Undergrad How does observation affect reality

[bob012345]

That would give the same outcome everytime - groundhog electrons.

Actually, using a wave packet to model the electrons takes into account the inevitable inaccuracy in the initial conditions.

I don't think so since the electron path is unpredictable so running same one over and over should be the same as running different electrons one after another. It's not about slight differences in initial conditions, that's Chaos, it's more fundamental.

 

[PeterDonis]

Are you satisfied with the word "explained" in the above sentence?

Sure, why not?

 

[Wallis]

I don't at all understand what you're saying. Could you take some very simple example and demonstrate what you mean?

An excited atom emits a photon and the electron descends to a lower orbital, removing energy from the atom into the photon (or it's wave function.) This energy is stored in the wave function of the photon (for where else can it reside as the wave function spreads out and evolves over time?) Now, this atom was on the surface of a star, and the wave function (carrying the energy) spreads out through the Universe, subtending a considerable volume of likely interactions. The wave function interacts with another atom, exciting it (this time in my retina) and, on its decay, and I "see" the light owing to the collapse of the wave function within the atom in my eye. This means the little green man 500 light years away cannot see the photon now, even though he is "equidistant" from the star's atom.

However, the same happens when the photon attempts to warm two rocks, lying near me, and near the little green man. So, the collapse of the wave function does not need an observer, or any other special situation. It simply needs to transfer energy. Once the energy is transferred, the wave function must necessarily instantaneously collapse everywhere in the Universe, all at one, otherwise there would be no such concept as the conservation of energy. Ipso-facto, as we say.

If this were not the case, there would be no such thing as the conservation of energy, and believe me, the Universe would be a very different place.

Therefore, I propose that there is no such thing as the "observer effect." There is just plain old physics, doing the conservation of energy thing.

As for the "aspect style experiment", I was proposing that the community explore the nature of collapse. Is there a case where the photon has so much energy, it can excite both atoms? This is the sort of "exploration" I am proposing. Is the transfer "sharp" in time, or has it a "shape" in space-time that can be explored?

 

[Zafa Pi]

Sure, why not?

(I start with an off topic parenthetical comment: you're a smart and educated guy and I think that if you were offered 100 grand to write how I would respond to your question you would most likely succeed.)

Your signature hero, R.F., says in a lecture on the double slit experiment, "Nobody has succeeded in an explanation"

In an article on the D.S.E. he says, "If this seems very mysterious, you are not alone. Understanding what is going on here is in some sense equivalent to understanding Quantum Mechanics. I do not understand Quantum Mechanics." Feynman admitted that he never understood Quantum Mechanics. It may be true that nobody can understand Quantum Mechanics in the usual meaning of the word "understand."
(I actually think he is referring to the quantum phenomena of nature rather than QM as a theory)

If asked my girl friend why my toaster wasn't working and she said, "Because it's not plugged in." That would be an explanation and I would understand.

If I asked her why two masses attract one another, and she said because of Newton's Law of gravitation, that would not be an explanation for me (and it wasn't for Newton as well). That would be a mathematical rule as to how to predict their motion. (And, BTW, adding on the gravitational field or moving onto GR wouldn't help.)

So I think you see my reticence with the word "explained". I would have used "predicted". I don't think this is quibbling.

 

[vanhees71]

Feynman is of course a big entertainer, and if anybody has understood QT it was him. What he says in his textbook in the first few pages is, in my opinion, the best advice: QT works, because it is successful in describing all objective observations so far. There's a formalism and, via Born's rule, a way to relate it to the observations, and that's it. Anything beyond that is philosophicy gibberish, und you might know, what Feynman thought about philosophy making even simple things difficult to understand, and QT is difficult to understand even without the philosophers' confusion bought up about apparent "problems in understanding the measurement process". Evidence proves the opposite: Both experiment and theory are so well able to test the "weirdest predictions" of QT that it reaches amazing accuracy (in some quantities like the anomalous moment of the electron agreement between theory and experiment in 12 significant digits), that it is hard to understand, what these problems should be.

The "problems" and appararent "weirdness" is due to the fact that our intuition is trained by our usual environment consisting of macroscopic many-body systems. To understand them we only need such a "coarse-grained" view that almost all "weird" QT effects are averaged over (decoherence is very effective to the dismay of people who like to construct quantum computers).

So, to our best knowledge today, the problem is not a physical but a pedagogical one, i.e., to teach QT in a way that students can build an intuition about what's going on. The price to be paid is that this intuition is way more abstract than in classical physics, which seems to directly reflect our daily experience about the behavior of macroscopic bodies and the classical electromagnetic field much closer than the probabilistic interpretation of quantum states and the description of observables with self-adjoint operators on a Hilbert space.

The heuristic link between the classical and the pictures are symmetry principles. My advice to students of QT thus is, not to bother too much about the philosophical and metaphysical issues first, but rather to invest your time in learning about the mathematics of Lie groups and their (ray) representations on Hilbert space! For the beginners of QT (and also advanced practitioners in research, bothering with understanding observations in the real world, which is what physics finally is all about), indeed Feynman's "shutup-and-calculate approach" towards interpretational issues is the best advice he could give, and he did it indeed in a very enertaining and charming way in his textbooks (especially in the Feynman Lectures).

 

[zonde]

There's a formalism and, via Born's rule, a way to relate it to the observations, and that's it.

Born's rule is part of formalism not bridge to observations.
Bridge to observations is probabilistic interpretation of numbers given by Born's rule.

 

[vanhees71]

For me the probabilisitic interpretation is Born's rule. So we agree in this point :-).

 

[zonde]

For me the probabilisitic interpretation is Born's rule. So we agree in this point :-).

Hmm, but then how do you call the "amplitude squared" operation? Or you do QM calculations without use of "probability amplitudes"?

 

[vanhees71]

Ok, to be specific, for me Born's rule is:

The state of the quantum system is represented by a positive semidefinite self-adjoint operator $\hat{\rho}$ with $\mathrm{Tr} \hat{\rho}=1$. If $\hat{A}$ is the self-adjoint operator representing the observable $A$ and $|a,\beta \rangle$ is a complete orthonormal set of eigenvectors to the eigenvalue $a$ of $\hat{A}$, then the probability to find the value $a$ when measuring $A$ is
$$P(a|\hat{\rho})=\sum_{\beta} \langle a,\beta|\hat{\rho}|a,\beta \rangle.$$
A state is called pure if and only if $\hat{\rho}=|\psi \rangle \langle \psi |$. Then
$$P(a|\psi)=\sum_{\beta} |\langle a,\beta|\psi \rangle|^2.$$
In this sense the "wave function" $\psi(a,\beta)=\langle a,\beta|\psi \rangle$ is a "probability amplitude", but that are just words. The physics essence is Born's rule.

 

[PeterDonis]

Your signature hero, R.F., says

The fact that he is in my sig doesn't mean I accept him as an authority on everything.

However, in this particular case I agree with what I think he is saying, although what I think he is saying might not be what you think he is saying. I think he is saying that there isn't actually any difference between "predicted" and "explained"--if you have a model that makes good predictions, that's the best you can do. Trying to look for an "explanation" in addition to that is a fool's errand.

If asked my girl friend why my toaster wasn't working and she said, "Because it's not plugged in." That would be an explanation and I would understand.

On what basis? On the basis that you know electricity is required for the toaster to work and it has to be plugged into get electricity. But how do you know that? On the basis of a comprehensive theory that is ultimately based on Maxwell's Equations. And what are Maxwell's Equations? They are mathematical rules that tells you how to predict how electromagnetic fields, charges, and currents behave. Just like Newton's Laws, which you said were not explanations, but just mathematical rules.

This is why I don't see any significant difference between "predicted" and "explained". The things you are saying are valid "explanations" boil down to having mathematical rules that make good predictions, and I don't see any basis other than arbitrary choice for you saying that some mathematical rules don't count as "explanations" while others do.

So I think you see my reticence with the word "explained". I would have used "predicted".

In other words, you are not satisfied with the word "explained". You would prefer the word "predicted". I have no objection to the word "predicted".

However, you didn't ask me what you would or would not be satisfied with. You asked me if I was satisfied with the word "explained". I am. I don't see any significant difference between "predicted" and "explained", as I said above. But if you do, and you would rather use the word "predicted" for this discussion, that's fine with me. The choice of words does not affect the physics.

 

[Zafa Pi]

However, in this particular case I agree with what I think he is saying, although what I think he is saying might not be what you think he is saying. I think he is saying that there isn't actually any difference between "predicted" and "explained"--if you have a model that makes good predictions, that's the best you can do. Trying to look for an "explanation" in addition to that is a fool's errand.

Feynman has explained many things and I think he is a great explainer. But he says he has no explanation for the double slit or the correlations when measuring entangled entities. So I think he is making a distinction.

When I said, "If I asked my girl friend why my toaster wasn't working and she said, "Because it's not plugged in." That would be an explanation and I would understand."

On what basis? On the basis that you know electricity is required for the toaster to work and it has to be plugged into get electricity. But how do you know that? On the basis of a comprehensive theory that is ultimately based on Maxwell's Equations. And what are Maxwell's Equations? They are mathematical rules that tells you how to predict how electromagnetic fields, charges, and currents behave. Just like Newton's Laws, which you said were not explanations, but just mathematical rules.

I showed her your response and she (being no physicist) asked if Maxwell's Equations were good to the last drop.

14th century Europeans had an excellent model for predicting when the sun would rise and set over a year. It was a table gleaned from past measurements. Looking for an explanation is not a fool's errand. At what level one is satisfied with an explanation is a personal choice. Knowing the toaster wasn't plugged in was good enough for me.

I guess this is getting a bit philosophical (a Feynman bane) and off topic, but interesting.

 

[PeterDonis]

At what level one is satisfied with an explanation is a personal choice.

Exactly.

 

[OCR]

I showed her your response and she (being no physicist) asked if Maxwell's Equations were good to the last drop.

Then, I suppose it's time to start the... experiments ...? ..

 

[vanhees71]

The aim of the natural sciences is not to "explain" anything but to "describe" quantitatively and as accurately as possible phenomena in Nature and to compare the "descriptions" to quantitative and accurate observations. If a physicist says, s/he "explains" a phenomenon, s/he means that it is describable using more or less "fundamental theories" or well-justified approximations to it. To "explain" your toaster's function for a physicist means to understand it in terms of the adequate model applicable here, and that's finally the quasistationary approximation of Maxwell's equations, which themselves are an effective classical many-body approximation of QED. For an electrician all that matters here FAPP should be Ohm's Law, and that's the right level of description and "explanation" you need to get it working.

The same holds true for the electron and the double-slit experiment. QM is the right level of description here with the slit being described simply as boundary conditions for the Schrödinger equation. There's no "deeper explanation" than QT since it's the most fundamental theory we have today about nature. Also the electron itself is already a "most fundamental" building block in the sense of the Standard Model notion of what's called an "elementary particle" (a Dirac field realizing an irreducible representation of the Poincare group, including space reflections). So there's no "deeper explanation" of the electron's behavior than the probabilistic content of QT since there's no more fundamental level of description than QT. Maybe, one day, we'll have a "more fundamental explanation" for QT, which then may become an effective model in the sense of an approximation to the more fundamental theory (as are Maxwell's equations of classical electromagnetics are an effective theory for macroscopic many-body QED).

 

[stevendaryl]

The aim of the natural sciences is not to "explain" anything but to "describe" quantitatively and as accurately as possible phenomena in Nature and to compare the "descriptions" to quantitative and accurate observations.

I don't agree with that. As @Zafa Pi said, you can describe things arbitrarily accurately by having lots of tables.

The idea that the point is accurate description is not the reason anyone actually becomes a scientist. Kids ask: "Why does the moon go through phases?", they don't ask: "Exactly how many hours are there between successive full moons?" What the kid wants is understanding, not accurate predictions.

To me, making accurate falsifiable predictions is the way that we test our understanding of nature---it isn't a goal in itself, or at least, it's not the only goal. I'm talking about "goals" in the sense of "why anyone wants to study science, in the first place".

 

[vanhees71]

Perhaps, I should have said "theoretical descriptions". "Stamp collecting" (as Rutherford put it) is of course not, what I aim at as a theoretical physicist. However, I also believe that without suficient empirical input, we can't make progress in finding better and better theories. That's why the HEP community is so eager to finally find empirical facts about "physics beyond the Standard Model", because we can speculate a lot theoretically (like inventing SUSY, technicolor, and what not) but without empirical guidance we are lost in a huge parameter space even within some class of possible models like SUSY.

 

[stevendaryl]

I don't agree with that. As @Zafa Pi said, you can describe things arbitrarily accurately by having lots of tables.

The idea that the point is accurate description is not the reason anyone actually becomes a scientist. Kids ask: "Why does the moon go through phases?", they don't ask: "Exactly how many hours are there between successive full moons?" What the kid wants is understanding, not accurate predictions.

To me, making accurate falsifiable predictions is the way that we test our understanding of nature---it isn't a goal in itself, or at least, it's not the only goal. I'm talking about "goals" in the sense of "why anyone wants to study science, in the first place".

Often in science, having a very accurate description of some phenomenon is the beginning of the quest for a satisfying theory, not the end. The spectral lines of hydrogen were described very accurately by the Balmer series, prior to quantum mechanics. The point of the Bohr model was not to get a more accurate description, but a way of deriving the Balmer series, which was already known. The point of Heisenberg's and Schrodinger's work on QM was to get a less ad-hoc and more general way to get the results of the Bohr model. The Lorentz transformations and I think even the E = mc² of SR were known before Einstein. The point of his investigations was to understand how to reconcile Newtonian physics with the constancy of the speed of light, not to give a more accurate description of how energy and momentum relate to velocity.

Scientists (or at least, the famous ones) are always striving for more complete understanding of nature. The hope is that searching for understanding will also give us better, more accurate predictions, but more accurate predictions are not what motivates them.

 

[vanhees71]

Of course, I agree with that. All models and theories are always preliminary. One day, there might be better and better theories. The Balmer series was an empirical law, i.e., it was found by looking at data and fitting a mathematical formula to it. Nevertheless it was a very important step towards building a model, because it provided a clear evidence for a systematic pattern. Bohr's model was an ad-hoc description, and many physicists tinkered around with it (most famously the Sommerfeld school) using a wealth of spectroscopic data for many different atoms becoming less and less satisfied by it and finally discovering modern QT. Whether modern QT or not is the "final word" is not known yet. Maybe there's a better/more comprehensive theory one day (including the quantum theory of gravitation perhaps).

Still, it's not the aim of all this progress in theory building to "explain" nature but to "describe" it with as consistent as possible mathematical models. The very fact, why this works at all, is not explainable. It's just an empirical fact that it works with an astonishing success!

 

[Wallis]

Feynman's explanation of how mirrors work is a delight. (As I remember it, in Six Not So Easy Pieces, but please correct me.) But back to the topic, I claim observation cannot possibly affect reality. Only the transfer of energy from one place to another affects reality. There you have it, no observer effect whatsoever. The Universe seems to work fine unobserved. When we look millenia later, it seems to have got on fine without us.

There's a probability I understand the magnitude of the wave equation, but the real and imaginary components phase me :-)

 

[vanhees71]

It's the modulus squared of the wave function which gives probability distributions, nicely real and positive.

The point is, you cannot observe anything without exchanging energy between the measurement device and the system you measure. While for macroscopic systems you can make the influence of the measurement device negligibly small, that's not possible for microscopic objects like an electron. To probe it you need other particles or em. waves to scatter at it, and there's nothing "smaller" than an elemetary particle as the electron itself to make the influence of the measurement device arbitrarily small. That's why observation and measurement are so much more emphasized in QT vs. in classical theory.

 

[bob012345]

The aim of the natural sciences is not to "explain" anything but to "describe" quantitatively and as accurately as possible phenomena in Nature and to compare the "descriptions" to quantitative and accurate observations.

If that were true there would be little enthusiasm for the push for more basic unification theories. Separate theories adequately describe nature right now if that's all we want. Epicycles worked very well for describing celestial motion. But Copernicus, Kepler and Newton gave us more insight.

 

[PeterDonis]

Separate theories adequately describe nature right now if that's all we want.

Only if we assume we will never see any evidence that contradicts them. Physicists anticipate that we will: that at some point, when we are able to do sensitive enough experiments, we will see evidence that, for example, General Relativity is not exactly correct at arbitrarily large spacetime curvatures. That is why they are trying to get ahead of the game by working out possible theories that might be testable in such a regime.

Epicycles worked very well for describing celestial motion.

Not once Tycho Brahe got accurate enough observations. That's why Kepler was forced to develop his theory using ellipses: because Tycho's observations convinced him that there was simply no way to make an accurate enough model using circles alone.

 

[bob012345]

Only if we assume we will never see any evidence that contradicts them. Physicists anticipate that we will: that at some point, when we are able to do sensitive enough experiments, we will see evidence that, for example, General Relativity is not exactly correct at arbitrarily large spacetime curvatures. That is why they are trying to get ahead of the game by working out possible theories that might be testable in such a regime.
Not once Tycho Brahe got accurate enough observations. That's why Kepler was forced to develop his theory using ellipses: because Tycho's observations convinced him that there was simply no way to make an accurate enough model using circles alone.

Good points! Glad you agree that the theories need to be testable. I assume you mean beyond computer 'experiments'.

 

[PeterDonis]

I assume you mean beyond computer 'experiments'.

Yes. Computer simulations aren't data.

 

[atyy]

Of course, I agree with that. All models and theories are always preliminary. One day, there might be better and better theories. The Balmer series was an empirical law, i.e., it was found by looking at data and fitting a mathematical formula to it. Nevertheless it was a very important step towards building a model, because it provided a clear evidence for a systematic pattern. Bohr's model was an ad-hoc description, and many physicists tinkered around with it (most famously the Sommerfeld school) using a wealth of spectroscopic data for many different atoms becoming less and less satisfied by it and finally discovering modern QT. Whether modern QT or not is the "final word" is not known yet. Maybe there's a better/more comprehensive theory one day (including the quantum theory of gravitation perhaps).

Still, it's not the aim of all this progress in theory building to "explain" nature but to "describe" it with as consistent as possible mathematical models. The very fact, why this works at all, is not explainable. It's just an empirical fact that it works with an astonishing success!

You miss the point about explaining nature. The measurement problem is not about explaining nature, but describing nature. The observer is also a part of nature, yet the quantum state cannot describe the observer and the system together.

 

[vanhees71]

Well, since I don't see any measurement problem beyond the challenge for the experimentalists to invent ever better devices, I can't follow this argument. The apparent measurement problems you are talking about are, in my opinion, outside of the realm of the natural sciences but rather in the realm of the philosophy of science.

The one mind-boggling theoretical problem in contemporary fundamental physics, in my opinion, is to find a consistent description of quantum gravity. Maybe related to it is the enigma of the value of the cosmological constant ("dark energy") or the question, whether there's really "dark matter" or whether one needs a new theory of gravitation.

 

[AlexCaledin]

"We do not belong to this material world that science constructs for us. We are not in it; we are outside. We are only spectators. The reason why we believe that we are in it, that we belong to the picture, is that our bodies are in the picture. Our bodies belong to it. Not only my own body, but those of my friends, also of my dog and cat and horse, and of all the other people and animals. And this is my only means of communicating with them."

https://www.amazon.com/dp/1107431832/?tag=pfamazon01-20 page 96

http://www.azquotes.com/author/13142-Erwin_Schrodinger

 

[bob012345]

It's the modulus squared of the wave function which gives probability distributions, nicely real and positive.

The point is, you cannot observe anything without exchanging energy between the measurement device and the system you measure. While for macroscopic systems you can make the influence of the measurement device negligibly small, that's not possible for microscopic objects like an electron. To probe it you need other particles or em. waves to scatter at it, and there's nothing "smaller" than an elemetary particle as the electron itself to make the influence of the measurement device arbitrarily small. That's why observation and measurement are so much more emphasized in QT vs. in classical theory.

It seems to me that you are describing the 'Observer Effect' which is not the same as quantum limits to measurment due to the HUP. Can you elucidate your view on the matter (sorry for the pun!)? Thanks.

 

[vanhees71]

The HUP is no limitation for measurements but for preparation. We have discussed this very often in this forum. Just use the search function!

 

[vanhees71]

"We do not belong to this material world that science constructs for us. We are not in it; we are outside. We are only spectators. The reason why we believe that we are in it, that we belong to the picture, is that our bodies are in the picture. Our bodies belong to it. Not only my own body, but those of my friends, also of my dog and cat and horse, and of all the other people and animals. And this is my only means of communicating with them."

https://www.amazon.com/dp/1107431832/?tag=pfamazon01-20 page 96

http://www.azquotes.com/author/13142-Erwin_Schrodinger

Hm, poor Schrödinger...