Paul,

>It is here argued that the photon has no physical existence at all,

Photons convey momentum and energy.

They also have discrete properties such as polarization states.

This, at least for me, passes the test for realism.

(viz Bishop Berkely: photons can kick me!)

James,

Would you agree that we only obtain information when photons are emitted or absorbed, rather than in between those events? We can't intercept them "in flight" without absorbing them.

While I certainly agree that you do gain its momentum on absorbing a photon, if there is zero 4-interval between you and whatever is the source of the momentum (e.g. a nucleus undergoing γ-decay) then what can exist, in Minkowski's space-time, between the source and a sink? 

In this view, the wave description is how we account for the momentum transition when we regard the relationship between space and time as a Newtonian relation.

Polarisation states, as far as I'm aware, are either established at emission via a filter, or again at the point of absorption.

Thanks for replying - Paul 

(And when Samuel Johnson kicked a stone to refute Berkeley's immaterialism, he was kicking an object with mass.)

Photons can be scattered too.

Dear Ravi,

No doubt photons can be scattered, but again, I suggest, they are only evidenced at the points of interaction with scatterers and thus traceable to the entities that bound their interactions.

Any scatter process can be analysed into a series of interactions from an initial emitter, via a scattering, which I suggest can be described as an absorption and re-emission. So I believe it still conforms to my initial point (adjusted to the case of scattering):

• "properties ascribed to it [in this case the scattered photon] should be traced to the emitter and scattering-absorber/re-emitter and subsequent absorber (or further scattering) experiencing the interaction (or transfer) that we characterise as a photon."

The only weak point I can see in my argument might be a case in which photon-photon scattering is allowed, but as I read the link below, photon-photon scattering doesn't occur.

Compton scattering, of course, doesn't contradict my argument.

If I'm being bone-headed, then someone please tell me why. 

Thanks for your contribution, Ravi.

http://physics.stackexchange.com/questions/54323/what-would-the-collision-of-two-photons-look-like

Michael

Agreed re mass; however, as I'm sure you know, photons are conventionally described as transferring momentum. Your other points are also agreed, as standard SR.

Thanks - P

Is the path not important? But light travels along a path. What if I put an opaque object in the path? Will photon reach its destination?

What about photon - photon interaction? Say in an interference pattern generation.

Ravi,

Regarding your first point: the photon's 'path': By interposing an opaque object you will of course "absorb the photon" (or possibly reflect it - i'.e.' absorb and re-emit it). But now you've simply changed the experiment. In principle, it's no different from the previous situation - you've just replaced the original absorber with one that's closer to the source. My assertions apply just as well to the new absorber.

As for your 2nd point, interference: I think we're mixing two ways of thinking about light. Interference is the result of a linear superposition of wave patterns, simply summing amplitudes. I don't think the waves affect each other directly, although we're free to re-analyse the combined waves as we see fit. And as argued above, the wave view is how we account mathematically for the effects linking emission and absorption of light at e.g. photographic emulsion or CCDs so it appears that we're forced back to the photonic interpretation to account in detail for the illumination of the absorber (you can measure light intensity, but ultimately it comes down to material excitations, whether atomic, nuclear or merely molecular excited states).

Aside from accidental reflections by e.g. dust in a laser beam, we don't "see" light in transition, only its eventual absorption.

Thank you for stimulating suggestions.

Paul

Paul,

"Would you agree that we only obtain information when photons are emitted or absorbed, rather than in between those events? "

Yes.

What happens between absorption and emission, well, I think that little can be said with certainty about that. To dwell on something that is (by definition) unobservable seems to me, well, a good waste of coffee-drinking time.

YMMV

"Polarisation states, as far as I'm aware, are either established at emission via a filter, or again at the point of absorption."

Yes.

"he was kicking an object with mass."

Strictly, he was scattering his foot off it - semi-elastically.

How do you experimentally prove your hypothesis?

Michael,

I’m not sure I understand you.

I’ve not intended to suggest, anywhere, that photons have mass – certainly not rest-mass. if I have implied it, please point it out.

Using the conventional photon description, we agree they carry momentum, h/λ , and as I’m sure you know at least as well as I do, they can therefore be ascribed a relativistic mass, hv/(c^2).

http://physics.stackexchange.com/questions/2229/if-photons-have-no-mass-how-can-they-have-momentum

Ravi,

What you call my "hypothesis" was posed only as a question – seeking clarification from those more expert than myself.

I am simply puzzled by what, in my chemist's relative ignorance, appears to be the clear implication of SR that the passage of what we describe as a photon is only an interaction of emitter and absorber that are, according to Feynman, in direct “4-contact” – no finite space-time interval separates them. The body of my question is my attempt to get clarification on this seeming puzzle.   

However, if I agree that my question does embody a hypothesis, then the reply to your latest question is: it’s the photon that’s the hypothetical entity!

And if I try to answer your question as stated, then the hypothesis could be:

      "massless particles have no material existence" ?

How does one prove, at least to 5σ, a negative? Maybe we have to turn it around and try to prove that massless particles exist? 

But, I still maintain that no-one ever saw a photon and I cannot conceive any way to observe one without absorbing or reflecting it. we only register changes in the supposed emitter and absorber (or absorber & re-emitter). And as for what happens in-between - it's all mathematics to explain what's observed at the end-points of the interaction.

I have to admit that this is a more positivist argument than I'm usually comfortable with, but I see no alternative.

Someone - convince me that I'm in error, please! 

In FRET expts, there is transfer of radiation non-radiatively between a donor molecule and an acceptor molecule mediated by dipole-dipole interactions. One excites the donor and the excitation energy is transferred by Forster mechanism to the acceptor non-radiatively. Of course for this radiation transfer to occur certain criteria need to be satisfied like distance, orientation factor and overlap integral.

I hope this supports your hypothesis. Correct me if I am wrong.

I'll have to look up Förster Resonance Energy Transfer.

Hopefully, "transfer of radiation non-radiatively" will turn out not to be self-contradictory. 

Thanks, Ravi,

Relatedly, I've just looked up non-linear optics. 

" the branch of optics that describes the behavior of light in nonlinear media, that is, media in which the dielectric polarization P responds nonlinearly to the electric field E of the light. This nonlinearity is typically only observed at very high light intensities (values of the electric field comparable to interatomic electric fields, typically 108 V/m) such as those provided by lasers. Above the Schwinger limit, the vacuum itself is expected to become nonlinear. In nonlinear optics, the superposition principle no longer holds."

Can I explain how dielectric polarisation in a nonlinear medium might be consistent with my view? I'll get back, but maybe not immediately.

Thanks for everyone's interest so far.

Paul

Dear Christian,

Thanks for your comment.

"with respect to massless particles, the relation simply is E²=p²c². "

I took the point about relativistic mass from a highly up-voted link on Stack Exchange (link, used earlier, above, but repeated below), part of which I'm copying in directly, here:

• "The theory of Special Relativity, proved in 1905 (or rather the 2nd paper of that year on the subject) gives an equation for the relativistic energy of a particle;
• E2=(m0c2)2+p2c2
• where m0 is the rest mass of the particle (0 in the case of a photon). Hence this reduces to E=pc. Einstein also introduced the concept of relativistic mass (and the related mass-energy equivalence) in the same paper; we can then write
• mc2=pc
• where m is the relativistic mass here, hence
• m=p/c
• In other words, a photon does have relativistic mass proportional to its momentum."

However, that's the 'top' answer, but only as it's gained the approval of the OP, by being an answer that satisfies him/her.

Underneath it, I've now noticed, is a 2nd contribution to that Stack Exchange, with an even higher number of up-votes (but not the OP's mark of approval); and that one supports your line, as does the final paragraph of Feynman's Lectures, Vol I, Ch 17 on p. 17.8.

Thanks for joining the conversation

Paul

PS While the discussion to date has focused primarily on photons in a vacuum I'm working on a reply concerning photons in material media, and especially how the question relates to non-linear optical effects.

http://physics.stackexchange.com/questions/2229/if-photons-have-no-mass-how-can-they-have-momentum

@Paul

As far as I can see after having scanned the discussion, nobody has mentioned the fact that you completely misunderstood the Feynman text you cite in your original question.

Being 'separated by a zero interval' is explained in  Feynman's Eq (17.5) and does by no means imply that  'events which are separated by a zero interval' are identical. (Well, you can call them  '4-interval coincident' but this is only words - it suggests properties of 'coincidence' that simply not hold in that case.) Since the whole section 17-2 is on 'space-time intervals' you could not have missed the point if you had read the whole section and not only a few sentences.

Any text on relativity will tell you what your experience would be if you would travel along the space-time path of a photon in an approximate way, namely at a velocity of , say, 0.99999999999999999999999999999999999999999 c.

Ulrich

Thank you - I will re-read the rest of the Feynman section and ponder your point.

Paul

Photon has physical existence.Please see our common papers :

https://www.researchgate.net/publication/302260015_Adaptive_Review_of_Three_Fundamental_Questions_in_Physics?ev=prf_pub

https://www.researchgate.net/publication/303988070_Generalization_of_the_Dirac%27s_Equation_and_Sea

Article Adaptive Review of Three Fundamental Questions in Physics

Article Generalization of the Dirac's Equation and Sea

Ulrich

With reference to your first post, about 4 posts above;

I will now agree that space-time

'events which are separated by a zero [4-]interval' are not identical.

Someone’s kindly pointed out to me that the concept of distance under a Lorentzian/pseudo-Riemannian (including Minkowski) metric doesn’t conform to the intuitive notion of distance found in a metric defined in what he/she calls “a strict mathematical sense”.

As mentioned in the original Question, it was already clear to me that points in 3-space and time, separated by a zero 4-interval, have different coordinates, but having re-read the whole of the Feynman 17-2 section as suggested, I still fail to see ‘what [I] could not have missed’ in there indicating that

"Being 'separated by a zero interval' … does by no means imply that 'events which are separated by a zero interval' are identical."

I’d appreciate instruction.

You wrote also:

“Any text on relativity will tell you what your experience would be if you would travel along the space-time path of a photon in an approximate way, namely at a velocity of , say, 0.99999999999999999999999999999999999999999 c.”

Yes – Clocks slow and rulers shrink; so, if I’ve got it the right way around, the end-point seems, to the near-light-speed traveller, already very close. At light-speed, therefore, a massless entity will experience the passage of no time and no distance (there is a Stack Exchange, by one of its more frequent and highly voted contributors, who says (my bold):

“ Just like you can figure out the amount of time between two events going by a particle's clock (the proper time) using that formula Δt2−Δx2/c2−−−−−−−−−−−√Δt2−Δx2/c2, you can also find the distance the particle sees the universe traveling past it by multiplying the proper time by the particle's speed. For a photon, since the proper time difference is always zero, the distance traveled is also zero. [snip]: the photon travels zero distance by its own measurement, in a nonzero time by an external observer's measurement. – David Z♦ May 5 '12 at 14:47”

So, for the present, I still cannot see other than that the interaction between emitter and absorber occurs ‘as one’ from a light-speed particle’s view (even though I now acknowledge that the end-points cannot be described as ‘identical’), as no time interval, no space interval (to paraphrase Bob Marley) has been experienced, and so for the moment I still believe it’s possible to argue that there is no such physical entity as a photon,

Thank you for helping me clarify at least one aspect of my question.

http://physics.stackexchange.com/questions/27794/is-a-photon-fixed-in-spacetime

While a photon itself experiences zero time during its flight, the cosmic background radiation cools as the Universe expands. The cooling of the cosmic background radiation is a measure of the passage of cosmic time during a photon's flight. It may be asked "If the spacetime distance of a photon is always zero, then why can't we observe light from a distant star immediately upon emission?" This problem is discussed by Cornelius Lanczos in his book "The Einstein Decade" and in the reference cited in this discussion. Apparently the spacetime distance isn't everything. Just as the comoving frame, in which the cosmic background radiation is isotropic, corresponds to a preferred measure of time, cosmic time, it also corresponds to a preferred measure of ordinary spatial distances. For example, the time since the Big Bang, or from the Big Bang to the Big Crunch in an oscillating Universe, is a maximum when measured in the comoving frame; in any other frame it is less. Thus while a photon has no rest mass, if one has to choose a frame in which its relativistic mass hv/c^2 has the truest value, this frame is the comoving frame.

Paul,

the point where you would 'appreciate instruction'  tried to express my expectation that your misconception concerning coincidence of light-like separated space-time points, --- which looks understandable to me if a single sentence of the Feynman text is considered ---  would hardly form in the brain of a reader of the whole section 17-2. So nothing what is new to you is involved --- and no instruction necessary.

Your picture of a travel with near light speed misses an important point: in the very short timespan of the travel as experienced by the traveller he would notice many stars along his track. He could have equipment on board which allows to identify the stars and which tell him when he left the solar system, our galaxy, our local galactic cluster .... Assigning to these celestial objects the coordinates that they have in our astronomical catalogues and observing the  change of coordinates of identified objects as a function of the travellers proper time he will compute a 'cruising speed' which exceeds c by a large factor. When we try to extrapolate this experience to a travel with exactly the speed of right it is obviously a poor statement to say that 'the distance travelled by the photon is zero'.

One may hold different opinions on the strength of arguments based on the hypothetical world view of a photon. By no means these arguments should be considered convincing enough to make irrelevant all the observations concerning motion of photons in the laboratory. You are probably familiar with the famous double slit experiment. For a given point source for photons and a fixed array of detectors (solid state matrix such as CCD or multi-channel plate) the detector patten depends on the location of the slits which proves that in the experiment there is not only involved the emission and detection of photons but also a propagation between emitter and detector. The science of quantum optics has detailed descriptions of this propagation process in terms of time dependent photon states.

That these quantum optical photon propagation models perfectly agree with experimental data show that 'something is going on in nature along these lines'.

Simply ignoring all aspects of photon propagation is not an acceptable solution.

Ravi,

I see FRET involves an r^(-6) dependence, which brings to mind the attractive long-range van der Waals force, or dispersion force described by one of the terms in the Lennard-Jones potential for intermolecular interaction. And as FRET is described as a non-radiative dipole-dipole coupling, then it seems to be, by definition, an electromagnetic but non-photonic interaction.

According to R M Clegg's History of FRET (2006), which I've just ccme across:

“Thus he [Förster] was able to develop a quantitative theory … of the rate of transfer from an excited donor molecule to a ground state acceptor molecule in terms of what is now known as the overlap integral…The overlap integral represents the probability that the two molecular transition dipoles will have the same frequency. This was a major conceptual step, because these spectroscopic transitions can be measured experimentally, independent of the FRET measurement. It is also important to realize that no "spectroscopic transition" takes place; that is, there is neither the emission of a photon nor the absorption of a photon in the FRET process.”

So FRET shows there are electromagnetic energy transfer processes between molecules that aren't accounted for by photon exchange. But logically that doesn't impinge on the question of the existence of photons either. Nice try!

Thanks - Paul

http://photobiology.info/PDF/History%20of%20FRET.pdf

 Jack,

According to Plebanski & Krasinski’s “Introduction to General Relativity and Cosmology”, even GR admits geodesic arcs of “zero length” (as they call it), and so I wonder how consideration of an expanding universe alters, fundamentally, the notion of spacetime distance as you question it.

Unfortunately, I can’t get to see a copy of the Lanczos reference at present. But if an erstwhile colleague of Einstein discussed the question of why we can't “observe light from a distant star immediately upon emission”, even in the context of cosmic time, then light’s zero 4-interval, and perhaps intuitions about it, seem to have been of some interest to far more accomplished thinkers than I could ever claim to be.

In a comparatively local (non-cosmic) context, it’s obvious that light’s emission and absorption points have sufficiently different coordinates that we can know it takes, e.g., eight minutes for the Sun’s light to reach 150m km to Earth, or a dozen-or-so nano-seconds to run from one end of a lab bench to the other, rather than being “immediately observable”. (The Minkowski measure of light paths may be zero, but as posted in my 2nd reply to Ulrich (i.e. just before yours), the concept of distance under the Minkowski measure is unlike normal space, presumably as the signature is not homogeneous.)

Regarding co-moving frames and cosmic time from the Big Bang, Rugh & Zinkernagel (Studies in History and Philosophy of Science, Part B: Vol 40 (1), January 2009, pp. 1–19; arXiv copy attached) suggest that:

“ ...the concept of a time scale in ‘very early’ universe cosmology lacks a physical basis or, at least, that the time scale will have to be based on speculative new physics.”

Thanks - Paul 

And incidentally, it’s reported (1303.5087.pdf attached) that our system’s peculiar velocity is only (369 ± 0.9) km/s relative to the co-moving frame. 

http://arxiv.org/pdf/1303.5087.pdf

Thanks to all for the interest, some good insights and, after a key correction along the way, I’ve now reached something close to the understanding I first sought.

One final contribution: the recent literature includes a paper in Quantum Stud.: Math. Found. (2016) 3:147–160 (copy attached) which also argues for the photon’s non-existence, but via the semi-classical approach, and which on first examination seems to account for all observations.

Although it goes into the matter in far greater depth and detail than I could achieve, I believe it is consistent with the view that, ultimately, the photon's characteristics might best be sought in the properties of the emitter and absorber, rather than being attributed to what might be nothing more than a convenient figment of the maths…

Or, as a very bright high school student (and potential maths Olympian) suggested in conversation yesterday – “Does it really matter?”

Regards - Paul

Paul,

Thank you for your thoughtful, informative, and detailed note.

The more complete Lanczos reference is Cornelius Lanczos, "The Einstein Decade" (Academic Press, New York, 1974), the section entitled "Einstein and the future on pp. 31-35. Lanczos also wrote a technical paper, Vector Potential in Riemannian space, Foundations of Physics, Vol 4, Issue 1 (1974), pp. 137-147, concerning the topic of spaceime distance versus spatial distance.

Thank you for Rugh and Zinkernagel's paper, and for the Planck paper, and for the reference to Plebanski & Krasinski’s “Introduction to General Relativity and Cosmology”. [I haven't seen the latter. But Dr. Wolfgang's Rindler's book "Relativity: Special, General, and Cosmological", 2nd Edition (Oxford University Press, 2006) is excellent.] But even if cosmic time is not perfect, it is an excellent approximation to absolute time (or at least the closest thing there is to absolute time) --- the best approximation that exists. Even if clocks, or even the cores of clocks, could not exist 10^-5 seconds after the Big Bang rather than the Planck time 10^-44 seconds after the Big Bang, since it is now 10^17 seconds after the Big Bang this should result in an "error" of only 1 part in 10 ^22. Also time as measured on Earth is only negligibly slower than cosmic time owing to the motion of Earth (V = 384 km/sec) relative to the comoving frame, slower by a fraction of  V^2/2c^2 = 1 part in 10^6.

Moreover, one can imagine a Universe in which bound systems, and hence clocks or even the cores of clocks as construed by Rugh and Zinkernagel, cannot exist at ANY time. Consider a Universe containing only cosmic background radiation (CBR). Since photons do not bind to each other, CBR photons cannot be used to construct a clock or even the core of a clock. Yet the CBR will cool as the Universe expands (and heat if it contracts if it is oscillating); thus the temperature of the CBR would still serve as a clock measuring cosmic time in such a Universe.

All the best.

Sincerely yours,

Jack

As written in my post 3 days before this one, I’ve now reached something close to the understanding I first sought.

However, as stated in the original post that started this thread, it was started in order to stimulate some discussion after getting votes hovering around zero on Stack Exchange, but getting no explanatory comments, on an answer I had just posted there to the question “What exactly is a photon?”

Subsequently, as mentioned in my 2nd reply to Ulrich in this thread, someone’s kindly pointed out to me on Stack Exchange that the concept of distance under a Lorentzian/pseudo-Riemannian (including Minkowski) metric doesn’t conform to the intuitive notion of distance found in a metric defined in what he/she calls “a strict mathematical sense”. This undermined my ‘short-cut’ justification for doubting the independent existence of photons, but not my fundamental position on this issue.  

Therefore, I have revised my argument, initially in order to respond to a later comment on Stack Exchange. As there’s no way to post a lengthy comment on Stack Exchange, I’ve attached the file as a PDF to the original post starting this thread, above. It explains and summarises the position I now take on this question, so it might be of slight interest to anyone still following this thread.

Thank you.