Why does light travel at the speed it does?

[diazona]

No, I wasn't reasoning about logical memory aids. I was trying to establish a logical reason why a certain amount of momentum/energy would result in a certain wavelength with constant velocity. It does seem logical that energy would be expressed as wave frequency if velocity was a given, but that still doesn't answer the question of why the velocity is given at the speed it is given.

OK, well, here's a reason why a certain momentum/energy corresponds to a certain wavelength: perhaps you know that a light wave is made up of oscillating electric and magnetic fields. So when this wave hits (and is absorbed by) a (free-ish) charged particle, the electromagnetic field will cause the particle to move around. If the particle is confined to a straight line (like an antenna) parallel to the electric field, it will oscillate in a sine wave; if it's completely free, it may undergo some more complicated motion.

Anyway, the higher the frequency of the radiation, the faster the EM field oscillates, and thus the faster the particle will move. If the particle moves faster, it has more energy; therefore energy is directly related to frequency. (This argument doesn't tell you that energy is proportional to frequency, just that when one gets bigger, so does the other)

Also, the higher the frequency of the radiation, the less time it takes for one complete cycle of the wave to pass through a given point. Assuming that the wave moves at a constant speed, if it takes less time for one cycle of the wave to pass through a point, the distance covered by one cycle of the wave (i.e. the wavelength) will be shorter. Thus higher frequency correlates to shorter wavelength; frequency is inversely related to wavelength. (This argument doesn't tell you that frequency is inversely proportional to wavelength, but if you know that velocity = distance / time it's pretty straightforward to figure out)

Combining the conclusions from the previous two paragraphs:
high energy = high frequency = short wavelength

This is too arbitrary. It would be nice to have a reason that correlates to the relationship between momentum and some force, e.g. electron momentum and strong nuclear force. I.e. something should explain the relationship between electron-nuclear-gravitation and the speed of light in a vacuum because otherwise there is no logical relationship between force, energy, and space. Material motion and radiation propagation are related in the speed of light, so there must be some logical relationship between matter and energy that explains the relationship.

What, like E=mc²? (actually E² = m²c² + p²c⁴) Although I doubt that that's the relationship you're looking for - it's another equation where the speed of light enters only as a unit conversion factor.

I don't mean to sound patronizing, but it really sounds like you're grasping at straws here. I don't know of anything that could be the relationship you're talking about and I don't even understand why you think there has to be one.

And the number 299792458 that humanity has chosen to represent the speed of light in SI units is arbitrary, no way around it, because our choice of units is arbitrary. Just look at how many different unit systems there are in the world: SI, CGS, imperial, atomic, astronomical, cosmological, probably plenty that I've never heard of...

Planck units have something to do with the minimum amount of energy transferred by a given frequency of radiation, which in turn seems to have something to do with the amount of energy released by a unit of electron motion change, right?

No, no, that's Planck's constant. Planck units are a system of units (like SI units) that are based only on the properties of free space, and thus in some sense are the most "fundamental" or "natural" units to do physics in.

So, this seems to have something to do with the inertia of the electron vis-a-vis the attractive force of a proton, no?

No, I don't see how that comes into it at all.

Interesting. This makes me wish I could read equations better qualitatively. This equation looks like the result of loads of data processing and attempts as fitting the data with predictive equations. Or was there a eureka moment of qualitative logic in there somewhere?

There are two detailed derivations of the equation in the Wikipedia article I linked to. This was not a result of data analysis, nor was it a random inspiration - its origins are well-grounded in electromagnetic theory.

How could it be random? Some factor must govern why and how a molecule "decides" whether to transfer KE to another molecule via contact or radiation, no?

Nope. There really is random chance at work at the most basic level of interparticle interactions. Quantum mechanics specifies that, to put it very simply, in situations where multiple outcomes are allowed by the laws of physics, there is no factor that predetermines which outcome will actually occur. Collisions between particles are of this sort; there are a few restrictions imposed by the laws of conservation of momentum and energy, but within the possibilities allowed by those, it's a random choice.

Note that quantum effects are small, so they're most noticeable on very small scales, generally the size of an atom and smaller (roughly speaking). When you work up to molecules, once you take into account the orientation and relative position of the molecules as they collide, often one possible result becomes overwhelmingly more likely than the others.

aha, thanks. I didn't know how an electron could travel a distance corresponding to the length of a radio wave, but I can see how a free electron in an antenna could.

OK, cool But actually how far the electron travels really determines the amplitude of the wave (more or less), and how fast it travels determines the frequency (and thus wavelength). It's possible to arrange for an electron to travel a long distance really fast (large amplitude, short wavelength) or a short distance really slowly (small amplitude, long wavelength).

I'm not sure, but it seems to me that time is relative to the wave, because the wave has no fixed time interval in and of itself. So the wavelength can vary according to how the time interval is defined. If a second is longer, red-shifted light would contain the same number of waves as its pre-shift predecessor, right?

Well, you don't use the same wave whose frequency you're trying to measure to define the unit of time! If you did that, then the wave would always appear to have the same frequency. But you would find that other physical processes, which normally always take a specific time, would take longer or shorter, and it'd be difficult or impossible to develop a consistent physical theory.

If that second is constant, then the red-shifted light would contain less waves-per-second (lower frequency) than its predecessor, right?

Right.

How would red-shift be distinguishable from the light slowing down within a constant time interval? If all light waves shifted by the same amount, how would the shift be identifiable as a frequency-shift and not a velocity-shift?

Because other time standards (besides light waves) don't change. For example, an atomic clock. You can set up an atomic clock and a ruler next to a redshifted wave and use them it to measure the wave's frequency and wavelength respectively, and you will always find that the wave travels at the speed c = 299792458 m/s. But you may find that its frequency has changed relative to some other location where you did the same measurement.

The only reason, I guess, would be the inability of the waves to decelerate due to lack of inertia.

Well, as DaleSpam pointed out, light actually does have inertia, because it has energy. I've been sort of glossing over that point.

I have read that in QM electron position is probabilistic, but that is imo like saying human height is variable.

That's what everybody thinks at first, but it's really not the same. The true probabilistic nature of QM takes some getting used to.

In other words, I don't think it changes the mechanics of how any given electron interacts with its nucleus. I think it's just impossible to specify the exact parameters, such as the exact mass of a given electron, the exact mass of its corresponding nucleus, etc.

I'm not quite sure what you're getting at with this...

 

[brainstorm]

Combining the conclusions from the previous two paragraphs:
high energy = high frequency = short wavelength

Great explanation, but I knew this.

What, like E=mc²? (actually E² = m²c² + p²c⁴) Although I doubt that that's the relationship you're looking for - it's another equation where the speed of light enters only as a unit conversion factor.

I've actually been figuring this out via another thread that discusses the units for measuring momentum in comparison to energy.

I don't mean to sound patronizing, but it really sounds like you're grasping at straws here. I don't know of anything that could be the relationship you're talking about and I don't even understand why you think there has to be one.

Couldn't you look at an atom/molecule as a tiny radio transmitter? If so, wouldn't the diameter of the electron orbit determine the amplitude of waves emitted? If the amplitude was fixed, wouldn't the wavelength also be fixed according to the rate at which the electron oscillated around the nucleus, which would be determined by the force-distance ratio between the nucleus and electron? If amplitude and frequency were determined in this way, wouldn't the velocity of the waves be the result of how much energy was expressed in the wave? I.e. if the wave traveled any faster or slower with the same frequency and amplitude, wouldn't it transmit a different amount of energy than it was originally endowed with?

And the number 299792458 that humanity has chosen to represent the speed of light in SI units is arbitrary, no way around it, because our choice of units is arbitrary. Just look at how many different unit systems there are in the world: SI, CGS, imperial, atomic, astronomical, cosmological, probably plenty that I've never heard of...

Fine, units are arbitrary. But whatever physical mechanics that governs the speed of light relative to, say, gravitation wouldn't be, would it?

Nope. There really is random chance at work at the most basic level of interparticle interactions. Quantum mechanics specifies that, to put it very simply, in situations where multiple outcomes are allowed by the laws of physics, there is no factor that predetermines which outcome will actually occur. Collisions between particles are of this sort; there are a few restrictions imposed by the laws of conservation of momentum and energy, but within the possibilities allowed by those, it's a random choice.

Well, it seems to me that energy transfers due to collisions are not that distinct from those transferred through radiation. Maybe the big difference is that the energy transferred during a collision involves a disturbance in the relationship/distance between the electrons and the nucleus. Actually, that doesn't make sense because the electrons change distance from the nucleus when absorbing or emitting radiation, too right? So what IS the difference between two particles bouncing off each other or the electrons bouncing an EM wave to another particle that receives it as momentum?

Because other time standards (besides light waves) don't change. For example, an atomic clock. You can set up an atomic clock and a ruler next to a redshifted wave and use them it to measure the wave's frequency and wavelength respectively, and you will always find that the wave travels at the speed c = 299792458 m/s. But you may find that its frequency has changed relative to some other location where you did the same measurement.

But why couldn't you just say that the atomic clock is measuring one time while the redshifted wave is actually existing at another time-rate, which it was emitted at?

Well, as DaleSpam pointed out, light actually does have inertia, because it has energy. I've been sort of glossing over that point.

Wouldn't that mean it would decelerate due to friction?

That's what everybody thinks at first, but it's really not the same. The true probabilistic nature of QM takes some getting used to.

I'm not quite sure what you're getting at with this...

I don't know what "true probabilistic nature" means. To me there are two ways of treating phenomena involving multiplicities. One is to recognize the individual elements in the multiplicity as being engaged in unique interactions, which nonetheless can be probabilistically predicted in terms of patterns. The other is to treat multiplicities as themselves collective entities, which I don't like to do because I find it confounding with regards to the behavior of elements at the individual level.

 

[diazona]

Great explanation, but I knew this.

OK, sorry. I kind of have to guess at what you know and don't know

Couldn't you look at an atom/molecule as a tiny radio transmitter? If so, wouldn't the diameter of the electron orbit determine the amplitude of waves emitted? If the amplitude was fixed, wouldn't the wavelength also be fixed according to the rate at which the electron oscillated around the nucleus, which would be determined by the force-distance ratio between the nucleus and electron?

For a while (up until the early 1900s) this is exactly what people thought, and it is what you'd expect based on classical mechanics. But it doesn't correspond to reality - experiments and theoretical considerations show that atoms don't give off radiation in this way, even though it seems like they should. This discrepancy was one of the main inspirations for quantum mechanics. See Wikipedia's article on the Bohr model, for example.

If amplitude and frequency were determined in this way, wouldn't the velocity of the waves be the result of how much energy was expressed in the wave? I.e. if the wave traveled any faster or slower with the same frequency and amplitude, wouldn't it transmit a different amount of energy than it was originally endowed with?

No, it wouldn't. The amount of energy transferred is determined by the frequency and amplitude. Even if the wave got faster or slower, the amount of energy wouldn't change unless the frequency and/or amplitude changed.

Fine, units are arbitrary. But whatever physical mechanics that governs the speed of light relative to, say, gravitation wouldn't be, would it?

No, you're right. That's the part that physics doesn't have a good answer for. As I might have said already, a lot of people hope that a "theory of everything" would shed some light on this, but current theories don't provide any explanation. Plenty of people are trying to figure it out, of course.

Well, it seems to me that energy transfers due to collisions are not that distinct from those transferred through radiation. Maybe the big difference is that the energy transferred during a collision involves a disturbance in the relationship/distance between the electrons and the nucleus. Actually, that doesn't make sense because the electrons change distance from the nucleus when absorbing or emitting radiation, too right? So what IS the difference between two particles bouncing off each other or the electrons bouncing an EM wave to another particle that receives it as momentum?

Good insight, the two processes do work pretty much the same way. You actually can elevate an electron in an atom to a higher energy level (further away from the nucleus) by hitting it with another electron, just as you could by hitting it with an EM wave. The difference is that only the EM wave can be absorbed. In the other (former) case, the impinging electron would have to come back out, though possibly with less energy than it had going in.

Incidentally, if you look closely enough, two electrons will never actually collide, they'll just get really close to each other and then "bounce" back due to their electrical repulsion. The "message" of that electrical repulsion is transmitted by an EM wave.

But why couldn't you just say that the atomic clock is measuring one time while the redshifted wave is actually existing at another time-rate, which it was emitted at?

What do you mean by a time-rate?

Wouldn't that mean it would decelerate due to friction?

There is no friction at the level of subatomic particles (or EM waves). Friction actually arises from the electromagnetic interactions of large numbers of atoms, and it's related to the roughness of a physical surface.

I don't know what "true probabilistic nature" means.

Well, here's an attempt (probably not the greatest) to clarify what I meant by that. If you start with a large number of people with different heights, you can do all sorts of statistical stuff like determining the height distribution, and if you pick a random person and measure their height, you'll get a result equivalent to a random number taken from that distribution. But all the randomness comes from how you make your choice of person. Once you've picked out a person, everything is completely deterministic. You measure their height and get some result, and then you can measure it again and again and reliably get the same result. Or you could look at their health records (well, if you had access) and read off their height without actually measuring it.

For the quantum equivalent, consider a large number of hydrogen atoms that are in the ground state. Again, you can do statistical stuff like determining the distribution of the electron's orbital radius, and if you pick a random atom and measure its orbital radius, you'll get a result equivalent to a random number taken from that distribution. But unlike people, the result is not determined by your choice of atom! If you measure the orbital radius, then wait a little while and measure it again, you might get a completely different, random result. And if you measure it yet again, you might get another completely different, random result. Make enough measurements on the same atom and you'll eventually start to see the same distribution you got from the whole ensemble of atoms in the beginning.

To me there are two ways of treating phenomena involving multiplicities. One is to recognize the individual elements in the multiplicity as being engaged in unique interactions, which nonetheless can be probabilistically predicted in terms of patterns. The other is to treat multiplicities as themselves collective entities, which I don't like to do because I find it confounding with regards to the behavior of elements at the individual level.

Indeed, and you'll find both those viewpoints used in physics. (Way #1 is used in classical mechanics, basic quantum mechanics, classical electromagnetism, quantum field theory, etc. Way #2 is used in statistical mechanics, thermal physics, fluid dynamics, solid state physics, etc.)

 

[brainstorm]

No, it wouldn't. The amount of energy transferred is determined by the frequency and amplitude. Even if the wave got faster or slower, the amount of energy wouldn't change unless the frequency and/or amplitude changed.

But frequency literally means how many waves per unit time. So a wave slowing down takes longer to transmit the same number of waves, thus reducing the frequency.

Good insight, the two processes do work pretty much the same way. You actually can elevate an electron in an atom to a higher energy level (further away from the nucleus) by hitting it with another electron, just as you could by hitting it with an EM wave. The difference is that only the EM wave can be absorbed. In the other (former) case, the impinging electron would have to come back out, though possibly with less energy than it had going in.

Thanks. I don't see why you call the EM wave's absorption as a difference, though. If the EM wave is just carrying the energy of a distant electron that was its source, then its "absorption" is no different from the absorption of one electron's momentum by another during a collision, right?

Incidentally, if you look closely enough, two electrons will never actually collide, they'll just get really close to each other and then "bounce" back due to their electrical repulsion. The "message" of that electrical repulsion is transmitted by an EM wave.

I knew electrons repelled each other due to charge, but I'd never thought of this as an EM wave, but I guess it is an electric field in motion, isn't it? But since it remains attached to the electron, it travels slower than C, right?

What do you mean by a time-rate?

Um, in other words you could say that the wave slowed down and this is what caused it to appear redshifted to a clock that measured it in terms of a time faster than itself.

There is no friction at the level of subatomic particles (or EM waves). Friction actually arises from the electromagnetic interactions of large numbers of atoms, and it's related to the roughness of a physical surface.

Maybe friction is the wrong concept. What I was addressing was the earlier point that light doesn't decelerate or accelerate in being emitted or changing direction. I presume objects/particles do this because they are subject to work (force over distance) instead of themselves being work, as radiation seems to be. So radiation can't accelerate itself because it is a means to accelerate. Does this make sense?

If you measure the orbital radius, then wait a little while and measure it again, you might get a completely different, random result. And if you measure it yet again, you might get another completely different, random result. Make enough measurements on the same atom and you'll eventually start to see the same distribution you got from the whole ensemble of atoms in the beginning.

Right, this is where I wonder why physicists deem it necessary to have the electron(s) of particular particle behaving in regular patterned ways. To me it seems logical that electrons are constantly brushing by other electrons in their own particle and others, which causes them to shift motion and direction in all sorts of quirky (not quarky) ways. So, just because electrons don't follow a fixed trajectory doesn't mean their behavior isn't the result of mechanical determination. The electron interactions could just be very complex, making them seem chaotic.

Indeed, and you'll find both those viewpoints used in physics. (Way #1 is used in classical mechanics, basic quantum mechanics, classical electromagnetism, quantum field theory, etc. Way #2 is used in statistical mechanics, thermal physics, fluid dynamics, solid state physics, etc.)

I don't like statistical methods because they tend to substitute abstract/processed data for direct observational data. They also substitute automatic validity testing for reasoning-based analysis. There are numerous reasons I dislike statistics - it's mostly similar to the reason I like manual transmissions, windows, door locks, etc. more than automatic ones.

 

[diazona]

But frequency literally means how many waves per unit time. So a wave slowing down takes longer to transmit the same number of waves, thus reducing the frequency.

Not necessarily, because the wavelength can (and does) change. (Technical note: frequency is cycles per unit time)

Thanks. I don't see why you call the EM wave's absorption as a difference, though. If the EM wave is just carrying the energy of a distant electron that was its source, then its "absorption" is no different from the absorption of one electron's momentum by another during a collision, right?

Well, yeah. But the EM wave completely ceases to exist once it is absorbed by the atom. A colliding electron couldn't cease to exist.

I knew electrons repelled each other due to charge, but I'd never thought of this as an EM wave, but I guess it is an electric field in motion, isn't it? But since it remains attached to the electron, it travels slower than C, right?

No, it doesn't remain attached to the electron, and it does travel at the speed of light.

Um, in other words you could say that the wave slowed down and this is what caused it to appear redshifted to a clock that measured it in terms of a time faster than itself.

You could say that its frequency slowed down (i.e. got smaller), and that's what caused it to appear redshifted, sure. But that's totally different from saying that its speed slowed down.

Maybe friction is the wrong concept. What I was addressing was the earlier point that light doesn't decelerate or accelerate in being emitted or changing direction. I presume objects/particles do this because they are subject to work (force over distance) instead of themselves being work, as radiation seems to be. So radiation can't accelerate itself because it is a means to accelerate. Does this make sense?

Objects/particles accelerate because they have mass. That's pretty much all there is to it. I don't think I'd call radiation a means to accelerate, although maybe you could say that about the energy it carries. (Of course, particles can carry energy too)

Right, this is where I wonder why physicists deem it necessary to have the electron(s) of particular particle behaving in regular patterned ways.

We (they) do? Maybe I'm missing something...

To me it seems logical that electrons are constantly brushing by other electrons in their own particle and others, which causes them to shift motion and direction in all sorts of quirky (not quarky) ways. So, just because electrons don't follow a fixed trajectory doesn't mean their behavior isn't the result of mechanical determination. The electron interactions could just be very complex, making them seem chaotic.

Well, but what I was saying applies equally well if there's just one electron and one proton and nothing else (except the measuring apparatus), so you can't use interactions between particles as an excuse.

For the first 40 or 50 years of quantum mechanics, many people thought the same way you do, that the behavior of electrons was mechanically determined and that the apparent randomness they showed was just due to some sort of complex interaction or other factor that we weren't able to observe. In 1964, John Bell published a paper describing an experiment that could be performed on pairs of particles, together with a particular condition (Bell's inequality) that had to be satisfied by any physical model in which measurements are mechanically determined. The experiment was first performed in 1972, and has been repeated many times since then, and the results violate Bell's inequality, thus proving that "mechanical determination" is insufficient to explain reality.

I don't like statistical methods because they tend to substitute abstract/processed data for direct observational data. They also substitute automatic validity testing for reasoning-based analysis. There are numerous reasons I dislike statistics - it's mostly similar to the reason I like manual transmissions, windows, door locks, etc. more than automatic ones.

If you're not a Linux user, you should be But that's not really what statistics is about. It's really about describing large amounts of data, and extracting useful information which you might never notice by looking at the individual values.

 

[brainstorm]

Well, yeah. But the EM wave completely ceases to exist once it is absorbed by the atom. A colliding electron couldn't cease to exist.

No, but my point is that the electron emitting the radiation doesn't cease to exist once its emission is absorbed by its destination particle. All I'm trying to say is that radiation is a teleported particle collision, for goodness sake!

No, it doesn't remain attached to the electron, and it does travel at the speed of light.

Huh? An electron's negative charge doesn't remain with the electron?

You could say that its frequency slowed down (i.e. got smaller), and that's what caused it to appear redshifted, sure. But that's totally different from saying that its speed slowed down.

How would you differentiate speed decreasing from frequency shift? How would speed-decrease of a beam consisting of waves be differentiated from a frequency decrease in the beam due to the waves stretching out?

Objects/particles accelerate because they have mass. That's pretty much all there is to it. I don't think I'd call radiation a means to accelerate, although maybe you could say that about the energy it carries. (Of course, particles can carry energy too)

Exactly, any expression of energy can be treated as a potential source of energy for another object/particle. Only, radiation isn't subject to deceleration or acceleration due to gravity, is it? I know gravity can change the direction of radiation, but not its momentum, right?

We (they) do? Maybe I'm missing something...

Well, why else would it be necessary to treat electron position probabilistically - if you didn't expect it to behave regularly in the first place?

Well, but what I was saying applies equally well if there's just one electron and one proton and nothing else (except the measuring apparatus), so you can't use interactions between particles as an excuse.

And nothing is interacting with the atom and its electron?

the results violate Bell's inequality, thus proving that "mechanical determination" is insufficient to explain reality.

I'll have to study this. Thanks.

If you're not a Linux user, you should be But that's not really what statistics is about. It's really about describing large amounts of data, and extracting useful information which you might never notice by looking at the individual values.

I acquired a distaste for statistics from social science. I'm aware of the claimed benefits of statistical modeling. I dislike them because they substitute empirical realities with means. They promote forms of thinking that avoid exploring the full range of interactive possibilities between individuals. This could be true of humans or particles, I assume.

 

[diazona]

No, but my point is that the electron emitting the radiation doesn't cease to exist once its emission is absorbed by its destination particle. All I'm trying to say is that radiation is a teleported particle collision, for goodness sake!

Right, I'll agree with that. (except: "teleported" always seems to give people the wrong impression, maybe I'd say "long-distance") I think we've been saying the same thing in different ways. I'm going to stop clarifying

Huh? An electron's negative charge doesn't remain with the electron?

I'm talking about the EM wave, not the charge. The EM wave doesn't remain with the electron, and it does travel at the speed of light. But the electron's charge does remain with the electron, and it doesn't travel at the speed of light.

The charge and the wave are not the same thing. You can think of the charge as the source that generates the wave, the same way a boat generates a wake.

How would you differentiate speed decreasing from frequency shift? How would speed-decrease of a beam consisting of waves be differentiated from a frequency decrease in the beam due to the waves stretching out?

As I've been saying, you'd directly measure the wave's speed and frequency. If the speed decreases, you'd see it in the measurement. If instead the frequency decreases, you'd see that in the measurement. (Technical point: I'd think of it as the frequency decrease causing the waves to stretch out, rather than the other way around.)

Exactly, any expression of energy can be treated as a potential source of energy for another object/particle. Only, radiation isn't subject to deceleration or acceleration due to gravity, is it? I know gravity can change the direction of radiation, but not its momentum, right?

You're right that radiation never changes its speed, even in the presence of gravity. But gravity actually can change the momentum (and energy) of radiation, in both magnitude and direction, because that momentum (energy) is related to the radiation's frequency, not its speed. This is called gravitational redshift.

Here's an example (based on an actual experiment): suppose you have a device that repeatably emits EM waves of a certain frequency, and put it at a certain height above Earth's surface. If you measure that frequency right next to the device, you'll get a certain result. Alternatively, you could measure the momentum transferred through the wave, and you'd get a certain result. Now aim the EM waves down a vertical shaft and measure them at some lower height, for example 10m or 100m lower. You'll get a higher frequency, and a higher momentum.

Well, why else would it be necessary to treat electron position probabilistically - if you didn't expect it to behave regularly in the first place?

Hm, we may have a different definition of "regular". Or rather, I'm not really sure what you mean by "regular". Could you explain?

Anyway, physicists use probability theory to model the electron position because it works. Perhaps one could say that they "deem it necessary to have the electron(s) of particular particle behaving in regular patterned ways" because that's what they see.

And nothing is interacting with the atom and its electron?

Nothing except the measurement device, which could be as simple as an EM wave (a photon). And that only because you can't even measure the position without having something interact with the electron.

I acquired a distaste for statistics from social science. I'm aware of the claimed benefits of statistical modeling. I dislike them because they substitute empirical realities with means. They promote forms of thinking that avoid exploring the full range of interactive possibilities between individuals. This could be true of humans or particles, I assume.

Ah, I see how you could come away that attitude I'm no expert on any social science, but as I understand it, what seems to happen a lot there is that you're working with data that is fundamentally qualitative, e.g. people's opinions. In order to apply math to social problems, you first have to quantify the system, i.e. come up with some way to use numbers to describe whatever you're studying, and inevitably a lot of useful and/or interesting information gets lost when you do that. You might get some interesting conclusion from a statistical analysis, but it's possible that that conclusion misrepresents the actual situation because of the way you chose to describe it numerically.

That's not the case in physics, though. In physics we deal with measurements and formulas, so everything is quantitative right from the start. So when you apply a statistical analysis, you don't lose information by converting things to numbers one way or another. For this reason, statistical results are a lot more reliable in physics than in pretty much any other science.

And anyway, the bottom line is, physicists have developed a system where you make predictions and test them using math (including statistics). It works really really well. So it doesn't seem that there's some empirical reality that's beyond the ability of math to describe, as there is in the social sciences.

 

[brainstorm]

I think you've moved from reasoning about the logic of the facts to simply recapitulating the facts as you know them. The topic of the OP is WHY light travels at the speed it does, not what the established facts of light behavior are.

Hm, we may have a different definition of "regular". Or rather, I'm not really sure what you mean by "regular". Could you explain?

Regular = the way you think about the Earth tracing the same orbital pattern each year with little if any deviation from last year's path. Physicists expected this from electron orbits in the Bohr model because of narrow assumptions about orbital motion, I think.

Anyway, physicists use probability theory to model the electron position because it works. Perhaps one could say that they "deem it necessary to have the electron(s) of particular particle behaving in regular patterned ways" because that's what they see.

Yes, it may also work to predict height by sex but that doesn't explain why a particular man or woman attained the height they did.

Nothing except the measurement device, which could be as simple as an EM wave (a photon). And that only because you can't even measure the position without having something interact with the electron.

Maybe not, but you could theorize it and theorize the potential effects and observe those.

Ah, I see how you could come away that attitude I'm no expert on any social science, but as I understand it, what seems to happen a lot there is that you're working with data that is fundamentally qualitative, e.g. people's opinions. In order to apply math to social problems, you first have to quantify the system, i.e. come up with some way to use numbers to describe whatever you're studying, and inevitably a lot of useful and/or interesting information gets lost when you do that. You might get some interesting conclusion from a statistical analysis, but it's possible that that conclusion misrepresents the actual situation because of the way you chose to describe it numerically.

And not only that, but you have created an individual picture from the average of a population of individuals. Thus when you look at causes and effects of particular aspects of the population, you tend to forget that the population doesn't actually exist as an entity but is purely an analytical construction.

That's not the case in physics, though. In physics we deal with measurements and formulas, so everything is quantitative right from the start. So when you apply a statistical analysis, you don't lose information by converting things to numbers one way or another. For this reason, statistical results are a lot more reliable in physics than in pretty much any other science.

That's exactly what social statisticians would claim about their methods vis-a-vis other methods.

And anyway, the bottom line is, physicists have developed a system where you make predictions and test them using math (including statistics). It works really really well. So it doesn't seem that there's some empirical reality that's beyond the ability of math to describe, as there is in the social sciences.

But statistics allows predictions about synthetic population means to be analyzed according to other means. Without a specific example, I couldn't show you the problem exactly, but if I could if I had all the details of the analysis in front of me.

 

[diazona]

I think you've moved from reasoning about the logic of the facts to simply recapitulating the facts as you know them. The topic of the OP is WHY light travels at the speed it does, not what the established facts of light behavior are.

Hey, I've been just stating the facts as I know them this whole time, in response to your questions and statements. (Except for the little bit I've said about how I don't think the fundamental nature of light speed can be described by physics, but that's speculative and is all I have to say on the topic)

Regular = the way you think about the Earth tracing the same orbital pattern each year with little if any deviation from last year's path. Physicists expected this from electron orbits in the Bohr model because of narrow assumptions about orbital motion, I think.

Aha. Physicists would know this as a periodic orbit. You're probably right about the assumptions - I guess they assumed this sort of behavior in the Bohr model (and the Rutherford model) because up until that point, the only sort of orbit anyone knew about was planetary orbits.

Maybe not, but you could theorize it and theorize the potential effects and observe those.

Well, theory predicts that you wouldn't measure anything without a measuring device Seriously though, in quantum mechanics you really have to take into account the effect of whatever you're using to do the measurement, since it influences the results. Quantum theory doesn't make predictions about what happens in the total absence of interactions.

And not only that, but you have created an individual picture from the average of a population of individuals. Thus when you look at causes and effects of particular aspects of the population, you tend to forget that the population doesn't actually exist as an entity but is purely an analytical construction.

Ah, I see. That's generally not a problem we have in physics. For instance, when physicists talk among each other about the radius of a hydrogen atom, they all know what they really mean is \sqrt{\langle\psi_{100}\rvert r^2 \lvert\psi_{100}\rangle} (the root mean square expectation value in the ground state), and that it's merely one statistic that characterizes a distribution, rather than The Radius.

However, we do often have that problem when trying to explain these things to people who haven't had it beaten into their heads by years of study in other words, non-physicists. When you say "radius of a hydrogen atom," someone who hasn't studied quantum mechanics might understandably think of it as a little solid sphere with that radius, and may not understand that there's a statistical distribution behind that. Or here's another one I've seen a couple times: when you talk about the half-life of some radioactive material, a layperson might think that all of the radioactive material will have disappeared after twice that time, and wouldn't understand that it's an exponential decay which never completely reaches zero.

That's exactly what social statisticians would claim about their methods vis-a-vis other methods.

Hmmm... well, that's a debate for another day.

But statistics allows predictions about synthetic population means to be analyzed according to other means. Without a specific example, I couldn't show you the problem exactly, but if I could if I had all the details of the analysis in front of me.

Well if you find a good example and are inclined to share, I'd be interested. (You meant an example from social science, not from physics, right?) Although perhaps it's best not to expand the scope of this discussion further than necessary

 

[JDługosz]

...and no physical change would change the number. You could redefine the second to be an hour and the number we get for c in m/s would be exactly the same. No matter how you change the fine structure constant (if you could change it) it would not change the number used to describe c, using the new definition of the meter.

You would get inconsistent results for the resulting length, depending on how you measured the Second. I don't mean the number, which goes into the definition, but the resulting length that it spits out that should be a standard Meter. If you changed "something" in physics, perhaps a Second based on the fine structure constant (e.g. atomic clock) would come out the same, but a Second based on gravity or the strong force or the weak force (half-life of a neutron) would give a different interval of time and thus a different length.

We agree that if you changed everything, you've changed nothing. That means keeping all the unitless constants the same, not just one of them. They ultimately relate different aspects of physics, and every such relationship must be represented.

 

[JDługosz]

Yes, but I think it is necessary to be a little more specific. This is the speed of light as measured by pendulum clocks and rods. The gravitational coupling constant enters in because I deliberately chose a pendulum clock for the measurement of time. If you used an ectromagnetic means to measure time then you would only have the fine structure constant.

I thought so. So if you measured the second using the Weak force, you would need to include that coupling constant as well, right? And to do so, you could observer something involving the Neutral Current (Z) reaction, or look at the statistical curve of neutron decay.

 

[JDługosz]

Interesting exchange we're having. It's particularly appealing to me that I haven't been called an idiot yet, either implicitly or explicitly.

That's probably because you write in complete sentences with proper grammar and spelling, separate paragraphs, etc.; and each post shows interaction with the respondents (e.g. you took in and processed what was said).

I note this so people in "the other boat" might look at this thread as an example of how to do better.

 

[JDługosz]

So this explains why its speed is limited and not infinite. But what determines the limit, then, as the OP asks?

As you can see worked out in http://arxiv.org/abs/physics/0302045" and other similar papers, you can start with the concepts of reciprocity and symmetry and work out the general form that the velocity addition must have in order to satisfy those constraints. What falls out is basically SR. The general form has a variable you can choose: negative doesn't work. Zero is no motion at all. Infinite is Newton's laws. And some positive number is SR. And, as the more philosophical posts here have been exploring, it doesn't matter what that number is as long as it's some positive and finite value. It just "is".

The ultimate speed "just is". What matters to you is how it compares to other things. And that says more about those other things than about light.

 

[jackmell]

Why not it's that fast simply because of the way the Universe was created? And if the Universe was created with different parameters, then the speed of light and other "constants" of nature would likewise be different. To me, that makes sense: nothing is constant. It all depends on which constructor was used for instantiation (what the parameters were when the Universe was created).

 

[mugaliens]

One of the things I learned a few years ago was how inextricably intertwined the speed of light happens to be with the Gravitational constant G, the Reduced Planck constat h-bar, the Coulomb constant k_e, and Boltzmann's constant k_B.

The relationships between these fundamental physical constants can be related by means of the Base http://en.wikipedia.org/wiki/Planck_units" of length, mass, time, electric charge, and temperature. Furthermore, many other Derived Planck units eminate from the Base Planck units, all of which happen to coincide quite nicely with known and observed limits, both large and small, of our universe.

This is key, as by means of multiplication and division, many of the key equations found throughout all works of physics can be normalized.

In summary, all of physics hinges on the Base http://en.wikipedia.org/wiki/Planck_units" themselves.

Working through that entry help me tremendously towards understanding the bigger picture of how all of this fits together, as well as why c is c. It sure put a grin on my face! :)

 

[Dale]

So if you measured the second using the Weak force, you would need to include that coupling constant as well, right?

Yes, exactly.

 

[Dale]

One of the things I learned a few years ago was how inextricably intertwined the speed of light happens to be with the Gravitational constant G, the Reduced Planck constat h-bar, the Coulomb constant k_e, and Boltzmann's constant k_B.

Sure, that is because it is the dimensionless combinations of these constants that are actually physically meaningful.

 

[brainstorm]

Well if you find a good example and are inclined to share, I'd be interested. (You meant an example from social science, not from physics, right?) Although perhaps it's best not to expand the scope of this discussion further than necessary

Well, this should be another thread but there's no point because people will just post a bunch of reasons that statistics has good predictive ability. I notice examples of what I am saying in various scientific expressions all the time, but it's easier when they're directly on hand. A classical example would be Durkheim's study of suicide where he said that a group's suicide rate would be the propensity of any given individual to commit suicide. So, e.g. if he found 1% of Catholics committed suicide (made up number by the way), then he believed any given Catholic would have a 1% chance of becoming suicidal. It extrapolates generalizations from a model to individuals without considering the actual mechanical parameters influencing any single individual.

That's probably because you write in complete sentences with proper grammar and spelling, separate paragraphs, etc.; and each post shows interaction with the respondents (e.g. you took in and processed what was said).

I note this so people in "the other boat" might look at this thread as an example of how to do better.

I always do this, but maybe more explicitly in this thread. Glad to be used as a positive example for a change, though:)

One of the things I learned a few years ago was how inextricably intertwined the speed of light happens to be with the Gravitational constant G, the Reduced Planck constat h-bar, the Coulomb constant k_e, and Boltzmann's constant k_B.

The relationships between these fundamental physical constants can be related by means of the Base http://en.wikipedia.org/wiki/Planck_units" of length, mass, time, electric charge, and temperature. Furthermore, many other Derived Planck units eminate from the Base Planck units, all of which happen to coincide quite nicely with known and observed limits, both large and small, of our universe.

This is key, as by means of multiplication and division, many of the key equations found throughout all works of physics can be normalized.

In summary, all of physics hinges on the Base http://en.wikipedia.org/wiki/Planck_units" themselves.

Working through that entry help me tremendously towards understanding the bigger picture of how all of this fits together, as well as why c is c. It sure put a grin on my face! :)

I would guess Planck units have something to do with why c is c. It's too bad you can't remember or say what you came up with. It would do a lot for the OP topic, I think.

Still, a mathematical relationship would not really satisfy me. I would like to see some qualitative reason why the speed of light relates to its momentum, for example, which I think would also have to entail some explanation for why a photon or EM wave carries the amount of energy it does. E.g. if a photon must be generated with a particular amount of energy because of its conditions of emission, then maybe it needs to travel at a particular speed to result in the correct amount of momentum in its moment of absorption when it reaches its destination.

 

[camel_jockey]

However, IMHO, the discussion you're interested in having is better had in philosophy than physics.

I have two points to make on this.

1) Firstly I agree that this is somewhat philosophical, and I love that! So interesting! Are there any other Lorentz scalars in any of our theories which describe motion with respect to time?

Unless there are such scalars, the speed of light is the only "speed-like" scalar to our knowledge and therefore must be the only fundamental, special, priviledged speed. We are then left with a situation where all other less fundamental speeds must be talked of in terms of the only special speed, the speed of light!

2) Therefore, c can be assigned any value. Preferably 1. Why it in this set of units carries no dimension, is yet to be understood by me! Is it just a way of saying "speed shows up more commonly than distance or time in our equations, so let's make speed the most fundamental of the three quantities and express the other two in terms of speed" ?

 

[brainstorm]

How would it change the question to ask what the speed of light is really dependent on and how? Is it, for example, always constant in a vacuum regardless of the level of gravitation? Does it even make sense to talk about light as having travel time? For example, when you hear that light takes 8 minutes to get from the sun to Earth, how can you say that there is something happening simultaneously on the sun and Earth that nevertheless takes 8 minutes to get from one to the other?

 

[Dale]

2) Therefore, c can be assigned any value. Preferably 1. Why it in this set of units carries no dimension, is yet to be understood by me!

I think you are referring to Geometrized Units where time is given dimensions of length meaning that speeds are dimensionless numbers (slopes):
http://en.wikipedia.org/wiki/Geometrized_unit_system

This is decidedly different from Planck Units where c is also numerically set to 1 but is still considered to have dimensions of L/T:
http://en.wikipedia.org/wiki/Planck_units

Afaik, the Planck units are much more widely accepted than geometrized units.

 

[brainstorm]

Could anyone answer the following question? It seems relevant to this thread to me for some reason:

If you took the amount of energy emitted as radiation by a particular electron's moment of momentum, and then applied that amount of energy to a free electron traveling through a vacuum, what velocity would the electron reach?

 

[pallidin]

Could anyone answer the following question? It seems relevant to this thread to me for some reason:

If you took the amount of energy emitted as radiation by a particular electron's moment of momentum, and then applied that amount of energy to a free electron traveling through a vacuum, what velocity would the electron reach?

I could be wrong, but I don't think a free electron "absorbs" photons.
Manipulated momentum through either an electric or magnetic field? Yes.

 

[diazona]

Regarding Durkheim's suicide study, that would make an interesting discussion. But this is not the place

Is it, for example, always constant in a vacuum regardless of the level of gravitation? Does it even make sense to talk about light as having travel time? For example, when you hear that light takes 8 minutes to get from the sun to Earth, how can you say that there is something happening simultaneously on the sun and Earth that nevertheless takes 8 minutes to get from one to the other?

Yes; yes, because light has a finite speed; what is it that is supposed to be simultaneous on the sun and the Earth? And who (what observer) is claiming that they are simultaneous?

If you took the amount of energy emitted as radiation by a particular electron's moment of momentum, and then applied that amount of energy to a free electron traveling through a vacuum, what velocity would the electron reach?

What do you mean "moment of momentum"?

I think it might be possible for a free electron to absorb a photon. You'd have to do a bit of math to see whether it's allowed by energy/momentum conservation. (I don't remember offhand for sure)

 

[brainstorm]

I'm not specifying that the free electron has to be propelled by a photon. What I'm trying to do is compare a given amount of energy in terms of how fast it causes a free electron to go. So let me be specific about the two quantities I want to compare:

1) The amount of momentum needed for a given orbital electron to generate a certain amount of light
2) The velocity attained by a free electron in a vacuum given the same amount of momentum from whatever source.

I guess what would be really interesting would be to plot one in terms of the other and look for a pattern, but I'm putting too much on my xmas list now. I'd like to just understand the relationship between quantities of radiation energy and electron velocity through a vacuum given the same amount of energy. Does this question make sense now?

 

[pallidin]

I think it might be possible for a free electron to absorb a photon. You'd have to do a bit of math to see whether it's allowed by energy/momentum conservation. (I don't remember offhand for sure)

Just as a suggestion: I think that question/answer needs to be firmly established before we can go responsibly further.
My own understating is that only "bound" electrons can "absorb" a photon, and that a free electron cannot. Again, I could be wrong.

 

[diazona]

I'm not specifying that the free electron has to be propelled by a photon. What I'm trying to do is compare a given amount of energy in terms of how fast it causes a free electron to go. So let me be specific about the two quantities I want to compare:

1) The amount of momentum needed for a given orbital electron to generate a certain amount of light
2) The velocity attained by a free electron in a vacuum given the same amount of momentum from whatever source.

I guess what would be really interesting would be to plot one in terms of the other and look for a pattern, but I'm putting too much on my xmas list now. I'd like to just understand the relationship between quantities of radiation energy and electron velocity through a vacuum given the same amount of energy. Does this question make sense now?

It makes a bit more sense, but it still needs to be clarified. Namely: by what physical process is the light produced? What do you mean by "amount" of light? What momentum specifically are you asking for? i.e. the momentum of what object? Before or after the production of the light? Or do you want a difference between two momenta?

The speed of an electron given a certain amount of momentum, assuming the electron started from rest, is given by solving the equation
p = \frac{mv}{\sqrt{1-\frac{v^2}{c^2}}}
for v. I get
v = \frac{p}{\sqrt{m^2 + \frac{p^2}{c^2}}}

Just as a suggestion: I think that question/answer needs to be firmly established before we can go responsibly further.
My own understating is that only "bound" electrons can "absorb" a photon, and that a free electron cannot. Again, I could be wrong.

Actually I think you're right, since now I remember that the basic QED vertex (electron + photon → electron) is not a permitted physical process, due to momentum nonconservation. I guess I should do the math: let p_iμ be the initial 4-momentum of the electron, p_fμ be its final 4-momentum, and q_μ be the 4-momentum of the photon. Conservation states that
p_{i\mu} + q_{\mu} = p_{f\mu}
Squaring that:
p_i^2 + 2p_i\cdot q + q^2 = p_f^2
and since 4-momentum squared = ±mass squared (depending on metric),
±m_e^2 + 2p_i\cdot q + 0 = ±m_e^2
So this is only possible if p_i\cdot q = 0.
p_{i\mu} = (\gamma_i m c, \gamma_i m v \cos\theta_i, \gamma_i m v \sin\theta_i, 0)
q_{\mu} = (E/c, E/c, 0, 0)
so
p_i\cdot q = E\gamma_i m - E\gamma_i m \tfrac{v}{c}\cos\theta_i = E\gamma_i m\Bigl(1 - \tfrac{v}{c}\cos\theta_i\Bigr)
Since v/c < 1 and |cosθ_i| ≤ 1, this will never be equal to zero. Therefore a free electron can't absorb a photon without side effects.

If I messed something up there, please do point it out.

 

[brainstorm]

It makes a bit more sense, but it still needs to be clarified. Namely: by what physical process is the light produced? What do you mean by "amount" of light? What momentum specifically are you asking for? i.e. the momentum of what object? Before or after the production of the light? Or do you want a difference between two momenta?

Your equation conversions seem complicated. I'm sorry I can't appreciate them more with my skill level. The momentum I'm thinking of, I guess, is maybe the amount that goes into bumping the electron to get it to produce the photon. So, let's say you can come up with the acceleration of an electron that results in a certain amount of radiation-emission. I guess I would like to know how much energy the electron gained and how much it lost to generate the given photon. Then I would want to know how fast an electron would travel if it was given that amount of energy as straight-line momentum in a vacuum.

Honestly, I'm embarrassed here because I'm getting lost in what I want to know. Basically, I want to compare electron-velocity with radiation-wavelength for a given amount of energy.

Maybe a better question to ask would be how long is the radiation beam emitted with one Planck unit of energy (is this the unit that is supposed to be the absolute smallest amount of energy that can be carried by an EM wave at a given frequency?)

 

[samuelwang]

I don't know a lot about the relativity theories yet, but here is my interpretation about this phenomena. I remember someone told me that gravity has effect on the time, and somehow I think the speed of light is relate to time, so may be this constant is a consequence as an equilibrium of all the gravity in this universe, may be?

 

[diazona]

Your equation conversions seem complicated. I'm sorry I can't appreciate them more with my skill level. The momentum I'm thinking of, I guess, is maybe the amount that goes into bumping the electron to get it to produce the photon. So, let's say you can come up with the acceleration of an electron that results in a certain amount of radiation-emission. I guess I would like to know how much energy the electron gained and how much it lost to generate the given photon. Then I would want to know how fast an electron would travel if it was given that amount of energy as straight-line momentum in a vacuum.

Sorry, still not clear. I think it'd be really important to specify what physical process is producing the photon, e.g. an atomic energy level transition, brehmsstrahlung ("braking radiation," produced when an electron is decelerated by electrical interactions), synchrotron radiation, etc. Once you figure that out, you can talk about what momentum/energy/whatever you're looking for, and what additional information you might need to figure it out.

Honestly, I'm embarrassed here because I'm getting lost in what I want to know. Basically, I want to compare electron-velocity with radiation-wavelength for a given amount of energy.

That's a little clearer. To be precise, you can calculate these two quantities:

• The speed of an electron which has an amount of kinetic energy E is
  v = c\sqrt{1 - \biggl(\frac{mc^2}{E + mc^2}\biggr)}
• The wavelength of a photon (quantum of EM radiation) with energy E is
  \lambda = \frac{hc}{E}

In both equations c is the speed of light and h is Planck's constant. m is the mass of the electron.

Maybe a better question to ask would be how long is the radiation beam emitted with one Planck unit of energy (is this the unit that is supposed to be the absolute smallest amount of energy that can be carried by an EM wave at a given frequency?)

"Planck unit of energy" is something different, but I get your meaning. Assuming that by "how long" you mean the wavelength, that is. If you have a radiation "beam" at a given frequency f, and that beam contains the minimum possible energy it can have with that frequency, its energy will be E = hf and its wavelength will be given by the formula I wrote above, λ = hc/E.