B How Does the Observer Paradox in Anthropology Reflect on Cultural Understanding?

[bertrandrussell]

TL;DR Summary

Is there really an observer paradox?

When an anthropologist analyzes a culture he/she might influence it and so does not get an accurate understanding of that culture. Does that mean that there was no specific culture before the anthropologist arrived? NO! Similarly, why would someone say that a particle has no position before being measured?

 

[PeroK]

I guess that's curtains for QM? Not quite!

The concept that dynamic quantities like position, spin and polarisation have no well-defined values until measured becomes critical when you look at quantum entanglement. There's a thing called Bell's Theorem which puts a limit on the maximum correlation between measurements of entangled particles, given the assumption that they really have defined values before measurement (known as local hidden variables).

QM, however, predicts a stronger (classically impossible) correlation.

Tests of Bell's Theorem have shown that the predictions of QM hold - hence there cannot be local hidden variables at work.

 

[DrChinese]

Similarly, why would someone say that a particle has no position before being measured?

Quantum mechanics and anthropology are not as similar as you give credit. And in fact, your argument is a tautology. You assume that you seek to prove. There is no known requirement in QM that position and momentum exist and have well-defined values at all times. And in fact the Heisenberg Uncertainty Principle (HUP) strongly implies that they do not. Although the HUP by itself (i.e. without Bell) is not enough to answer this question.

In addition to PeroK's excellent comments: there are some interpretations of QM that allow for particles to have well-defined (although unknown) positions at all times. The Bohmian interpretation (BM) is one such. I won't discuss it here as there is a separate subforum for discussing foundations and interpretations. The thing some folks don't like about it is that it explicitly includes nonlocal interactions.

https://www.physicsforums.com/forums/quantum-interpretations-and-foundations.292/

 

[weirdoguy]

Does that mean that there was no specific culture before the anthropologist arrived? NO! Similarly, why would someone say that a particle has no position before being measured?

Why do you even compare those things? On what basis?

 

[Keith_McClary]

 

[Nugatory]

Similarly, why would someone say that a particle has no position before being measured?

That's how the math of quantum mechanics works.

It is very natural to think that's just a limitation of the math, that the particle has a position but we don't know what it is until we measure it just as your hypothetical tribe has a culture even if it's never visited by an outsider . It turns out, however, that there are subtle statistical differences between "the particle has a position but we don't know where yet" and "the particle has no position", these differences can be tested experimentally, the experiments have been done, and the results have decisively confirmed the quantum mechanical model.
For more about this, google for "Bell's Theorem" and review some of the 93 megabazillion other threads we have on the subject.

Be aware also that none of this involves the presence of a conscious observer, and the underlying cause is not the impossibility of measuring something without perturbing it; these are urban legends based on decades-old misunderstandings from the days when physicists were first trying to make sense of QM. Often when someone uses the term "observer effect" in a QM discussion, they've been misled by these urban legends.

 

[atyy]

Similarly, why would someone say that a particle has no position before being measured?

One can measure momentum also, and be tempted to say that the particle has a momentum before being measured. However, it turns out that in general it is not possible for the particle to have a certain sort of classical position and momentum before being measured. It might have one or the other, or neither. Since we don't know the position or momentum of a particle before measurement, and don't need this knowledge to make good predictions, we don't talk about those properties just to be practical.

In the standard, practical understanding of quantum mechanics, the observer has a special status. From the point of view of common sense, this seems incomplete, since the obserber should also be a physical system, and so should not have any special status. Thus there are attempts (none fully successful so far) to formulate a theory where the observer has no special status, and that is consistent with quantum mechanics in some regime.

 

[vanhees71]

I would formulate it a bit differently: Within quantum theory a particle always has the usual observables (position, momentum, angular momentum,...), but it depends on the state which observables take determined values. In general all there is are probabilities for the outcome of a measurement of some observable, given by the state the particle is prepared in. Some observables never take exactly determined values. That's the case particularly all observables which can take continuous values like position and momentum. For the latter the famous Heisenberg-Robertson uncertainty relation $\Delta x \Delta p_x \geq \hbar/2$ tells you that there's no state where $x$ is precisely determined, but you can prepare a particles $x$ component to be as precisely determined as you like, i.e., you can make the standard deviation $\Delta x$ arbitrarily small. Then the uncertainty relation tells you that the momentum component $p_x$ is very uncertain, i.e., it's standard deviation $\Delta p_x \geq \hbar/(2 \Delta x)$ gets very large.

 

[Tendex]

The concept that dynamic quantities like position, spin and polarisation have no well-defined values until measured becomes critical when you look at quantum entanglement. There's a thing called Bell's Theorem which puts a limit on the maximum correlation between measurements of entangled particles, given the assumption that they really have defined values before measurement (known as local hidden variables).

QM, however, predicts a stronger (classically impossible) correlation.

Tests of Bell's Theorem have shown that the predictions of QM hold - hence there cannot be local hidden variables at work.

That's how the math of quantum mechanics works.

It is very natural to think that's just a limitation of the math, that the particle has a position but we don't know what it is until we measure it just as your hypothetical tribe has a culture even if it's never visited by an outsider . It turns out, however, that there are subtle statistical differences between "the particle has a position but we don't know where yet" and "the particle has no position", these differences can be tested experimentally, the experiments have been done, and the results have decisively confirmed the quantum mechanical model.
For more about this, google for "Bell's Theorem" and review some of the 93 megabazillion other threads we have on the subject.

Be aware also that none of this involves the presence of a conscious observer, and the underlying cause is not the impossibility of measuring something without perturbing it; these are urban legends based on decades-old misunderstandings from the days when physicists were first trying to make sense of QM. Often when someone uses the term "observer effect" in a QM discussion, they've been misled by these urban legends.

This is all fine and I gave likes to both answers but some questions could be relevant here:
What's wrong(if anything) with interpreting the quantum statistics with different correlations not achievable classically as a hint that quantum math is simply better at getting those predictions than the classical way with classical probabilities, and that classical observables could use some modernization, or that rather than getting into philosophical cul-de-sacs about having or not realistic properties between measurements we just interpret it as a different kind of change in flight that doesn't have to be classical?

 

[PeroK]

This is all fine and I gave likes to both answers but some questions could be relevant here:
What's wrong(if anything) with interpreting the quantum statistics with different correlations not achievable classically as a hint that quantum math is simply better at getting those predictions than the classical way with classical probabilities, and that classical observables could use some modernization, or that rather than getting into philosophical cul-de-sacs about having or not realistic properties between measurements we just interpret it as a different kind of change in flight that doesn't have to be classical?

You'd need to come up with an alternative to classical probability theory. That isn't QM with its complex probability amplitudes.

 

[Nugatory]

What's wrong(if anything) with ...classical observables could use some modernization, or ... a different kind of change in flight that doesn't have to be classical?

That's Bell's theorem.
(1) The predictions of any theory that relies on classical observables, no matter how "modernized", must obey a particular set of inequalities. Quantum mechanics predicts and experiments confirm, that these inequalities are violated. Therefore any such theory fails.
(2) A change in flight that is determined only by the state of the particle as it heads towards the detector is just another case of #1. To violate Bell's inequality the two particles would have to change in flight in a way that is coordinated with the detector settings at the moment the particles reach their respective detectors. Not only is there no remotely plausible communication channel available (we do the experiments with different types of particles so we'd need a different mechanism for each type - changing the spin of an electron is something completely different than changing the polarization of a photon) but these experimentsare routinely done with the two detectors positioned in a such a way that the communication would have to faster than light.

 

[Tendex]

You'd need to come up with an alternative to classical probability theory. That isn't QM with its complex probability amplitudes.

Could you elaborate on the mathematical reason why it isn't?

 

[Nugatory]

Could you elaborate on the mathematical reason why it isn't?

Google for "Quantum mechanics negative probability" will get you started, but be aware that this will very quickly take you beyond the a PhysicsForum B-level thread.

 

[vanhees71]

You'd need to come up with an alternative to classical probability theory. That isn't QM with its complex probability amplitudes.

This I don't understand. QM clearly provides the probabilities for the outcome of measurements. What else is it then than an alternative to classical probability theory?

 

[PeroK]

This I don't understand. QM clearly provides the probabilities for the outcome of measurements. What else is it then than an alternative to classical probability theory?

That's @Tendex 's homework!

 

[Tendex]

That's @Tendex 's homework!

Well, I did my homework long ago. I suppose I also misinterpreted what you meant like vanheees71 did. The point being you must has have thought (like Nugatory) that I was referring to "negative probabilities", but no, QM doesn't have "negative probabilities", that's what Born's rule makes sure of. What I was aimint or hinting at with my questions was not a theory of "negative probabilities", but rather that we could interpret the improved way of obtaining probabilities by using complex amplitudes in the quantum way as a "clue" that there is a way to handle local measurements probabilities that might actually be at odds with considering probabilities themselves in a classical physics way.

 

[PeroK]

... there is a way to handle local measurements probabilities that might actually be at odds with considering probabilities themselves in a classical way.

You can look at Bell's theorem as well as I can: what alternative is there? Let's see your ideas! Personally, I don't see a lot of room for manoeuvre.

 

[Tendex]

You can look at Bell's theorem as well as I can: what alternative is there? Let's see your ideas! Personally, I don't see a lot of room for manoeuvre.

At this thread and subforum for sure there is absolutely no room, probably not in the foundations subforum either though I might follow this discussion opening a thread there. Anyway my ideas are not interesting in the least, I only have questions about some different scenarios other than the usual ones, sorry.

 

[cmb]

Summary:: Is there really an observer paradox?

When an anthropologist analyzes a culture he/she might influence it and so does not get an accurate understanding of that culture. Does that mean that there was no specific culture before the anthropologist arrived? NO! Similarly, why would someone say that a particle has no position before being measured?

It is a fundamental part of the scientific method to try to resolve this. There may always be argued to be an interference in the measurement of a thing when being measured. This is the way of things.

For example, I am currently designing a circuit to measure the resonant response of a circuit to very small parasitic impedance effects, but my measurement device will have, itself, a small parasitic impedance effect! Unavoidable. How am I to deal with this? (I mention this in case anyone has any good generic solutions to this!? ;) )

What you describe is everywhere in science and engineering, it is not a new discussion. The 'art' of good science is to know when you have derived a clear signal of the thing you are trying to get a clear signal of. One might even include 'observer bias' in the same discussion because the influence an observer has on results may extent to beyond the time the results are recorded.

 

[DrChinese]

It is a fundamental part of the scientific method to try to resolve this. There may always be argued to be an interference in the measurement of a thing when being measured. This is the way of things. ... One might even include 'observer bias' in the same discussion because the influence an observer has on results may extent to beyond the time the results are recorded.

This has nothing to do with "noise" or anything like your circuit problem. In a discussion of quantum observables, the "observer paradox" (if you want to call it a paradox) is this:

How an observer chooses to measure shapes quantum reality. A decision to measure position Qx with precision renders the value of momentum Px correspondingly imprecise, regardless of any previous momentum value. In fact, the [new] momentum could be any of a wide range of values, including very large ones. And similarly for ANY conjugate pairs of quantum observables. Note that this observer paradox does NOT apply - at all - to quantum observable pairs that are not conjugate, where both may be simultaneously measured to a very high level of precision without affecting the precision of the other. (An example of such a pair is Qx and Py.)

Just to be clear: there are many, many experimental demonstrations of the above.

Is it really a paradox that an observer shapes reality by choice of measurement basis? Entangled systems are extended in spacetime, and cannot be separated into quantum subcomponents until measurements are made (collapse if you use a collapse model). There is no paradox within QM itself, the paradox only appears when certain other assumptions are added to the mix. See the EPR paper (1935) with those assumptions, and the Bell paper (1964) attacking them.

 

[cmb]

This has nothing to do with "noise" or anything like your circuit problem. In a discussion of quantum observables, the "observer paradox" (if you want to call it a paradox) is this:

How an observer chooses to measure shapes quantum reality. A decision to measure position Qx with precision renders the value of momentum Px correspondingly imprecise, regardless of any previous momentum value. In fact, the [new] momentum could be any of a wide range of values, including very large ones. And similarly for ANY conjugate pairs of quantum observables. Note that this observer paradox does NOT apply - at all - to quantum observable pairs that are not conjugate, where both may be simultaneously measured to a very high level of precision without affecting the precision of the other. (An example of such a pair is Qx and Py.)

Just to be clear: there are many, many experimental demonstrations of the above.

Is it really a paradox that an observer shapes reality by choice of measurement basis? Entangled systems are extended in spacetime, and cannot be separated into quantum subcomponents until measurements are made (collapse if you use a collapse model). There is no paradox within QM itself, the paradox only appears when certain other assumptions are added to the mix. See the EPR paper (1935) with those assumptions, and the Bell paper (1964) attacking them.

I was reflecting on the general nature of science; it is a truism of the scientific method itself that the observer interferes with the observation, not only quantum physics, as the OP himself alluded to with the anthropologist example.

 

[vanhees71]

This is an a bit misleading formulation. Of course, measurements work by the interaction between a measurement apparatus with the object to be measured, and thus there's always an influence of the measurement on the object.

The uncertainty relation, however doesn't say anything about the ability measure precisely quantities but about the possibility to prepare quantum states. The Heisenberg-Robertson uncertainty relation for position-vector components and momentum in some direction, $\Delta x \Delta p_x \geq \hbar/2$ tells you that you cannot prepare a particle which has at the same time well-prepared positions and momenta. It doesn't limit in any way the ability to measure the one or the other observable with any precision you like.

It's also in no way surprising that you have to decide which quantity you measure when you plan your experiment. That's nothing specific to quantum theory.

 

[DrChinese]

I was reflecting on the general nature of science; it is a truism of the scientific method itself that the observer interferes with the observation, not only quantum physics, as the OP himself alluded to with the anthropologist example.

The OP is asking about the quantum world: do well-defined values for position and momentum (or any conjugate pairs) exist independently (simultaneously) of the act of observation? Without getting into the philosophical side, or the interpretations side: the short simple/general/standard answer is NO.

That is NOT the case with any other scientific observables I am aware of. My mass remains the same whether measured or not, and does not change with a measurement of my height or anything else about me.

 

[DrChinese]

To use vanhees71's example: if you prepare a group of particles with a specific momentum to a very tight range, then their positions will be extremely varied. Keeping their momentum tight, it is not possible for you to do anything to constrain the variability of their positions. That limitation ONLY applies to non-commuting quantum observables, and does not apply to cultures/anthropologists or other measurements.

That does not mean there are no other observer effects in science - there are, and those exist in the quantum world too. It just means that the Heisenberg Uncertainty relations have no analog in other areas of science other than the quantum world.

 

[cmb]

The OP is asking about the quantum world:

Actually, he asked two questions, going on the number of question marks, relating quantum physics to another scientific discipline. 'Asking about the quantum world' was the second question. I was just commenting on the relationship between the two questions. The general nature of the principle is reflected across science as a whole.

 

[DrChinese]

Actually, he asked two questions, going on the number of question marks, relating quantum physics to another scientific discipline. 'Asking about the quantum world' was the second question. I was just commenting on the relationship between the two questions. The general nature of the principle is reflected across science as a whole.

No, it's not, because the Heisenberg Uncertainty Principle is not a general principle related to science a whole. Either there is a question (which I answered) about particle position when not being observed (and the observer paradox, also the title of the thread, which I explained); and/or there is a separate question about indeterminacy in science generally. The second question then has nothing to do with quantum physics, and therefore does not belong in the Quantum Physics forum.

Trying to tie the two together is exactly the thinking that is wrong, and why I keep saying: unless a particle is being constrained to a very small volume of spacetime, generally it does not have a well-defined position. And if it is constrained to a very small space, it definitely does not have a well-defined momentum. Which is in fact pointing to the observer effect: the observer's choice of basis to measure creates the particle's state. It cannot be said to have had that state previously. And in fact there are an infinite number of non-commuting bases (and mixtures thereof) on which a quantum particle can be observed, each leading to a different resulting state. The observer chooses the reality, as opposed to simply revealing a state that is pre-existing (again, ignoring interpretations/philosophy which are discussed elsewhere here). If an observer chooses to observe in the one state that it could have been in previously (say a position eigenstate yielding a specific position eigenvalue), then it will get the same result as before. In any other of the [infinity minus 1] bases, the position and momentum will be correspondingly less defined. None of that has anything to do with noise, limits on testing equipment, contributions from the interaction between the observer and the observed, etc.). And importantly, it ONLY relates to conjugate observables and not to any others.

There are many terms that have specific and/or different meanings in Quantum Physics, and we make it a point to segregate their usage so we don't get into useless semantics debates. "Uncertainty" is one of those, so is "position/momentum" and the connection between the two.

 

[edmund cavendish]

Referring to the OP surely the point is not that the particle has no position before being measured but that it has multiple super positions. Shrodingers equation functioning quite happily whoever may or may not be looking. Enjoying the posts.

 

[Nugatory]

but that it has multiple super positions

And all that means is that in some problems it is convenient to write the function that gives the probability of finding the particle at a given location as the sun of other functions... and because position is a continuous observable, the wave function is never not such a superposition.