User talk:Sbyrnes321/Archives/2012

Classical vacuum as a "thing"?

Hi Steve: You've said:

"There is no such thing as a 'classical vacuum'. There is only 'a perfect vacuum as understood in classical physics'. This is because the word 'thing' is conventionally understood to mean 'something that could, at least hypothetically, exist in the universe'."

I find this statement hard to follow. It seems to say that whatever meaning might attach to 'classical vacuum', it is not a "thing", presumably because it does not exist, and so isn't a "thing". I hesitate to take this too seriously, and maybe it isn't exactly what you wanted to say?.

This statement, at face value, appears to try to limit the kinds of abstractions one can use in a theory.

I wonder if you ever took a look at The Grand Design (book)? It deals with model-dependent realism. It's a complicated subject, and this book is probably not the best place to find a clear enunciation. However, one of the tenets is that theoretical constructs like quarks, or maybe the Higgs boson, are real even if they are not observable, provided the whole fabric of the theory leads to verified experimental consequences:

QCD [Quantum chromodynamics] also has a property called asymptotic freedom, which we referred to, without naming it, in Chapter 3. Asymptotic freedom means that the strong forces between quarks are small when the quarks are close together but increase if they are farther apart, rather as though they were joined by rubber bands. Asymptotic freedom explains why we don’t see isolated quarks in nature and have been unable to produce them in the laboratory. Still, even though we cannot observe individual quarks, we accept the model because it works so well at explaining the behavior of protons, neutrons, and other particles of matter [Emphasis added]. Hawking/Mlodinow, p. 110

So, I suppose, if a medium with properties defined by c, c₀, μ₀ and the subsidiary derived properties with exact values ε₀ and Z₀ proves useful, even though it is not realizable even in principle, it could be considered a "thing", from some metaphysical stance, eh? If you want to call this thing "classical vacuum", well, what's in a name? It may subsume 'a perfect vacuum as understood in classical physics', but its defining property is not 'absence of matter', but its electromagnetic properties.

This is all a bit of a digression. Just wondered what you really meant here? Brews ohare (talk) 17:31, 20 December 2011 (UTC)

I mean, there is no circumstance in which it is helpful to refer to "classical vacuum" as a kind of thing that you can imagine existing in the world. It cannot exist in the world, because vacuum fluctuations are everywhere in the universe and always will be and always must be.

Contrast with, say, "a box with zero molecules of gas in it". This doesn't exist, but it is entirely possible to imagine it existing in the world. Indeed, someday it may exist. A "box at exactly absolute zero" can never be created (in a finite amount of time), but if it did exist, the microscopic laws of physics are quite capable of saying in great detail how it would behave. Maybe there's "no such thing as a box at exactly absolute zero", but it's still very useful to refer to it and to imagine it existing.

My suggestion: Whenever you want to refer to "classical vacuum", please just refer instead to "vacuum, and by the way I am approximating things with classical physics right now". You can still discuss all the same concepts. But you will not be asking readers to imagine sucking the virtual photons out of a box. Likewise, the readers will not start wondering whether an electron entering this "classical vacuum" would acquire an infinite charge because it's not shielded by virtual photons, or lose its mass because there's no Higgs field. And they will not wonder whether an atom entering the "classical vacuum" box would convert into a planetary Bohr model, etc. etc. --Steve (talk) 21:57, 20 December 2011 (UTC)

Thanks for the answer. It is clear that this is a discussion about semantics. It seems you understand the term "classical vacuum" in one way, and I in another. For you, the words 'classical vacuum' invokes 'something that you can imagine existing', while both of us agree that an "electromagnetic medium defined by c, c₀, μ₀ and the subsidiary derived properties with exact values ε₀ and Z₀" is not realizable because it doesn't allow vacuum fluctuations.

I don't see why a connotation of realizability is a necessary attribute of the words 'classical vacuum', and some published sources don't find that to be a necessity either.

Could we not simply define 'classical vacuum' as a technical term referring to an unrealizable electromagnetic medium defined by c, c₀, μ₀ and the subsidiary derived properties with exact values ε₀ and Z₀? What is so special about these words as to put them off limits?

Would you prefer the old-fashioned terminology "free space" used in the terms characteristic impedance of free space, permittivity of free space, permeability of free space? Brews ohare (talk) 23:01, 20 December 2011 (UTC)

The terminology I really like is

"vacuum, and by the way I'm using the classical electromagnetism approximation right now".

The difference is that when you describe it that way, you're not implying that we need to thoroughly understanding the properties of a box which does not contain quantum fluctuations. That's good because classical physics is not a self-consistent theory! Forget about agreeing with experiments, it's not even logically possible to postulate a box without quantum fluctuations. For example, if I put a hot object in a "classical vacuum", does it radiate an infinite amount of power? (see Ultraviolet catastrophe).

All the books you found that use the words "classical vacuum", what they mean is "vacuum, and by the way I'm using the classical electromagnetism approximation right now". But you mean something different, something more "real"...in the sense that you imply that it can be described in any level of detail without self-contradiction. So I guess that's a better answer to your original question. "There's no such thing as a classical vacuum" because as soon as you look at it in detail you find self-contradictions. There is no self-contradiction when describing at any level of detail, say, a system at absolute zero, or a hypothetical quantum vacuum without even a single air molecule in it. --Steve (talk) 18:11, 23 December 2011 (UTC)

Hi Steve: Thanks for getting back to this. I do not think of 'classical vacuum' as real at all. I think of it as a hypothetical medium with defined electromagnetic properties c, c₀, μ₀ and the subsidiary derived properties with the exact values ε₀ and Z₀. I fully agree that this medium cannot be realized, even in principle because of virtual particles and vacuum fluctuations. I also think of this hypothetical medium as a baseline reference state to which the permittivity and permeability of all media are compared, including those of any real vacuum, be that partial vacuum, or some theoretical vacuum like QCD vacuum. So maybe we agree after all? Brews ohare (talk) 17:37, 24 December 2011 (UTC)

You don't see any problem with defining things via a "reference state" which is internally inconsistent and self-contradictory? I see that as a big, deep problem. Again, if it is a hypothetical medium, please tell me its hypothetical properties. What radiation field is emitted by a hot object in it? What is the charge of an electron in it?

Do you agree with the statement: "QED is not a comprehensive theory of electromagnetism because it is impossible to understand material permittivity just within QED. You also need classical electromagnetism to fully understand electromagnetism in our universe." (I would disagree but it seems you would agree.)

Another note: I have been trying to make the point "It is conceivable for c0 in meters per second to be a fixed exact number, yet at the same time c0 is not exactly constant. Just like, the mass of the IPK in kilograms is a fixed exact number, yet at the same time the mass of the IPK is not exactly constant." This is a basic, important point. Do you understand and agree with it? I'm just saying, it's conceivable, we can discuss later whether or not it's actually the case. --Steve (talk) 17:55, 26 December 2011 (UTC)

←outdent Hi Steve: I have a lot of trouble understanding your remarks. Let me try to explain what I understand, and please try to address things from that stance so I can see what you are talking about.

As I see it, the reference state of 'vacuum' is entirely specified by its defined electromagnetic properties c, c₀, μ₀ and the subsidiary derived properties with exact values ε₀ and Z₀. It is not a real medium. It cannot be realized in a laboratory or anywhere else. Any medium you can imagine that has these electromagnetic properties along with any other properties you like, is equivalent electromagnetically.

Within classical electromagnetism, one adopts Maxwell's equations and sets the source terms to zero. Then D=ε₀E and B=μ₀H and you have a theoretical version of this medium.

Within field theory, one has to quantize the fields and one then finds different constitutive relations with D=εE and B=μH, despite having no explicit charge and current densities, and now one no longer has this medium: ε ≠ ε₀ and μ ≠ μ₀.

In the lab we know the latter situation is closer to reality than the first, so we know the first is inadequate and no laboratory version of the medium can in fact be formed.

I expect that the inconsistencies etc that you refer to find their origin not in the reference state but in the theory underlying what happens when matter is introduced, or when the fields are quantized. Of course, the reference state precludes any phenomena that would inevitably interfere with its defined properties, so to deal with such departures one has to introduce a different medium, and discuss that medium. Whether a theory of that medium is inconsistent or whatever is about that theory, not about the reference state with defined electromagnetic properties. Brews ohare (talk) 01:03, 27 December 2011 (UTC)

Concerning the speed of light, which I suppose is the physical speed limiting (e.g.) the propagation of information, why does this subject come up here? Brews ohare (talk) 01:03, 27 December 2011 (UTC)

OK, here goes.

• Nothing in the BIPM/NIST literature implies that they think there is a hypothetical medium with exactly fixed permittivity, speed of light, etc. "But," you object, "Didn't they say 'vacuum permeability is exactly 4pi * 1E-7 N/A2'"? Yes they did! But as I explained above,

Quantity X is an exact number in terms of Unit Y does not logically imply that Quantity X is exactly fixed.

This is just simple logic. If you cannot wrap your mind around the logic involved, think again about what would happen if the IPK gained or lost a few atoms. So we cannot take the statement "vacuum permeability is exactly 4pi * 1E-7 N/A2" to be evidence that BIPM thinks there is a hypothetical medium called "vacuum" (or any other term) with exactly fixed permeability. (Likewise permittivity, speed of light, etc.) Indeed, there is no evidence whatsoever that BIPM thinks this.

• Some aspects of the BIPM/NIST literature imply that they think "vacuum" contains quantum fluctuations. For example, the definition of the ampere discusses the force between two parallel wires in a "vacuum". But a classical vacuum cannot even hypothetically have matter in it, because it leads to logical contradictions. More fundamentally, every experimental measurement of anything ever has used an apparatus full of quantum fluctuations, and it is not a priori obvious that these have no effect on (say) the propegation of low-intensity light; yet BIPM has never discussed how or whether to do corrections for quantum fluctuations altering low-intensity light propegation.
• It is quite possible and practical to define the meter without invoking a classical vacuum: Sure, maybe the speed of light in a quantum vacuum varies with intensity, but such effects are too small to measure. So the definition would be:

One meter is the length that light travels in 1/299792458 s through a real-world best-practices vacuum.

• It is quite possible and practical to discuss the issue of quantum vacuum polarization without invoking a classical vacuum. If someone learned QED without ever learning classical electromagnetism, surely you agree that they will nevertheless have all the tools to fully understand vacuum polarization. After all, QED is a more accurate and more complete theory than classical electromagnetism, so it does not need to "lean on" classical electromagnetism for explaining any phenomenon. So, here's my definition of ε0:

ε0 is the ratio of D to E in the case when P=0.

(Remember, D=ε0*E+P.) P is the polarization density, i.e. the density of dipoles. When an intense electric field is in a vacuum, the electric field pulls the virtual electrons to the left and the virtual positrons to the right (for example) creating electric dipoles out of the vacuum, and then P is the density of those dipole moments. When there is no field, there are no dipole moments, so P=0. When the field is very weak, P is almost 0, so D/E is almost ε0. So we now have a theoretical definition of ε0 and an operational way of realizing it. It's a complete picture without requiring any mention or reference to the classical vacuum. (Of course, ε0 isn't really meaningful anyway; it doesn't even exist in gaussian units.) So again, there is no need to invoke classical electromagnetism if you want to define permittivity or permeability, nor if you want to discuss the electromagnetic properties of the (quantum) vacuum.

• Therefore, there is never a need to invoke the classical vacuum when discussing SI units, electromagnetic phenomena, or anything else in the real world. When you invoke the classical vacuum, you are introducing something entirely unnecessary into the discussion of what is happening in our universe.
• The classical vacuum is a bizarre and foreign thing even if you think about it hypothetically. If you put any matter or charge into a classical vacuum, you get inconsistencies. Therefore, a classical vacuum is a hypothetical medium where charge and matter cannot exist even hypothetically. It's impossible, even hypothetically, to create a pulse of light in a classical vacuum. It's impossible, even hypothetically, to detect a pulse of light in a classical vacuum. Therefore, "measuring the speed of light in a classical vacuum" is something that you could not do even if you were God and had the power to change the laws of physics. I'm saying this to (hopefully) impress on you that the classical vacuum is so bizarre that sensible people don't waste time talking (even hypothetically) about it...let alone defining important things to be based on specific properties of a classical vacuum. Other than you, I don't think anyone spends time thinking seriously about the "classical vacuum", and I think that the vacuum article would do well to not even mention the "classical vacuum" anywhere.
• Invoking the classical vacuum is not even helpful for defining a real-world quantity. Let's say that BIPM, following your advice, said "One meter is the length that light travels in 1/299792458 s through a classical vacuum." OK, now I, an experimentalist, want to measure a certain length in meters. I cannot suck out the quantum fluctuations from a box. What to do? I look through my textbooks and see that the standard model is approximated by Maxwell's equations in a certain limit (large but not too-large fields, photons in a coherent state, etc.). Maybe I can measure in that limit? No, that's not a classical vacuum, that's a quantum vacuum which coincidentally is behaving almost classically. What to do? I throw out my QFT textbooks and imagine that my box, the size of a small table, is a "classical vacuum", where the laws of physics are Maxwell's equations. OK, but how long does it take for light to travel through this box? The box can be a "classical vacuum" where it takes light a few nanoseconds to travel through it, or the box can be a "classical vacuum" where it takes light a few years to travel through it. Both can equally well satisfy Maxwell's equations. So I'm stuck. The BIPM prescription for a meter is useless to me.
• So the best approach... is to say that a "partial vacuum" is what you get in the laboratory, and a "perfect vacuum" is what you get in the ideal with no air molecules, walls at absolute zero, and perfect shielding of all fields (but still quantum fluctuations). If you don't need high accuracy, you have permission to discuss the perfect vacuum using the approximations of classical physics, but let's be clear that you are discussing approximate properties of a quantum vacuum, not exact properties of something different called a "classical vacuum". If you are BIPM/NIST, you have no need or desire to ever mention the "classical vacuum"; you can and should base everything on the "perfect (quantum) vacuum". --Steve (talk) 15:21, 31 December 2011 (UTC)

Steve: You have made quite an exposition here. There are many details that could be discussed, but there is, I believe an over-arching disconnect between our views that might be resolvable in broader terms if we can identify it carefully. My impression of many of your comments is that they involve what we might call "hands-on" aspects: what can be done in a lab, what is happening in the real universe, and so forth. It seems that you see my comments as impinging upon such matters.

I'd suggest that in fact my observations have little bearing upon such things.

BIPM has defined exact values for c₀, μ₀ and also suuplied definitions like ε₀=1 ⁄ (μ₀c₀²) and Z₀ = (μ₀ ⁄ ε₀)^1/2 that lead to exact values for these quantities as well. Whatever the meaning ascribed to these postings, they are there for sure.

What is the role of the BIPM posted values?

It seems that BIPM has supplied these values for use as a reference, for example, as the units (so to speak) in which permittivities and permeabilities are to be defined as, for example, the relative permittivity ε ⁄ ε₀.

It seems clear that theoretical calculations of the electromagnetic properties of media follow this recommendation and express relative permittivities as ε ⁄ ε₀.

It does not appear that we have to ascribe any physical meaning to these BIPM supplied properties; they do not have to be seen as properties of a fictitious medium, or as limiting properties of some sequence of real media. They can be seen simply as conventions. They could be chosen as unity were it not that the speed of light then would be 1 m/s which would cause some practical inconvenience in retooling.

Steve, does it seem to you that this is the nub of the differences in our views? Namely, that you feel the 'vacuum' of BIPM has to be related to something "real", while I'd say the properties of BIPM 'vacuum' are no more significant to reality than, say, the length of the king's arm, the old standard for the yard. Brews ohare (talk) 19:04, 31 December 2011 (UTC)

I think the disagreement is more basic...I probably disagree with your premise "BIPM has defined exact values for c0...", but let's find out. I assume by c0 you mean "the speed of light in vacuum" (where the definition of "vacuum" is to be debated).

When you say "BIPM has defined an exact value for c0" do you mean:

(A) "BIPM has defined an exact value for "c0 in m/s" "

(c0 is a speed, but "c0 in m/s" is a dimensionless numerical value, i.e. a real number. It is exact and fixed if it is always the same exact real number.)

Or do you mean:

(B) "BIPM has defined an exact value for c0"

(c0 is a speed, not a dimensionless number. What does it mean for a speed to be fixed at a certain exact speed? For example, if the speed of snails is fixed at a certain exact speed, then the acceleration of a snail is always exactly zero, and a race between two snails is always exactly a tie.)

(My view: (A) is true because of the definition of the meter. I see no reason to believe that (B) is true.) --Steve (talk) 14:02, 3 January 2012 (UTC)

←outdent I agree that c₀ is an exactly defined number. As such it is unrelated to reality and to measurement in particular. As you know, should it happen that the physical speed of light changed (maybe because it relates to something like the expansion of the universe) c₀ would not change, but (other things held fixed) the metre would change because light would travel a different distance each elapsed second.

So this is not a point of disagreement. The disagreement apparently relates to the role of the other BIPM postings, μ₀, ε₀=1 ⁄ (μ₀c₀²) and Z₀ = (μ₀ ⁄ ε₀)^1/2. These are historically referred to as the permittivity of vacuum, the permeability of vacuum, and the characteristic impedance of vacuum. I take them to refer to the same BIPM 'vacuum' mentioned in the definition of c₀. I also take their nature as defined exact values to indicate they also do not refer to any physically real medium and as not being subject to measurement. Is that your take on these posted exact values? Brews ohare (talk) 16:10, 3 January 2012 (UTC)

You say "I agree that c0 is an exactly defined number". Well, I disagree. I think c0 is a speed, not a number. Do you? I don't believe statement (B) above. Do you? If so, why? --Steve (talk) 16:17, 3 January 2012 (UTC)

Sorry for my sloppy language: I agree with 'A', in detail, that definition of c₀ is an exactly defined speed, with no necessary relation to laboratory measurement. Likewise, the other BIPM postings μ₀, ε₀=1 ⁄ (μ₀c₀²) and Z₀ = (μ₀ ⁄ ε₀)^1/2 are exactly defined dimensional quantities with no necessary connection to measurements. Brews ohare (talk) 16:41, 3 January 2012 (UTC)

Speed of light questions

One more time: c0 is "the speed of light in vacuum" (for some definition of "vacuum" to be debated). You agree with that, right? Therefore, c0 is a speed, not a unitless number. (B) is the statement that this speed is exactly constant. On the other hand, "c0 in meters per second" is a unitless number, not a speed. (A) is the statement that this unitless number is exactly constant. When you say "c0 is an exactly defined speed", it sounds like you're saying (B) not (A). But you seem to think that you're talking about (A). So I guess you believe both (A) and (B)? If you believe (B), please tell me why. (Do you believe that (A) implies (B)?) Do you understand how there is a difference between "c0" (a speed) and "c0 in m/s" (a unitless number)?

Think about the difference between "mass" and "mass in kilograms". The "mass of an electron" is the same today as it was last year. But the "mass of an electron in kilograms" can change randomly from year to year as the IPK gains and loses atoms. If you can't keep straight the difference between "mass" and "mass in kilograms", you would think that an electron is occasionally violates conservation of energy by gaining or losing mass. --Steve (talk) 17:42, 3 January 2012 (UTC)

←outdent Applying the analogy about the mass of an electron, the physical speed of light is a measurable property of the universe, and may vary as the universe evolves (so far, no evidence to within experimental error). It might be a constant, but we may never know that if the changes remain below our measurement error. The speed c₀ is of a size chosen by committee as 299*** m/s, and they can choose any number that appears palatable to them. Until the committee changes their minds it is a fixed speed and it tells us nothing about the physical speed of light, because that speed, whatever it is, sets the metre, which elastically adjusts to make the transit time down its length 1/299*** s. Steve, I repeat this only because you seem to think I might have some other idea about what is going on. I don't. Brews ohare (talk) 19:47, 3 January 2012 (UTC)

Why do you say c0 "is a fixed speed"? Are you using the logic: "c0 is chosen by committee to be 299... m/s. Therefore, c0 is a fixed speed."? If that's your logic, it's bad logic. ("The mass of the IPK is chosen by committee to be 1 kg. Therefore, the mass of the IPK is a fixed mass.") Again, if you believe that c0 is a "fixed speed", please tell me why you think so. --Steve (talk) 20:06, 3 January 2012 (UTC)

I am indeed saying "c₀ is chosen by committee to be 299... m/s. Therefore, c₀ is a fixed speed." Do you see some subtlety here I am missing? What is the logical problem here? If BIPM says the symbol c₀ denotes a speed of 299 792 458 m/s, then this is the definition of c₀ and it is fixed by convention and can be no other speed. The number of m/s is defined, and not measured, and is exact, not a value to within some uncertainty. What are your thoughts, Steve? Brews ohare (talk) 20:16, 3 January 2012 (UTC)

In the case of the electron mass, m_e is not a defined quantity, and neither is the physical speed of light. Although both are constant to within experimental error, we know they are constant only to within some measurement uncertainty. The standard kg is a physical embodiment of the standard and subject to variations according to its environment and environmental hazards. On the other hand c₀ is not a physical embodiment but a particular defined speed 299 792 458 m/s that never changes and is not environmentally sensitive. The metre, on the other hand, has to be realized, and that involves a real medium and a clock, both of which are subject to uncertainties. Brews ohare (talk) 20:29, 3 January 2012 (UTC)

I'll invent a new symbol "K0". This denotes the mass of the IPK. It is exactly 1kg. (I hope you agree!) This is the definition of K0, and it is fixed by convention and can be no other mass. The number of kg is defined, and not measured, and is exact, not a value to within some uncertainty. Therefore, it is perfectly clear that K0 is a "fixed mass", right? Oh wait, hold on, K0 is not a fixed mass, because the IPK can gain a few atoms and then K0 becomes a larger mass.

I hope you spend a moment thinking about this example. Everything you said about c0 -- exactness, conventions, etc. -- applies equally well to K0...yet K0 is not a "fixed mass"! How is that possible? Do you understand what's happening here?

I'm pointing out a logic error here. You say "c0 is chosen by committee to be 299.. m/s. Therefore c0 is a fixed speed." Now I'm pointed out that the same logic gives faulty conclusions in a different case: "K0 is chosen by committee to be 1kg. Therefore K0 is a fixed mass." You're not allowed to reply, "Well, K0 is different from c0 in various ways..." You see, I'm pointing out a logic error. If the logic is sound, it will work in all cases. If some other aspect of c0 is relevant, it should be built into the logical argument, and the logical argument should not work without that assumption. So let's start again: What is your logical argument that c0 (as before I mean "c0 the speed", not "c0 in m/s, the dimensionless number") is fixed?

(I'll help you out. "1 kg" is not always the same mass, because the IPK can gain or lose atoms. Therefore, the SI kilogram is not really "a unit of mass", it's a quasi-unit, an ill-defined thing that is only approximately "a unit of mass". By the same token "1 m/s" might always be the same speed...or it might not. Therefore, "meters per second" might or might not really be "a unit of speed". You seem very confident that "meters per second" is exactly "a unit of speed" (at least m/s as it relates to c0), definitely not a quasi-unit. It seems you're so confident that you don't even realize that it's an assumption. I want to know why you believe that the "m/s" in the definition of c0 is really truly exactly "a unit of speed" and not a quasi-unit like the kg.) --Steve (talk) 21:10, 3 January 2012 (UTC)

Hi Steve: I guess some very great care is needed in usage here, not my forte. If I think of a "unit of speed" as a real physical speed, then a m/s is only such a unit to the extent that the physical speed of light is fixed by nature and the second is a time interval fixed by nature, neither of which statements can be known to hold true with certainty. I am not sure that this is the point that I am talking about. When I said that c₀ is a "fixed speed" all that I wanted to say was that m/s has dimensions of speed and 299 792 458 is a fixed number, so a clearer statement is to leave "speed" out of the discussion and just say that 299 792 458 m/s is a fixed number of m/s. There is no implied connection to any physical speed or any physical measured quantity. Likewise, for the other posted electromagnetic properties. No connection of ε₀, μ₀ to measurement or to reality. How's that sound? Brews ohare (talk) 00:03, 4 January 2012 (UTC)

OK, great. "299792458 m/s is a fixed number of m/s", we can agree on that. Moving on, do you agree with this: "The fact that c0 is an exact fixed number of m/s does not, in itself, rule out the possibility that c0 is a not-exactly-constant physical speed in the real world." (Just like, "the fact that the mass of the IPK is an exact fixed number of kilograms does not rule out the possibility that the mass of the IPK is a not-exactly-constant physical mass in the real world".) (Maybe you have other unrelated reasons for believing that c0 cannot be a not-exactly-constant physical speed in the real world, but we can discuss those afterwards.) --Steve (talk) 13:54, 5 January 2012 (UTC)

I'd rather reserve c₀ as a notation for the expression 299 792 458 m/s. That would be convenient. Then how do I interpret your question? Is the question that the physical speed of light might not be constant in time. I agree that is possible to within present day measurement uncertainty. If that happens, then the metre as the distance light travels in 1/299792458 s will correspond to a different real-world length at different times, according to c/c₀ with c the real physical speed of light and c₀=299792458 m/s. Likewise, if the second changes, the real-world distance corresponding to the metre will change accordingly. The definition will not change, but will remain the distance light travels in 1/299792458 s.

Is the question whether 299792458 m/s is non-constant in time? That is a bit more subtle. As a defined value, it could be changed only by the standards organizations, if they saw fit.

Is the question: might there be a phenomena in nature that propagates in the real world at 299792458 m/s? I suppose so, although one could establish that only to within a measurement uncertainty. It's possible, I guess, that 299792458 m/s might slightly exceed the real-world relativistic limiting speed, as that speed is not known exactly, in which case a propagation at 299792458 m/s could not happen.

Is the question that given some phenomenon propagating in the real world at 299792458 m/s today, might it propagate at a different speed tomorrow? Of course, that might happen.

Is the question whether the metre might be a different real length in the future and so the measured speed of propagation of the phenomena that was measured as 299792458±Δ m/s today could be measured as something else tomorrow, even though the phenomena was actually propagating at the same speed at both times? Yes, that is possible, within some experimental uncertainty.

Are any of these questions what you are asking, Steve? Brews ohare (talk) 15:21, 5 January 2012 (UTC)

It might be interesting to describe how the real-world physical speed of light c might be established. Of course, the metre being defined as how far light travels in 1/299792458 s, the speed of light in SI units is 299792458 m/s regardless of the real-world speed of light. So we have to introduce a length that is not based itself upon the speed of light, perhaps a wavelength, as in the old definition of the metre in terms of so many wavelengths of some particularly suitable radiation? Or, possibly, the spacing between lattice planes in a specimen of some easily controlled crystalline substance? Brews ohare (talk) 16:40, 5 January 2012 (UTC)

Before I say anything else, I'm confused by: "I'd rather reserve c₀ as a notation for the expression 299792458 m/s". Doesn't c0 mean "the speed of light in vacuum"??? (For some definition of 'vacuum' to be debated.)

Anyway, this is helpful, although I'm focused more on the source of your belief that BIPM must be referring to a classical vacuum, not a real vacuum.

You said earlier: "The BIPM web listings of c, c0, μ0 and the subsidiary derived properties with exact values ε0 and Z0 are exact values, not measurements with measurement errors. So with the notion of a real 'vacuum', these exact properties make no sense.... The speed of light c0 is just one of several electromagnetic properties of the 'vacuum' referred to in the definition of 'vacuum', as listed above. Of course, the prescription of exact values for a medium exceeds the capacity of any measurement on any real medium. Measurement always is subject to uncertainty. Thus, no real medium can be known with certainty to be a realization in the laboratory of the defined 'vacuum'."

The logic here, if I understand, is something like:

• PREMISE 1: c0 is an exact number of meters per second
• PREMISE 2: The speed of light in a real-world quantum vacuum varies with intensity
• CONCLUSION 1: c0 cannot be the speed of light in a real-world quantum vacuum.
• PREMISE 3: BIPM says "c0 is the speed of light in vacuum"
• CONCLUSION 2: Therefore BIPM means "classical vacuum", not "real-world quantum vacuum"

I'm trying to argue that the path from premises 1&2 to conclusion 1 is not sound. Instead, here's my logic:

• (A): The speed of light in a real-world best-practices quantum vacuum varies slightly with intensity
• (B): c0 is "the speed of light in the real-world best-practices quantum vacuum". (More precisely, it is not a speed but a slight range of different speeds.)
• (C): 1 meter is the length that light travels in 1/299792458 s in a real-world best-practices quantum vacuum. (More precisely, it is not a length but a slight range of different lengths.)
• (D): Nevertheless, it is OK to say "c0 (the speed of light in real-world best-practices quantum vacuum) is exactly 299792458 m/s".

I don't see any problem or contradiction here. Therefore, (D) does not disprove (B). In other words, the "exactness of c0 in SI units" does not prove that c0 is not a property of a real-world medium.

From what you wrote, it seems you would agree with "the physical speed of light might vary with time, and therefore the meter might vary with time, yet it is always true that the physical speed of light is exactly 299792458 m/s." Is that right? If so, what is the problem with saying "c0 is the physical speed of light"? --Steve (talk) 17:19, 5 January 2012 (UTC)

Hi Steve: Maybe there is light at the end of this tunnel. Here is my world view: "out there" in the "real universe" there is such a thing as a light wave, and it propagates at a certain speed that depends upon the medium. In the SI units this speed is c₀/n, n being the index of refraction. So if it were possible to construct a medium that one knew with complete certainty n=1, one would know with certainty that light propagated at speed c₀ in that medium. Of course, even in principle, one cannot know that a medium with n=1 has been achieved with certainty.

Also, in my world view, there is a physical speed of light in this unrealizable vacuum. In SI units it is always 299792458 m/s even if it happens that the physical speed of light is a time-varying entity. So you see, my notion of the "physical speed of light" is different from the SI units value 299792458 m/s in principle, though it may be a distinction only in principle. Here is the difference: my physical speed of light can change, in principle if not in fact, while 299792458 m/s never changes. If one wished to determine whether my physical speed of light was in fact changing with time, maybe a way to do it would be to look at a particular diffraction pattern over time. If this pattern changed with time, and all obvious changes like temperature, the Earth's position wrt to the other celestial bodies, whatever, could not account for the change, a candidate would be that the speed of light c was changing, altering wavelength via λ=c/f. Of course, it would add to one's confidence in drawing this conclusion if one had a theory that predicted a change in the physical speed of light with time, say by connecting it to the volume of the universe, or to the political party in the White House. If one adopts the notion that the physical speed of light varies with time, it has no effect upon the speed of light in SI units, which remains 299792458 m/s. Brews ohare (talk) 19:08, 5 January 2012 (UTC)

If there is a semantic difference between us, it seems to be point (B) above. I'd take c₀ as related to the speed of light c in a real-world best-practices quantum vacuum by the relation c = c₀/n, with n the refractive index of the real-world best-practices quantum vacuum. I believe this view is also that of these sources, which in effect calculate n for field-theoretic vacuums. Brews ohare (talk) 19:45, 5 January 2012 (UTC)

Previously you were focusing on an argument like "the electromagnetic properties of the 'vacuum' in SI units had no error bars, therefore the 'vacuum' cannot possibly be a real physical thing, because real physical things have error bars". Again, "the mass of the IPK in kilograms" is a simple example that this is bad logic. I'm happy to see that you're now emphasizing this argument less, and I hope you're really abandoning it for good.

It seems you're focusing now instead on a different argument:

It is possible, in principle, to calculate (using QFT methods) the index of refraction of a quantum vacuum, and it is not necessarily 1. Therefore c0 cannot possibly be the speed of light in a quantum vacuum.

Let's talk about, say, the electric repulsion between two electrons. The repulsion is screened by virtual photons in the vacuum. In very high energy collisions (in other words, very close distances), the screening becomes less and less important...you might say the dielectric constant of vacuum is reduced. So my question for you: Is ε0 approximately equal to the relative permittivity of vacuum at very high energy (where the vacuum screening is relatively less important)? Or, is ε0 approximately equal to the relative permittivity at very low energy (where the vacuum screening is a much more important effect)?

(Based on what you've been saying (ε0 is the permittivity of a classical vacuum, and a classical vacuum has no quantum fluctuations), I expect you to say that the high-energy collision gets closer to ε0. But the value people actually use for ε0, with BIPM's endorsement, is definitely the low-energy one.)

So, what is "the permittivity of a vacuum" as discussed in QFT? Well, take a look at Bertulani's book. Eq. (1.64) gives a formula for "the permittivity of the vacuum in QED" as something equivalent to:

ε (as function of r and r0) = [(e-e repulsion at distance scale r0)*r0^2]/[(e-e repulsion at distance scale r)*r^2]

In other words, calculate the charge e based on interactions at distance scale r0, then calculate the dielectric constant to correct for the fact that interactions are different at distance scale r. Both the numerator and denominator are in the QED vacuum. Bertulani most certainly does not use the definition

ε = (e-e repulsion in classical vacuum at distance scale r) / (e-e repulsion at distance scale r) <--- NOT RIGHT!

Really, in Bertulani's calculation, no classical vacuum is involved. He is calculating the ratio of one aspect of the quantum vacuum to another aspect. Moreover, there is no answer for "What is the permittivity of the vacuum?" (unless you specify both r and r0).

Likewise, in papers about the refractive index of the quantum vacuum for high-intensity light, they define it as:

n^-1 = (speed of high-intensity light)/(speed of very-low-intensity light).

No classical vacuum is involved, they are calculating the ratio of one aspect of the quantum vacuum to another aspect.

(By the way, I don't see any of your sources "calculating n for field-theoretic vacuums", can you be more specific?) --Steve (talk) 22:20, 5 January 2012 (UTC)

Steve: Your italicized statement is one I'd agree with. With some modifications: “It is possible, in principle, to calculate (using QFT methods) the index of refraction of a quantum vacuum, and it is not necessarily identically 1. Therefore c₀=299 792 458 m/s cannot possibly be the speed of light in a quantum vacuum. Instead, the speed of light in quantum vacuum isc=c₀ ⁄ n.”

I'll look into the details you describe later. For the moment, I'd say the refractive index is closely related to the relative dielectric permittivity of a medium. The latter is calculated as ε/ε₀. Brews ohare (talk) 23:41, 5 January 2012 (UTC)

Although in my mind the easiest way to look at things is to regard BIPM 'vacuum' as a medium with specific defined properties, that is not essential. Instead one can look upon ε₀, μ₀ as simply defined constants with dimensions of permittivity and permeability used to express relative permittivity and relative permeability as per the ratios ε/ε₀ and μ/μ₀. Brews ohare (talk) 23:54, 5 January 2012 (UTC)

QED vacuum

Steve: If you wish to go into the details of some of the permittivity and permeability calculations for field-theoretic vacuums, that's fine. However, it would be well to have an objective before plunging into a lot of detail, which may prove a diversion unless we have a road map. The details will not establish that these vacuums have indices of refraction that are identically unity, or that are field-strength & frequency & wavelength & polarization independent, and so the details certainly cannot lead to a speed of light c₀/n that is c₀ = 299 792 458 m/s. If that is the goal of a detailed assessment, I'll go through the analysis with you, but that objective will not be realized. Brews ohare (talk) 18:50, 6 January 2012 (UTC)

For example, this source says "In QED, vacuum nonlinearities appear near the Schwinger critical field for pair creation" This source says the QED vacuum exhibits birefringence and dichroism. This source points out that QED vacuum is nonlinear and allows light scattering by light. These phenomena don't happen in a medium with the electromagnetic properties of the reference vacuum of SI units. There are literally dozens of sources along these lines. Brews ohare (talk) 16:48, 7 January 2012 (UTC)

Sure, here is a roadmap:

• Vacuum fluctuations really do screen weak electric fields. This is not a small effect, it is a huge effect, way way above measurement precision.
  • At infinite distance, with all the vacuum screening, the apparent (screened) electron charge is what we call e. If you get closer to the electron, you're partly inside the shell of screening, and you can see that the real electron charge is significantly bigger, maybe you see 10% or 20% more charge when you get within 1E-32 meters (GUT scale). (Look up "running coupling constants" to find the exact formula and graph, something like this.)
• Therefore, the "relative permittivity of the vacuum", relative to no screening by quantum fluctuations, is not 1, it's at least 1.1 or 1.2. Maybe even infinity. (When you get as close as the Planck distance, who knows...)
• Screening by vacuum fluctuations is never "corrected for" in experimental measurements and calibrations
• Therefore, if we're really supposed to compare to "classical vacuum", then loads of "precision" measurements are actually more than 10% off.
• 10% errors would not go unnoticed in the precision metrology community, which often claims parts-per-million or better accuracies.
• Therefore, we can be sure that BIPM does not intend for ε0 to be the permittivity before screening by vacuum fluctuations. Instead, we infer that ε0 is the permittivity with vacuum fluctuations, in the long-distance / low-energy limit.

That's my roadmap. Where do you start disagreeing?--Steve (talk) 02:31, 8 January 2012 (UTC)

I find none of the above remarks have any connection to the topic at hand. Brews ohare (talk) 05:53, 8 January 2012 (UTC)

Again, you can find (and have found) lots of papers on the interesting electromagnetic properties of the vacuum at high fields. As far as I can tell, whenever anything is calculated in those papers, it is a comparison of one property of the quantum vacuum to a different property of the quantum vacuum. For example, the speed of light at high intensity divided by the speed of light at low intensity. Or the static permittivity in a certain strong field divided by the static permittivity in the limit of arbitrary weak fields. Or whatever. I have not seen anything that was defined or described by comparing it to a classical vacuum with no quantum fluctuations. --Steve (talk) 02:31, 8 January 2012 (UTC)

These calculations of permittivity etc. all are expressed (if done in SI units) as multiples of ε₀ or μ₀. Brews ohare (talk) 05:53, 8 January 2012 (UTC)

On the other hand, now it sounds like maybe you're moving back towards the older argument: "The "speed of light in vacuum" is a certain number of meters per second, therefore it is inconceivable that the "speed of light in vacuum" is intensity-dependent." (And similar with permittivity, etc.) Are you saying that? As I've been saying over and over, this is bad logic. "The mass of the IPK is a certan number of kilograms, therefore it is inconceivable that the mass of the IPK is dependent on the time of day." --Steve (talk) 02:31, 8 January 2012 (UTC)

Steve: You have not grasped the point of a reference medium, and repeatedly confuse it with a realizable medium, which it is not. Brews ohare (talk) 05:53, 8 January 2012 (UTC)

I refer you to my comments of 19:08, 5 January 2012 (UTC), which you did not address. Brews ohare (talk) 15:29, 8 January 2012 (UTC)

One way to look at this reference medium is that it is the vacuum that results from QED vacuum as ℏ → 0. Brews ohare (talk) 18:15, 9 January 2012 (UTC)

OK, I'm interested in the specific relationship between quantum vacuum and classical vacuum. Knowing this relationship is a prerequisite for the classical vacuum to function as a reference medium. For example, if you say "the classical vacuum is an imaginary universe where Maxwell's equations hold", that's not enough, it's too vague. It would be impossible to take the ratio of, say, the speed of light in the quantum vacuum to the speed of light in the classical vacuum, because so far all we know is that the classical vacuum is an imaginary universe, and we have not yet been given any basis for comparing a speed in this imaginary universe to a speed in the real universe. So we need a specification of how you go between quantum and classical vacuum. (Obviously it can be a hypothetical procedure, even an impossible procedure, but most important is that it is specific.)--Steve (talk) 04:42, 12 January 2012 (UTC)

I'd guess that any calculation of the relative permittivity or relative permeability of a medium, including quantum vacuum, if made in SI units, will incorporate the parameters ε₀, μ₀. So when the calculation is finished, one has something like the permittivity of the medium is ε=κ_rε₀, which makes clear the relation to the 'reference vacuum' where κ_r≡1. Of course, the relation ε=κ_rε₀ may involve some more complicated expression like an integral over frequency and wavelength with some kernel, but the factor ε₀ will be there. Brews ohare (talk) 18:50, 12 January 2012 (UTC)

So here are two possibilities, which are specific enough to get off the ground and work with:

1. "If you start with a quantum vacuum, and turn off all the quantum fluctuations, then you have the classical vacuum." Well, now we've ruled this out: If you turned off the quantum fluctuations, you would be left with something where the permittivity is not ε0, it is 0.9*ε0 or less. (Do we agree?) So that's not "the classical vacuum", it's something else (or maybe it's nonsense). --Steve (talk) 04:42, 12 January 2012 (UTC)

I don't see why letting "ħ→0", which means the fields Π and A commute, removing fluctuations, causes a change in permittivity to something other than ε₀. Brews ohare (talk) 16:06, 12 January 2012 (UTC)

2. "The classical vacuum is the ħ->0 limit of the quantum vacuum". I actually like this! Of course, ħ being a dimensional parameter, ħ->0 doesn't make sense literally (e.g. in natural units ħ=1). But any physical situation will involve quantities with units of action (or angular momentum), and we can certainly take the limit that those quantities are all much larger than ħ. (This is how "ħ->0" is usually interpreted in my experience.) So that would be, I suppose, the limit of large distances, low energies, low field strengths, etc. --Steve (talk) 04:42, 12 January 2012 (UTC)

"ħ→0" means the fields Π and A commute, removing fluctuations. $\left[\Pi _{i}(\mathbf {x} ,\ t),\ A_{j}(\mathbf {x'} ,\ t)\right]=-i\hbar \delta _{ij}\delta (\mathbf {x-x'} )\ .$ Brews ohare (talk) 16:06, 12 January 2012 (UTC)

I do believe and I am delighted to say the following:

"In the limit of large distances, low energies, low field strengths, etc., the quantum vacuum of our universe is believed to approach, in its electromagnetic behavior, a "classical vacuum" exactly satisfying Maxwell's equations. The speed of light in this limit (called c0) is exactly 299792458 m/s (by definition of the meter), the permeability in this limit (called µ0) is exactly 4piE-7Vs/Am (by definition of the ampere), the permittivity (called ε0) is ... etc. The "relative" permittivity of a material (or the quantum vacuum itself) is "relative" to ε0, and likewise with permeability, index of refraction, etc." --Steve (talk) 04:42, 12 January 2012 (UTC)

It isn't clear to me that the fields commute under the proposed limits. Brews ohare (talk) 16:06, 12 January 2012 (UTC)

What do you think of this? One point we can discuss is that you seem 100% sure that the quantum vacuum in the limit ħ->0 exactly satisfies Maxwell's equations. ("One way to look at this reference medium is that it is the vacuum that results from QED vacuum as ħ->0.") How do you know that? Isn't it conceivable that the quantum vacuum has a ħ->0 limit slightly different from Maxwell's equations in some respect? For my part, I phrased my above paragraph to be slightly more cautious ("...is believed to approach..."). --Steve (talk) 04:42, 12 January 2012 (UTC)

I suspect the question of validity of Maxwell's equations is a separate topic, not so far discussed. Insofar as one views QED vacuum as a theoretical model, one might presume that it assumes Maxwell's equations. Likewise, there is no problem demonstrating that in the "ħ→0" limit it is exactly the vacuum of classical electromagnetism. That leaves open the question of how well any real vacuum is described by QED vacuum (or QCD vacuum or quantum gravity vacuum). Brews ohare (talk) 22:30, 12 January 2012 (UTC)

There are other points to be discussed too. But for now I'm interested in your thoughts. :-) --Steve (talk) 04:42, 12 January 2012 (UTC)

There are a few ways to phrase the same question: "How is ε0 defined?" "How can I calculate the "relative permittivity" (relative to the SI ε0) of a medium or vacuum in QFT?" "How can I translate a QFT calculation into SI units?" "What is the exact specific relationship between the 'classical vacuum' and the quantum vacuum?" I'm happy for you to answer any of these, as long as you don't answer one by assuming that the answer to a different one is obvious. :-)

You said: "I don't see why letting "ħ→0", which means the fields Π and A commute, removing fluctuations, causes a change in permittivity to something other than ε₀."

I'll address the "removing fluctuations" part. I thought I answered this above, maybe I wasn't clear. (Incidentally, I'm not ready to agree that "removing fluctuations" and "ħ→0" and "[Π,A]=0" are all equivalent.)

First, some background. (I believe you already know this stuff well but I want to be sure we're on the same page.) Let's say you put a point charge +Q into a block of glass with ε=10ε0. The charge polarizes the glass, drawing negative bound charge towards it. After equilibrium is reached, the total charge at the center of the block is +0.1Q, which consists of the original +Q added onto the -0.9Q of negative bound charge (electrons from glass molecules). If I want to know the relative permittivity of glass: First, look at the total charge at the center of the block: It's +0.1Q. Second, look at the charge at the very center (excluding the shell of screening glass electrons): It's +Q. Finally, take the reciprocal of the ratio of those two charges: It's 10, as expected.

Now let's do the same procedure for an electron in vacuum. If you draw a macroscopic (say, 1cm) sphere around an electron and say "What's the total charge inside this sphere?", the answer is what people call "-e". This net charge includes the negative "bare charge" of the electron itself, plus the positive charge associated with screening by quantum fluctuations. If you draw a much smaller sphere around the electron, say 1E-32 meters, then you are excluding part of the positively-charged shell of screening fluctuations, and the charge inside the sphere is more negative, something like -1.1e or -1.2e. If you draw the sphere even tighter than 1E-32 meters, the charge changes some more, but we don't know what happens at the Planck length so the final result is a mystery. Let's just say the bare charge is -1.2e, that's as good a guess as any. (Certainly, it's a better guess than -e. Incidentally, in a naive extrapolation of QED, the bare charge is "minus infinity".)

Now let's do the procedure above to calculate the relative permittivity of the quantum vacuum, relative to no screening by vacuum fluctuations. First, I look at the total charge in the vicinity of the electron: It's -e. Second, I look at the charge at the very center, inside the screening shell: It's -1.2e. Third, I take the reciprocal of the ratio: It's 1.2. Therefore the relative permittivity of the quantum vacuum (relative to no screening by vacuum fluctuations) is 1.2.

You keep saying, ε0 is the permittivity with no screening vacuum fluctuations. Well, then the permittivity of the quantum vacuum is 1.2ε0 in the everyday limit of relatively weak fields, etc. [Of course, I would say that the permittivity with no screening vacuum fluctuations is (1/1.2)ε0, and the permittivity with screening vacuum fluctuations is ε0.] :-) --Steve (talk) 17:37, 15 January 2012 (UTC)

Steve: I don't know if we are capable of discussing the same thing. If we focus upon classical electromagnetism for the moment, and we think about not electrons but ideal, stationary point charges, I'd say that the force law was Coulomb's law with ε₀. I'd say that when there are no charges, the fields are zero.

On the other hand, if one quantizes the fields, the variances of the fields are not zero, virtual particles can come about and vacuum fluctuations. In this case the medium has relative permittivities and permeabilities that are not identically unity.

Hence my view that when quantization is not performed, classical vacuum results, and when it is performed QED vacuum results.

Admittedly, this is a very restricted situation. It doesn't deal with realizable vacuum in either case. It doesn't deal with real particles like electrons. Brews ohare (talk) 01:54, 16 January 2012 (UTC)

Brews, I understand, you believe that if you turn off vacuum fluctuations, then you end up with a permittivity of ε₀. Then I just explained in excruciating detail why, if you turn back on vacuum fluctuations, the permittivity increases by 20%, to 1.2 ε₀ (at least, in the vicinity of an electron, in the everyday-low-energy-large-distance limit). So do you agree: "The relative permittivity of QED vacuum is 1.2ε₀" (in the specified situation)? Or not?

It seems to me that you want to believe in various things that contradict each other. You cannot ignore this problem by saying that you don't want to talk about electrons! After all, I can do the calculation, in the context of an electron, for "what is the relative permittivity of QED vacuum". Why should I not get a correct result? What is it that I am calculating, if I'm not calculating "the relative permittivity of QED vacuum"? If you can think of some way I can modify this calculation to more correctly answer the question "what is the relative permittivity of QED vacuum relative to no vacuum fluctuations", please tell me, and I'm happy to see whether the answer is still 1.2, or whether it is some different number. (As I mentioned above, "Just do your QFT calculations in SI units" is not a helpful response. I don't know how to do QFT calculations in SI units as you understand them, so you'll have to explain the procedure.) --Steve (talk) 15:45, 20 January 2012 (UTC)

Steve: Your use of 'vacuum' in your explanation instead of identifying exactly which 'vacuum' you were designating left me confused about your intentions. Now that I see you are talking about QED vacuum which is a quantized vacuum state, I agree that the permittivity of QED vacuum is not ε₀, of course, as I have said myself repeatedly. Whether your estimate of its permittivity is accurate or not is not material here, but such calculations can be found in the published literature as cited in the article QED vacuum. Have you some point to make based upon your estimate, or can the same point be made based upon the permittivity calculations described in the sources cited in QED vacuum?

My understanding at this point is that you wish to say that if quantization is turned off, QED vacuum does not revert to a medium with permittivity ε₀ but to something else. That belief, which may not be yours (I am unsure about your belief), contradicts the sources in QED vacuum. Brews ohare (talk) 15:54, 21 January 2012 (UTC)

I have in mind this source and this and this as indicative of Quantum vacuum reverting to a medium with permittivity ε₀.

Quantum state

If you don't mind, could you please look at this article and read here. Its written a little oddly. It has been posted to the project talk page, but no one has responded since then. Thanks a lot. -- F = q(E + v × B) 11:08, 22 January 2012 (UTC)

Please forget it. I'll just sort it out myself in time, with the others at the talk page.-- F = q(E + v × B) 00:52, 23 January 2012 (UTC)

Permittivity of 'real vacuum'

When I say "vacuum" or "quantum vacuum" I mean "the real-world vacuum". Of course, the electromagnetic properties of the real-world vacuum are primarily determined by QED, unless you're looking towards the Planck scale.

I am interested in the exact value of permittivity that you think the real-world vacuum has in the everyday limit (low energy, long distances). Let's call it ε_{EverydayWorld}. What value do you think ε_{EverydayWorld} is? Well, I don't care so much about the exact value as:

• Is the ratio ε_{EverydayWorld}/ε₀ different from 1 by more than 1 part per million?

If you believe that this is true, or even that it might be true, then I can definitively prove--based on specific results in hundreds or thousands of published papers--that you are using a different definition of ε₀ than is NIST and the scientific community. (We can cross that bridge when we get to it.)

I happen to believe that ε₀ is defined to be ε_{EverydayWorld}, although I understand that you disagree. Therefore I believe ε_{EverydayWorld}/ε₀ = 1. You claim to believe that "When you turn off vacuum fluctuations, the resulting permittivity is ε₀". Therefore, you ought to believe that ε_{EverydayWorld}/ε₀ ~ 1.2, as I explained above.

You seem to want me to find a book calculating the value of ε_{EverydayWorld}/ε₀. Sorry, I don't know of any. (Do you?) Again, I believe that scientists take this to be 1 by definition. I do, however, believe that it is possible to calculate what this ratio would be using your definition of ε₀: The ratio of the bare charge to the dressed charge of an electron or other charged particle. Bertulani's book agrees that this is an appropriate way to think about how vacuum fluctuations affect permittivity. However, Bertulani's book does not give any formula for the permittivity, he gives permittivity as a function of two distance parameters, which corresponds to my definition if I take one parameter to be infinity and the other to be the Planck length. --Steve (talk) 17:54, 22 January 2012 (UTC)

Hi Steve: It appears we are approaching the state of "irreconcilable differences". As you may suspect, I really don't have any interest in the value of ε_{EverydayWorld}, or how slightly it may depart from ε₀. Moreover, I view QED vacuum as a model of "real vacuum", which may provide an excellent approximation to ε_{EverydayWorld} in some regimes, and not in others. So the behavior of QED vacuum in limiting cases is an entirely theoretical question about the mathematics of the model, and has nothing to do with ε_{EverydayWorld} in principle, although, of course, its adequacy as a model may be found suspect if its predictions in limiting cases do not accord with experiment.

You identify ε_{EverydayWorld} to be the permittivity of "real vacuum", which I take to mean a quantity found in principle by measurement, although measurements may have to improve to get a good value.

At the same time you seemingly take ε₀ to be "defined" to be exactly the same thing.

So I am perplexed, inasmuch as ε₀ is taken by NIST and BIPM as the CODATA value of (1/μ₀ c₀²)^1/2, with c₀ and μ₀ defined values. See "Electric constant, ε₀". NIST reference on constants, units, and uncertainty: Fundamental physical constants. NIST. Retrieved 2012-01-22. The formula determining the exact value of ε₀ is found in Table 1, p. 637 of PJ Mohr, BN Taylor, DB Newell (April–June 2008). "Table 1: Some exact quantities relevant to the 2006 adjustment in CODATA recommended values of the fundamental physical constants: 2006" (PDF). Rev Mod Phys. 80 (2): 633–729.{{cite journal}}: CS1 maint: date format (link) CS1 maint: multiple names: authors list (link). Brews ohare (talk) 20:34, 22 January 2012 (UTC)

Yes, ε_{EverydayWorld} is the permittivity of "real vacuum", meaning the vacuum of the universe that we live in. I agree, we can generally use QED (or better yet, the full standard model) to accurately understand the real vacuum, although it's always possible that the standard model is not accurately describing the real vacuum.

You use the word "slight" to describe the departure of ε_{EverydayWorld} from ε₀. Does that mean you believe that they are almost equal, and definitely not 20% different? If you believe that, why?

I hope you understand that if ε_{EverydayWorld}=1.2ε₀, then every permittivity value ever measured in SI units by any scientist on earth is 20% wrong!! (If you don't believe me, think about how, in practice, an engineer would measure "the permittivity of glass".) Given that, I cannot fathom how you are not interested in the value of ε_{EverydayWorld}. You don't care a fig whether or not millions of measurements, textbooks, handbooks, and datasheets are 20% wrong???

Again, according to your understanding, ε_{EverydayWorld} is a crucial parameter, without which it is impossible to make almost any accurate electrical measurement in SI units. Given that, isn't it strange that the physics literature has nothing whatsoever to say about the numerical value of this parameter? Isn't it strange that BIPM has not issued a recommendation on what value metrologists should use for ε_{EverydayWorld}? Isn't it strange that QFT textbooks do not say what the numerical value of this parameter is thought to be, or how it is estimated? Really, you don't see anything strange or implausible here??? --Steve (talk) 22:28, 22 January 2012 (UTC)

If I thought there really was a 20% error, that would be a curious situation, but still irrelevant to this discussion. If you like, you can look at how the refractive index of air is determined. See this. I don't see where you get the idea that the QED permittivity is not calculated: the sources mentioned earlier do just that. And you haven't addressed the role of the posted ε₀. Is it your opinion that it is just a curiosity? Please fill me in. Brews ohare (talk) 01:37, 23 January 2012 (UTC)

In particular, how can the defined ε₀ be identified as exactly the same concept as the experimentally measured ε_{EverydayWorld}?? Brews ohare (talk) 17:26, 23 January 2012 (UTC)

I suspect, Steve, that you may have inadvertently misspoken in this confusion between measurable entities and models that make predictions that can be compared to measurable entities? Brews ohare (talk) 16:04, 24 January 2012 (UTC)

Can you please show me specifically the source that calculates "the permittivity of the QED vacuum"? It should be a specific real number, or I suppose a specific real number times ε₀. A formula that can be evaluated to get a specific real number is also OK, I have a calculator and can plug things in.

For example Bertulani's book gives ε_QED(r,r0), a function of two distance parameters "r" and "r0". What values should I plug in for r and for r0 to get the QED approximation to ε_{EverydayWorld}? (Based on your definitions, I think it should be r0=infinity, r=Planck length, as I explained above. What do you think? Or am I looking at the wrong book?)

If the value of ε_{EverydayWorld} is really 1.2ε₀ (as I believe it is according to your definitions), that would be extremely relevant, because then I can find very easily find hundreds of sources that clearly and specifically disprove your theory about how ε₀ is defined. All I have to do is find a source that says, for example "If you experimentally measure the permittivity of air, you should find that it gets closer and closer to ε₀ as the air pressure is decreased to towards 0." That proves that your understanding is wrong, because if your understanding was correct, then the source would have to correctly say: "If you experimentally measure the permittivity of air, you should find that it gets closer and closer to 1.2ε₀ as the air pressure is decreased towards 0." Do you understand? If you want to argue that your understanding of ε₀ is consistent with sources like that, it is essential that you check the numerical value of ε_{EverydayWorld}. --Steve (talk) 04:49, 25 January 2012 (UTC)

The source you just gave for the index of refraction of air is a great example of exactly what I am talking about. It is assumed that 1 is the limit of refractive index as the air pressure decreases towards 0 with quantum fluctuations still present. If the speed of light is different with quantum fluctuations versus without (BTW I don't believe it is), then this page would contradict your definition of refractive index and c0. --Steve (talk) 04:49, 25 January 2012 (UTC)

Posted value of electric constant

On a different topic, you asked me what I think is the role of the posted value of ε0. Well, I see that NIST has a website listing the numerical values of various quantities in SI units. I don't see anything curious about that! I see that if the numerical value of a quantity in SI units happens to be known exactly, then the website makes a note of that fact. Nothing curious about that either! Finally, I see that ε0 is one of the quantities whose value in SI units is known exactly. This puts it in the same category as "the fine-structure oscillation period of caesium" and "the mass of the IPK", among others. I don't see anything curious about ε0 being in that category. For everything in this category, "you cannot 'measure' its value in SI units because the SI units are defined in terms of it". For everything in this category, "experimentalists can and do study and reproduce the quantity in laboratories". I hope by now you agree that there are real-world quantities whose numerical value in SI units is known exactly (even if you don't agree that ε0 is one of them), and I hope you understand how exactly a real-world quantity can be in this category. After all, I have brought up the "mass of the IPK" example over and over. I wonder what you think about the fine-structure oscillation period of caesium. "How can the defined oscillation period of caesium be identified as exactly the same concept as the experimentally measured oscillation period of caesium??" How would you answer that question? --Steve (talk) 04:49, 25 January 2012 (UTC)

Hi Steve: You say that ε₀ ≡ (1/μ₀ c₀² )^1/2 is beyond measurement because of the use of a defined value for c₀ and μ₀. I'd agree with that. You also say that for that same reason, the exact value of ε₀ does not mean it isn't a property of real vacuum. So I guess this puts matters back on the same footing as the statement that c₀=299***m/s is the 'real' speed of light?

As I've said before, if the speed of light c that appears in special relativity were measured in crystal lattice spacings/s, that number could be measured over time and in principle could change. But c₀ never changes. Only the number of crystal lattice spacings in a metre changes. So I'm left with the idea that the speed of light in special relativity is a concept that allows c to vary in principle, but in SI units any such variation can be discovered only by going outside the SI units to introduce a standard of length like crystal lattice spacings.

So, to follow this through, I guess one could express ε₀ using c₀ in crystal lattice spacings/s? Then it would become a measurable entity with an error bar? I haven't thought this through. Brews ohare (talk) 12:59, 25 January 2012 (UTC)

Even if the parameters entering the formula for ε₀ were measured in some other units system instead of using defined quantities, ε₀ still would be defined as (1/μ₀ c₀² )^1/2, so it still would be an evaluated quantity, not an independently measured quantity. In particular, there would be no reason to expect it to be measurable using a capacitor in 'real vacuum', because that would measure ε_Everyday, the permittivity of a medium which has a dependence upon wavelength, frequency, intensity and so forth. Brews ohare (talk) 13:20, 25 January 2012 (UTC)

Barnstars!

It really is only fair to give you these, a trillion thanks for your hard work in:

		The Content Creativity Barnstar
		Your style and contribution of stirring images, animations, boxes and mainstream text together (those wavefunction animations + explanations simply couldn't be clearer!). =) F = q(E + v × B) 18:16, 1 February 2012 (UTC)

		The E=mc² Barnstar
		Re-transforming the physics articles: into simple language for lay-readers, yet technical enough for precision and knowledgeable readers, without dumbing down content, while maximizing the amount of content per paragraph. =) F = q(E + v × B) 18:16, 1 February 2012 (UTC)

In particular, I'm really happy that you re-wrote my completely inferior introduction sections to Schrödinger equation, and feel a little guilty that it may have been a lot of work! -- F = q(E + v × B) 18:16, 1 February 2012 (UTC)

Disambiguation link notification

Hi. When you recently edited Quasiparticle, you added a link pointing to the disambiguation page Effective mass (check to confirm | fix with Dab solver). Such links are almost always unintended, since a disambiguation page is merely a list of "Did you mean..." article titles. Read the FAQ • Join us at the DPL WikiProject.

It's OK to remove this message. Also, to stop receiving these messages, follow these opt-out instructions. Thanks, DPL bot (talk) 11:02, 2 February 2012 (UTC)

I fixed it for you ("effective mass" of electron in solid state physics - right?).-- F = q(E + v × B) 12:25, 2 February 2012 (UTC)

Oscillation period of caesium

Steve: You ask:

"How can the defined oscillation period of caesium be identified as exactly the same concept as the experimentally measured oscillation period of caesium??" How would you answer that question?

I'd assume that this question is related to the definition of the second

The second is the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom.

I'd take it that you are suggesting that it is not possible to measure the period of the cesium atom in seconds, because it is defined in terms of the period. Of course that is true.

It also happens that this unit is defined in terms of a physical event that repeats in time, and not in terms of some other property entirely, like a length, or like the speed of light. Naturally, a unit of time can be chosen as the period of any repeated event, like the rotation of the Earth as another old example.

I don't think this example has any bearing upon ε₀, which in the view of CODATA is defined by a formula involving μ₀ and c₀ and is not defined in terms of the permittivity of 'vacuum' or any real medium, as you seem to think. It is defined in terms of μ₀ and c₀ in Table I, p. 1191, and is not related to any physical realization. Brews ohare (talk) 19:06, 26 January 2012 (UTC)

Maybe I don't have your question clear? The phenomena of the oscillation of cesium can be observed, but to measure it one has to compare it to another periodic phenomenon. If the standard for timing events is the oscillation of cesium, then it cannot be measured. Of course, one could compare a number of different periodic phenomena to see whether they agree over long time intervals, whether one was easier to use, whether one was more precise, and so forth. Those comparisons would decide whether cesium was the best choice.

IMO, this exercise in deciding what is the best unit for time has no parallel in setting up ε₀. Perhaps you disagree? Brews ohare (talk) 18:00, 27 January 2012 (UTC)

To recap where we seem to be...The two issues we seem to be discussing are (1) You think I must be crazy to think that ε₀ is by definition the same as ε_{EverydayWorld}, (2) I think you are crazy to think that ε₀ is different from ε_{EverydayWorld}. Within (2) is: (2a) If this were true, then the numerical value for ε_EW/ε₀ would be an incredibly important parameter in real life, calculated in many textbooks and papers; however, in the real world, I still have not seen any numerical estimates besides my own rough guess of 1.2. (2b) All the hundreds of thousands of physicists and engineers in the world have "voted with their feet" that ε_EW and ε₀ are equal, by measuring permittivities in a way that references them to "vacuum with quantum fluctuations", rather than "vacuum without quantum fluctuations" (which may be ~20% different).

Out of those two issues, right here we're focused on (1): Am I crazy to think that ε₀ could possibly be the same (by definition) as ε_EW? Well, I thought that your argument was "ε_EW is a kind of real-world thing that has something to do with experimental measurements. On the other hand, CODATA says that ε₀ expressed in SI units is an exact numerical quantity. Therefore ε₀ and ε_EW cannot possibly be the same." I was rebutting this argument by saying that there are real-world things that have something to do with experimental measurements, that are nevertheless exact numerical quantities when expressed in SI units. The oscillation period of caesium is a good example, the mass of the IPK is another, the mass of a mole of carbon-12 is yet another, etc.

It seems your argument was subtler than that. If I understand now, you are saying: "Yes, it is possible for a quantity to have an exact value in SI units, but nevertheless to be experimentally realizable (at least in principle). This funny situation only occurs when the SI unit is defined directly in terms of the quantity. For example, it is possible for the oscillation period of caesium to be an exact fixed number of seconds, because the second is defined as a multiple of the oscillation period of caesium. As another example, it is possible for the mass of the IPK to be an exact fixed number of kilograms, because the kilogram is defined as a multiple of the mass of the IPK."

"But," you say, "ε₀ is not like this, because there is no SI unit defined as a multiple of ε₀ or otherwise in terms of ε₀. Without that trick, it is impossible to think that ε₀ could simultaneously be exact in SI units, and experimentally realizable (at least in principle)."

"Finally," you say, "CODATA says ε₀=1/c₀²μ₀, not approximately equal but exactly equal. If ε₀, c₀, μ₀, were separately defined as three different parameters describing the real-world, experimentally-realizable vacuum (at least in principle), it would be impossible to say with certainty that they satisfy any exact relation."

I will hold off on responding until you can confirm that I am correctly understanding and summarizing your arguments. :-) --Steve (talk) 19:22, 27 January 2012 (UTC)

Looks largely correct. The definition of ε₀≡1/c₀²μ₀ clearly precludes any measurement of ε₀, because we know its value exactly and no measurement of permittivity can change that. Moreover, ε₀ does not refer to any realizable "unit" of permittivity (say, in terms of the capacitance of some standard capacitor) in the way the period of cesium refers to an actual unit of time.

In the event the logic of the matter is obscure, I appeal to the fact that all models that might apply to real vacuum, such as QED vacuum, demonstrate nonlinearity, dispersion, nonlocality and whatever, while ε₀ shows none of this behavior. Consequently, it seems likely that when experiment rises to the occasion where these things can be demonstrated, ε₀ will not be a candidate to describe any real vacuum. I understand photon-photon scattering has already been observed in real vacuum, showing nonlinearity is an experimental fact as well as a theoretical prediction.

The case of the speed of light is an interesting one. Choosing c₀ as the unit of speed might refer to a real speed as entertained by relativity, and if so, it does make it unobservable in SI units, just like the second makes the period of Cesium unobservable. And like that case, this speed can be compared to other speeds to decide what is the best choice for a standard. The experiments supporting relativity indicate it has some undisputed advantages in reproducibility etc., although in practice people will use the speed of light in air or in helium-filled chambers and correct for the medium using c₀/n

In an exactly similar fashion to realizing the standard speed in air, it may turn out that the refractive index of no real vacuum has identically n≡1. In which case choosing the speed of light as a constant independent of frequency, wavelength, polarization, intensity, etc. makes this choice a convenient fiction, which nonetheless can be used here as well to refer to the speed of light in real vacuum as c₀/n. Brews ohare (talk) 21:20, 27 January 2012 (UTC)

Hi Steve: Have you decided we are at an impasse at this point, or have you decided that in fact we are on the same page? Brews ohare (talk) 14:06, 31 January 2012 (UTC)

I've been busy, sorry.

• YOUR ARGUMENT 1: "all models that might apply to real vacuum, such as QED vacuum, demonstrate nonlinearity, dispersion, nonlocality and whatever, while ε₀ shows none of this behavior."
• MY BELITTLING REPHRASE 1: "There are some people--not CODATA but other people--who use the term "vacuum permittivity" for the quantity ε₀. They do not call it "vacuum permittivity in the limit of weak fields", they just call it "vacuum permittivity". Therefore we must require that ε₀ is the permittivity of "vacuum" (whatever that is) at any field whatsoever, no matter how high. This proves that "vacuum" cannot be "real-world vacuum", because the permittivity of a real-world vacuum changes at extremely intense fields."
• MY RESPONSE 1: I think that ε₀ is the permittivity of the real-world vacuum in the limit of very weak fields but not in the limit of extremely intense fields. I don't see any reason to be bothered by that or any suggestion that SI authorities think differently.
• YOUR ARGUMENT 2 (as rephrased by me): "CODATA says ε₀=1/c₀²μ₀, not approximately equal but exactly equal. If ε₀, c₀, μ₀, were separately defined as three different parameters describing the real-world, experimentally-realizable vacuum (at least in principle), it would be impossible to say with certainty that they satisfy any exact relation."
• MY RESPONSE 2: OK, fine, we can take ε₀≡1/c₀²μ₀ to be a definition. (It is certainly the definition in the CODATA paper.) Then I would say μ₀=μ_{EverydayWorld} by definition, c₀=c_{EverydayWorld} by definition, and ε₀≈ε_{EverydayWorld} insofar as Maxwell's equations hold in the EverydayWorld limit, which is probably "they hold exactly in this limit", and certainly "they hold within parts-per-billion in this limit", and definitely not "they are 20% wrong in this limit". [Again, EverydayWorld means the real-world vacuum in the limit of weak fields, long distances, removal of every last particle, etc.] We can shift the debate, therefore, to whether μ₀=μ_{EverydayWorld} exactly by definition (which I believe), or whether μ₀=μ_{EverydayWorld}*1.2 (which is the value you would get based on the QED calculation of vacuum fluctuations; "the QED vacuum is diamagnetic, with relative magnetic permeability < 1", as you put it.)

So I'm sorry about the unnecessary diversion into ε₀. You can now please try to explain to me why "Steve, you would have to be crazy to believe that μ₀=μ_{EverydayWorld} exactly by definition!" and also why "There is nothing at all implausible or troubling in my belief that μ₀=μ_{EverydayWorld}*1.2." Again, for the latter, there is (1) the fact that all the engineers of the world have "voted with their feet" that μ₀ and μ_EW are not 20% different; (2) the fact that CODATA has explicitly endorsed measurements in the literature that use the assumption μ₀=μ_EW; (3) the fact that no one on earth has ever given a numerical estimate for the extremely-important ratio μ₀/μ_{EverydayWorld} except for my own estimate of 1.2 right here; etc. --Steve (talk) 21:33, 15 March 2012 (UTC)

Disambiguation link notification

Hi. When you recently edited Bra-ket notation, you added a link pointing to the disambiguation page Ket (check to confirm | fix with Dab solver). Such links are almost always unintended, since a disambiguation page is merely a list of "Did you mean..." article titles. Read the FAQ • Join us at the DPL WikiProject.

It's OK to remove this message. Also, to stop receiving these messages, follow these opt-out instructions. Thanks, DPL bot (talk) 11:18, 9 February 2012 (UTC)

The discussion continues

Steve: I wondered if you noticed this exchange. It would seem that this issue is not clear from the present intro, and is not understood even by Blackburne, who says the "exact value" is a consequence of advanced concepts from special relativity. Could it be that there is no English language formulation possible, or could it be that the climate on this page is so tempestuous that no agreement can be found? Brews ohare (talk) 18:05, 14 February 2012 (UTC)

I stopped watching that page a long time ago...was getting too time-consuming...maybe I'll look. I owe you a response on this page too but I've hit a very busy patch at work....well, that hasn't quite totally stopped me from procrastinating via wikipedia, but still, I'm trying to minimize it :-) --Steve (talk) 18:10, 14 February 2012 (UTC)

A point you raised...

on the WikiProject Physics talk page on March 15th, the new subsection

5.1 - {{template:Physics equations navbox}} "here to stay just because of user:F=q(E+v^B)"????

sparked an argument and then settled down, concerning the pointless {{template:physics equations navbox}}. I also find no use with the template. Just thought I'd let you know. Keep up the good + hard work, cheers, Maschen (talk) 17:54, 1 April 2012 (UTC)

I agree with your revert at Planck units...

... but Frank Wilczek used the term "Natural (Planck) units", for example. I don't like that particle physics uses that general term, particularly when their unit voltage is the SI volt, hardly a natural quantity to reference. A natural unit system is one that requires some definitive meaning: it should be possible for the aliens on the planet Zog can come up with the same system. There is no possibility that the Zoglings will come up with the eV for their unit energy. Planck units do have a special claim to "naturalness", moreso than the eV-based "natural units" of particle physics. And the article mentions that, but I also don't think it should be "a.k.a. Natural units". 71.169.179.168 (talk) 03:13, 25 April 2012 (UTC)

Glad you agree with the edit. Whether or not the particle-physics community is justified in calling something "natural units" when it has eV's is an unrelated issue. The more relevant issue here is that Zoglings may not be using exactly Planck units. Maybe they set h=1 instead of hbar=1, to take just one example. The Frank Wilczek quote is interesting, but I do think he's using "natural" as a describing adjective, rather than as a specific terminology...notwithsanding the capitalization. :-) --Steve (talk) 04:05, 25 April 2012 (UTC)

I consider it quite salient that particle-physics community calls the eV a "natural unit" of energy. It just isn't, and the name they attach to that system of units is not appropriate. Nonetheless, they use it.

So I'll emphasize a word A natural unit system is one that requires some definitive meaning: it should be possible for the aliens on the planet Zog to come up with the same system. Zoglings could choose to normalize ħ (or they could h), but they cannot choose to normalize the eV as the unit of energy. So whether it's ħ or h, it's a "natural" system. But it's not if it's the volt. BTW, I think that if the Zoglings come to a similar understanding of the laws of nature, they will be thinking about the ħ vs. h issue and will likely choose ħ. And I think they would normalize 4πG rather than G. But that, of course, is idle speculation. 71.169.179.168 (talk) 15:01, 26 April 2012 (UTC)

I don't think you'll find anyone who says "the eV is THE unit of energy in natural units". Once you set $\hbar =c=1$, then everything is expressed in powers of a single unit, but the phrase "natural units" does not speak to what that single unit is. Usually it is eV or keV or MeV etc. But if a particle physicist said

The particle's energy is 20 nanometers^-1 (by the way, I'm obviously using natural units here).

then I think he or she would be understood quite clearly and would not be saying anything unusual or incorrect. Again, I think the phrase "natural units", as it is used in particle physics, just refers to $\hbar =c=1$ and nothing else. eV is merely a common and conventional choice of unit that supplements "natural units", rather than being one of the natural units.

If you agree that "natural units" means $\hbar =c=1$ and nothing else, then I think your objection is based on the word "units" rather than the word "natural". Maybe you think it should be called "some natural units" so that it does not imply a complete system of units? :-) --Steve (talk) 15:25, 26 April 2012 (UTC)

I think any system of units must be complete in that any quantity can be expressed with that system. As best as I can tell, that means you don't need any mole or candela as long as you have length, time, mass, electric charge, and temperature or some other independent combination. Like you can eliminate either length, time, or mass from the list if you include energy or momentum in its stead (or like SI, you can use electric current instead of charge as a base unit). Temperature is largely considered a scaling factor applied to energy per particle.

So a system of natural units must define more than ħ and c as natural units. It must pick another universal quantity in nature (perhaps the mass or rest energy of an elementary particle if not G) just to get to the mechanical units, and it must pick up something regarding the EM interaction (perhaps e or ϵ₀) to define a natural unit of charge. And I cannot think of any other constant, other than the Boltzmann constant, k_B, to use for a natural definition of a temperature unit. If there is any unit defined by an anthropocentric physical quantity, the system that uses that unit is not a "natural system of units". 71.169.179.168 (talk) 17:11, 26 April 2012 (UTC)

Yea, I think that's why people usually say "natural units" rather than "natural system of units". There are exceptions. (Incidentally, this is an example where it is stated very clearly that the extra unit is not part of the phrase "natural units", and also an example of eV not being used.) here's another example where the word "system" is used like you say. But again, 90% of the time or more, the word "system" is not used. I agree with you that the word "system" should not be used. "System" implies "Complete system". :-) --Steve (talk) 17:37, 26 April 2012 (UTC)

Chemical potential

My draft is now available at http://en.wikipedia.org/wiki/User:Petergans/sandbox Please feel free to comment, amend etc. Watch out for typos, there may be lots of them which I don't see! Petergans (talk) 13:05, 2 May 2012 (UTC)

See User_talk:Petergans/sandbox :-) --Steve (talk) 13:54, 2 May 2012 (UTC)

Intolerable behaviour by new user:Hublolly

Hello. This message is being sent to inform you that there is currently a discussion at WP:ANI regarding the intolerable behaviour by new user:Hublolly. The thread is Intolerable behaviour by new user:Hublolly. Thank you.

I know you haven’t encountered him but see [1][2] for what I mean (I had to include you by WP:ANI guidelines, sorry...)

F = q(E+v×B)⇄ ∑_ic_i 07:27, 10 July 2012 (UTC)

Please desist from whipping up so much opposition. I apologized already. Hublolly (talk) 12:07, 10 July 2012 (UTC)

RE:Infoboxes

Hey! Thanks for your comments, I understand them completely. I agree with the whole "derivations" thing you were talking about - its helpfulness is limited, and I was merely trying to make it so that the infobox template itself was filled out as completely as possible. For the most part, I was attempting to add infoboxes to the articles about various physical quantities for the sake of organization. It is generally useful to have an infobox at the beginning of the page which says the name of the quantity, has an image, and proper SI units, symbols, etc. rather than just having nothing at the top of the page but the main template of all the concepts (that is, the entire Electromagnetism template or the entire Classical mechanics template).

However, just as you pointed out, we run into trouble with the "derivations" part. As you were getting to in your message, it may even be more prudent to leave that section out entirely. In short, I respect your critique and suggestions, and would add that you should edit the pages however you see fit. I just visited the Voltage article, and I agree that your edits to the infobox may have been an improvement.

Thank you!

--JSquish (talk) 19:32, 15 July 2012 (UTC)

Great! Yes, I certainly agree that it's better to have an on-topic image at the top of the article rather than one of those navigational templates, and it's great that you've been doing that.

If I were making the physical units template I would not have put "derivations" as one of the options. But I didn't make the template and it is awfully hard to change it now! :-) --Steve (talk) 20:22, 16 July 2012 (UTC)

spin-orbit interaction...

This sentence is from you, if I traced that correctly: In quantum physics, the spin-orbit interaction (also called spin-orbit effect or spin-orbit coupling) is any interaction of a particle's spin with its motion.

I agree, that there is a interaction a particle with its motion in a magnetic field, like in the described spin-orbit interaction effects, but I can not find that this is also called spin-orbit interaction in a case were you do not have an orbit. Can you give a book or article where this definition is used. Where is the orbit, if we just assume a free particle with magnetic moment and an electric field. --Do ut des (talk) 16:26, 28 August 2012 (UTC)

I know that the term "spin-orbit" is frequently used in solid-state physics (eg spintronics), where it is applied to valence electrons moving over mesoscopic distances. [3] [4] etc. etc. etc. These electrons are not "orbiting" anything.

For a particle in vacuum, I agree, the spin can affect motion, and I agree, nobody seems to calls that effect by the term "spin-orbit". You should change it....sorry for my mistake... --Steve (talk) 21:26, 28 August 2012 (UTC)

Quasiparticle

Hi,

why did you undo my version of "quasiparticles"?. If you think it is not clear, then you could put your own version of why they are different from "standard" particles because it is not clear from the article.

Rolancito (talk) 11:02, 21 October 2012 (UTC)

Thanks for asking, sorry I didn't leave you a note. I explained at Talk:Quasiparticle (scroll to the bottom).

You're right, it's a good idea to have some basic discussion that precedes the mention of many-body quantum mechanics. I will try based on your example, and you can let me know if I am at all successful :-) --Steve (talk) 15:28, 21 October 2012 (UTC)

OK I tried ... did that help at all??? --Steve (talk) 16:42, 21 October 2012 (UTC)

Thanks for your consideration, I answered your post in the talk page. Regards, Rolancito (talk) 16:24, 22 October 2012 (UTC)

Faraday's Law / Maxwell-Faraday eqn

You reverted my correction to the wrong statement "a changing magnetic field creates an electric field". I appreciate that my wording could be improved, but that is no excuse for retaining a mistake. If you want to revert, please provide (as I did) an explicit reference in the scientific literature supporting "a changing magnetic field creates an electric field"; i.e., one with an explicit causality-based reasoning, not just the statement itself. — Preceding unsigned comment added by 2001:630:12:10C6:217:8FF:FE2A:A8ED (talk) 09:01, 23 October 2012 (UTC)

I did not revert it. "Reverting" means undoing an edit word-for-word to restore what was there before. I didn't do that. Your reference convinced me that the article should not say that a changing magnetic field creates an electric field, therefore I did not put that statement back.

In my new version, the text neither asserts nor denies that either thing causes the other. It just doesn't discuss it. This is in accordance with standard textbook discussions, which do not analyze causality in detail. Indeed, it is possible to be an electrical engineer or physicist with an excellent working knowledge of Faraday's law, the ability to apply it in all situations to better understand the world, etc., despite having never considered the issue of which side of the equation is the cause and which side is the effect (in the sense of the paper that you cited). I know of no one who has seriously considered the issue before 2011 (the date of your citation), surely a sign that it is not vitally important to understand this issue. In your version, the causality analysis is discussed before even writing down what the equation is! Surely that's undue emphasis. We shouldn't distract from the most important points.

For the record, I don't think that the technical definition of "cause" used in the paper you cite is the only possible valid definition. I think it is used in a looser sense in everyday life, related to intentions etc. For example, when I rotate a bar magnet with my hand, Paul Kinsler would say: "You're creating a curl in the electric field, which in turn is changing the magnetic field." But I would say "My spinning the magnet is the cause of both the curl in the electric field and the changing magnetic field." It is strange and confusing (in my opinion) to imagine that the immediate cause of the changing magnetic field is anything but my decision to spin the magnet. I am not a philosopher, I am only trying to picture what a typical person would find confusing versus helpful.

I hope that clarifies what I was up to... Let me know your thoughts.  :-) --Steve (talk) 12:26, 23 October 2012 (UTC)

I misread the details of your change, for which I apologise.

I agree my change was arguably inelegant, but there is value in explicitly rebutting the (not uncommon, but wrong) view of cause and effect as contained in the original statement - which after all has persisted for some time without comment. Perhaps the clarification could put in a footnote?

As regards a failure to consider causality being typical of textbooks, that is certainly true in my experience; but why should wikipedia inherit that failing? The prior use of "creating" already strongly implied a certain (incorrect) cause and effect - if that was OK, why is a different (but now systematically justified) statement of cause and effect not OK? Note that the EJP paper is not claiming novelty in its approach to causality (see e.g. the existing wording of http://en.wikipedia.org/wiki/Causality#Engineering ).

In regards to "original causes", such as your argument where you decide to move a magnet, Faraday's law has no mechanism to include decisions to wave magnets about - it only has dB/dt = curl E. Thus within the narrow scope of Faraday's law, curl E is the cause of a temporal change in B - there is nothing else there to consider (although there might be in some putuative extension, to be later known as the Byrnes-Faraday Law :-) — Preceding unsigned comment added by 2001:630:12:10C6:217:8FF:FE2A:A8ED (talk) 17:10, 23 October 2012 (UTC)

I am working on an edit and I am struggling with the wording. Can you help?

This is in the context of a very basic introduction to Faraday's law, and how it applies to an electric generator. Last week I would have said something like

Rotating a bar magnet changes the magnetic field around it. According to Faraday's law, this changing magnetic field creates ("induces") an EMF in the wire.

Today, I'm not sure what to say. Maybe something like

If a bar magnet is rotated, the magnetic field around it changes. According to Faraday's law, a changing magnetic field is always accompanied by an EMF in the wire.

Or something like

Rotating a bar magnet changes the magnetic field around it. According to Faraday's law, this changing magnetic field creates ("induces") an EMF in the wire.[Note 1]

[Footnote 1:] Strictly speaking, it is incorrect to say that the magnetic field is the cause and the EMF is the effect. The rotating magnet is the true cause of both. It has been argued that it is valid to say that a curl in the electric field is the (proximal) cause of a changing magnetic field, but not the other way around. See Kinsler, P. (2011)....

I like the third best, although I imagine you would object to it. What do you think? Is there a better way?

The trouble is, it is intuitively obvious to everyone that rotating bar magnet should change the magnetic field, but it is not intuitive or obvious why a rotating bar magnet should directly create a curl in the electric field. --Steve (talk) 18:41, 23 October 2012 (UTC)

If you want to wave a bar magnet about, then you need to start with the source of the magetization (eg current loops). But since that seems a bit over the top, we might instead simply specify a time varying B(t). What changes might that induce? For "changes" (aka effects) we need a time derivative, so we see that the equation to use when we have a pre-specified B(t) is not dB/dt=curl E but instead dD(rt)/dt=curl B(rt), as we can then calculate induced changes in E field using E = D/epsilon. That is, we shouldn't use the Maxwell-Faraday eqn. at all, but Maxwell-Ampere (!) I suspect that in practice the typical interpretation works because the two curl Maxwell's eqns are coupled and support waves, thus in many cases a solution to one is a solution to the other.

It might be best to go with your (3) for the moment, but I'd strengthen the wording, e.g.:

Rotating a bar magnet changes the magnetic field around it. A common usage of Faraday's law in this situation is to say that this changing magnetic field creates ("induces") an EMF in the wire (but see[Note 1])

[Footnote 1:] Strictly speaking, the Faraday-Maxwell equation treats the electric field as a cause of a changing magnetic field; it is the Maxwell-Ampere's equation that has the magnetic field causing changes in the electric field (See Kinsler, P. (2011)....).

Although I think my footnote wording has the advantage of being more specific, it does beg the question as to why the existing interpretation is not right, even if in principle we imagine all things are answered in the EJP. And since that existing view is to an extent embedded in the rest of the article, it's hard to make a localised fix, as opposed to recasting the entire article in the more careful way. 2001:630:12:10C6:217:8FF:FE2A:A8ED (talk) 09:33, 25 October 2012 (UTC)

Hmm ... food for thought ... will respond later ... --Steve (talk) 12:24, 25 October 2012 (UTC)

This chain of events is quite strange. I start rotating the bar magnet. Rotating the bar magnet causes the magnetic field to change ("obviously" - or is this inference not allowed??) Now I am supposed to use Ampere's law:

$\nabla \times B=(1/c^{2}){\frac {\partial E}{\partial t}}$

The answer is supposed to be "it creates an EMF", i.e. the curl of E becomes nonzero.

Wait - I know Faraday's Law refers to curl E, but EMF is not required by definition to have a curl. Do we need curl E nonzero for the basic physical argument we want to follow? 2001:630:12:10C6:217:8FF:FE2A:A8ED (talk) 13:20, 5 November 2012 (UTC)

Let me try ... from Ampere's law:

${\frac {\partial }{\partial t}}(\nabla \times E)=c^{2}\nabla \times \nabla \times B=-c^{2}\nabla ^{2}B$

I want to argue that the left-hand-side is nonzero, so I suppose I should be arguing that the right-hand-side has to be nonzero. But why? It's not clear to me in any way!

Well, I find myself back where I started. If we know that one side of the Faraday's law equation is nonzero, then we can always infer that the other side is nonzero. If it is obvious from external events that one side has to be nonzero, we can speak loosely of that side "causing" the other side to be nonzero. Generations of physicists and engineers have imagined that a changing magnetic field "causes" a curl in the electric field, and never has it caused anyone to design a faulty transformer, or even come to an incorrect answer in a homework problem. Finally, now, when I tried to explain an extremely simple phenomenon in a way that rigorously follows the EJP cause-and-effect restrictions, I found that it was impossible (at least, too hard for me).

If you mean Faraday's Law / induction as the "extremely simple phenomenon" here, then the answer is that it's hard because you picked an example overloaded with a standard non-causal interpretation. If you want simple, then how about "acceleration causes a change in velocity"? Or would you like to say "velocity causes a change in acceleration"? 2001:630:12:10C6:217:8FF:FE2A:A8ED (talk) 13:51, 5 November 2012 (UTC)

So what's the point? I am back to thinking that we should acknowledge and even embrace the fact that people can use cause-and-effect in a loose, flexible way in their everyday lives and when learning and understanding physical phenomena. Another example: Talk to an expert analog circuit designer. When a resistor is driven by a current source, they will have no problem saying "there is a current through this resistor, which causes a voltage drop across it". When a resistor is driven by a voltage source, they will have no problem saying "there is a voltage drop across this resistor, which causes a current to flow through it". Again, they are describing logical relations. The EJP article is fun, but is not really relevant to anything, because it is discussing cause-and-effect in a highly specific, technical way, not related to how people discuss cause-and-effect in real life. That explains why humanity was extremely successful at using the laws of electromagnetism even before the EJP article was published in 2011. --Steve (talk) 14:24, 31 October 2012 (UTC)

(Why curl E? I am trying to explain the situation where a rotating bar magnet induces an EMF in a nearby stationary wire loop. Sorry I left out the word "stationary" above. An EMF in a stationary wire loop can only come from a curl in E.) --Steve (talk) 13:24, 5 November 2012 (UTC)

I have been following this conversation with interest, although I missed the beginning. I hope I can add something useful to the conversation. It is my understanding of Faraday's Law is that it does not imply any cause and effect (although in some cases, like a transformer you can force the value of one side in a particular way and the other side will follow along). Rather, what it means is that however the EM field is changed, it will be changed in a way that Faraday's Law is valid. But let me add something useful. In situations such as these, I find it useful to use the magnetic vector potential (the A field). If you can compute A, you just take its time derivative and you have the E field without even thinking about curl. You do have to integrate over all the currents, but there is this insight: a small element of current, dJ contributes to A in the same direction as the current. So, if all the current is in the X direction then A is in the X direction. Most of the time, currents follow a closed path; the current density field has curl. It is no surprise then that if the current density has curl then A would have curl also. And since E = -dA/dt (assuming that the charge distribution is everywhere zero) it is no surprise that E has curl.Constant314 (talk) 03:45, 6 November 2012 (UTC)