The Quintessential Quantum Question

[Jeff Root]

I think this may be the quintessential question about quantum theory.

Why can a photon be detected only after the entire photon has
interacted with the detector?

That appears to be true whether the "detector" is a long antenna
wire or a single electron.

-- Jeff, in Minneapolis

[John Jaksich]

This my attempt--but may sound overly simplistic or naive.

It is a result of Heisenberg's uncertainty principle.

[Grey]

I think this may be the quintessential question about quantum theory.

Why can a photon be detected only after the entire photon has
interacted with the detector?

That appears to be true whether the "detector" is a long antenna
wire or a single electron.I think this has the same problem as any "why" question in physics. We don't know why the universe is the way it is, we just know that it is that way. Electromagnetic interactions are quantized, that's just the way the world works. So you can't interact with half a photon. Any interaction that takes place is an interaction with a whole photon. And even in the case of an antenna picking up a radio signal, I'm pretty sure that if you look at the interaction on a quantum level it would still be a series of individual radio photons interacting with individual electrons in the wire.

[John Jaksich]

I think this has the same problem as any "why" question in physics. We don't know why the universe is the way it is, we just know that it is that way. Electromagnetic interactions are quantized, that's just the way the world works. So you can't interact with half a photon. Any interaction that takes place is an interaction with a whole photon. And even in the case of an antenna picking up a radio signal, I'm pretty sure that if you look at the interaction on a quantum level it would still be a series of individual radio photons interacting with individual electrons in the wire.

I certainly agree but if my memory serves me correctly---there is an interpretation of QM which speaks of how a photon will be detected in a two slit experiment. Probablity amplitudes (?) of a photon can be detected in one of two slits depending on whether one or the other slit is closed.

[Strange]

Why can a photon be detected only after the entire photon has
interacted with the detector?

Because that is what "quantized" means?

[HenrikOlsen]

I certainly agree but if my memory serves me correctly---there is an interpretation of QM which speaks of how a photon will be detected in a two slit experiment. Probablity amplitudes (?) of a photon can be detected in one of two slits depending on whether one or the other slit is closed.
That's the attempt at figuring out which slit the photon goes through. It fails because those photons going through when one of the slits is closed act according to the probability distribution from the much smaller interference effect from the edges of the slit (nearly the same as if they were particles) and only those going through when both slits are only adds to the interference pattern of the main probability distribution.

The effect of blinking one of the slits is a pattern which is two (or three if both slits blink) different property distributions overlaid, with the relative intensity of the patterns proportional to the relative amount of time of each slit configuration.

[HenrikOlsen]

Why can a photon be detected only after the entire photon has
interacted with the detector?
Because that's what they do.

Sorry, that's as close to an answer you'll get to a "why" question.


This my attempt--but may sound overly simplistic or naive.

It is a result of Heisenberg's uncertainty principle.
Horse before carriage, Heisenberg's uncertainty principle is a consequence of quantized interactions, not the other way around.
First the physical world, then humans models for it. Not the other way around.

[EigenState]

Greetings,


Because that is what "quantized" means?

Better to say that is what it means to be a quantum.

Best regards,
EigenState

[John Jaksich]

Because that's what they do.

Sorry, that's as close to an answer you'll get to a "why" question.


Horse before carriage, Heisenberg's uncertainty principle is a consequence of quantized interactions, not the other way around.
First the physical world, then humans models for it. Not the other way around.

Thanks for the clarificatiions

--john

[Jeff Root]

Why can a photon be detected only after the entire photon
has interacted with the detector?
I think this has the same problem as any "why" question in
physics. We don't know why the universe is the way it is, we
just know that it is that way.
Why does smoke rise?
Why do we need to eat?
Why does rubber stretch?
Why do lightbulbs burn out?
Why are candle flames yellow?
Why does lightning make thunder?
Why is iron stronger than aluminum?
Why is the Sun brighter than the Moon?
Why is ocean water saltier than lake water?
Why does the Sun's spectrum have dark lines in it?
Why does a magnifying glass make things look bigger?

Perhaps my question can only be answered, "That's just
the way it is", but if so, it is absolutely, definitely, certainly
NOT because it contains the word "why"!



And even in the case of an antenna picking up a radio signal,
I'm pretty sure that if you look at the interaction on a quantum
level it would still be a series of individual radio photons
interacting with individual electrons in the wire.
That's a very interesting notion! I'll have to research it!

-- Jeff, in Minneapolis

[Jeff Root]

Why can a photon be detected only after the entire photon
has interacted with the detector?
Because that is what "quantized" means?
Interesting.

A photon can only be detected after the entire photon has
interacted with the detector because "quantized" means
"a photon can only be detected after the entire photon has
interacted with the detector."

That isn't really what you meant to say, is it?

Why can the cart only be behind the horse?

Because that is what "harnessed" means.

Hmmm. Doesn't quite work for me.

-- Jeff, in Minneapolis

[korjik]

Whether it works for you or not is utterly irrelevant.

You cannot detect half a photon. You cannot detect any part of a photon. You can only detect whole photons.

Even if you dont like it.

[Ken G]

A photon can only be detected after the entire photon has
interacted with the detector because "quantized" means
"a photon can only be detected after the entire photon has
interacted with the detector."Your objection reminds me of people who object to "survival of the fittest" on the grounds that "fittest" is defined as "those that survive." What they miss is that "survival of the fittest" is not a statement about what survives, it is a statement about what matters. Evolution happens because some survive and some don't. Photons interact the way they do because they are quantized. So yes, this is the quintessential question behind quantum mechanics, but it's answer is that this is why we have quantum mechanics. If you ask, why do things fall, and someone says "because of gravity", you can then ask "but isn't gravity essentially the observation that things fall?" Yes, it is, and that's how we do physics. We observe that things fall, so we have gravity, and we observe that the whole photon has to interact, so we have quantum mechanics. So you are not asking a question about quantum mechanics, you are asking what quantum mechanics is.

[Jeff Root]

korjik,

I didn't say or imply that I don't like the fact that we can only
detect whole photons. If you think I did, you misunderstood
what I said.

What I said was that saying photons are quantized becaused
"quantized" means "quantized" doesn't tell me anything I didn't
already know. Just like saying that "harnessed" means the
horse is in front of the cart doesn't explain why the horse needs
to be in front of the cart. Strange's suggestion/comment/question
was pointless. Not wrong, but pointless.

So was your comment. My original post includes the assertion
that we can only detect whole photons. All you did was repeat
what I had already asserted.

-- Jeff, in Minneapolis

[grapes]

I may be just shooting in the dark, here, but I'd like to help Jeff answer his questioning. What about that detector? Even in a one slit experiment, with a detector, how does the detector affect the experiment? How does the detector (which kind of detector?) keep from affecting the experiment? What precautions have been taken, what advances over the years have been made in detection technology?

[EigenState]

Greetings,


I may be just shooting in the dark, here, but I'd like to help Jeff answer his questioning. What about that detector? Even in a one slit experiment, with a detector, how does the detector affect the experiment? How does the detector (which kind of detector?) keep from affecting the experiment? What precautions have been taken, what advances over the years have been made in detection technology?

I believe the insertion of interference into the discussion is a point of confusion. The original question posed was:


Why can a photon be detected only after the entire photon has
interacted with the detector?

That really has nothing to do with interference phenomena--the slits in such experiments are not detectors. It has everything to do with the fact that the photon is a single, indivisible quantum of energy which by definition cannot be partially used. Photon interactions with matter are an all or nothing event. Even in the case of scattering, be it elastic or inelastic, the incident photon is annihilated and the scattered photon is created.

Regarding interference phenomena, the nature of the detector is not relevant. Nor does the relative spatial positions of the detectors matter. All that matters is if one attempts to determine which slit the photon passed through. Do that, by any conceivable approach, and the interference pattern is lost.

Best regards,
EigenState

[Strange]

That isn't really what you meant to say, is it?

Pretty much (although EigenState has said it better). A photon is a quantum; which means it is indivisible; which means it all interracts in one go.

A better horse and cart analogy might be: why do we have to have a whole number of horses pulling the cart, why can't we have 2.3 horses?

reading it again, the way you phrase the question ("after the entire photon has interacted") it sounds a bit like you imagine that the photon starts interacting, more and more of it interacts, eventually it has all interacted and then we see the result. Which isn't how it works (as Grey says).

[profloater]

Why does smoke rise?
Why do we need to eat?
Why does rubber stretch?
Why do lightbulbs burn out?
Why are candle flames yellow?
Why does lightning make thunder?
Why is iron stronger than aluminum?
Why is the Sun brighter than the Moon?
Why is ocean water saltier than lake water?
Why does the Sun's spectrum have dark lines in it?
Why does a magnifying glass make things look bigger?

Perhaps my question can only be answered, "That's just
the way it is", but if so, it is absolutely, definitely, certainly
NOT because it contains the word "why"!


That's a very interesting notion! I'll have to research it!

-- Jeff, in Minneapolis

Aha there is more than one "why", Aristotle was onto that and he had only a smattering of English. Why did the stick hit me over the head? It was his mother's idea.

[Strange]

Perhaps my question can only be answered, "That's just
the way it is", but if so, it is absolutely, definitely, certainly
NOT because it contains the word "why"!

True. But the answers to any of these can be followed by another "why" (or "how") until it bottoms out and all you can say is either "that's the way it is" or "we don't know".

[Jeff Root]

That isn't really what you meant to say, is it?
Pretty much (although EigenState has said it better). A photon
is a quantum; which means it is indivisible; which means it all
interracts in one go.
Which is completely useless as a reply to my question.

Q: Why is a saw-edged blade used to cut bread?
A: Because that is what "bread knife" means.

Q: Why can't we see light with wavelengths longer than red?
A: Because that is what "infrared" means.

Q: Why can a photon be detected only after the entire photon
has interacted with the detector?
A: Because that is what "quantized" means.



A better horse and cart analogy might be: why do we have to
have a whole number of horses pulling the cart, why can't we
have 2.3 horses?
That's an analogy to my question. My horse and cart analogy
was an analogy to your response to my question, like those
above.

Q: Why do we have to have a whole number of horses pulling
the cart, why can't we have 2.3 horses?
Q: Because that is what "horse" means.

No useful information is conveyed by such a reply.



reading it again, the way you phrase the question ("after the
entire photon has interacted") it sounds a bit like you imagine
that the photon starts interacting, more and more of it interacts,
eventually it has all interacted and then we see the result.
I make no assumption about that. My question allows for all
possibilities. If one could not allow the possibility that interaction
might take some time, then the question would not be askable.
I asked the question to learn the answer, not to present a logical
paradox.



Which isn't how it works (as Grey says).
Is there actually evidence that photon-charge interactions are
instantaneous? I don't think there is. On the contrary, I'm pretty
sure it takes time for an electric charge to emit or absorb a photon,
and the time can be measured statistically, like the wavelength of
monochromatic light can be measured statistically.

-- Jeff, in Minneapolis

[Grey]

Jeff, nobody knows the answer. Nobody is sure why certain things in the universe are quantized. Planck introduced the quantum hypothesis pretty much as a mathematical trick to fit the blackbody radiation curve. From what I've read, he was planning on doing the standard calculus-style trick of separating something into discrete segments, and then letting the size parameter go to zero so that it's back to a continuum. It was something of a surprise to find out that it only fit the observations when the size parameter was a specific value. Since then, we've found Planck's constant cropping up in pretty much every small-scale phenomenon that we've investigated. It is a mystery why that's the case, or why that constant should have the specific value it does. When particles interact via the electromagnetic force, those interactions occur in little quantized bits, and nobody really know why.

I'm sorry that you seem really unhappy with that state of affairs, but that's the way of it. We also don't know why electric charge comes in little discrete bits of a certain size, or why the number of baryons is conserved in particle interactions. There's a whole bunch of stuff that we just don't know. Maybe someday we'll replace quantum theory with something else, and we'll be able to "explain" why electromagnetic energy is quantized. But it's a pretty sure bet that even if we do, there will be features of that new theory that aren't explainable, that we just have to accept as part of how the theory works. Until then, all anyone can do is keep repeating, "we don't really know".

[Strange]

Which is completely useless as a reply to my question.

I guess I don't really understand what it is that you are asking then (this happens to me a lot!).

Are you actually asking "why are photons quantized?" ?

[Jeff Root]

Jeff, nobody knows the answer. Nobody is sure why certain things
in the universe are quantized. ...
That reminds me of the joke about the census taker explaining to a
person being canvassed that they are trying to find out how many
people live in the US. The person replies, "Why ask me? I have
no idea!"



I'm sorry that you seem really unhappy with that state of affairs,
Even if you can't explain why photons are quantized, maybe you
can explain to me why I seem unhappy with that state of affairs.
What did I say that gave you that impression? Is it just from
something I said in this thread? Is it based on something I said
elsewhere? Or what? Please be as specific as possible. I have
no idea what gave you the impression that I'm in any way unhappy
about photons or anything else being quantized.

-- Jeff, in Minneapolis

[Grey]

Even if you can't explain why photons are quantized, maybe you
can explain to me why I seem unhappy with that state of affairs.
What did I say that gave you that impression? Is it just from
something I said in this thread? Is it based on something I said
elsewhere? Or what? Please be as specific as possible. I have
no idea what gave you the impression that I'm in any way unhappy
about photons or anything else being quantized.It's not that you seem unhappy that photons are quantized. It's that you seem unhappy that nobody can give you a good answer for why that's the case. It seems like everyone in this thread has given some variant on the only answer we have: "we don't know, that's just the way things seem to be". And you keep responding with things like "[that] doesn't tell me anything I didn't already know" or that such a response is "completely useless as a reply to my question".

We know that the electromagnetic force is quantized, and we call those quantized bits photons. We have no idea why the world should work that way, and we have no suggestion of an underlying mechanism that should make it that way. So we can't provide any better explanation than that. If you really already know all that, why are you asking the question, and then getting annoyed at people when they give you the only answer we can give? It's admittedly not much of answer, but there's no better one available, and there won't be a better one unless and until we replace quantum theory with something else.

[Jeff Root]

I guess I don't really understand what it is that you are asking then
(this happens to me a lot!).

Are you actually asking "why are photons quantized?" ?
Yes.

I asked it in the same way that I imagine someone who knows
nothing about quantum theory would ask. Say, any physicist in
1905. Asking it that way is more general than asking why photons
are quantized, because the latter assumes that we all agree on
exactly what "quantized" means. I was asking for an explanation
of the phenomenon, not a name for it.

The basic fact, as I understand it, is that an observer doesn't
know that a photon is being or has been emitted or absorbed
until it has been completely emitted or absorbed. I'm looking
for a description / explanation of why it works that way rather
than some other way. Maybe if you consider what I'm asking
for to be a description instead of an explanation, it will seem
more tractable.

-- Jeff, in Minneapolis

[Jeff Root]

Grey,

Thank you.

My complaint about your first reply wasn't that you couldn't
supply an answer-- it was that you pinned the problem on my
question being a "why" question. My question may well be
unanswerable with current knowledge-- and might even be
unanswerable with any knowledge-- but it isn't because the
question is asked with a "why".

My complaint about Strange's reply was that it was completely
circular. It didn't say anything that wasn't in the original post.

Maybe you guys were just trying to say "We don't know" in
a way that sounds more interesting and helpful than that.

-- Jeff, in Minneapolis

[EigenState]

Greetings,

Perhaps this might help some. Again, it is much better to state that a photon is a quantum rather than it is quantized. The terms are really not interchangeable. Being a quantum, it is the smallest known packet of electromagnetic energy. It is the fundamental unit of electromagnetic energy. It is therefore indivisible. That is our definition of a photon. Based upon that definition, one cannot postulate about a photon not interacting as a unit--it is axiomatically all or nothing.

As far as stimulated or spontaneous transitions, the absorption or emission of the photon is instantaneous. The easier case to imagine is spontaneous emission. The lifetime of the excited state is certainly finite. But when it does radiate, the creation of the emitted photon is indeed instantaneous.

I cannot answer the question why those are the observed and theoretical results. But they are.

Best regards,
EigenState

[HenrikOlsen]

Which is completely useless as a reply to my question.
Q: Why can a photon be detected only after the entire photon
has interacted with the detector?
A: Because that is what "quantized" means.
A: It's what they do, we don't know why.

Would you have preferred that?

[Grey]

My complaint about your first reply wasn't that you couldn't
supply an answer-- it was that you pinned the problem on my
question being a "why" question. My question may well be
unanswerable with current knowledge-- and might even be
unanswerable with any knowledge-- but it isn't because the
question is asked with a "why".Perhaps. But anytime someone answers a "why" question in science, you can ask another "why" about the answer to that one, and eventually you get to a point where you just have to accept that the universe seems to work that way. So there are always a huge number of unanswered "why questions: Why was there a big bang? Why is there gravity? Why are there three kinds of quark color, two kinds of electrical charge, and only one kind of mass? And so forth. I felt pretty sure (from interacting in other threads) that you knew enough about the situation to be aware that your question was more like "why was there a big bang?" than like "why does smoke rise?". I apologize if I made a false assumption.

[korjik]

Grey,

Thank you.

My complaint about your first reply wasn't that you couldn't
supply an answer-- it was that you pinned the problem on my
question being a "why" question. My question may well be
unanswerable with current knowledge-- and might even be
unanswerable with any knowledge-- but it isn't because the
question is asked with a "why".

My complaint about Strange's reply was that it was completely
circular. It didn't say anything that wasn't in the original post.

Maybe you guys were just trying to say "We don't know" in
a way that sounds more interesting and helpful than that.

-- Jeff, in Minneapolis

Actually, it did supply your answer. You cannot detect a photon before it finishes interacting with the detector because before it has finished interacting, nothing has happened. You cannot be half-interacted because you cannot split a photon in half.

[Jeff Root]

Gray,

In trying to understand anything, a person usually asks
questions until they reach a point at which they feel
satisfied with the answers they have so far obtained.
At a later time, they may no longer feel satisfied with
those answers and want to go beyond them. I must
have no longer been satisfied. Or maybe I just wanted
to tell BAUT readers what occurred to me might be the
quintessential question about quantum theory: The
question which underlies all other questions about
quantum theory.

The only way to know whether a question can or
cannot be answered is by trying to answer it. There
is no obvious reason that light has to be quantized,
and there is no obvious reason that there should be
no explanation of why that is the case, so it is a
natural target for questioning.

EigenState,

Light is not quantized by definition-- it is observed to be
quantized. The definition of a photon as a quantum of light
follows from observations, so the definition is not immutable.
A photon is not whatever the definition says it is. It is the
other way around. The definition has to fit the object.

-- Jeff, in Minneapolis

[EigenState]

Greetings,


EigenState,

Light is not quantized by definition-- it is observed to be
quantized. The definition of a photon as a quantum of light
follows from observations, so the definition is not immutable.
A photon is not whatever the definition says it is. It is the
other way around. The definition has to fit the object.

You are again misusing the word "quantized". All of the internal degrees of freedom of a diatiomic molecule--electronic, vibrational, rotational, and spin--are quantized. That does not mean that that diatomic molecule cannot be subjected to say photodissocation. The photon is a quantum. That quantum cannot be broken down into smaller units.

No, light is not an ensemble of quanta by our definition--rather by Nature's definition which we happened to have discovered. Once we discovered that fact, we named the quantum of electromagnetic radiation the photon. Ergo, the photon is by definition a quantum of electromagnetic radiation.

Best regards,
EigenState

[EigenState]

Greetings,


I think this may be the quintessential question about quantum theory.

Why can a photon be detected only after the entire photon has
interacted with the detector?

Actually that is not even close to being the quintessential question about quantum theory. An infinitely more fundamental question is associated with the notion of the collapse of the wavefunction upon measurement. Note that the Schrödinger equation is linear. Collapse of the wavefunction is anything but linear. There is no doubt that the concept of the collapse of the wavefunction does provide quantitative predictions of physical observables, but it is a very mysterious process and one that is far from being resolved.

Best regards,
EigenState

[HenrikOlsen]

Greetings,
You are again misusing the word "quantized". All of the internal degrees of freedom of a diatiomic molecule--electronic, vibrational, rotational, and spin--are quantized. That does not mean that that diatomic molecule cannot be subjected to say photodissocation. The photon is a quantum. That quantum cannot be broken down into smaller units.

No, light is not an ensemble of quanta by our definition--rather by Nature's definition which we happened to have discovered. Once we discovered that fact, we named the quantum of electromagnetic radiation the photon. Ergo, the photon is by definition a quantum of electromagnetic radiation.

Best regards,
EigenState
I think the two of you are actually in violent agreement.

[Boratssister]

I always thought that the only logical reason why is because its the only way it can be........

Does quantized mean ''that's the limit we can measure''?
If it does we have to entertain the thought that we may never know ''what'' never mind ''why''.

If this is the only way it can be then every photon is exactly at the right place at the right time, otherwise that particular photon would not exist.

And I would gamble that if you can't nudge a photon then that photon is not a quantized thing.

[EigenState]

Greetings,


Does quantized mean ''that's the limit we can measure''?
If it does we have to entertain the thought that we may never know ''what'' never mind ''why''.

No, quantized means that the system is characterized by discrete-valued rather than continuous-valued states.

Best regards,
EigenState

[HenrikOlsen]

Limits to measurement is a consequence of this discreteness though, because measurements require interactions and these interactions, because of the discreteness, can't be arbitrarily small.

[Ken G]

Imagine this conversation:
Evolution doubter: "Why do you think evolution occurs?"
Darwinist: "Via natural selection-- survival of the fittest."
Evolution doubter: "But why do the fittest survive?"
Darwinist: "Because that's what it means to be fit-- there is not actually a question there."
Evolution doubter: "So you are saying that the reason we have evolution is that survivors survive? That's a circular answer to my question."

The circularity of the answer "photons are quantized because that's what it means to be a photon" is just like the circularity of the answer "the fittest survive because that's what it means to be the fittest." The reason we can't answer your question in any way that doesn't sound circular is because it isn't the right question to put to a scientist. The scientific question to ask about evolution is how does a model that invokes survival of the fittest work to correctly describe the evolutionary process. That's it, that's the scientific question to ask there, not "but why is there death", or "but why is there competition", those just aren't questions that science is built to answer. Same with why are photons quantized-- it's not the right question for science, ask your local clergyman or philosopher. The right question for science is "how does a model that invokes quantized photons work to correctly describe the behavior of light." We have no idea if there really are any such things as photons, but we can talk about how the theory works, and what observations it is successful with-- just like with evolution.

[Boratssister]

Greetings,



No, quantized means that the system is characterized by discrete-valued rather than continuous-valued states.

Best regards,
EigenState

Greetings
is not time continuous?The System may be charactarized by discreet- valued quantities but those quantities may be made up of an infinity of discrepencies. So yes.

[EigenState]

Greetings,



is not time continuous?

To be determined. Many hold the opinion that time will ultimately prove to come in discrete units.



The System may be charactarized by discreet- valued quantities but those quantities may be made up of an infinity of discrepencies. So yes.

I do not understand what you mean by "an infinity of discrepencies".

Best regards,
EigenState

[HenrikOlsen]

Greetings
is not time continuous?The System may be charactarized by discreet- valued quantities but those quantities may be made up of an infinity of discrepencies. So yes.
Whether time is discrete or not has no bearing on what happens when things that have only discrete-valued properties interact.

[Ken G]

Discrete, folks. (Pet peeve, forgive the indiscretion.)

[HenrikOlsen]

I got it right in all the other posts, then got it messed up by quoting Boratssister. Fixed now. Yet another data point for the "can spell checkers replace learning to spell" debate.

[Ken G]

It's just my personal crusade to protect the words "discrete", "gantlet", and "led", as opposed to discreet, gauntlet, and lead. Pay no mind.

[HenrikOlsen]

It's just my personal crusade to protect the words "discrete", "gantlet", and "led", as opposed to discreet, gauntlet, and lead. Pay no mind.
As I have previously asked Gillianren, if you notice a misspelling in something I wrote, please tell me.

[Jeff Root]

All righty, then. Henrik, could you expand on your opinion
that EigenState and I are in agreement? I suspect that you
will need to craft one explanation that I can understand and
one that EigenState can understand, since we evidently
speak different languages.

There are other directions that I could steer the thread in
which would more directly attack my question, but this is
easier, and may be more fruitful in the short term.

-- Jeff, in Minneapolis

[Strange]

All righty, then. Henrik, could you expand on your opinion
that EigenState and I are in agreement? I suspect that you
will need to craft one explanation that I can understand and
one that EigenState can understand, since we evidently
speak different languages.

Well, you both seem to agree that light (EM radiation) is bundled in indivisible little, what shall we call them... "quanta" or "photons". These are, well, indivisible.


There are other directions that I could steer the thread in
which would more directly attack my question, but this is
easier, and may be more fruitful in the short term.

Your question seems to be "but daddy, why does light come in indivisible photons/quanta?"

Is there a reason you are not happy with the two options provided:
a) We don't know (perhaps with the implication there might be a deeper reason and one day we might understand it - but that will just lead to another "why" that we probably can't answer)
b) That's just the way it is (with the implication there is no deeper reason, and the "why" questions have to stop there)

[Jeff Root]

Strange,

I'm starting from the statement which I heard or read many
years ago, and have no memory of where I heard or read it,
that a photon can only be detected after the entire photon has
interacted with the detector. So my first question, asked many
years later, is "Why can a photon be detected only after the
entire photon has interacted with the detector?"

It is that simple.

If there is no known answer to that question, then it is
obviously a very interesting target for research.

As I said at the top of this page, a person stops asking
questions when they are satisfied with the answers they have
got so far ... or when the answerer runs out of answers, or
runs out of patience. The context of the situation determines
at what point the questioner is satisfied and stops asking
questions for the time being.

It is not apparent that interactions between electric charges
and photons occur instantaneously. I am not aware of any
evidence that they do. There is no obvious reason that one
should have to wait for the entire energy of a particle of light
to interact with a detector in order for the particle to be
detected. So it is natural to ask why that is. It's all I've
done so far.

-- Jeff, in Minneapolis

.

[HenrikOlsen]

All righty, then. Henrik, could you expand on your opinion
that EigenState and I are in agreement? I suspect that you
will need to craft one explanation that I can understand and
one that EigenState can understand, since we evidently
speak different languages.

There are other directions that I could steer the thread in
which would more directly attack my question, but this is
easier, and may be more fruitful in the short term.

-- Jeff, in Minneapolis
It looks like you're both in agreement about how things apparently do behave, the disagreement is in what words to use when talking about it.

Photons themselves are discrete, they only interact or not, one whole photon at a time, they are quanta.

But, as Eigenstate uses the word, photons are not quantized. Which is just a way of saying that they don't have properties that can only have discrete values1, photons can exist with any amount of energy and there is a continuum of possible wavelengths.

1) apart from existing or not, you can't have half of one, though you can have another one with half its energy.

[EigenState]

Greetings,



I'm starting from the statement which I heard or read many
years ago, and have no memory of where I heard or read it,
that a photon can only be detected after the entire photon has
interacted with the detector. So my first question, asked many
years later, is "Why can a photon be detected only after the
entire photon has interacted with the detector?"

It is that simple.

If there is no known answer that question, then it is obviously
a very interesting target for research.

A photon can be detected only after the entire photon has interacted with the detector because the photon is the smallest unit of electromagnetic energy. For it to be otherwise, the photon could not be the smallest unit of electromagnetic energy. There is nothing to research on this question.



It is not apparent that interactions between electric charges
and photons occur instantaneously. I am not aware of any
evidence that they do. There is no obvious reason that one
should have to wait for the entire energy of a particle of light
to interact with a detector in order for the particle to be
detected. So it is natural to ask why that is. It's all I've
done so far.

Oh, but it is apparent. This might not satisfy your operational definition of "instantaneous" but it will give you an experimentally defined upper bound on the time required for a photon interaction, and it will quantitatively illustrate the point I made earlier that the spontaneous radiative lifetime of an excited state is not determined by the temporal requirements of the actual act of photon emission. For electric-dipole allowed transitions, the spontaneous radiative lifetime of an electronic excited state is typically of the order of 10-8 seconds. By comparison, time-resolved LASER excitation experiments are conducted utilizing LASER pulses of the order of 10-15 seconds (femtoseconds). That is a difference of the order of 107.

For time-resolved LASER excitation experiments to be possible, the interaction of the photon with the sample must be less than the temporal width of the photon pulse. Thus there is very direct experimental evidence that the photon interacts with the material sample in less than 10-15 seconds. The femtosecond regime is limited not by the photon interaction dynamics, but by our current ability to generate ultrashort LASER pulses.

From a different perspective, there is neither experimental evidence nor theoretical reason to believe that the photon interaction is anything but instantaneous. Presumably that would change if time ultimately proves to be discrete.

Best regards,
EigenState

[Jeff Root]

EigenState,

The diameter of an atom is on the order of 10-10 meter.
In 10-15 second, light travels almost 3x10-7 meter, or about
3000 times the diameter of an atom. You would need to show
that a photon interacts with an atom in significantly less than
3x10-19 second to support the idea that it interacts with the
atom instantaneously. Your figure of 10-15 second tends to
support the idea that photons do not interact with atoms
instantaneously.

The fact that light progresses more slowly through matter than
through vacuum tends to support the idea that photons do not
interact with atoms instantaneously.

-- Jeff, in Minneapolis

[EigenState]

Greetings,

I would like to revisit the question of whether or not a photon-target interaction is instantaneous, and the conclusions that can be drawn from such considerations.

To our best understanding, a photon is a massless, dimensionless point particle propagating at c. Were the photon-target interaction to require some finite time interval, the photon would by logical necessity have to be localized during the time interval of that interaction. Newton and Wigner have shown (reference cited in an earlier post) definitively that the photon cannot be localized. The sole logically compelling conclusion is that photon-target interactions are indeed instantaneous.

A direct consequence of the conclusion just drawn is that a photon can be "detected only after the entire photon has interacted with the detector" because of the intrinsic physical properties of the photon: the photon is not only the indivisible unit of electromagnetic energy, but also the photon cannot be localized.

Best regards,
EigenState

[Jeff Root]

Saying that a photon cannot be "localized" appears to say
that we cannot know where the photon is at a given time.
That being the case, you cannot know whether it interacts
instantaneously or not, and the assertion that it does interact
instantaneously has no basis in either observation or theory.

-- Jeff, in Minneapolis

[HenrikOlsen]

The diameter of an atom is on the order of 10-10 meter.
In 10-15 second, light travels almost 3x10-7 meter, or about
3000 times the diameter of an atom. You would need to show
that a photon interacts with an atom in significantly less than
3x10-19 second to support the idea that it interacts with the
atom instantaneously. Your figure of 10-15 second tends to
support the idea that photons do not interact with atoms
instantaneously.
Erhm, it's the other way around, someone has to show that it definitely takes longer that a certain amount of time in order to say that is ISN'T instantaneous.
The 10-15s number is an upper bound, it says that we have measured that it takes at most 10-15s, it doesn't tell anything about a lower limit to interaction time, only that out engineering skills currently aren't good enough to make shorter light pulses.


The fact that light progresses more slowly through matter than
through vacuum tends to support the idea that photons do not
interact with atoms instantaneously.
The explanation given for that is that photons do interact instantaneous with matter to get annihilated to transfer their energy to the matter, then the creation of new photons to pass that energy on happens after a short interval.
The interactions are instantaneous, the slight delay between interactions isn't.

[EigenState]

Greetings,


Saying that a photon cannot be "localized" appears to say
that we cannot know where the photon is at a given time.
That being the case, you cannot know whether it interacts
instantaneously or not, and the assertion that it does interact
instantaneously has no basis in either observation or theory.

That the photon cannot be localized without its being annihilated means that the photon cannot be described in terms of position eigenstates.

As our colleague HenrikOlsen suggested, perhaps you would be kind enough to provide us with some evidence, experimental or theoretical, that is consistent with either of your speculations. Preferably some evidence that is also consistent with quantum mechanics and quantum electrodynamics.

Best regards,
EigenState

[Jeff Root]

The diameter of an atom is on the order of 10-10 meter.
In 10-15 second, light travels almost 3x10-7 meter, or about
3000 times the diameter of an atom. You would need to show
that a photon interacts with an atom in significantly less than
3x10-19 second to support the idea that it interacts with the
atom instantaneously. Your figure of 10-15 second tends to
support the idea that photons do not interact with atoms
instantaneously.
Erhm, it's the other way around, someone has to show that
it definitely takes longer than a certain amount of time in order
to say that is ISN'T instantaneous.
Why? Is everything assumed to be instantaneous unless
shown to be otherwise?



The 10-15s number is an upper bound, it says that
we have measured that it takes at most 10-15s, it
doesn't tell anything about a lower limit to interaction time,
only that out engineering skills currently aren't good enough
to make shorter light pulses.
Understood and agreed. The fact that this time interval is
long compared to the time it takes light to go past an atom
tends to support the idea that the interaction takes time.
That is very weak support, but it is support. If the measured
upper bound were less than 3x10-19 second, it would
tend to support the idea that the interaction is instantaneous.
If it were much less than 3x10-19 second, the support
for instantaneous interaction would be strong.




The fact that light progresses more slowly through matter than
through vacuum tends to support the idea that photons do not
interact with atoms instantaneously.
The explanation given for that is that photons do interact
instantaneously with matter to get annihilated to transfer their
energy to the matter, then the creation of new photons to pass
that energy on happens after a short interval.
The interactions are instantaneous, the slight delay between
interactions isn't.
Most interesting! I imagine that there is no "explanation" of
how, when light passes through a transparent material, the
newly-created photons are emitted in the same direction as
the photons which were absorbed some time earlier, during
which interval the electrons have had time to bounce around
their atoms thousands of times?

-- Jeff, in Minneapolis

[EigenState]

Greetings,


Why? Is everything assumed to be instantaneous unless
shown to be otherwise?

Two fundamental reasons have been articulated that negate your ideas. You seem to reject those physical principles out of hand. Thus, I do believe that other readers can reasonably request that you provide some evidence in support of your ideas. That is the way the scientific community functions.



Most interesting! I imagine that there is no "explanation" of
how, when light passes through a transparent material, the
newly-created photons are emitted in the same direction as
the photons which were absorbed some time earlier, during
which interval the electrons have had time to bounce around
their atoms thousands of times?


Indeed there is a rigorous explanation. It goes by the name of Quantum Electrodynamics.

Best regards,
EigenState

[caveman1917]

Why can a photon be detected only after the
entire photon has interacted with the detector?"
{snip}
There is no obvious reason that one
should have to wait for the entire energy of a particle of light
to interact with a detector in order for the particle to be
detected.
(my bold)

I think the problem you're having with this is because of a hidden assumption in your thinking.
There's no sense in placing the qualifier "entire" with something unless it makes sense to talk about "parts" of that something.
You're starting from a premisse that the photon (or its energy) is divisible, which naturally gets you to a contradiction. The resolution of your conundrum lies in letting go of that hidden premisse, otherwise it's only natural that no explanation given here will seem satisfactory to you.

[Ken G]

It might help you to recognize that there is no such thing as photons, other than the demonstrably true statement that a "photon" is a semantic element of a physical theory. Every theory that invokes the photon concept must define what that concept is. Then we test if the theory works. Your question is, "why does a theory work that involves discrete interactions between quanta?" It's not just that we don't know why the theory works, it's that science never knows why its theories work. The theories work because they capture some aspect of the true reality, presumably, and nothing more can ever be said. Why does electroweak theory work by invoking virtual bosons? Why does general relativity work by invoking spacetime curvature? I don't see your question as any different from those, can you explain why you think there is something in your question that is not in those?

[noncryptic]

Interesting.

A photon can only be detected after the entire photon has
interacted with the detector because "quantized" means
"a photon can only be detected after the entire photon has
interacted with the detector."


Quantized means there are no smaller parts - of a photon, i this case.
So you can't first have a part of a photon interact with the detector and after that have another part of the photon interact with the detector.

[Jeff Root]

Why can a photon be detected only after the
entire photon has interacted with the detector?"
{snip}
There is no obvious reason that one should have to wait
for the entire energy of a particle of light to interact with
a detector in order for the particle to be detected.
(my bold)

I think the problem you're having with this is because of a
hidden assumption in your thinking.

There's no sense in placing the qualifier "entire" with
something unless it makes sense to talk about "parts"
of that something.

You're starting from a premise that the photon (or its energy)
is divisible, which naturally gets you to a contradiction.
It is neither an assumption nor is it hidden.

I used the word "entire" deliberately. But it does not imply
that a photon has parts. Instead, it results from the fact that
energy is observer-dependent. A particular quantum of light
does not have any particular energy. The energy observed
depends in part on the motion of the observer, but the entire
energy must be observed in order for any of it to be observed.
The question of this thread asks why it is that way, since it
isn't obvious that it has to be.

I make no assumption that a photon is divisible into any
parts. On the contrary, I have never wavered from the
assumption that it does not. Nevertheless, the question
still presents itself: Why can a photon be detected only
after the entire photon has interacted with the detector?



The resolution of your conundrum lies in letting go of
that hidden premise, otherwise it's only natural that no
explanation given here will seem satisfactory to you.
It has been said several times that no explanation is
known, and that no explanation is possible. Do you
think that an explanation has been gven here?

-- Jeff, in Minneapolis

[Jeff Root]

It looks like you're both in agreement about how things apparently
do behave, the disagreement is in what words to use when talking
about it.

Photons themselves are discrete, they only interact or not, one
whole photon at a time, they are quanta.

But, as Eigenstate uses the word, photons are not quantized.
Which is just a way of saying that they don't have properties that
can only have discrete values1, photons can exist with any
amount of energy and there is a continuum of possible wavelengths.

1) apart from existing or not, you can't have half of one, though
you can have another one with half its energy.
Thanks. That was just simple enough for me to be able to
understand, and I think it ought to be complex enough for
EigenState to understand, too.

It appears that I didn't consider the difference between the
terms "quantum" and "quantized" to have physical significance,
just grammatical significance. I can't promise to use them any
better than I have, but I'll try to watch out for situations where
the ambiguity seems likely to cause problems, and choose
the more appropriate term.

-- Jeff, in Minneapolis

[EigenState]

Greetings,



I used the word "entire" deliberately. But it does not imply
that a photon has parts. Instead, it results from the fact that
energy is observer-dependent. A particular quantum of light
does not have any particular energy. The energy observed
depends in part on the motion of the observer, but the entire
energy must be observed in order for any of it to be observed.
The question of this thread asks why it is that way, since it
isn't obvious that it has to be.

Asked and answered. A motional frequency shift does not alter the fundamental nature of the quantum--it remains the smallest unit of electromagnetic energy. That "the entire energy must be observed in order for any of it to be observed" is inherent to the very concept of the quantum.



It has been said several times that no explanation is
known, and that no explanation is possible. Do you
think that an explanation has been gven here?


Perhaps the more physically substantive question would be: Why is electromagnetic radiation comprised of quanta?

Best regards,
EigenState

[EigenState]

Greetings,



Thanks. That was just simple enough for me to be able to
understand, and I think it ought to be complex enough for
EigenState to understand, too.


Thank you for your concern, but I understood it already.

Best regards,
EigenState

[Jeff Root]

Yes, I was depending too much on my memory for the context.
Henrik basically just explained to me what you were saying, so
there wasn't anything that he needed to explain to you about it!

The thread is apparently long enough now that I can't remember
it all well enough to depend on my memory.

-- Jeff, in Minneapolis

[Jeff Root]

I think we're past this, but the post happens to be sitting
in front of me, and I now have a reply to it.




Which is completely useless as a reply to my question.
Q: Why can a photon be detected only after the entire photon
has interacted with the detector?
A: Because that is what "quantized" means.
A: It's what they do, we don't know why.

Would you have preferred that?
Yes, I probably would have let such a reply go without comment.

The answer Strange suggested contains the relevant fact that
"a photon can be detected only after the entire photon has
interacted with the detector" *is* what "quantized" means.
But saying that doesn't answer the question, so starting his
reply with the word "Because" made it wrong. Also, it implied
that I didn't understand that fact, when it should have been
clear from the context that I did understand it.

-- Jeff, in Minneapolis

[Frog march]

if interactions with photos happen instantaneously then why does gravitational time dilation affect the frequency of the photon emitted?


also, you can have a quantum of cart pulling, ie 1 horse, but that doesn't mean that the horse isn't made of lots of parts.

[HenrikOlsen]

if interactions with photos happen instantaneously then why does gravitational time dilation affect the frequency of the photon emitted?
If you look at it as affecting the energy available to the photon instead, the mystery goes away.

[Jeff Root]

... anytime someone answers a "why" question in science,
you can ask another "why" about the answer to that one, and
eventually you get to a point where you just have to accept
that the universe seems to work that way. So there are always
a huge number of unanswered "why questions: Why was there a
big bang? Why is there gravity? Why are there three kinds of
quark color, two kinds of electrical charge, and only one kind
of mass? And so forth. I felt pretty sure (from interacting in
other threads) that you knew enough about the situation to be
aware that your question was more like "why was there a big
bang?" than like "why does smoke rise?". I apologize if I made
a false assumption.
I'm not sure I've ever made a distinction.

As I indicated in an earlier post, I see myself and others
stop asking questions when the answer seems satisfying, or
when the answerer has no more answers, or the answerer loses
patience with the questioner. Sometimes the questioner will
pick up where he left off at a later time. There are any number
of unanswered questions floating around, but they are generally
unanswered because they are historical in nature and the
answers have apparently been lost, or because they are
impractical to answer, or because an answer hasn't turned
up yet, or because the question doesn't seem important
enough to bother with just now.

The question of why there was a Big Bang is being worked on.
Since it is a particularly big question, I don't expect an
answer any time soon. I could hope that an answer emerges in
my lifetime. I've given a little thought to it, but not much.

The question of why there is gravity is one I don't understand,
so I haven't asked it and wouldn't know how to investigate it.
The question certainly sounds interesting, though.

I might put those questions into different categories: The first
into the category of questions now being researched, and the
second into the category of questions I don't understand.

The question of why a photon can be detected only after the
entire photon has interacted with the detector is one that I
might categorize as one I am currently trying to understand
better.

-- Jeff, in Minneapolis

[Jeff Root]

Why can a photon be detected only after the entire photon has
interacted with the detector?


Perhaps the more physically substantive question would be:
Why is electromagnetic radiation comprised of quanta?
I like my question, and I don't much like yours. It reminds me
of the question my English composition instructor came back
with after I posed a question to him, in 1971.

I had made a short list of questions that I thought interesting,
and showed it to him outside of class. It included the question,
"How do I recognize beauty?" He responded, "Maybe first you
should ask, 'Is there such a thing as beauty?'" That eventually
became my prime example of a "stupid question".

First, there are only two possible answers to his question:
"Yes" and "No". So in effect, I already know the answer, I
just don't know which of those two possibilities it is.

Now, how would I go about answering his question? Well, In
order to answer his question of whether there is such a thing
as beauty or not, I'd have to have some way of recognizing
beauty, so I need to know how to recognize it. Oooop. That
was my original question. In order to answer his question, I
first have to answer mine. There is no way to determine
whether there is such a thing as beauty without first having
a way to recognize it. And it got to that point in just one step.

Your question isn't quite as bad. Rather than needing to answer
my question before yours can be answered, the two questions
are probably equivalent. But I think mine is clearer and less
ambiguous in what it is asking.

-- Jeff, in Minneapolis

[baskerbosse]

If there are no parts to a photon, there is nothing like a 'partial' detection, surely.
If that were to be the case, -what would be detected? A potential photon? A partial photon?

EigenState's question "Why is electromagnetic radiation comprised of quanta?" is very interesting.
Is not the energy that can be contained in a photon linear?
Redshifting is not in discrete steps, is it?


Peter

[Ken G]

I'm still wondering if you see your question any differently from these: "Why does electroweak theory work by invoking virtual bosons? Why does general relativity work by invoking spacetime curvature?"

I don't see your question as any different from those, can you explain why you think there is something in your question that is not in those? Or would you call them the "quintessential question of electroweak theory" and the "quintessential question of general relativity" too?

[caveman1917]

Redshifting is not in discrete steps, is it?

No it isn't, it's continuous. I may be wrong on this, but i think you're supposed to take the energy in the rest frame of the atom it's interacting with, and either it all interacts or nothing interacts.

[Jeff Root]

... You cannot detect a photon before it finishes interacting with
the detector because before it has finished interacting, nothing
has happened. You cannot be half-interacted because you cannot
split a photon in half.
I agree entirely with the last bit: A photon is indivisible.

However, I see no reason to claim that "nothing has happened"
before the photon has finished interacting. (I take responsibility
for introducing and using the expression "finished interacting".)
I am still not aware of any evidence that interactions between
photons and electrons are anywhere close to instantaneous.
The measurement EigenState quoted seems to say that the
interaction time has only been narrowed down to 10-15 second,
which is so long that a photon could pass through thousands
of atoms in that time if it did not interact with them. While not
convincing evidence that interactions take time, it also is not
evidence that interactions are instantaneous. It merely limits
the interactions to about 3000 times as long as one would
expect them to take.

-- Jeff, in Minneapolis

[Jeff Root]

Ken,

I don't have any opinion about those other questions right now.

The question I asked is interesting to me at the moment, and it
still seems to me to be expressed in a way that is most clear and
least ambiguous.

-- Jeff, in Minneapolis

[baskerbosse]

No it isn't, it's continuous. I may be wrong on this, but i think you're supposed to take the energy in the rest frame of the atom it's interacting with, and either it all interacts or nothing interacts.

I suspected as much. Otherwise the expansion of the universe would have to be in discrete steps, wouldn't it? That would be bizarre. ;-)

I guess it follows from this, that any lacking or additional energy (not absorbed by the electron) would be added or subtracted from the kinetic energy of the atom?
(-or is this the wrong way of thinking? Does maybe the photon only react with an atom that already has the appropriate speed (red/blue shift) compared to the photon?)
Is this related to how you make Bose-Einstein condensates?

Peter

[Ken G]

I am still not aware of any evidence that interactions between
photons and electrons are anywhere close to instantaneous.There is no reason to think the interaction is instantaneous, any more than there is a reason to, think there is a specific interaction in the first place (in some literal sense). Everything that happens in the quantum domain is modeled as a sum over many possible things that could have happened, none of which can individually be said to have actually happened. So an "interaction" is actually quite a nebulous thing-- in quantum mechanics, you have what enters the black box, and what comes out of the black box, and you can predict the statistical tendencies of what comes out, but you can never say what happened in that box. The individual possibilities are all instantaneous, that's the "quantum" element, but the sum over all the possibilities involves a spread in time, so one would not say the interaction is instantaneous any more than one would say that one of those possible interactions "is the one that actually happened, we just don't know which."

Here's the general rule: if the reality is not set up to pose a particular question, then the realitiy does not answer the question either. The reality is simply not set up to pose the question of what happened to the photon until the interaction is over, and that's why the interaction must be completed before we can say what happened. But note that by "what happened", I mean the outcome of what happened, not what went on in the black box that the reality is simply not set up to answer. So you can see the same circularity in this answer-- you can't say what the outcome is until the outcome is an outcome. But this is the general law of quantum behavior, which is fundamental and therefore not answerable. Reality is not set up to answer your question, this is the lesson of quantum mechanics.

[Jeff Root]

i think you're supposed to take the energy in the rest frame of the
atom it's interacting with, and either it all interacts or nothing interacts.
Unless you are trying to learn something about the atom or the
neighborhood of the atom from which the light came, the frame
is irrelevant. If you are interested in redshift or blueshift because
you want to determine the radial speed of the emitting atom, or
because you want to identify the isotope, then you'll be comparing
photons from the moving atoms with photons from atoms at rest
with respect to you. Either way, the photons behave the same
as far as their quantumness goes.

I don't think I'm saying anything you don't know every bit as well
as I do, so maybe your uncertainty was motivated by something
that went completely over my head. I just wanted to clarify it for
any readers who know even less about it than I know.

-- Jeff, in Minneapolis

[HenrikOlsen]

I suspected as much. Otherwise the expansion of the universe would have to be in discrete steps, wouldn't it? That would be bizarre. ;-)

I guess it follows from this, that any lacking or additional energy (not absorbed by the electron) would be added or subtracted from the kinetic energy of the atom?
That would be my understanding too.

[Jeff Root]

Henrik,

Could you go back and reply to the questions at the top of my
post #56, pleeeease? Thank you. :)

-- Jeff, in Minneapolis

[Strange]

Also, it implied
that I didn't understand that fact, when it should have been
clear from the context that I did understand it.

That certainly wasn't clear fom the context (perhaps because of the word "entire", which as others have said implies there is/could be partial interaction). I assumed you know it from previous conversations though (hence the question mark).


But I think mine is clearer and less
ambiguous in what it is asking.

Yours is highly ambiguous. Maybe because of the word "entire"; maybe because it mixes up both the issue of em radiation coming in quanta and the way quanta interact.

p.s. I think the question 'Is there such a thing as beauty?' is a really good one. For one thing, it doesn't have a simple yes or no answer. I'm sure one could write an essay, or even a book, just about that question before you even get on to the 'recognizing' one...

[HenrikOlsen]

Henrik,

Could you go back and reply to the questions at the top of my
post #56, pleeeease? Thank you. :)

-- Jeff, in Minneapolis
I'm guessing you mean this one:

Why? Is everything assumed to be instantaneous unless
shown to be otherwise?
We're quite certain, based on experimental data, that the photon interacting with matter is an all or nothing event, with no way the photon can interact halfway.

To me this means that assuming it happens over time leads to a contradiction because, if it did happen over time, i.e. it's an event with a start end end time that aren't the same, there'd be a time between the start and the end of the interaction where it had either

only interacted partially, which contradicts that interactions can only be all or nothing
hadn't interacted yet, which contradicts the assumption that this is after the start
has interacted already, which contradicts the assumption that this is before the end

To me, this means that either it IS instantaneously so the start and end are the same; or it isn't actually correct that a photon can only interact all-or-nothing; or time is discrete and the start and end are at most one tick apart, making it impossible to pick a time between start and end; or something else is preventing us from picking a time between the start and end.

One thing's for certain, if experimental evidence shows that it does take time to happen, it's going to result in some quite interesting revisions to a lot of assumptions.

[EigenState]

Greetings,


I guess it follows from this, that any lacking or additional energy (not absorbed by the electron) would be added or subtracted from the kinetic energy of the atom?

Considering photon absorption, in the case of a resonant transition (between discrete, quantized bound energy levels), there is no additional energy. All of the energy of the photon goes into the internal degrees of freedom of the atom or molecule. In the case of excitation to the continuum (photoionization) then the amount of energy of the incident photon that is in excess of that required to reach the photoionization threshold (ionization potential) ends up as kinetic energy of the freed electron.



Does maybe the photon only react with an atom that already has the appropriate speed (red/blue shift) compared to the photon?)


For a single photon or ideally monochromatic source of photons, then yes the radial speed of the atom relative to the direction of photon propagation would matter because you need to satisfy the resonant condition.

If you have an excitation source in which the photon energy can be scanned over some bandwidth--for example a tunable LASER--then yes indeed the motion of the target relative to the direction of photon propagation matters. In the laboratory, tunable LASERS are utilized to measure excitation spectra and the random thermal motions of the target atoms results in Doppler broadening. There are ways to overcome Doppler broadening of course, but they are well beyond the scope of this thread.

Best regards,
EigenState

[EigenState]

Greetings,



I like my question, and I don't much like yours. It reminds me
of the question my English composition instructor came back
with after I posed a question to him, in 1971.

I had made a short list of questions that I thought interesting,
and showed it to him outside of class. It included the question,
"How do I recognize beauty?" He responded, "Maybe first you
should ask, 'Is there such a thing as beauty?'" That eventually
became my prime example of a "stupid question".


With all due respect, I never asked you to like the question that I posed. All I said is that it is a more physically substantive question.

Nor is it in any way similar to the question posed by your English instructor. Beauty is a subjective attribute. The nature of electromagnetic radiation is objective.

Best regards,
EigenState

[EigenState]

Greetings,


There is no reason to think the interaction is instantaneous, any more than there is a reason to, think there is a specific interaction in the first place (in some literal sense). Everything that happens in the quantum domain is modeled as a sum over many possible things that could have happened, none of which can individually be said to have actually happened. So an "interaction" is actually quite a nebulous thing-- in quantum mechanics, you have what enters the black box, and what comes out of the black box, and you can predict the statistical tendencies of what comes out, but you can never say what happened in that box. The individual possibilities are all instantaneous, that's the "quantum" element, but the sum over all the possibilities involves a spread in time, so one would not say the interaction is instantaneous any more than one would say that one of those possible interactions "is the one that actually happened, we just don't know which."

Is the Hamiltonian your "black box"? Are you invoking a "Many Worlds" perspective?

Best regards,
EigenState

[Ken G]

No, many worlds is an interpretation, I'm talking about basic quantum mechanics. A good example is the double slit experiment-- there the "black box" is the two slits. Not only do we not say which slit the photon went through, the reality itself takes no position on the question, to whatever extent reality mirrors the approach of quantum mechanics.

[EigenState]

Greetings,


No, many worlds is an interpretation, I'm talking about basic quantum mechanics. A good example is the double slit experiment-- there the "black box" is the two slits. Not only do we not say which slit the photon went through, the reality itself takes no position on the question, to whatever extent reality mirrors the approach of quantum mechanics.

Thank you for that clarification. However, interference phenomena are substantively different from the question of the detection of a single photon in which that single photon is annihilated.

The question that I was leading up to is predicated upon your statement, with emphasis added:


There is no reason to think the interaction is instantaneous, any more than there is a reason to, think there is a specific interaction in the first place (in some literal sense). Everything that happens in the quantum domain is modeled as a sum over many possible things that could have happened, none of which can individually be said to have actually happened. So an "interaction" is actually quite a nebulous thing-- in quantum mechanics, you have what enters the black box, and what comes out of the black box, and you can predict the statistical tendencies of what comes out, but you can never say what happened in that box. The individual possibilities are all instantaneous, that's the "quantum" element, but the sum over all the possibilities involves a spread in time, so one would not say the interaction is instantaneous any more than one would say that one of those possible interactions "is the one that actually happened, we just don't know which."

Would you care to elaborate on the "the sum over all the possibilities involves a spread in time" within the context of a single particle "interaction"? Given that you are thinking within the framework of the Copenhagen interpretation of a collapsing wavefunction, you seem to be implying that the collapse of the wavefunction to any of the various potential outcomes is somehow time-dependent.

Best regards,
EigenState

[Ken G]

Thank you for that clarification. However, interference phenomena are substantively different from the question of the detection of a single photon in which that single photon is annihilated. No, detection experiments are exactly like interference experiments. Everything in quantum mechanics is an interference experiment, because they all involve wave phenomena. That's what the wave function is all about.

Would you care to elaborate on the "the sum over all the possibilities involves a spread in time" within the context of a single particle "interaction"? Tell me what single particle interaction you have in mind, and I'll tell you the sum over possibilities that is involved. Most likely, it will be a sum that involves the wave functions of the particles that are interacting.


Given that you are thinking within the framework of the Copenhagen interpretation of a collapsing wavefunction, you seem to be implying that the collapse of the wavefunction to any of the various potential outcomes is somehow time-dependent. I'm not invoking any interpretation at all, merely how quantum amplitudes get calculated. The calculation will inevitably refer to a spread in times, as per the Heisenberg uncertainty principle.

[Jeff Root]

EigenState,

My understanding is that, even when thermal Doppler shifting is
completely controlled, photon emission due to a transition between
energy levels is a Gaussian distribution, or something very close to
a Gaussian distribution. It will be very narrow, but it will still be a
distribution. If for no other reason, such a distribution must occur
as a result of uncertainty. Is that correct? Can you locate a graph
showing the shape of the curve for an example transition? I don't
know what to search for.

-- Jeff, in Minneapolis

[Jeff Root]

Nor is it in any way similar to the question posed by your
English instructor. Beauty is a subjective attribute. The
nature of electromagnetic radiation is objective.
That is one difference. Even if you listed twenty thousand
differences, there could still be enough similarity between
your proposed question and my English instructor's proposed
question for one to remind me of the other.

And even if there weren't *that* much similarity, there could
still be *some* similarity.

-- Jeff, in Minneapolis

[Jeff Root]

The diameter of an atom is on the order of 10-10 meter.
In 10-15 second, light travels almost 3x10-7 meter, or about
3000 times the diameter of an atom. You would need to show
that a photon interacts with an atom in significantly less than
3x10-19 second to support the idea that it interacts with the
atom instantaneously. Your figure of 10-15 second tends to
support the idea that photons do not interact with atoms
instantaneously.
Erhm, it's the other way around, someone has to show that
it definitely takes longer than a certain amount of time in order
to say that is ISN'T instantaneous.
Why? Is everything assumed to be instantaneous unless
shown to be otherwise?


We're quite certain, based on experimental data, that the photon
interacting with matter is an all or nothing event, with no way the
photon can interact halfway.
Okay.



To me this means that assuming it happens over time leads to a
contradiction because, if it did happen over time, i.e. it's an event
with a start and end time that aren't the same, there'd be a time
between the start and the end of the interaction where it had either

only interacted partially, which contradicts that interactions can only be all or nothing
hadn't interacted yet, which contradicts the assumption that this is after the start
has interacted already, which contradicts the assumption that this is before the end

To me, this means that either it IS instantaneously so the start
and end are the same; or it isn't actually correct that a photon can
only interact all-or-nothing; or time is discrete and the start and end
are at most one tick apart, making it impossible to pick a time
between start and end; or something else is preventing us from
picking a time between the start and end.
I don't see a contradiction.

If at some point, the interaction is only partially complete, there is
no conflict with the fact that incomplete interactions are never seen.
How would you observe a partially completed interaction? You
wouldn't, because there is no mechanism by which such an
observation could be made. Only completed interactions can be
seen. That doesn't mean the interactions are instantaneous.

I still am not aware of any evidence that such interactions are
instantaneous.

-- Jeff, in Minneapolis

[Ken G]

Guys, the interaction is not instantaneous, insofar as it receives contribution from a sum of possible interactions, the latter of which are instantaneous but occur over a spread in times. This is the Heisenberg uncertainty principle. This is indeed the reason that the interaction cannot be with half a photon, that would only be half the sum that you need to do.

[EigenState]

Greetings,


No, detection experiments are exactly like interference experiments. Everything in quantum mechanics is an interference experiment, because they all involve wave phenomena. That's what the wave function is all about.

The photon has no wavefunction within the Schrödinger formalism. That was established long ago by Wigner and I have posted the reference earlier within this thread.


Tell me what single particle interaction you have in mind, and I'll tell you the sum over possibilities that is involved. Most likely, it will be a sum that involves the wave functions of the particles that are interacting.

Single photon, resonant absorption to a bound atomic state will do.


I'm not invoking any interpretation at all, merely how quantum amplitudes get calculated. The calculation will inevitably refer to a spread in times, as per the Heisenberg uncertainty principle.

I understand your point regarding the probability amplitudes. But I am talking about the collapse of the wavefunction that takes place at measurement.

Best regards,
EigenState

[Ken G]

The photon has no wavefunction within the Schrödinger formalism. That was established long ago by Wigner and I have posted the reference earlier within this thread.
And what is the photon interacting with? There's your wave function.


Single photon, resonant absorption to a bound atomic state will do.
Then the atomic state you are imagining provides the wave function I'm referring to. That interaction is not instantaneous, it is exactly what I've been saying-- a sum over all the possible times that such an interaction could occur, based on the atomic state wave function. Normally, a calculation like that involves what is known as a "matrix element" in the "position basis", which involves a sum over all the possible locations where the atomic state has a non-negligible amplitude, each of which corresponds to a different time of interaction with the photon (which itself does not have a known time of arrival, so there is a time uncertainty relation involved in both the photon and the atom).



I understand your point regarding the probability amplitudes. But I am talking about the collapse of the wavefunction that takes place at measurement.That is also not instantaneous, because any measurement requires a finite time to "decohere" the amplitudes and weed out all the possible outcomes that are not actualized. Indeed, that is just exactly what a measurement is.

[Jeff Root]

This is the Heisenberg uncertainty principle.
Ah! *That's* the way to put it! All I could think of was,
"This is due to the Heisenberg uncertainty principle", or
something equivalent, and I didn't like that at all, so I
didn't say anything about it.

-- Jeff, in Minneapolis

[Ken G]

Right, it's not due to the HUP, it is the HUP.

[EigenState]

Greetings,



My understanding is that, even when thermal Doppler shifting is
completely controlled, photon emission due to a transition between
energy levels is a Gaussian distribution, or something very close to
a Gaussian distribution. It will be very narrow, but it will still be a
distribution. If for no other reason, such a distribution must occur
as a result of uncertainty. Is that correct? Can you locate a graph
showing the shape of the curve for an example transition? I don't
know what to search for.


You are talking about the natural line profile which is Lorentzian not Gaussian. The two distributions are very similar in shape if you use the same linewidth to calculate the profiles and normalize them, but the Lorentzian is wider in the wings. The natural linewidth is the result of the Uncertainty Principle as it applies to the energy uncertainties of both states involved in a transition.

It should be recognized however that for a sample of He gas at STP, the Doppler broadening from the random thermal motions of the He atoms will exceed the natural linewidth by more than a factor of 10.

Best regards,
EigenState

[EigenState]

Greetings,


And what is the photon interacting with? There's your wave function.
Then the atomic state you are imagining provides the wave function I'm referring to. That interaction is not instantaneous, it is exactly what I've been saying-- a sum over all the possible times that such an interaction could occur, based on the atomic state wave function. Normally, a calculation like that involves what is known as a "matrix element" in the "position basis", which involves a sum over all the possible locations where the atomic state has a non-negligible amplitude, each of which corresponds to a different time of interaction with the photon (which itself does not have a known time of arrival, so there is a time uncertainty relation involved in both the photon and the atom).

Of course the atomic wavefunction is a time-dependent linear super position of the eigenfunctions which can however be prepared in a functionally time-invariant pure state. However, the interaction is governed by the Hamiltonian which is not explicitly time-dependent. You are talking about the atom, I am talking about the Hamiltonian.

As for the collapse of the wavefunction versus decoherence, that is a question that remains unresolved.

Best regards,
EigenState

[Ken G]

Of course the atomic wavefunction is a time-dependent linear super position of the eigenfunctions which can however be prepared in a functionally time-invariant pure state. However, the interaction is governed by the Hamiltonian which is not explicitly time-dependent. You are talking about the atom, I am talking about the Hamiltonian.Arguing time independence is pretty much the opposite of instantaneity, you realize. In the limit of time-independent wave functions, we have complete ignorance of any time when anything happens. That is indeed the opposite of instantaneity. However, if you wanted to calculate when a photon would be detected, you would not use time-independent wavefunctions, and your prediction of when the photon will be detected will always have a time spread, consistent with the HUP. Even more importantly, any effort to measure the time that the photon was detected would also have an uncertainty consistent with the HUP, or there would be no HUP. The time independence of the Hamiltonian is irrelevant to this question, it would be like saying that if my car is always my car, it must collide with things instantaneously.

The key point in all this is that in quantum mechanics, things simply don't happen at specific times. Instead, the wave function exists at a given time, but things that happen involve matrix elements, which are essentially "dot products" of wave functions. A dot product involves expanding the wave function on a basis that is physically relevant to the happening in question, which the wave function itself most likely does not have a specific meaning in regard to. Once you've done the expansion on the new basis, you have a way of talking about what happened there, but at the cost of having to do the sum over all the possible ways that it could happen. That sum is the dot product.

[EigenState]

Greetings,

I have always told my students that what is correct matters, while who was correct does not. Time to live up to that.

I have yet to digest the following references, but they appear to make a very strong argument favoring the decoherence perspective vis a vis measurements.

Theory (60 pages and not easy going at all): Wojciech H. Zurek, Decoherence, einselection, and the quantum origins of the classical (http://arxiv.org/abs/quant-ph/0105127), Reviews of Modern Physics 75, 715 (2003).

Experiment: M. Brune, E. Hagley, J. Dreyer, X. Maitre, A. Maali, C. Wunderlich, J. M. Raimond, and S. Haroche, Observing the Progressive Decoherence of the “Meter” in a Quantum Measurement (http://www.uni-siegen.de/fb7/quantenoptik/forschung/publikationen/publis/sk_prl.pdf), Physical Review Letters, 77 (24), 4887-4890 (1996). NB: the link is to a direct download of a PDF file of the article.


A mesoscopic superposition of quantum states involving radiation fields with classically distinct phases was created and its progressive decoherence observed. The experiment involved Rydberg atoms interacting one at a time with a few photon coherent field trapped in a high Q microwave cavity. The mesoscopic superposition was the equivalent of an “atom 1 measuring apparatus” system in which the “meter” was pointing simultaneously towards two different directions—a “Schrödinger cat.” The decoherence phenomenon transforming this superposition into a statistical mixture was observed while it unfolded, providing a direct insight into a process at the heart of quantum measurement.

Best regards,
EigenState

[baskerbosse]

Isn't it possible for it to neither take any specific time, nor happen at a specific time?

Seems to me that the argument is now that, -if it doesn't happen at a specific time, then it has a duration?

Peter

[Ken G]

I have yet to digest the following references, but they appear to make a very strong argument favoring the decoherence perspective vis a vis measurements.

Theory (60 pages and not easy going at all): Wojciech H. Zurek, Decoherence, einselection, and the quantum origins of the classical (http://arxiv.org/abs/quant-ph/0105127), Reviews of Modern Physics 75, 715 (2003).

Experiment: M. Brune, E. Hagley, J. Dreyer, X. Maitre, A. Maali, C. Wunderlich, J. M. Raimond, and S. Haroche, Observing the Progressive Decoherence of the “Meter” in a Quantum Measurement (http://www.uni-siegen.de/fb7/quantenoptik/forschung/publikationen/publis/sk_prl.pdf), Physical Review Letters, 77 (24), 4887-4890 (1996). I agree that these kinds of decoherence experiments really do have a lot to tell us about the quantum/macro interface, I'm glad they are being carried out.

[Ken G]

Isn't it possible for it to neither take any specific time, nor happen at a specific time?

Seems to me that the argument is now that, -if it doesn't happen at a specific time, then it has a duration?
Not quite-- I'm saying that if the reality is not set up to say some particular thing about time, then that particular thing about time has no meaningful answer. So if the reality doesn't establish a moment, there is no moment, and if it doesn't establish a duration, there is no duration. Concepts like moments and durations are macro concepts.

[EigenState]

Greetings,


I agree that these kinds of decoherence experiments really do have a lot to tell us about the quantum/macro interface, I'm glad they are being carried out.

Absolutely!

And thanks for the kick in the shins.

Best regards,
EigenState

[Ken G]

You took it gracefully enough to leave no bruise!

[HenrikOlsen]

Guys, the interaction is not instantaneous, insofar as it receives contribution from a sum of possible interactions, the latter of which are instantaneous but occur over a spread in times. This is the Heisenberg uncertainty principle. This is indeed the reason that the interaction cannot be with half a photon, that would only be half the sum that you need to do.
Ken, the "integrate over all possible paths and see what the result is", which was one of Feynman's contributions to quantum (electro dynamics I think it was at the time) is a mathematical trick which is just one way of calculating what's going on. I don't really see where you get the idea from that this is how the universe does its thing.

[Ken G]

I never claimed that this is how the universe does its thing-- indeed, my approach has always been that this is how the physicist does his/her thing. But so is time-- time is how the physicist does his/her thing. And this thread is about time, about when things happen in quantum mechanics. Quantum mechanics, after all, is nothing but a prescription for doing exactly the calculation you just mentioned.

[EigenState]

Greetings,



My understanding is that, even when thermal Doppler shifting is
completely controlled, photon emission due to a transition between
energy levels is a Gaussian distribution, or something very close to
a Gaussian distribution. It will be very narrow, but it will still be a
distribution. If for no other reason, such a distribution must occur
as a result of uncertainty. Is that correct? Can you locate a graph
showing the shape of the curve for an example transition? I don't
know what to search for.

Comparison of normalized Lorentzian and Gaussian line profiles for the same center frequency and linewidth.

Best regards,
EigenState

[HenrikOlsen]

I never claimed that this is how the universe does its thing-- indeed, my approach has always been that this is how the physicist does his/her thing. But so is time-- time is how the physicist does his/her thing. And this thread is about time, about when things happen in quantum mechanics. Quantum mechanics, after all, is nothing but a prescription for doing exactly the calculation you just mentioned.
I'm confused, are you saying the interaction takes non-zero time because it is a sum on possible interactions, or are you saying it isn't a sum of possible interactions?

To me it definitely read as if you argued for the non-instantaneousness of the interaction because of the mathematical trick used to calculate things.

[Ken G]

I'm confused, are you saying the interaction takes non-zero time because it is a sum on possible interactions, or are you saying it isn't a sum of possible interactions?
Neither. I've made no attempt to provide a "because" to the reason that the interaction takes non-zero time, nor have I asserted what "actually happens" in the interaction. Instead, I've asserted that what actually happens is undefined, beyond the outcome of the interaction. I've also said that the interaction is not instantaneous, and we can use quantum mechanics calculations to establish that it is not instantaneous. The "because" is a "because the calculation that works says it isn't", not "because the calculation is what actually happens."

Indeed I've taken pains, here and elsewhere, to point out that a key feature of quantum mechanical calculations is that they never assert what "actually happens", so within our best understanding, which is quantum mechanics, there is simply no physical meaning to what actually happened, so we certainly could not say that the interaction was actually instantaneous. What we know is that when we attempt to give a mathematical description of what happens, suitable for making testable predictions (i.e., science), we need to include a sum over a range of possibilities that occur over a range of times. We also know that if we attempt to ascertain an instant that the "interaction" occured via observations, we will observe instead a range of times, despite identical initial conditions. Interestingly, the range we will observe is precisely the range that appears in the mathematical calculation.

All of this has a very simple label: it's called the Heisenberg uncertainty principle, and is responsible for familiar quantum mechanical phenomena like two-slit diffraction patterns. If the "interaction" between a photon and the slits were "really instantaneous", you would never get two-slit patterns, nor most of the other observed phenomena we talk about in quantum mechanics. It is on that basis that I assert the interaction is known to not be instantaneous, but the mathematical calculation is the way we understand this observed fact.

To me it definitely read as if you argued for the non-instantaneousness of the interaction because of the mathematical trick used to calculate things.I hope the above clarifies what I am saying. Nothing controversial there at all, it's bread-and-butter quantum mechanics-- it is the concept of the wave function in a nutshell. If physical interactions were instantaneous, we'd have no HUP. Instead, we do have a HUP, and we can mathematically analyze non-instantaneous interactions via a sum over all the possible elementary interactions that are themselves instantaneous, they are just not what "actually happens." What actually happens is the outcome, nothing else. Or at least, asserting that anything else happens is not physics, it is philosophy.

[EigenState]

Greetings,

Based upon the experimental paper by Brune, et. al. cited above, the "time factor" appears to depend upon the detection apparatus. I say "appears" because my understanding is far from complete. Thus any gross misrepresentations are mine alone.

They treat the simplest model of a quantum measurement in terms of a two-level atom coupled to a quantum oscillator. The quantum oscillator is the detection apparatus. They employ the terminology "mesoscopic" to describe the combined system of the microscopic atom and the macroscopic detection apparatus. There ideal measurement corresponds to quantum entanglement of the atomic states with the phase of their detector oscillator. The relevant wavefunction is therefore a superposition of atomic and detector states.

The "time factor" of the measurement depends upon the time required for that superposition of atomic and detector states to decohere. The qualitative result is interesting and depends upon the mean number of oscillator quanta n, that constitute the detection apparatus.


Mesoscopic superpositions made of a few quanta [n ≈ 1] are expected to decohere in a finite time interval ..., while macroscopic ones (n ≫ 1) decohere instantaneously and cannot be observed in practice.

My initial impression is that conventional laboratory detection apparatus are much more likely to fit the limit of n ≫ 1.

Best regards,
EigenState

[Ken G]

It's still the HUP. The timeframe in which decoherence occurs, so an interaction with a macro detection apparatus, depends on the energy associated with the coupling between the detector and the quantum state. If E is that energy, then h/E is the time spread over which nothing specific can be said about the state of the detection. When E is significant, that spread might be so short that no electronics can isolate it, in which case it would be functionally "instantaneous." But the OP of this thread is framed on the very short timescale of this interaction-- it asks why the photon has to finish interacting, why can't it be detected at some intermediate moment during that interaction. So we must resolve the timescale of the interaction to answer that, no matter how short. Thus the question is actually easier to address in situations when the interaction time can be resolved, which relates to the "low n" case from your quote. I agree that in actual practice, only specially designed detectors will be able to see this, but that just means it requires specially designed detectors to address the OP experimentally. The mathematics of QM is the way to address it theoretically.

So maybe a better answer to the OP than "science doesn't answer that" is "the interaction must be over before we can even talk about the outcome of the interaction, so the photon has to interact completely before we can be in any position to say if a detection occured, or anything else about the interaction for that matter." Science doesn't answer why that is the answer, it is basically a semantic truth about how we do science and what are the meanings of the terms "interaction" and "detection" in the context of quantum mechanics.

[EigenState]

Greetings,


... But the OP of this thread is framed on the very short timescale of this interaction-- it asks why the photon has to finish interacting, why can't it be detected at some intermediate moment during that interaction. So we must resolve the timescale of the interaction to answer that, no matter how short. Thus the question is actually easier to address in situations when the interaction time can be resolved, which relates to the "low n" case from your quote. I agree that in actual practice, only specially designed detectors will be able to see this, but that just means it requires specially designed detectors to address the OP experimentally. The mathematics of QM is the way to address it theoretically.

Agreed. And we are going to need more than a new PM tube for those experiments. :lol: The Brune et. al. experiments were very far from trivial!


So maybe a better answer to the OP than "science doesn't answer that" is "the interaction must be over before we can even talk about the outcome of the interaction, so the photon has to interact completely before we can be in any position to say if a detection occured, or anything else about the interaction for that matter." Science doesn't answer why that is the answer, it is basically a semantic truth about how we do science and what are the meanings of the terms "interaction" and "detection" in the context of quantum mechanics.

Agreed again.

Best regards,
EigenState