mraltair wrote:For my level of understanding, it's both wave and particle.

I have a sneaking suspicion that it's neither, but that both are approximations of different aspects of the behaviour of light.

That reminds me, I really must get on with that essay on the various formulations of QM at some point.

twistor59 wrote:
Incidentally, do you think that two electrons really exchange photons when paticipating in the electromagnetic interaction ?

zaybu wrote:
That's what QED says. Do you have a better theory?

twistor59 wrote:
QED says nothing of the sort. It merely says that one way (perturbation theory) of computing the effects of the interaction is to treat the system as if virtual quanta are exchanged. However this is merely a calculational device. If you could solve the nonlinear interaction exactly, there would be no need to invoke virtual particles. The electromagnetic field, however, does allow real excitations, namely the photons that we can measure.

zaybu wrote:
QM is based on several assumptions such as operators are observables, or that the probability is the square of the amplitude, etc. The point is that this assumption of exchange particles explains many phenomena. The thousands of Feynman diagrams that explain millions of interactions observed in the many supercolliders across the planet wouldn't make any sense without that assumption. That's why it's called high energy particle physics, and the Standard Model describes the fundamental particles of nature.

There is a common misunderstanding of what a Feynman diagram represents. It doesn't represent a process that is literally occurring in spacetime. For example, looking at the diagram for electron scattering:
the only components that are detectable by any experimental means are the in and out states, represented by the electron lines. The virtual photon represented by the internal line is not detectable by experimental means. The diagram just represents one of many contributions to the scattering of the two electrons in perturbation theory.

If QED were an exactly solvable theory, there would be no need to do perturbation theory, draw Feynman diagrams and talk about virtual particles. There would just be an interaction between electrons and the electromagnetic field which we could compute exactly. "Exactly solvable" is a requirement on the mathematical model, not on the physics.

hackenslash wrote:

mraltair wrote:For my level of understanding, it's both wave and particle.

I have a sneaking suspicion that it's neither, but that both are approximations of different aspects of the behaviour of light.

That reminds me, I really must get on with that essay on the various formulations of QM at some point.

It is indeed neither - it's a quantum field !

twistor59 wrote:

zaybu wrote:

twistor59 wrote:

If you think that applying first quantization to photons is "elementary", then you have a different definition of elementary to the one I have !

It's elementary in the sense that there hasn't been any new development in QM since the 1930's. Omitting the advancement in QFT from 1930 to 1970 would be like studying math without studying any math from calculus and on. Your grasp of the subject would be quite limited. How could you discuss renormalization, gauge invariance, higgs mechanism or Wilson loop? Yet these are the backbones of the subject matter.

(I'll split the responses to the previous post into separate posts, otherwise the paragraphs are going to be very hard to keep track of.)

But the issue under discussion, which is essentially something along the lines of "what does a photon look like - how should I think of it?" does not need the advances you describe (with the exception, perhaps, of the gauge idea, which has been around since Hermann Weyl, and which is important in identifying the fundamental degrees of freedom). Although you can get by without asking simplistic questions like the one I'm addressing, you cannot suppress the urge to try to answer such questions, and such questions are surprisingly difficult to answer.

QM deals with low energies. At those scales the wave model has validity. However, if you deal with smaller and smaller distances, meaning higher and higher energies, then QM is inadequate, and one must deal with QFT. At those higher scales of energies, the wave model is redundant. Take any textbook on QFT, and what you'll get is particles, particles, particles. Even a ''field'' is thought as an exchange of particles.

Now if you want to restrict the discussion to QM, I'm saying you're missing the point.

twistor59 wrote:
There is a common misunderstanding of what a Feynman diagram represents. It doesn't represent a process that is literally occurring in spacetime. For example, looking at the diagram for electron scattering:

MollerScattering.jpg

the only components that are detectable by any experimental means are the in and out states, represented by the electron lines. The virtual photon represented by the internal line is not detectable by experimental means. The diagram just represents one of many contributions to the scattering of the two electrons in perturbation theory.

If QED were an exactly solvable theory, there would be no need to do perturbation theory, draw Feynman diagrams and talk about virtual particles. There would just be an interaction between electrons and the electromagnetic field which we could compute exactly. "Exactly solvable" is a requirement on the mathematical model, not on the physics.

First of all, perturbation theory is something you'll find in a QM textbook.That's not how the Feynman diagrams were worked out. It was Dyson who ingeniously combined the Heisenberg picture and the Schroedinger picture to form what is known as the interaction picture. It is in that picture that the Hamiltonian gives out an integral that can only be solved by a Taylor expansion. Feynman was able to translate that Taylor expansion into a series of his eponymous diagrams. Schwinger was able to do the same but instead of using diagrams, he used correlation functions. As in many instances in physics, you have two methods yielding the same results, tho Schwinger loathed the Feynman diagrams!

If you say that these interaction don't represent an exchange of particles then I can easily say that there are no waves passing through the double-slit experiment. You can't have it both ways. We assume a model, and then check if that model explains the data. Yet you are willing to accept readily the wave model in QM but not the QFT model of exchange particles when interactions take place, you are not being honest in your acceptance. What I'm saying is that the wave model is fine at low energies, at higher energies and smaller distance scales, the particle model is all there is.

EDIT: String theorists will argue that at even smaller distances (Planck distances) you get strings. But their theory has a big insurmountable flaw: it cannot be verified.

Perhaps everyone can understand what is meant by particle and is meant by wave, and for those who cannot understandable analogies involving stuff from everyday life can be drawn. How should we imagine/understand this quantum field?

twistor59 wrote:It is indeed neither - it's a quantum field !

But as far as I understand it, you can say the same thing about electrons and pretty much any other particle, that under QFT they ultimately reduce to quantum fields.
What I don't understand is why you claim that the particle concept is not appropriate for the photon, whereas it is for the electron.
I can understand that of the fundamental particles, the photon is the one which behaves less particle like, due to the lack of rest mass and consequently the unlimited range of the interactions it mediates, but where do you draw the line there?

By the way, is it possible that there is a confusion going on here around the meaning of the word particle?
As I see it, under QFT particles are not defined in the same sense as you would define it when talking about the wave-particle duality. They rather seem to be defined as discrete ripples in the quantum field. Under this definition, the photon would be a particle as much as the electron is, and it doesn't really matter here if under some conditions they behave either like a 'particle' or a 'wave' as defined in classical mechanics.

zaybu wrote:

twistor59 wrote:

zaybu wrote:

It's elementary in the sense that there hasn't been any new development in QM since the 1930's. Omitting the advancement in QFT from 1930 to 1970 would be like studying math without studying any math from calculus and on. Your grasp of the subject would be quite limited. How could you discuss renormalization, gauge invariance, higgs mechanism or Wilson loop? Yet these are the backbones of the subject matter.

(I'll split the responses to the previous post into separate posts, otherwise the paragraphs are going to be very hard to keep track of.)

But the issue under discussion, which is essentially something along the lines of "what does a photon look like - how should I think of it?" does not need the advances you describe (with the exception, perhaps, of the gauge idea, which has been around since Hermann Weyl, and which is important in identifying the fundamental degrees of freedom). Although you can get by without asking simplistic questions like the one I'm addressing, you cannot suppress the urge to try to answer such questions, and such questions are surprisingly difficult to answer.

QM deals with low energies. At those scales the wave model has validity. However, if you deal with smaller and smaller distances, meaning higher and higher energies, then QM is inadequate, and one must deal with QFT. At those higher scales of energies, the wave model is redundant. Take any textbook on QFT, and what you'll get is particles, particles, particles. Even a ''field'' is thought as an exchange of particles.

Now if you want to restrict the discussion to QM, I'm saying you're missing the point.

QM deals with scenarios where particles are not created or destroyed. That is the distinction between

The comment about QFT textbooks is not correct, it should read "take any textbook on QFT written with the main aim of teaching particle physicists to calculate scattering amplitudes .....". There is another massive domain of QFT - quantum optics - where the wave picture gives a better description of the behaviour of the quanta.

What exactly do you mean when you say that a field is thought of as the exchange of particles ?

zaybu wrote:

twistor59 wrote:
There is a common misunderstanding of what a Feynman diagram represents. It doesn't represent a process that is literally occurring in spacetime. For example, looking at the diagram for electron scattering:

MollerScattering.jpg

the only components that are detectable by any experimental means are the in and out states, represented by the electron lines. The virtual photon represented by the internal line is not detectable by experimental means. The diagram just represents one of many contributions to the scattering of the two electrons in perturbation theory.

If QED were an exactly solvable theory, there would be no need to do perturbation theory, draw Feynman diagrams and talk about virtual particles. There would just be an interaction between electrons and the electromagnetic field which we could compute exactly. "Exactly solvable" is a requirement on the mathematical model, not on the physics.

First of all, perturbation theory is something you'll find in a QM textbook.That's not how the Feynman diagrams were worked out. It was Dyson who ingeniously combined the Heisenberg picture and the Schroedinger picture to form what is known as the interaction picture. It is in that picture that the Hamiltonian gives out an integral that can only be solved by a Taylor expansion. Feynman was able to translate that Taylor expansion into a series of his eponymous diagrams. Schwinger was able to do the same but instead of using diagrams, he used correlation functions. As in many instances in physics, you have two methods yielding the same results, tho Schwinger loathed the Feynman diagrams!

But everything you have described here is perturbation theory. The Wick/Dyson expansion is entirely a perturbative entity.

zaybu wrote:
If you say that these interaction don't represent an exchange of particles then I can easily say that there are no waves passing through the double-slit experiment. You can't have it both ways. We assume a model, and then check if that model explains the data. Yet you are willing to accept readily the wave model in QM but not the QFT model of exchange particles when interactions take place, you are not being honest in your acceptance. What I'm saying is that the wave model is fine at low energies, at higher energies and smaller distance scales, the particle model is all there is.

EDIT: String theorists will argue that at even smaller distances (Planck distances) you get strings. But their theory has a big insurmountable flaw: it cannot be verified.

I'm not claiming that there are waves passing through the slits in a DS experiment. The reason I think it gives a better mental picture than the particle one in this case is that we're dealing with a spatially extended region, whereas "particle" suggests a very localized object in the popular mind. But I've no idea what really passes through the slits, I just know how to calculate the outcome.

On the "interaction = exchange of virtual particles" picture, you will find many different ideas. The fact is that, in this picture, all you're trying to calculate is a scattering amplitude. You know details about the incoming particles, you know the details of the outgoing particles, but you don't know anything about what happens in between when the particles interact somehow. There is no way to confirm experimentally that a virtual photon has passed between a couple of electrons. For this reason I think of it as a mere calculational device.

Remember method of images in electrostatics ? You have a charged object next to a conducting plate. You have to compute the electric field. You do it by imagining that there's an opposite charge at the other side of the plate - a mirror image. And you compute the field between the charges and it gives the right answer. There isn't really another opposite charge there, it's just a calculational tool. That's how I think of virtual particles.

mizvekov wrote:By the way, is it possible that there is a confusion going on here around the meaning of the word particle?
As I see it, under QFT particles are not defined in the same sense as you would define it when talking about the wave-particle duality. They rather seem to be defined as discrete ripples in the quantum field. Under this definition, the photon would be a particle as much as the electron is, and it doesn't really matter here if under some conditions they behave either like a 'particle' or a 'wave' as defined in classical mechanics.

I'll try to explain this in a bit more detail soon when I have some time. RL is occupying a lot of time just now...

Forgive the intrusion by a perennial contender for village idiot, but if you got something that does things particles cannot, and also does things that waves cannot, aren't you left with the idea that we really don't have a fucking clue what's going on, it's clearly neither as hackenslash wisely stepped in with? Sure we have the equations and can predict to a hair's-breadth over 3000 miles, but do we have quantum of an idea [sorry] what it is we are calculating?

crank wrote:Forgive the intrusion by a perennial contender for village idiot, but if you got something that does things particles cannot, and also does things that waves cannot, aren't you left with the idea that we really don't have a fucking clue what's going on, it's clearly neither as hackenslash wisely stepped in with? Sure we have the equations and can predict to a hair's-breadth over 3000 miles, but do we have quantum of an idea [sorry] what it is we are calculating?

Yes it's absolutely true that these entities are neither waves nor particles as we know them from the classical world. In some scenarios they look more like waves and in some, more like particles. They really aren't either, and it's a shame that the word "quanta" isn't more universally used. But that's the way it is, "particles" has stuck.

I think we do know what it is we're calculating (in particle physics) - namely the relationship between the shit going into the collision with the shit coming out of it. That's what we compute with the 1 hair per 3000 mile accuracy. What we don't know is the internal mechanism of the scattering process itself. What zaybu and I are differing on is our interpretation of the mathematics used to do this calculation. I'm advocating that it's just mathematics that gives the right answer and zaybu, I believe, is advocating that the mathematics is representing processes which are actually occurring in order to carry out the scattering process.

twistor59 wrote:

crank wrote:Forgive the intrusion by a perennial contender for village idiot, but if you got something that does things particles cannot, and also does things that waves cannot, aren't you left with the idea that we really don't have a fucking clue what's going on, it's clearly neither as hackenslash wisely stepped in with? Sure we have the equations and can predict to a hair's-breadth over 3000 miles, but do we have quantum of an idea [sorry] what it is we are calculating?

Yes it's absolutely true that these entities are neither waves nor particles as we know them from the classical world. In some scenarios they look more like waves and in some, more like particles. They really aren't either, and it's a shame that the word "quanta" isn't more universally used. But that's the way it is, "particles" has stuck.

I think we do know what it is we're calculating (in particle physics) - namely the relationship between the shit going into the collision with the shit coming out of it. That's what we compute with the 1 hair per 3000 mile accuracy. What we don't know is the internal mechanism of the scattering process itself. What zaybu and I are differing on is our interpretation of the mathematics used to do this calculation. I'm advocating that it's just mathematics that gives the right answer and zaybu, I believe, is advocating that the mathematics is representing processes which are actually occurring in order to carry out the scattering process.

Thanks for the reply, I was trying to think of why the question bothered me, and has for awhile. When you first get exposed to science in grade school/high school you begin to see matter as molecules and atoms, you find that a chunk of the mass, the matter, the 'substance' that is an atom is actually binding energy, Hiroshima or Nagasaki is mentioned, de rigueur-ly , of course, ....you find out how huge is the ratio of 'stuff' to empty space in an atom. upwards of 10-100 trillion I believe, then quarks come along and you find out, again, that there is less 'stuff', less substance, than the weight-mass leads you to believe, here the binding energy exceeds the 'actual' mass I think, or is comparable. Now maybe it's strings, and they are described, whatever that may mean here, in terms of something, stringy, that has dimensions, and tension, etc, but are they really implying a stuff, a substance? This is my real ponder, is there ever any there there? Do the prevailing physicist brains think of matter, of substance, as really a process. a relationship kinda thing? Will it boil down at some point to reality is only information and relations? And we all know how ugly the relations can get [yeah, you Uncle Henry!].

twistor59 wrote:

QM deals with scenarios where particles are not created or destroyed. That is the distinction between

The comment about QFT textbooks is not correct, it should read "take any textbook on QFT written with the main aim of teaching particle physicists to calculate scattering amplitudes .....". There is another massive domain of QFT - quantum optics - where the wave picture gives a better description of the behaviour of the quanta.

You're still looking at low energy physics. At high energy physics, QM is useless and is replaced by QFT.

What exactly do you mean when you say that a field is thought of as the exchange of particles ?

Here's a thought experiment. Suppose two electrons interact. From far away, you see that they repelled each other, therefore, you think, there is a repulsive force. Were you to probe at small distances without disturbing the system, which in the real world you can't, but suppose we're Gods and we can! What we see is that the two electrons exchange a particle, a photon, each particle's momentum is then changed according to conservation laws, and off the two electrons go in different directions. Of course, we are not Gods, and the photon exchange cannot be seen, which is why it is labelled as a virtual photon. It's just a matter of scale: from far, a repulsive force; up close, a photon exchange.

zaybu wrote:

twistor59 wrote:

QM deals with scenarios where particles are not created or destroyed. That is the distinction between

The comment about QFT textbooks is not correct, it should read "take any textbook on QFT written with the main aim of teaching particle physicists to calculate scattering amplitudes .....". There is another massive domain of QFT - quantum optics - where the wave picture gives a better description of the behaviour of the quanta.

You're still looking at low energy physics. At high energy physics, QM is useless and is replaced by QFT.

Yes, but I was responding to your point that QFT was "particles particles particles". There are application domains where it is less useful to think in terms of particles and more useful to think in terms of waves (although neither waves nor particles give the true picture).

zaybu wrote:

twistor59 wrote:
What exactly do you mean when you say that a field is thought of as the exchange of particles ?

Here's a thought experiment. Suppose two electrons interact. From far away, you see that they repelled each other, therefore, you think, there is a repulsive force. Were you to probe at small distances without disturbing the system, which in the real world you can't, but suppose we're Gods and we can! What we see is that the two electrons exchange a particle, a photon, each particle's momentum is then changed according to conservation laws, and off the two electrons go in different directions. Of course, we are not Gods, and the photon exchange cannot be seen, which is why it is labelled as a virtual photon. It's just a matter of scale: from far, a repulsive force; up close, a photon exchange.

No, virtual particles are not measurable, even in principle by a God with perfect knowledge.

When you calculate that scattering problem in QFT, you’re answering the question “if I start with a pair of electrons in a given state (momentum, spin etc), what’s the probability amplitude that they’ll end up in another given state (momentum, spin etc)”. (the "8" is meant to be ∞ ) The only things measurable in a lab here are the initial and final states of the electrons. I cannot measure anything that happens in the cloud.

In quantum theory, in order to be able to measure something, it must have a state. The state encodes all the information about that something. It contains all that could be known by an omniscient deity. The quantity I measure is then an eigenvalue of the desired observable, computed in that state. Virtual particles do not have a state. There is no creation operator that allows me to write |Virt Particle State> = a†|0>. Therefore, since they’re not measurable I prefer not to attach the label “objectively real” to them.

What they are, however, are elements of a perturbation expansion that allows an approximate computation of the scattering amplitude. The approximation can be improved by going to higher orders, adding more diagrams.

twistor59 wrote:

zaybu wrote:

twistor59 wrote:

QM deals with scenarios where particles are not created or destroyed. That is the distinction between

The comment about QFT textbooks is not correct, it should read "take any textbook on QFT written with the main aim of teaching particle physicists to calculate scattering amplitudes .....". There is another massive domain of QFT - quantum optics - where the wave picture gives a better description of the behaviour of the quanta.

You're still looking at low energy physics. At high energy physics, QM is useless and is replaced by QFT.

Yes, but I was responding to your point that QFT was "particles particles particles". There are application domains where it is less useful to think in terms of particles and more useful to think in terms of waves (although neither waves nor particles give the true picture).

zaybu wrote:

twistor59 wrote:
What exactly do you mean when you say that a field is thought of as the exchange of particles ?

Here's a thought experiment. Suppose two electrons interact. From far away, you see that they repelled each other, therefore, you think, there is a repulsive force. Were you to probe at small distances without disturbing the system, which in the real world you can't, but suppose we're Gods and we can! What we see is that the two electrons exchange a particle, a photon, each particle's momentum is then changed according to conservation laws, and off the two electrons go in different directions. Of course, we are not Gods, and the photon exchange cannot be seen, which is why it is labelled as a virtual photon. It's just a matter of scale: from far, a repulsive force; up close, a photon exchange.

No, virtual particles are not measurable, even in principle by a God with perfect knowledge.

When you calculate that scattering problem in QFT, you’re answering the question “if I start with a pair of electrons in a given state (momentum, spin etc), what’s the probability amplitude that they’ll end up in another given state (momentum, spin etc)”. (the "8" is meant to be ∞ )

The attachment scat.jpg is no longer available

The only things measurable in a lab here are the initial and final states of the electrons. I cannot measure anything that happens in the cloud.

In quantum theory, in order to be able to measure something, it must have a state. The state encodes all the information about that something. It contains all that could be known by an omniscient deity. The quantity I measure is then an eigenvalue of the desired observable, computed in that state. Virtual particles do not have a state. There is no creation operator that allows me to write |Virt Particle State> = a†|0>. Therefore, since they’re not measurable I prefer not to attach the label “objectively real” to them.

What they are, however, are elements of a perturbation expansion that allows an approximate computation of the scattering amplitude. The approximation can be improved by going to higher orders, adding more diagrams.

But the "something happens" is not as nebulous as your diagram implies, it's more like: SEE ATTACHMENT

By considering all possible interactions that we can get very precise agreement between theory and experimental observation in QED.

well, I would like to consider this from conservation of energy point of view in the universe, there light gets red shifted, does it disappear ?.Or virtual particles popping into existence. If so, one wonders what we mean by a particle.

zaybu wrote:

But the "something happens" is not as nebulous as your diagram implies, it's more like: SEE ATTACHMENT

By considering all possible interactions that we can get very precise agreement between theory and experimental observation in QED.

No, those diagrams are only pictorial representations of a bunch of integrals you need to compute the scattering amplitude of the incoming and outgoing electrons. They're only a way of organising the mathematics. You only have to do it this way because you're using perturbation theory. We're only doing perturbation theory because we don't know how to set up and solve the equations of QED exactly.

The scattering amplitude is given by the sum of the contributions of all the diagrams to a given order. Each diagram (in configuration space) requires you to integrate over all the spacetime locations of each vertex. Only the sum has meaning. The electron never really emits a virtual photon.

Let me ask a question:

The diagram (a) in the picture represents a contribution of order α2, (where α is the fine structure constant), diagram (b) of order α4 etc...
To what order must we go in order to represent reality ? (just sticking to QED here - assuming no new physics, quantum gravity etc)

cavarka9 wrote:well, I would like to consider this from conservation of energy point of view in the universe, there light gets red shifted, does it disappear ?.Or virtual particles popping into existence. If so, one wonders what we mean by a particle.

Is energy conserved for the entire universe ?

Noether => time symmetry ~ energy conservation

Universe not time symmetric (no timelike Killing vector).