device used to measure itself what changes the results of the experiments.

Hence the confusion in the interpretation of my question. Either way is fine, I'm glad to take all the views in the subject since I'm new to QM.

 

[Ken G]

It seems to me that the hype that has been created around the double slit experiment has made a diffuse line between the scientific question and the philosophical question.

I agree. And personally, I'd prefer if the lines were kept drawn more clearly, but that's just not what happens.

It has led many to assume that it is the consciousness, PER SE, what collapses the wave function, and makes an electron appear in a determinate position. But from what I have learned in this discussion, it's not the observer but the act of measuring, it is the device used to measure itself what changes the results of the experiments.

Here's the problem. Decoherence can explain with no difficulty how different "pointer positions" of a macro instrument get coupled to certain "quantum outcomes," with no cross-talk (i.e., no interference) between different pointer positions. This is the state of affairs before we look at the outcome of the measurement-- we have only a statistical understanding of the possibilities. The problem comes when we actually look at the pointer ourselves, and at that point, something happens that there is no consensus about in quantum measurement theory. The Copenhagen school says that the statistical prediction is all we can use quantum mechanics for, and the actual looking is something different, something outside the ability of quantum mechanics to describe (expressly because it is outside the ability of quantum mechanics to predict, and Copenhagen likes to equate science with prediction). The many-worlds school says that all the pointer positions actually occur, and each sub-world spawns its own intelligent analyses of their own particular pointer, so each consciousness is trapped by, or born into if you prefer, a kind of coherent sub-world of the incoherent and non-interfering many worlds. You can see how it is a little hard to talk about the differences in these interpretations without talking about consciousnesses. Finally, the third main school is deBroglie-Bohm, which says that there is only one world, and it is computable and deterministic, we just don't have access to the information one needs to do the computation (which is called a "pilot wave" and is not directly observable at present).

To summarize, the role of consciousness is quite substantially different in the three interpretations. To Copenhagen, consciousness is paramount, because quantum mechanics is just a tool that the consciousness uses to predict outcomes, and some elements of the outcome are simply not describable so must be treated as random. To many worlds, the consciousness is a kind of minor player in the vast array of many worlds (and I do mean vast), because some of the worlds spawn consciousnesses and some don't, and the physics doesn't really care if there's a consciousness in there or not. To deBroglie-Bohm, the consciousness is neither paramount nor minor-- the deterministic physics is the "truth" of the situation, just as in many worlds, but now there is just one world that is being determined, and so that one world must be the home to all the consciousnesses.

The bottom line is, all these interpretations make the same successful predictions, so choosing between them (or ignoring them altogether) is really a matter of personal taste. The choice is very often motivated by how you like to think about the role of consciousness, and that's why consciousness continues to play a key role in, not the predictions of quantum mechanics about measurable outcomes, but in understanding what quantum mechanics is really describing, what it really is.

 

[BruceW]

Ken G, the standard interpretation of QM doesn't depend at all on consciousness.
The macro pointer instrument itself causes wavefunction collapse.
It doesn't require a human looking at it to cause collapse.
This is why human consciousness is not required for the standard interpretation.

 

[G01]

Here's the crux of my problem with the question labelled 2) in your post above. For reference,

2) "Is consciousness/awareness/intelligence/perception involved in giving meaning to the entire concept of doing a measurement that collapses a wave function?

This just seems to boil down to the old "If a tree falls in the woods..." question.

Of course, we need conscious observers to do science. Without conscious observers no one would be around to question how nature works. Yes, I get that.

However, I still think it's entirely beside the the point regarding the nature of consciousness in QM. Consciousness does not play a role in collapsing the wavefunction. No viable mainstream QM interpretation assigns human consciousness a role during measurement.

Thus, I don't see how changing the conversation into a conversation about the philosophical role of consciousness in doing science helps. I think they are separate conversations.

 

[thoughtgaze]

I think people should admit ignorance on this subject. The question inherently involves a physically ill-defined and ill-understood object, i.e. consciousness.

Does consciousness create our reality? Or is it the other way around? Do they coexist in some sort of "symbiotic" fashion? We don't know the answers to these questions. All people are doing is speculating on something they know very little about, reasonably so; consciousness is one of the most difficult things to understand even though it is the heart of our experience.

 

[Ken G]

This just seems to boil down to the old "If a tree falls in the woods..." question.

Well, you are taking us in this philosophical direction, but I did not because this has nothing to do with the usual question about the tree in the woods. That's the whole reason I introduced the concept of a hypothetical observer. Instead, what this means is we can say the tree falls if we can imagine what a hypothetical observer would experience. But if we cannot, then exactly what does "a tree falls in the woods" even mean? So everything we say about reality, including every equation of physics, comes through the filter of "what if we were there", that's the point here. This is why we call physics an empirical science, pure and simple.

However, I still think it's entirely beside the the point regarding the nature of consciousness in QM. Consciousness does not play a role in collapsing the wavefunction.

No, but it plays a role in what we mean by "collapsing the wavefunction." I described quite specifically how that phrase means three different things in interpretations that treat consciousness in three different ways.

Thus, I don't see how changing the conversation into a conversation about the philosophical role of consciousness in doing science helps. I think they are separate conversations.

You must be a realist, who sees a firewall between the observer and the observed. It's very hard to get a consistent view of quantum mechanics that way. Quantum mechanics is fundamentally a language, a language with a mathematical syntax that deals in testable predictions, but a language all the same. Your position is basically "I don't see how the mind of the speaker has anything to do with the language they are speaking." I don't think that's very likely.

 

[Ken G]

Ken G, the standard interpretation of QM doesn't depend at all on consciousness.
The macro pointer instrument itself causes wavefunction collapse.

There are two very different stages to "wavefunction collapse" that people often misconstrue, and only the first stage is independent of interpretation. It sounds like you are talking about simple decoherence, by which a pointer and an outcome become coupled, and pick up random or noncoherent phases with respect to all other pointer/outcome couplings. This is called a "mixed state" when you project onto either the pointer or the quantum alone, but it is still a unitary state when you consider the full wavefunction of the combined system. So far we have nothing but basic quantum mechanics, no need for any interpretation, and it describes the situation before anyone looks at the outcome of the experiment.

Enter the next stage of the "collapse", when someone does look at the outcome. Now we desperately need an interpretation, because we no longer have a unitary state. We just have one pointer result, and one quantum state. What happened to the rest of the unitary state-- it's gone! This is clearly the part that involves consciousness, because consciousness is the whole reason we need an answer to this part of the question. As I said above, if we adopt many-worlds, we solve the problem by saying the consciousness is in some sense "mistaken" or "under an illusion" that the unitary state is gone. The full state of "many worlds" is still there, the consciousness is just not aware of it. So absolutely yes, this is all about perception and consciousness.

Copenhagen also requires that we address the consciousness issue. Here, we do not say the consciousness is deluded or tricked, because we say that physics is ultimately a task for that consciousness. Instead we say the mathematics is not the reality. Bohm says the consciousness also gets the reality, but there's a different mathematics that is the fundamental reality-- though one we cannot directly interact with, it seems (some wonder if maybe we can).

So as you can see, consciousness is all over the hard problem of collapse, it's just not in the easy problem-- the decoherence.

 

[danR]

I believe the historical development of this question has shifted from some kind of special conscious observer required, to measurement, to mere interaction.

Given that particles are constantly interacting everywhere on some quantum level, could we say that in an ensemble of nucleons in a nucleus, their wave functions are being collapsed on a rather regular basis?

 

[Ken G]

You are also talking about the easy problem of collapse, the decoherence. That does not tell us how we get a single outcome, that question remains entirely unanswered and very much does require an interpretation choice, or you have to punt the question entirely (which is more or less choosing CI interpretation).

 

[G01]

You must be a realist, who sees a firewall between the observer and the observed. It's very hard to get a consistent view of quantum mechanics that way. Quantum mechanics is fundamentally a language, a language with a mathematical syntax that deals in testable predictions, but a language all the same. Your position is basically "I don't see how the mind of the speaker has anything to do with the language they are speaking." I don't think that's very likely.

I consider the observer to be part of environment the system interacts with. However, I do hold the view that the world works the same way when we're not observing it. In that sense, I am a realist. I agree that language and the mind are intertwined, but outside world we describe with that language won't change based on that language or that mind. Nature will do it's thing, regardless of how we try to describe it. (Now we are almost totally in the realm of philosophy.)

And, yes you are correct that a true description of the measurement process has to involve a some form of non-unitary evolution, but I don't see how this implies that consciousness must be involved. Yes, I agree that the description of quantum measurements is incomplete, but what scientific evidence points towards consciousness playing an active role? Why is the nonunitary evolution "clearly the part that involves consciousness." True, we don't understand this process yet, but why is consciousness the only way to solve this problem?

In other words, what is it about nonunitary evolution that screams consciousness, other than the fact that we don't yet have a mechanism for it?

 

[atyy]

Quantum mechanics is fundamentally a language, a language with a mathematical syntax that deals in testable predictions, but a language all the same. Your position is basically "I don't see how the mind of the speaker has anything to do with the language they are speaking." I don't think that's very likely.

But classical mechanics is also a language. Yet it doesn't seem to have a measurement problem. I guess that shows that only modern physicists are conscious;)

 

[Ken G]

I consider the observer to be part of environment the system interacts with. However, I do hold the view that the world works the same way when we're not observing it. In that sense, I am a realist.

I agree that the world works the same way when we are not observing it, my point is that when we are not observing it, we still have to imagine that we are, or else we don't even have a language to talk about what the world is doing that we are not observing! And if we have to imagine we are, we are automatically conjuring all the experiences we have when we are actually observing the world. So we never escape that interaction, the world on one side and us on the other, it informs all our concepts. Then we get the real kicker: the distinction between us and the world is artificial, as we are part of the world. So ultimately, the world is observing itself, in a way that involves our intelligence/perception/consciousness, and this is inescapable in any language that we even begin to use to establish physical truths.

I agree that language and the mind are intertwined, but outside world we describe with that language won't change based on that language or that mind.

There's nothing to "change" if there is not that language. The language doesn't change it, the language is it.

Nature will do it's thing, regardless of how we try to describe it.

So is physics about what nature is doing, or how we describe what nature is doing? That's the key point-- we have no idea what nature is doing, beyond our own efforts to describe it, and we are involved in that in a fundamental way. I like to say that science is a conversation between us and nature, not a movie staring nature. It just is, we can see this just by looking at it critically.

And, yes you are correct that a true description of the measurement process has to involve a some form of non-unitary evolution, but I don't see how this implies that consciousness must be involved.

The evidence for that is that the key interface between the easy decoherence part of the collapse, and the hard single-valued outcome part of the collapse, is the moment when the experiment goes from being just an apparatus coupled to a system, to an apparatus that someone has looked at the outcome of. At no earlier time do we have a single-valued result, instead we have a statistical range of outcomes. This is the moment that the reality goes from "we know it's one of these but we don't know which", which can easily be embedded into a unitary combined wavefunction, to "it's this one", which cannot (without many worlds, which seems like a high price to pay for unitarity).

In other words, what is it about nonunitary evolution that screams consciousness, other than the fact that we don't yet have a mechanism for it?

The above.

 

[BruceW]

Ken G, I simply disagree with you on this.
In my definition of collapse, collapse happens when a classical object makes a 'measurement'. (The definition of a classical object is a vague notion, so that is a bad point about this interpretation).
And then your definition of collapse requires a conscious observer. The idea of what is 'conscious' is even more physically vague.
Therefore your interpretation is less scientifically good.
Do you see the point I'm trying to make?
But either way, there is no experiment in the near future that will be able to distinguish between our two interpretations.

 

[Ken G]

And then your definition of collapse requires a conscious observer. The idea of what is 'conscious' is even more physically vague.

Actually, it follows immediately, there is no need to define conscoiusness at all. I showed there really are two stages to the collapse, and the second stage is when someone registers the outcome. That's all you need, the outcome to be registered by someone who is thinking about that outcome. You never need to define consciousness, because the person trying to understand the answer has no need to define their own consciousness to do so. But they do have to have it, or we can leave the situation in the decohered mixed state. Like a deck of cards that was shuffled and dealt but never played-- it remains a mixed state forever, until we at least imagine looking at the cards.

But either way, there is no experiment in the near future that will be able to distinguish between our two interpretations.

Every experiment we ever do results in an outcome, not the mixed state of quantum decoherence. Do you understand the difference between a mixed state and a known outcome? That is all that is required, I do not require any kind of interpretation, this is just an observed fact. The interpretation's job is to explain what is going on when we go from a decohered mixed state to a particular outcome, but one does not need an interpretation to recognize the problem.

 

[G01]

Actually, it follows immediately, there is no need to define conscoiusness at all. I showed there really are two stages to the collapse, and the second stage is when someone registers the outcome.

That's all you need, the outcome to be registered by someone who is thinking about that outcome. You never need to define consciousness, because the person trying to understand the answer has no need to define their own consciousness to do so. But they do have to have it, or we can leave the situation in the decohered mixed state. Like a deck of cards that was shuffled and dealt but never played-- it remains a mixed state forever, until we at least imagine looking at the cards.
Every experiment we ever do results in an outcome, not the mixed state of quantum decoherence. Do you understand the difference between a mixed state and a known outcome? That is all that is required, I do not require any kind of interpretation, this is just an observed fact. The interpretation's job is to explain what is going on when we go from a decohered mixed state to a particular outcome, but one does not need an interpretation to recognize the problem.

The measurement process happens in two steps, a unitary evolution, followed by as as yet not understood non unitary evolution. We all agree on this.

However, you are not making any scientific case whatsoever that consciousness must be involved or that it is better than any other possible explanation for a non unitary process.

You are assuming that the nonunitary process must coincide with someone looking at the measuring device. Nothing in quantum mechanics requires the non unitary process to happen when someone "looks at" the pointer needle on the measurement device. I could just as easily take the point of view that the non unitary part of the process would happen regardless of whether I, you, or anyone else is looking at the pointer needle on the device.

In order to say that consciousness must be involved you are going to have to show some reason why your interpretation is more scientifically valid that mine. Can you offer any new testable prediction not offered by any other interpretation's description of the measurement process?

If not, then BruceW's description of the problem above is valid. We are just replacing one poorly understood, vaguely defined concept with another and the only justification for that replacement is a interpretation of quantum mechanical measurements that we all don't agree on, and there is not any scientific reason why we should.

 

[Ken G]

The measurement process happens in two steps, a unitary evolution, followed by as as yet not understood non unitary evolution. We all agree on this.

It should be pointed out that some would not-- "many worlds" practitioners don't think that non-unitary step ever happens. They think it is an illusion of the conscious physicist, who sees only one world (but again, the role of consciousness is very clear). Also, Bohmian practitioners don't think there is ever the unitary evolution-- they think that not only is the unitary wave function deterministic, but also the underlying "classical trajectory" is also deterministic. So the unitary evolution of the wavefunction, to a Bohmian, is just a kind of cloak that is concealing the underlying deterministic evolution. But I think what you mean is that the current members of this discussion all seem to follow the Copenhagen-esque view that something non-unitary must be addressed in the final stage of the measurement, so that very much puts us on the same page.

However, you are not making any scientific case whatsoever that consciousness must be involved or that it is better than any other possible explanation for a non unitary process.

That is simply not true, I am making a logically bulletproof case that consciousness is involved. The reason is that without consciousness, it is perfectly clear that the non-unitary step is never needed. This even holds in the classical realm (since the step we refer to is after the measurement, so it survives the classical translation). Consider a deck of cards that is shuffled and dealt. There is absolutely no physical theory of those cards that will ever need to treat them in any way other than as a mixed state of statistical outcomes (a "density matrix", if you will) if the cards are never looked at by anyone. Do you dispute that? So there is nothing to explain if the cards are not looked at-- we have a mixed state, period.

When we deproject that mixed state into quantum observables, we find we have a perfectly good unitary evolution there. There just isn't anything nonunitary if no one looks at the cards, this is just demonstrably true-- that's what happens with decoherence and von Neumann's approach to measurement, you get a unitary state of couplings between macro pointers and quantum states, there's no correlations between the different pointings, but there's also nothing nonunitary there because you don't yet have a single pointer outcome.

Until you look at the outcome, then all of a sudden you have something nonunitary on your plate that you have to explain. The role of your consciousness in that story is inescapable, the problem simply never comes up without it.

Nothing in quantum mechanics requires the non unitary process to happen when someone "looks at" the pointer needle on the measurement device.

Correct, just like nothing in classical statistical mechanics requires that a particular outcome occurs when we roll a pair of dice. It isn't the theory that requires a particular nonunitary outcome, it is experience that requires that, and the experience is the experience of a conscious observer. You are looking for consciousness in the theory, I'm saying the consciousness is in the experience of the outcome. The conciousness is the reason the question even comes up experimentally.

To see that even more clearly, let's say I program a computer to analyze experimental outcomes. The computer is just an extension of the apparatus, there's still no reason to ever include anything but mixed-state statistical analyses. The "theory" of such an analysis would be entirely statistical, even in the classical realm, there would never be the slightest requirement for the theory to account for particular outcomes ever happening. This is just demonstrably true, I could actually do it easily if I had the necessary machinery. I could build a machine to analyze a trillion dice rolling all at once, analyze the statistics and make a prediction. Then I could build another machine to test the statistical predictions of the first machine by actually rolling a trillion dice, all at once, and sure enough, the theory works great. All this without ever having a consciousness enter, and the other thing that would never need to enter is any concept of a "particular outcome" in any individual die roll.

The theory was never designed to understand or predict a particular die roll, and so the concept never even appears in the theory. So it would be with quantum mechanics without any consciousnesses present to ask: "but why this outcome this time?" That is a question that the theory of quantum mechanics has no interest in and cannot address, it comes not from QM but from the experience of a conscious individual who has a question that QM cannot answer. Remove that consciousness and we have a unitary evolution once again. Put another way, we'd have the many-worlds interpretation that predicts every experiment correctly and without a hitch because there is no one to object "but that's not what I experience, I don't experience many worlds."

I could just as easily take the point of view that the non unitary part of the process would happen regardless of whether I, you, or anyone else is looking at the pointer needle on the device.

But you wouldn't need to take any such point of view, the question doesn't even come up-- the unitary piece would completely resolve everything that needs to be resolved in a universe with no consciousnesses. Many worlds would be the obvious interpretation because there'd be no reason to even imagine a different one. The shuffled cards are always a mixed state, there'd be no such thing as the "actual lay of the cards", no physics would require that concept without a person looking at their hand.

Can you offer any new testable prediction not offered by any other interpretation's description of the measurement process?

Yes. Shuffle a deck and pick out a card. Look at the card. I predict you will see a particular card. That prediction is not made by any interpretation of physics that does not include consciousness, because a mixed-state statistical treatment of the outcome of that process is always completely adequate until you actually have a consciousness there. Again, note that I never need to understand consciousness, or model it, or include it in my theory, I merely recognize the observational fact that physics needs no concept of a particular outcome, rather than a statistical distribution, until I have a consciousness involved.

 

[bugatti79]

During the measurement, the quantum system must interact with the measurement device. The measurement device is inherently classical, being built from many atoms. So, the effect is such that when the measurement happens, our quantum system interacts with the measuring device. This causes decoherence to set in in our quantum system and the wavefunction collapses. Or, in other words, the interaction with the measuring device "washes out" the the interference terms of the systems density matrix.

1) but what is this 'interaction' between the quantum system and the measuring device?

2) how does the quantum system know the measuring device is there?

 

[G01]

Yes. Shuffle a deck and pick out a card. Look at the card. I predict you will see a particular card. That prediction is not made by any interpretation of physics that does not include consciousness, because a mixed-state statistical treatment of the outcome of that process is always completely adequate until you actually have a consciousness there. Again, note that I never need to understand consciousness, or model it, or include it in my theory, I merely recognize the observational fact that physics needs no concept of a particular outcome, rather than a statistical distribution, until I have a consciousness involved.

We may just have to agree to disagree on this topic.

I see how, working in your interpretation, consciousness would be related to the nonunitary portion of the measurement, but I don't think you've shown that this interpretation is better than many worlds or the CI. Perhaps you have some philosophical arguments, but the above does not count as a new testable prediction.

If instead you said, "Look pick a card. The probability that you pick that card is X, the probability that you pick this other card is Y."

That's the type of prediction that right now, QM has trouble with. We want to be able to predict which classical state in which our system ends up, and the probability that it ends up in that state over any other. We need a model for this portion of the measurement interaction.

However, you are not supplying a model. You are taking the known fact that the system ends up in a certain state, and saying, "Hey, look. The system has to end up in a particular state when we look at it. Physics describes how systems end up in particular states, but is developed by our minds. Therefore physics is inseparable from our conscious minds. Our conscious minds must have something to do with this measurement."

That is not a scientific argument. It is an interpretation of a not completely understood area of QM that is supported only by the fact that it is consistent with your philosophy of science and language.

You don't provide any mechanism by which consciousness collapses the WF. You can't provide any quantitative prediction about which state the system will collapse into. You only offer the trivial statement that the system will end up in a particular state. That's not a useful scientific prediction. At best, it is a check that your are still consistent with nature.

Until those last two conditions are met, it's not true to say that consciousness must be involved in the hypothetical, predictive, scientific theory of quantum measurement.

I'm not saying I have the answers. I don't. Most of the time I tend to be a "shut up and calculate" kind of guy. When I do talk about this stuff, I tend to focus on the unitary part of the process that we understand. I'm as confused as the next guy. You philosophy of science and language even has some appeal to me.


However, it's not correct to say that something must scientifically be the case if that something provides no mechanism, no real predictive power, and whose support is derived from a philosophy of science that not everyone must agree on.

1) but what is this 'interaction' between the quantum system and the measuring device?

2) how does the quantum system know the measuring device is there?

Hi bugatti79. Sorry, we are talking in very abstract terms here. The actual interaction will depend on what your actually measuring. For example, if we are measurement spin moments of nuclei, the interaction would be magnetic in nature.

As far as how does the system "know" about the measuring device: Well it "knows" because of the interaction. Interacting with the measuring device is what we mean when we say, the system "sees" the measuring device or "knows about" the measuring device.

 

[Ken G]

We may just have to agree to disagree on this topic.

I see how, working in your interpretation, consciousness would be related to the nonunitary portion of the measurement, but I don't think you've shown that this interpretation is better than many worlds or the CI.

What I've tried to stress is that what I'm talking about is not an alternative interpretation than CI or many worlds, it is an inherent component to CI and many worlds. Indeed, without the consciousness of the physicist to register an experimental outcome, many worlds is not even an interpretation any more, it is just quantum mechanics, and there simply is no CI interpretation. This is because there is no CI without physicists, CI is all about what the physicist is doing, and what CI means by a physicist is someone conscious (whether it's the consciousness, or intelligence, or ability to register perceptions, is all the same in what I'm saying, I've made no effort to distinguish those concepts because we don't know how to).

If instead you said, "Look pick a card. The probability that you pick that card is X, the probability that you pick this other card is Y."

That's the type of prediction that right now, QM has trouble with.

No, that is exactly the kind of prediction that quantum mechanics has no trouble with, so much so that you don't even need an interpretation of quantum mechanics to answer that. What QM has trouble with, and why you need interpretations, is the next step, when you don't say the probability is X, but you say the card was actually X. That's where the problem appears, and that's also the step adjudicated by the consciousness of the experimenter.

However, you are not supplying a model.

The model to do that is already here, that's what quantum mechanics does, that's what the theory of decoherence does. All of that is before you even get to an interpretation, all the interpretations include that part, as does the things one learns in a textbook on quantum mechanical calculations.

You don't provide any mechanism by which consciousness collapses the WF.

That is correct, nor does the logic of my argument require any such mechanism. And if someone says "if there's no mechanism, it isn't science," then they just don't understand science. Newton gave no mechanism for why F = ma, after all. Science works from the other direction-- we first decide what we need to explain, and then we try to explain it with models, and sometimes the models involve enough other elements to be considered a mechanism. But the real point is, first we have what we must explain, then we try to explain it. What I've demonstrated is that what we must explain is how a consciousness comes up with a single outcome, when every theory of physics we have only tells us statistical distributions over an ensemble of outcomes. This is just plain true, all physics theories have that property, whether quantum mechanical or classical. You are welcome to suggest one that does not if you doubt it.

You can't provide any quantitative prediction about which state the system will collapse into.

Obviously, something like that would not only be a different theory from quantum mechanics, it would contradict quantum mechanics.

However, it's not correct to say that something must scientifically be the case if that something provides no mechanism, no real predictive power, and whose support is derived from a philosophy of science that not everyone must agree on.

That is just not true. There are a whole host of things that scientifically must be the case, yet provide no mechanism and no predictive power. That is indeed the very building block of science-- I would call it a "raw observation." And I'm saying that the role consciousness plays in obtaining a particular outcome, when all the theories of physics produce only statistical distributions (when you include the necessity of noise and uncertainty), is a raw observation. You are asking me to explain it before you will recognize it as true, and I'm saying that you are asking science to work backwards. We don't start with explanations, we start by recognizing what it is we need to explain.

 

[Fredrik]

I just want to say that I agree with Ken's comments about consciousness. The key point is that an interaction that doesn't produce a persistent record that a human can interpret as a specific result of the interaction wouldn't be considered a measurement. This isn't a magical property of consciousness, or some highly speculative pseudo-theory that tries to explain something. It's just a statement about what sort of interaction we would consider a "measurement".

 

[Ken G]

That's an excellent way to put it Fredrik, thank you. I think a lot of the problem was with all the magical properties attributed to consciousness, and I am referring only to the central issue that you also focus on: the ability for a consciousness to register an outcome. This is a very fundamental problem for physics, because physics is only good at predicting probability distributions when you call on it to achieve high precision, yet when we test those predictions, we do it with ensembles of fairly precise individual outcomes. There's a kind of disconnect there which exists even in purely classical contexts, yet is very reminiscent of the famed "Heisenberg gap" in quantum/classical interfaces.

 

[bugatti79]

We may just have to agree to disagree on this topic.

I see how, working in your interpretation, consciousness would be related to the nonunitary portion of the measurement, but I don't think you've shown that this interpretation is better than many worlds or the CI. Perhaps you have some philosophical arguments, but the above does not count as a new testable prediction.

If instead you said, "Look pick a card. The probability that you pick that card is X, the probability that you pick this other card is Y."

That's the type of prediction that right now, QM has trouble with. We want to be able to predict which classical state in which our system ends up, and the probability that it ends up in that state over any other. We need a model for this portion of the measurement interaction.

However, you are not supplying a model. You are taking the known fact that the system ends up in a certain state, and saying, "Hey, look. The system has to end up in a particular state when we look at it. Physics describes how systems end up in particular states, but is developed by our minds. Therefore physics is inseparable from our conscious minds. Our conscious minds must have something to do with this measurement."

That is not a scientific argument. It is an interpretation of a not completely understood area of QM that is supported only by the fact that it is consistent with your philosophy of science and language.

You don't provide any mechanism by which consciousness collapses the WF. You can't provide any quantitative prediction about which state the system will collapse into. You only offer the trivial statement that the system will end up in a particular state. That's not a useful scientific prediction. At best, it is a check that your are still consistent with nature.

Until those last two conditions are met, it's not true to say that consciousness must be involved in the hypothetical, predictive, scientific theory of quantum measurement.

I'm not saying I have the answers. I don't. Most of the time I tend to be a "shut up and calculate" kind of guy. When I do talk about this stuff, I tend to focus on the unitary part of the process that we understand. I'm as confused as the next guy. You philosophy of science and language even has some appeal to me.


However, it's not correct to say that something must scientifically be the case if that something provides no mechanism, no real predictive power, and whose support is derived from a philosophy of science that not everyone must agree on.




Hi bugatti79. Sorry, we are talking in very abstract terms here. The actual interaction will depend on what your actually measuring. For example, if we are measurement spin moments of nuclei, the interaction would be magnetic in nature.

As far as how does the system "know" about the measuring device: Well it "knows" because of the interaction. Interacting with the measuring device is what we mean when we say, the system "sees" the measuring device or "knows about" the measuring device.

I see, so measuring quantum systems involves some form of electromagnetic fields and hence its a 'physical' interaction.

Ok, thank you for the clarity.

 

[G01]

The key point is that an interaction that doesn't produce a persistent record that a human can interpret as a specific result of the interaction wouldn't be considered a measurement.

That's an excellent way to put it Fredrik, thank you.

OK. I agree with this. But I think I'm getting stuck on the use of the word consciousness. Let me give an example:

We build a robot that sets up a two state system in some superposition. The robot then measures the system, finding it in some eigenstate. It records the eigenstate. The wavefunction collapses when we read the data sheet the robot prints out then? Or the second, non-unitary portion of the measurement process happens when we read the data sheet?

The way I would have looked at this scenario before this conversation is as follows:

The whole measurement process happens when the robot interacts with the system. The system collapsed into a random eigenstate, though the robot is not conscious. It is in a definite eigenstate, though I can't know what one it ended up in.

 

[G01]

it is in a definite eigenstate, though i can't know what one it ended up in.

Yes. Shuffle a deck and pick out a card. Look at the card. I predict you will see a particular card. That prediction is not made by any interpretation of physics that does not include consciousness, because a mixed-state statistical treatment of the outcome of that process is always completely adequate until you actually have a consciousness there. Again, note that I never need to understand consciousness, or model it, or include it in my theory, I merely recognize the observational fact that physics needs no concept of a particular outcome, rather than a statistical distribution, until I have a consciousness involved.

ahhhhhhh...OK. I see your point. You are talking about a much broader problem in physics than I originally grasped.

 

[Ken G]

We build a robot that sets up a two state system in some superposition. The robot then measures the system, finding it in some eigenstate. It records the eigenstate. The wavefunction collapses when we read the data sheet the robot prints out then? Or the second, non-unitary portion of the measurement process happens when we read the data sheet?

Yes, the robot participates in the decoherence, but that just puts the robot into a state that physics would need to describe as a probability distribution, or mixed state. We don't have a robot, just like the original system, in a particular state until the consciousness shows up to register that.

The way I would have looked at this scenario before this conversation is as follows:

The whole measurement process happens when the robot interacts with the system. The system collapsed into a random eigenstate, though the robot is not conscious. It is in a definite eigenstate, though I can't know what one it ended up in.

Yes, I agree with all that with one caveat: where does the "definite eigenstate" come from? The equations of physics would never want to give you a definite eigenstate, they'd always want to give you a probability distribution of all of them. But we can't just leave it at that, even though it's what our equations give us, because of our consciousness, and our consciousness doesn't perceive that distribution. This is the only reason we're not all "many worlds" quantum mechanicians. I think you are seeing what I mean now.

 

[G01]

I think you are seeing what I mean now.

Yes, we were having completely separate conversations for most of page 2. Special thanks to Fredrick for his post. I had to reread it several times. Then, I think I finally realized what you guys were saying.

So, the part of the measurement process represented by the nonunitary evolution in quantum mechanics exists throughout all of physics, I see what you mean. We get specific results, even though our process of modeling nature only gives us probability distributions.

So, what your saying is that our method of modeling nature is incomplete or limited, and the measurement problem is a symptom of that?

 

[BruceW]

Ken G - you seem to have been confusing the term "mixed state" with the term "pure state".
Pure state means each particle is in a superposition of eigenstates. Mixed state is when each particle is in one eigenstate, but there is a certain percentage of particles in each state, which gives a probability.

 

[BruceW]

For example, when you said that decoherence puts the robot into mixed state, you mean pure state.
Decoherence doesn't put something from a pure state into a mixed state.
In fact, it is collapse that puts something from a pure state into a mixed state.

 

[Fredrik]

Pure state means each particle is in a superposition of eigenstates. Mixed state is when each particle is in one eigenstate, but there is a certain percentage of particles in each state, which gives a probability.

It's also when you're dealing with a single object that you know is in an eigenstate, but you don't know which one. This is the situation when your measuring device has just measured something, but you haven't yet examined the result. Since you don't know which of its eigenstates it's in, you would have to describe it as a mixed state.

 

[Ken G]

So, the part of the measurement process represented by the nonunitary evolution in quantum mechanics exists throughout all of physics, I see what you mean. We get specific results, even though our process of modeling nature only gives us probability distributions.

Exactly.

So, what your saying is that our method of modeling nature is incomplete or limited, and the measurement problem is a symptom of that?

Yes. Though I don't think I'd use the term "symptom", because that suggests there is something sick in our approach. Instead, I think there is not anything sick in our approach, I think the whole point of our approach is to be limited. We already have the real world, it is not the goal of physics to recreate it. It is the goal of physics to understand a simplified version of it, so the limitations of physics and the limitations of our minds are perfectly matched. That's a feature of the measurement problem, not a bug of it-- we need to pass reality through the filter of measurement if we are to have any hope of understanding what aspects of it we can understand. The question then is, is the wavefunction, or the statistical distribution classically, the real thing we are passing through the filter of individual measurement, or are those just the mathematical tools we use to analyze whatever survives that translation? I tend to see it the latter way, so I would say the measurement problem is a reflection of the purpose of physics, but "symptom" certainly carries much of the same connotations.

 

[Ken G]

Ken G - you seem to have been confusing the term "mixed state" with the term "pure state".

No, the issue is that a closed system evolves unitarily, so remains in a pure state even if a decohering measurement occurs within it. However, when you project this pure state of the full system onto the quantum subspace, you get a mixed state, which is a probability distribution of all the possible outcomes, with incoherent relative phases.

What you are talking about comes later, after we have witnessed the state of the pointer. Then we reject all other members of the mixed state, and only then do we get a pure state for the quantum subspace. That entire process is what we mean by "measurement", but this last phase is nonunitary, not described by any equation of physics, and heavily reliant on our consciousness, as described above.

 

[Ken G]

For example, when you said that decoherence puts the robot into mixed state, you mean pure state.

No, the robot is still a subspace of the entire system. There is no way to get nonunitary evolution of the entire non-conscious system, if quantum mechanics is a good theory. This means the robot, being a subspace of a pure state, is projected onto a mixed state. That's what a mixed state is-- a projection of a pure state onto a subspace that has nonzero components on multiple eigenstates of the subspace.

Decoherence doesn't put something from a pure state into a mixed state.

Yes, it certainly does. For example, if we start with a vertically spin-up particle, and send it through a Stern-Gerlach apparatus inclined horizontally, and don't look at the outcome, we have a mixed state of a photon with two different spins going two different ways. Usually, you hear it said that the photon goes one way or the other, and is in a pure state, but that's because they are implicitly assuming someone looked. If no one looks, one can imagine a hypothetical observer serving the purpose, and say it is a pure state and we just don't know which, but that still requires applying the conscious action of that hypothetical observer. If you simply don't do that final step, it remains in a mixed state, just like an ensemble does (the ensemble is exactly the configuation where no one is considered to have looked at the individual members).

In fact, it is collapse that puts something from a pure state into a mixed state.

If you reread the above, you will see there are two steps to "the collapse." The first step, the decoherence, destroys correlations between the eigenstates, but cannot choose an outcome for the final eigenstate because it is a unitary evolution. The second step, the choosing of the final pure state, is nonunitary, does not appear in any of the equations of quantum mechanics, and is imposed manually by our conscious need to agree with our experience. So collapse first puts a pure state into a mixed state, and then consciousness somehow necessitates that we consider the mixed state as collapsed into a registered pure state-- unless we don't imagine anyone looking, and just leave it as an ensemble, in which case it is still a mixed state of the quantum subsystem.

 

[BruceW]

I think you took the term "mixed state" to mean a quantum superposition of pure states. But this is not the case.
A mixed state and a quantum superposition of pure states are two different things.
After decoherence happens, the two things are no longer physically distinguishable.
But this doesn't mean a quantum superposition becomes a mixed state after decoherence. Collapse is required for the mixed state to actually happen.

 

[Ken G]

I think you took the term "mixed state" to mean a quantum superposition of pure states.

Certainly not, a quantum superposition has definite phase relationships, leading to interference. That's exactly what a mixed state does not have. My use of the term is standard, as confirmed by Fredrik. Is there some particular point that is relevant to what you are saying?

A mixed state and a quantum superposition of pure states are two different things.

Yes, nor did I ever suggest otherwise.

After decoherence happens, the two things are no longer physically distinguishable.

That isn't making much sense. A quantum superposition of pure states is easily distinguished from a mixed state. After decoherence, the quantum superposition is the entire system, which we cannot even write the wavefunction for except in pictorial terms a la von Neumann. The mixed state is the projected substates, such as what often gets called the state of the quantum, a statistical representation that embodies the information we have about the quantum.

But this doesn't mean a quantum superposition becomes a mixed state after decoherence. Collapse is required for the mixed state to actually happen.

I really don't understand what you are trying to say. What I said was correct, and I clearly separated what was meant by the various aspects of "the collapse." A quantum superposition (the treatment of the quantum subsystem) does become a mixed state after decoherence, that's more or less the definition of decoherence. Perhaps if you have a particular point that you are making in regard to what you view as corrections, it would be clearer why you believe corrections are needed.

ETA: for those who are uncertain about the issues of pure and mixed states, you are welcome to simply consult the generally reliable authority of the Wiki on the matter: http://en.wikipedia.org/wiki/Quantum_decoherence
"Specifically, decoherence does not attempt to explain the problem of measurement. Rather, decoherence provides an explanation for the transition of the system to a mixture of states that seem to correspond to those states we perceive as determinate." (my bold)

 

[BruceW]

For example, if we start with a vertically spin-up particle, and send it through a Stern-Gerlach apparatus inclined horizontally, and don't look at the outcome, we have a mixed state of a photon with two different spins going two different ways. Usually, you hear it said that the photon goes one way or the other, and is in a pure state, but that's because they are implicitly assuming someone looked. If no one looks, one can imagine a hypothetical observer serving the purpose, and say it is a pure state and we just don't know which, but that still requires applying the conscious action of that hypothetical observer. If you simply don't do that final step, it remains in a mixed state, just like an ensemble does (the ensemble is exactly the configuation where no one is considered to have looked at the individual members).

You say that the particle which comes out of the Stern-Gerlach apparatus is in a mixed state, but it is in a superposition of pure states. (which is why it looked like you got the terms confused). P.S. I think you meant electron, not photon.

To clarify the other stuff I was saying in my last post:
After decoherence happens, the quantum superposition of the outcomes is (approximately) equal to a mixed state. This is why collapse happens sometime after decoherence.

 

[BruceW]

In your interpretation, you have defined collapse to finish when a conscious observer realizes what the outcome was.
But generally, collapse could have finished at any time between decoherence and the point when the person consciously realizes the outcome.

(P.S. I would say the term 'collapse' meant the nonunitary evolution. But I've used your meaning of the word in this post, so I can get my point across better).

 

[Ken G]

You say that the particle which comes out of the Stern-Gerlach apparatus is in a mixed state, but it is in a superposition of pure states. (which is why it looked like you got the terms confused).

I forgot to add that we need a pointer to say which way the photon went, built into the system. That's what creates the decoherence that creates the mixed state, the Stern-Gerlach just separates them.