What does the current usage of the word photon in textbooks mean?

[Jano L.]

What do you think the word "photon" used in textbooks and papers at present mean?

- extended object which, when there are many of them, form light

- point-like particle which, when there are many of them, form light

- merely a short word for a _continuous_ change of energy of the field after a " quantum jump "
occurred

- merely a short word for a _jump-like_ change of energy of the field after a " quantum jump " occurred

- other

 

[Simon Bridge]

The use of the word depends on context.

Usually a "photon" is a single quanta of electromagnetism.
Do you have examples where the use has confused you?

 

[Jano L.]

There are many uses of that word which I think are quite incompatible with electromagnetic theory - the options above are the most usual I could think of.

single quanta of electromagnetism.

What do you mean by that?

 

[Simon Bridge]

Exactly what I said.

 

[Jano L.]

Can you explain the word " quanta " ? It is as mysterious as the word " photon ".

When you say electromagnetism, you are implying connection to classical electromagnetic theory. But there are no quanta there. So can you explain what " quantum " means? What are its properties? Does it have position? How does it connect to electromagnetic phenomena?

 

[Oudeis Eimi]

Can you explain the word " quanta " ? It is as mysterious as the word " photon ".

When you say electromagnetism, you are implying connection to classical electromagnetic theory. But there are no quanta there. So can you explain what " quantum " means? What are its properties? Does it have position? How does it connect to electromagnetic phenomena?

Quantum Electrodynamics.

 

[Dead Boss]

Particle.

 

[Simon Bridge]

There is a problem in that I can only describe words to you in terms of other words which are likely to also be mysterious. I'll give you an analogy:

Imagine I have no knowledge of fire and I come to you and ask, "What is this fire that people keep speaking about?" How do you answer me?

You are basically stuck with playing word association games until, after a lot of to-ing and fro-ing you feel confident that I'll recognize fire when I see it and have a fair idea about how it is likely to behave but until I do I won't actually know what the word "fire" actually means. When I do, I'll realize that I never really knew what it was even though I recognize it. What's more, I need many experiences of fire to really get it and why nobody want's fire that can be fitted nasally. It's a long process - you wouldn't expect me to get fire from a few paragraphs in an internet forum. But I'll eventually gain a solid understanding through theory and experience and then someone else asks me what fire is and I'm stuck playing the same word games you had to use for me.

If you are being honest with us and not just playing at philosophy 101, you are in the position of the person who does not know what fire/photon is and the rest of us are stuck with playing word games.

It's a little bit easier though since you just want to know what these textbooks are talking about - but you won't give us a specific instance from a textbook but instead supply these somewhat vague statements to puzzle over. I've seen stuff kinda like that in textbooks - but not actually like that... I'm left trying to guess where you are coming from which is why I asked for a specific example where you were getting confused: copy one out. But maybe there is enough to work with after all?

Physics textbooks are typically written from a world-view of empirical realism; namely that there is an objective reality "out there" independent of our senses, our sense data tells us useful things about it, and that the truth of any proposition made about this reality cannot be known a-priori. I hope you are familiar with things like empiricism and scientific method. If not: look them up.

Photons are usually written of as objects that have a real existence as "quanta" within the standard model for electromagnetism. In this model, many things come in packages or lumps - it is a particle model. These packages follow the Rules of Quantum Mechanics ... which is why they need a special name. Quanta are usually thought of as the smallest bit of something that you can get. eg. a quanta of charge would be the size of the charge on an electron (you can get smaller charges but not by themselves).

A photon is a quanta of electromagnetism - how it manifests depends on how it is detected. In this model "light" and "electromagnetism" are synonyms - but we like to avoid the word "light" due to excess baggage.

So a photon is kinda a small bit of light ... however, a whole lot of photons interacting at a surface would be called "illumination" rather than "light" and a bunch of photons going in a stream roughly the same way would be called a "flux" rather than a "beam of light". You can think of a photon as a bit of light the same way you can think of a grain as being a bit of sand if you like - in that description it would be the smallest bit of light you can get of a particular color - it's not the whole story: remember the fire?

You gave four examples.
The first one is kinda close - except the photon is not an "extended object". Photons arrive in one lump.

The second is close - except it leaves out the wave-like properties.

Which of these you experience depends on how you look - if you are careful though, you get to see some weird stuff which shows it's nature as neither of these things. It is itself.

The last two don't make much sense ... the photon given off in, say, an atomic de-excitation is a lump of light just the same. But it looks like you are struggling with "wave particle duality". Photons are neither classical waves nor particles ... the textbooks are just being inclusive while they break the news to you gently.

Until you have more experience with photons and photon models of nature, for you, a photon will be a step in a calculation that you do to make predictions in physics experiments. You'll just have to settle for understanding them in the abstract like the guy who doesn't know what fire is.

-----------------------
Caveat: all the above is intended in the context of what is solidly within established science using a simplified language. This science is constantly being tested so at any time there will be some evidence of some kind of exception to some of the things I have stated as fact above. That's just the nature of the beast - you get used to it. In the end we are all just making models. The models are just models - reality is reality, and the Universe knows far more physics than we do.

For more detail on the "fire" analogy, see Mahasamatman's "names are not important" speech in Roger Zelazney's Lord of Light.

 

[Jano L.]

Photons are neither classical waves nor particles ... the textbooks are just being inclusive while they break the news to you gently.

You say what the photon is not, ok, but that is no use.

The problem is not in philosophy nor in language, but with the definition and status of photon in theoretical physics. In have not found any textbook which would define the term in technical sense. However, many of them use the word freely assigning it quite different sense, leaving the imagination of the reader to work it out. The use of it among researchers is equally poetic.

Let's take total force acting on the body. This has an unambiguous definition as the time derivative of its momentum. There is no much difficulty with it.

Can you provide similar definition for the " quantum of electromagnetism" ?

 

[Jano L.]

...for you, a photon will be a step in a calculation that you do to make predictions in physics experiments. You'll just have to settle for understanding them in the abstract

Which step? It would be great to understand them at least in an abstract way, but can you provide a mathematical definition?

 

[Simon Bridge]

You say what the photon is not, ok, but that is no use.

Tough. I have also described it to you in terms of what it is... eg. It is an object that describes electromagnetism according to the rules of the standard model in physics. You don't have to like these descriptions but they are what's available.

Let's take total force acting on the body. This has an unambiguous definition as the time derivative of its momentum. There is no much difficulty with it.

OK - then in the same terms a photon is a wave-packet with total energy $E=h\nu$. Better?

But - again, from the same language you have been using: what is this word "momentum" that you use to describe "force"? You can tell me it is the product of mass and velocity but what are these? Where does it stop? For that matter - you have told me how to work out the amount of force from the effects (rate of change in momentum) but that does not tell me what a force is?

If you want to define force by its effects then a photon is defined as the fundamental source of all electromagnetic effects just the same. The standard model is much much more general than Newton's Laws of motion and so it's components have much greater and wider scope.

Which step?

[in a calculation] ... which calculation would you like to do? I'll point out the photon.

It would be great to understand them at least in an abstract way, but can you provide a mathematical definition?

It is not something that is defined, it is something that is. We just try to describe it.

However - those textbooks you have been reading are full of examples of calculations involving photons. Pick any of them. eg. in an electromagnetic interaction, there is a link between cause and effect ... this link is provided by a photon which can carry momentum and energy between two interacting bodies. It plays the same role as F in Newton's Laws.

Since you have ready access to mathematical examples, I suspect you are playing games. Please come up with a specific example that you don't understand. What is it about photons you do not understand?

 

[Jano L.]

I do not want to play games. I am just interested how other people interpret the word and what is the main stream tendency. That's why the poll.

But since you do seem to see any difficulty with photon, let's try an example.

One of the main reasons photon was accepted by physicists as something that IS was success of Compton's explanation of his experiments on the scattering X-ray radiation. He made a calculation in which he assumed that point-like electrons scatter point-like particles of light while the law of conservation of momentum was conserved, and the momentum of light quanta was defined as $hf/c.$
The resulting Compton's shift of wavelength agrees well with the measured wavelength of scattered radiation.

However, the photon of Compton has two contradictory properties:

1/ it collides with the electron in one point of space-time, so it has to be point-like
2/ it has associated frequency and wavelength, so the corresponding EM field exist with certain spatial extent

How do you think this is to be resolved in the theory where photons are not defined but are?

 

[Simon Bridge]

Those are only contradictions to the classical description of particles and waves. There is no contradiction in the quantum mechanics. A photon is a quantum mechanical description of electromagnetism, not a classical one.

I don't know what level I need to pitch this at. You seem to have got as far as wave-particle duality but not as far as QED. Quantum mechanics is a superset of classical mechanics so there are things in the standard model which are contradictory in the classical.

Wave particle duality is like blind men examining an elephant - you know the story? One gets the head and announces that an elephant is like a fat snake (trunk) and another gets the back and announces that an elephant is like a rope (tail) ... they put their heads together and announce that an elephant exhibits snake-rope duality. But an elephant is not a rope and it is not a snake: it is an elephant. A photon is not a classical wave and it is not a classical particle ... it is a photon. Before you can progress to understand what a photon is in terms other than what it is not, you must first get rid of these notions. You seem to be having a lot of trouble giving them up.

I usually show people Richard Feynman's lectures on wave-particle duality - they are in youtube - for a clear description for the layman. In short, the photon is a QM particle - the physical thingy always arrives as a lump. It is the statistics of them that has wave-like behavior ... which is just to say we have to use similar math to what we use to analyse waves.

Go hunt down the lecture series.

... watch all of them. He addresses wave-particle duality explicitly several times but he has to build the ideas slowly. At the end there's a Q&A where the question comes up again.

 

[Jano L.]

There is no contradiction in the quantum mechanics. A photon is a quantum mechanical description of electromagnetism, not a classical one.

You mentioned Feynman. One thing one should remember from his example is to not trust authorities and think things through again.

The problem does not vanish when electromagnetic field is described in terms of Heisenberg operators. These operators are defined on space-time and so have some resemblance to classical field quantities. Their evolution is described by differential equation and comes out to be continuous.

However, the standard calculation of Compton's effect (say that which Sakurai gives) rests on an additional incompatible idea that one infinite plane wave transforms itself to another (photon changing its momentum in the scattering event), a completely non-local process that is at variance with wave equation. Nor the time neither the occurence of the jump is calculable from Schroedinger's equation; it is a foreign process that survived from the Old Quantum Theory.

The contradiction between unitary evolution and jump process in QT is as problematic as the contradiction of a point-like particle and continuous wave in CT.

 

[Cthugha]

The problem does not vanish when electromagnetic field is described in terms of Heisenberg operators. These operators are defined on space-time and so have some resemblance to classical field quantities. Their evolution is described by differential equation and comes out to be continuous.

Quantum optics does quite fine using probability amplitudes...

However, the standard calculation of Compton's effect (say that which Sakurai gives) rests on an additional incompatible idea that one infinite plane wave transforms itself to another (photon changing its momentum in the scattering event), a completely non-local process that is at variance with wave equation. Nor the time neither the occurence of the jump is calculable from Schroedinger's equation; it is a foreign process that survived from the Old Quantum Theory.

The contradiction between unitary evolution and jump process in QT is as problematic as the contradiction of a point-like particle and continuous wave in CT.

The standard calculation as you call it uses plane waves as a simple basis. A plane wave is not normalizable and therefore obviously and trivially does not describe a state the em field can take. This is well known. But it is a good basis for constructing wavepackets which you need to consider in actual scattering problems. The formal way to do that can be found in e.g. Collision Theory by Goldberger and Watson, but it is quite cumbersome and lengthy. For a basic understanding using plane waves is enough and most people are aware that the plane waves do not actually correspond to physical elements of reality, but are just some basis.

 

[Jano L.]

Quantum optics does quite fine using probability amplitudes...

I suggest that for a while we keep the discussion on Compton's scattering which is from certain point of view simpler (no bound state present).

For a basic understanding using plane waves is enough and most people are aware that the plane waves do not actually correspond to physical elements of reality, but are just some basis.

You are right, but also missed the point. The problem remains with any representation of the field state. Take normalized waves in a cube. Still the projection postulate implies the state of the field would change by a jump in whole cube.

If you do not like plane waves and want to scatter localized packet of EM field instead, what dimensions would such a packet have? Could it be ascribed single frequency?

 

[Cthugha]

You are right, but also missed the point. The problem remains with any representation of the field state. Take normalized waves in a cube. Still the projection postulate implies the state of the field would change by a jump in whole cube.

The field will change quickly or instantaneously (depending on your way of tackling the problem) in a (maybe large) volume. Is possible non-locality a problem? This is why I mentioned probability amplitudes earlier on. Most people feel more comfortable considering these as non-local compared to fields they consider as entities having "physical reality".

If you do not like plane waves and want to scatter localized packet of EM field instead, what dimensions would such a packet have? Could it be ascribed single frequency?

The dimensions are given by its coherence volume, so coherence length and time are important. These depend strongly on the emission process and the emitter used. It will definitely not have single frequency in the strict sense of being perfectly monochromatic. No real light field is perfectly monochromatic.

 

[Jano L.]

The field will change quickly or instantaneously (depending on your way of tackling the problem) in a (maybe large) volume. Is possible non-locality a problem? This is why I mentioned probability amplitudes earlier on. Most people feel more comfortable considering these as non-local compared to fields they consider as entities having "physical reality".

Avoiding contradiction with relativity is possible only in non-relativistic theory of electrons. Schroedinger's wave function Psi(r1,r2,...) can be understood as a kind of probability distribution for the coordinates of particles and there is no problem with instantaneous changes of it (the theory ignores relativity right from the beginning).

In the theory including radiation, things are different. One uses different mathematical object to describe the radiation - a function of space and time coordinates x,t - a field. This field obeys Maxwell's equations and propagates continuously with finite velocity. I do not know of a theory where the field would be described by a wave function of some coordinates of " particles of light ".

In quantum optics, there are rather field operators in Heisenberg picture. These are far from describing probability distribution for particles.

The dimensions are given by its coherence volume, so coherence length and time are important. These depend strongly on the emission process and the emitter used. It will definitely not have single frequency in the strict sense of being perfectly monochromatic. No real light field is perfectly monochromatic.

Exactly, my point. There is no such thing as field with exact frequency or wave length. But then what attribute of " photon " remains ?

 

[Simon Bridge]

I call troll.
Jano is not interested in learning about photons - he already knows. He is interested in picking arguments.

 

[D H]

Thread closed, poll edited.