User talk:Martin Hogbin/Archive 5

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1 Archive 3 Archive 4 Archive 5 Archive 6 Archive 7 Archive 10

Since you participated in the last one, I'd appreciate your opinion in this final one aswell. Thanks, Nightw 12:56, 10 October 2011 (UTC)

Archiving assistance

Hi Martin, I noticed your talkpage is getting quite long. If you'd like, I could setup an archive bot which would automatically archive older threads (older than 90 days), and then you wouldn't have to worry about it anymore? Let me know, --Elonka 13:33, 11 October 2011 (UTC)

Elonka, is extraordinarily kind of you to offer to help me tidy my room but I thing I would prefer to do it manually. Some threads go on for a long time. Thanks again for the offer it is appreciated. Martin Hogbin (talk) 13:11, 13 October 2011 (UTC)
If you would like, I could easily change the date threshold to something else. For example, there are some threads here that have had no activity since January 2010. Just let me know where you'd like the cutoff to be (no activity in 6 months?) and I can set the bot accordingly. Another option is to manually post an advance date in some threads that you'd like to permanently "stick" to the page, and then the bot will leave those alone. --Elonka 15:26, 13 October 2011 (UTC)

comment on Tree Shaping Talk

Martin I have made a comment on the tree shaping talk page.?oygul (talk) 13:24, 7 November 2011 (UTC)

Talk:Pregnancy#RfC: Which photo should we use in the lead?

You are invited to join the discussion at Talk:Pregnancy#RfC: Which photo should we use in the lead?. You participated in the previous RFC on the lead image, Talk:Pregnancy/Archive 4#Lead image RfC. Nil Einne (talk) 15:00, 17 November 2011 (UTC)

Singular "they"

Thanks; that was interesting. Haven't decided whether to keep working on the issue or not. -- Jo3sampl (talk) 01:20, 26 November 2011 (UTC)

The problem is that most people have a concept of 'correct' English which is based on what they were taught between the ages of 5 and 18. Language is not as simple as that. Much of the grammar we were taught was manufactured by the Victorians and has never formed part of written or spoken English. Martin Hogbin (talk) 12:41, 26 November 2011 (UTC)

Vacuum

Hi Martin: I wonder what your opinion is about vacuum as specified by BIPM? Do you have doubt that they mean it to be the reference medium with electromagnetic properties dictated by c, c0, ε0, μ0? I agree that this reference medium can be described in other units as well, for example units where the relative permittivity and relative permeability are identically unity for all frequencies and wavelenghts. But even in such units that means that the reference medium still has no capacity to recognize vacuum fluctuations or virtual particles, which lead to relative permittivity and relative permeability that depart from unity to a degree, and to observed behaviors like the Casimir effect and the Lamb shift. Whatever the system of units, the reference state is there with defined properties that render it distinct from quantum vacuum, don't you agree? Brews ohare (talk) 19:17, 3 December 2011 (UTC)

I do not think it is for us to say what BIPM mean by vacuum if they choose not to say themselves. The whole subject is complicated and bulk properties such as μ0 and even speeds are not well defined in the quantum domain. When there is a real and practical need to specify what is meant by a vacuum in order to better define the metre, I an sure that the relevant standards bodies, guided by experts in the subject, will do so. Martin Hogbin (talk) 21:45, 3 December 2011 (UTC)
Hi Martin: I don't think that is quite the point: Of course the standards organizations will adapt to whatever comes along. But for the moment they have settled upon a reference state defined by parameters c, c0, ε0, μ0. For any experimental vacuum, including whatever turns out ultimately to characterize a "quantum vacuum", the present approach would be the same as for, say, air. You just measure the actual permittivity and permeability and present it relative to the standard reference vacuum. That is, there is no need to worry whether the reference vacuum is obtainable experimentally (it is not realizable even in principle, though at the moment any difference between the reference vacuum and quantum vacuum is found only indirectly through the Lamb shift or the Casimir effect, which don't exist even theoretically in the reference vacuum). It seems likely to me that the only thing that would cause a shift to a new standard is if it were found that some realizable vacuum were actually easier to use than making the corrections to refer to the present c, c0, ε0, μ0 standard. Don't you think that is so? Brews ohare (talk) 23:36, 3 December 2011 (UTC)
Do you have any evidence that the EM constants are measured or calculated when making allowance for the refractive index of the actual medium used for realisations of the metre. I would have thought that the refractive index would have been used directly. Martin Hogbin (talk) 12:48, 4 December 2011 (UTC)
Martin: The refractive index of air is found using an input vacuum wavelength and then calculating the various contributions that introduce a departure from ideal vacuum, based upon measured partial pressures of CO2, H2O, and so forth. Here is how NIST does it. However, this topic appears to me to be a digression, no? Brews ohare (talk) 13:13, 4 December 2011 (UTC)
That is more or less as I thought. The digression is the involvement of the EM constants; they seem to play no direct part in the calculation. Whether the ideal vacuum is considered a classical vacuum or a quantum vacuum is irrelevant; there are no quantum theories which suggest an abrupt change in bulk refractive index when an ideal vacuum is approached. Martin Hogbin (talk) 14:06, 4 December 2011 (UTC)
Martin: From an experimental approach, the refractive index is not the best way to decide whether experimentally one can distinguish between the c, c0, ε0, μ0 vacuum and a field-theoretic vacuum. These effects are very small: see John F. Donoghue, Eugene Golowich, Barry R. Holstein (1994). Dynamics of the standard model. Cambridge University Press. p. 47. ISBN 0521476526.{{cite book}}: CS1 maint: multiple names: authors list (link) The distinction is based more clearly upon the prediction of the Casimir effect, or the Lamb shift, or other such experiments. If one uses a model of c, c0, ε0, μ0 vacuum the prediction is that these effects do not exist, because the c, c0, ε0, μ0 vacuum does not allow for virtual particles or vacuum fluctuations. Thus, the c, c0, ε0, μ0 vacuum does not agree with these experimental observations. On the other hand, the vacuum of quantum electrodynamics does predict these effects. Thus, the c, c0, ε0, μ0 vacuum is an idealization not observed in experiment, but still useful as a reference state because theory is able to calculate corrections to this reference state that allow experimental observations to be predicted. Brews ohare (talk) 15:01, 4 December 2011 (UTC)
As you say, the practical difference between the two vacua is very small and in terms of the speed of light and the realisation of the metre it is insignificant compared to current experimental uncertainties. When metrology reaches the state where the difference in vacua starts to become important then the standards authorities will have to agree in exactly what vacuum they mean, just like they may have to specify other factors that turn out to be important. You may have an opinion on what would be the best choice but until the type of vacuum is specified by the standards authorities we cannot specify one in WP. Martin Hogbin (talk) 17:18, 4 December 2011 (UTC)
Well Martin, maybe we have to leave things like this. However, I don't understand your position, and it is unsourced. It appears to me that when the BIPM and NIST have specified that the electromagnetic properties of what they call "vacuum" are c, c0, ε0, μ0 there is simply no room for a more precise specification. That also is the view of Weiglhofer & Lakhtakia, who take the reference state of classical vacuum as one with ε0, μ0 and state further "It is important to realize that the vacuum is the only "medium" for which this identity in form holds" referring to the constitutive relations that use ε0, μ0. Other sources that agree are Cardarelli, Fleisch and many more. You may be a minority of only one person. Brews ohare (talk) 20:21, 4 December 2011 (UTC)
On the contrary,it is your view that is unsourced, unless you can find a source that says specifically that the vacuum mentioned in the BIPM definition of the metre is a classical vacuum, in those actual words. Martin Hogbin (talk) 22:02, 4 December 2011 (UTC)
Martin, you choose to ignore the BIPM and NIST postings at c, c0, ε0, μ0 which clearly refer to what they call vacuum, and what Weiglhofer & Lakhtakia call "classical vacuum" , which also has of course the properties c, c0, ε0, μ0. The adjective "classical" is appended, not to describe something different, but to separate it from QCD vacuum and the like, as you well know. Weiglhofer & Lakhtakia also say in their section The classical vacuum as reference medium "It is important to realize that the vacuum is the only "medium" for which this identity in form holds", referring to the constitutive relations involving ε0, μ0.
So, Martin, the situation is very clear: you haven't entertained the above facts, and you haven't any sources to support your view that there is ambiguity in whether BIPM and NIST have adopted vacuum as the c, c0, ε0, μ0 vacuum, or may have deliberately left unspecified whether the reference medium is QCD vacuum or some other field-theoretic vacuum. In other words, according to your view, we don't know whether the reference vacuum today is one that does not allow the Lamb shift, namely the c, c0, ε0, μ0 vacuum, or a different reference state that does allow the Lamb shift. In your view, NIST and BIPM have left this decision to await later developments. The Lamb shift has been known for over 50 years so, from your viewpoint, one might query what the standards agencies are waiting for.
However, the BIPM and NIST in actual fact have indeed made the choice of the c, c0, ε0, μ0 vacuum, as indicated on their web sites by the choice of electromagnetic properties for their vacuum of c, c0, ε0, μ0, and the secondary property Z0, "the characteristic impedance of vacuum". Various secondary sources make the same identification.
It's my suggestion that you reconsider this matter: if you have doubts, you might ask yourself what is the meaning of the c, c0, ε0, μ0 postings, and why do Weiglhofer & Lakhtakia refer to a vacuum with c, c0, ε0, μ0 as a reference medium? Brews ohare (talk) 19:04, 7 December 2011 (UTC)
You have missed the point Brews. Whatever you think ought to be the name used by BIPM is not important. The name they do use is 'vacuum'. Indeed only one, rather obscure, source even mentions the term 'classical vacuum'. Martin Hogbin (talk) 22:35, 7 December 2011 (UTC)

outdent Martin, sorry if I misunderstood you. I thought your remark "When there is a real and practical need to specify what is meant by a vacuum in order to better define the metre, I am sure that the relevant standards bodies, guided by experts in the subject, will do so." indicated that you felt the term "vacuum" was somehow vague. But apparently you have no problem with the BIPM and NIST term "vacuum" referring to the medium with c, c0, ε0, μ0. That being so, we agree. Brews ohare (talk) 23:58, 7 December 2011 (UTC)

My point was simpler than that; BIPM and NIST just use the term 'vacuum', so we should use exactly the same term in WP. Any ambiguity or imprecision (which I am sure is incredibly small) caused by the use of the unqualified term 'vacuum' is currently inherent in the standard. In is not up to us to interpret what they really meant or to try to reduce any ambiguity which we may see in the terms used.Martin Hogbin (talk) 16:54, 8 December 2011 (UTC)
Martin, while in common usage the term "vacuum" is vague indeed, the imprecision in the BIPM and NIST usage of the word "vacuum" is zero. They have specified to infinite precision (by using definitions) the values c, c0, ε0, μ0 as the electromagnetic properties of their "vacuum", and there isn't any room to wonder what they "really mean", or struggle to "reduce any ambiguity". In any event, assuming you agree that their vacuum uses c, c0, ε0, μ0, we don't need to argue over the implications of these choices. Brews ohare (talk) 16:12, 9 December 2011 (UTC)
Because the word "vacuum" in everyday use (for example, as understood by many WP readers) is vague, the link vacuum should not be used when vacuum in electromagnetism really is meant, as is the case in defining the metre. So I am happy you have agreed to use of the latter link. Brews ohare (talk) 16:33, 9 December 2011 (UTC)
You may wish to read this discussion. I find it odd that you would change the links on the page metre without any Talk page comment there, when we actually agreed about this earlier. And why no engagement here in this discussion on your Talk page, which you seemingly have abandoned? Brews ohare (talk) 16:48, 10 December 2011 (UTC)
Martin, what happened to agreement about links? Brews ohare (talk) 14:23, 11 December 2011 (UTC)
Sorry Brews (I also apologised in my edit summary) but I did not notice that you had linked to a specific section of the Vacuum article. Martin Hogbin (talk) 16:51, 11 December 2011 (UTC)

Fairground MHP

You are at a fairground where a game is played for small prizes you are asked this question:

Suppose you're on a game show, and you're given the choice of three doors: Behind one door is a prize; behind the others, nothing. You pick a door, say No. 1, and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?"

The host has an assistant who plays the game occasionally to drum up business but he wins every time. I have some inside information and I am told that the prize is placed randomly and the host chooses which door to open, within the rules, randomly. I have also been told that the the assistant has no information about the placement of the prize. How does he win every time? Martin Hogbin (talk) 13:23, 4 December 2011 (UTC)

I hope this will not distract you from the above discussion. Anyway: you did not say whether the assistant swaps choice every time? And, you say you've been told he ha no info. Does he indeed, at the moment of decision, have no information about the position of the price, other than within the game rules? If yes, then it is sheer luck, because I then exclude him being clairvoyant or tipped off in some way, i.e. by some secret agreement with the host about the way he wears his spectacles, or the words he uses when offering the switch, etc,Nijdam (talk) 21:14, 4 December 2011 (UTC)

You obviously know what I am getting at, which is that the words spoken could easily, in some circumstances, be used to give information about the location of the prize. You consider this unlikely, or at least not justified by the wording of the question, but you must agree that it is a possibility that might, in some circumstances need to be taken into account. Martin Hogbin (talk) 21:21, 4 December 2011 (UTC)

Yeah, yeah. And may be Whitaker was just lying about the car, or there were in fact 20 doors, or just one door. Or his aunt just had her period, and so on. This all has nothing to do with the MHP. Nijdam (talk) 23:35, 4 December 2011 (UTC)

Yes indeed, those any many other questions might need to be answered in a real life problem. An over-simplistic, 'mathematics-tells-us-so' approach to real life problems can have serious consequences. Martin Hogbin (talk) 22:53, 7 December 2011 (UTC)


Exam question

There are three boxes, numbered 1,2 and 3. One of the boxes contains a price, the others are empty. The price is put at random (uniformly, to avoi any misunderstanding) in one of these three boxes. You are unknown with the position of the price and you may choose one of the boxes and have what's in it. After you choose one of the boxes but before you open it, I open one of the not chosen boxes, showing it is empty. This game may be repeated, and I will always open an empty box after you have made your initial choice. If There are two empty boxes for me to choose from, I open one of them at random (uniformly ...).

  1. Give a relevant sample space.
  2. What is the probability the price is in box No. 1?
  3. If you did choose box No. 1, what then is the probability the price is in box No. 1?
  4. If you did choose box No. 1 and I opened box No. 3 showing it is empty, what then is the probability the price is in box No. 1?

In each case not only give the desired answer, give the derivation as well and the relation with the sample space.

Please Martin, care to solve it?Nijdam (talk) 10:32, 5 December 2011 (UTC)

That looks like the question we answered at User:Martin_Hogbin/Monty_Hall_analysis
I don't know, I'm looking forward to your solution. Nijdam (talk) 16:53, 8 December 2011 (UTC)

As you describe this as an exam question and the doors are numbered, I imagine that I am expected to chose a sample space with 18 elements of the type (PHN) where P is the door number hiding the prize, H is the door I choose and N is the door you open, with the elements not permitted by your rules being excluded.

The probability that the prize is behind door 1 is 1/3, whether or not I choose box 1. Am I on the right track? Martin Hogbin (talk) 00:13, 11 December 2011 (UTC)

I presume you mean box, where you say door. It's not about what I expect, but what you think. I take your answer to point 1 to be your mentioned 18 elements. Your answer to point 2 is, I guess; P(P=1)=1/3. Please formulate how you arrive at this conclusion and formulate the answers to the other points; use the sample space. Nijdam (talk) 17:50, 11 December 2011 (UTC)

It might be easier if you explained what you are getting at. The important point from my perspective is that you described this as an exam question, this means that I make certain assumptions. Martin Hogbin (talk) 21:01, 11 December 2011 (UTC)

Which assumptions did you make? And as this is supposed to be an exam question, it's up to you.Nijdam (talk) 22:01, 11 December 2011 (UTC)

Because you call it an exam question, I assume that all the information given is relevant and expected to be used in the answer and that no important information is missing.

To clarify P(P=1) = 1/3 at the start

I then set up the sample space, as described, using the given uniform distributions then condition it for the box originally chosen and the box you opened and conclude that

P(P=1|H=1) = 1/3

P(P=1|H=1,N=3)=1/3

I have calculated a conditional probability just the way you would like me to (I presume). Martin Hogbin (talk) 22:19, 11 December 2011 (UTC)

How precisely does this problem differ from the MHP?Nijdam (talk) 08:57, 12 December 2011 (UTC)

In the context, as I gave said all along. As I say above, I have chosen to assume, 'all the information given is relevant and expected to be used in the answer and that no important information is missing', because you described this as an exam question. Because this is an exam question I can determine the precise intent of the questioner. In other contexts this might not be so easy.

Here are the sentences one by one, and numbered for your convenience:
  1. There are three boxes, numbered 1,2 and 3.
  2. One of the boxes contains a price, the others are empty.
  3. T he price is put at random (uniformly, to avoi any misunderstanding) in one of these three boxes.
  4. You are unknown with the position of the price and you may choose one of the boxes and have what's in it.
  5. After you choose one of the boxes but before you open it, I open one of the not chosen boxes, showing it is empty.
  6. This game may be repeated, and I will always open an empty box after you have made your initial choice.
  7. If There are two empty boxes for me to choose from, I open one of them at random (uniformly ...).
Mention the numbers of the sentences which in your opinion differ form the MHP, and motivate your answer.

Let me ask you about Whitaker's question. Do you believe that Whitaker was interested in specific doors? Do you believe that he wanted vos Savant to consider the case that the host opens door 2 to reveal a goat to be somehow different from the case where the host opens door 3? What exactly do you think Whitaker wanted to know? Do you bilieve that Whitaker was not aware of the obvious symmetry with respect to door number? Martin Hogbin (talk) 22:23, 12 December 2011 (UTC)

A lot of questions again, but first it is your turn to respond to the question above. Nijdam (talk) 22:40, 12 December 2011 (UTC)

I have responded to your question. The way this question differs from the MHP is the context in which the natural language question is asked. Asked by you as an 'exam question' I am happy to formulate it mathematically in the way that you think correct. When the same, or a rather more vague, natural language question is asked in a popular magazine, I would formulate it differently (using the sample space {car, goat}). Martin Hogbin (talk) 14:40, 13 December 2011 (UTC)

Nijdam, you have not responded to my comment above. Are you overwhelmed with my stupidity or are you just taking a break? Martin Hogbin (talk) 10:35, 16 December 2011 (UTC)

Please Martin, mention the numbers where you think the MHP differs from my exam question. Nijdam (talk) 12:13, 20 December 2011 (UTC)

The question asked by Whitaker differs considerably from your question, which is far less vague (I can indicate them all if you wish) but that is not my main point, which is that even the same question may be understood differently depending on the context. Martin Hogbin (talk) 19:31, 20 December 2011 (UTC)

I do not wish you indicate them all, however, is it your opinion they all differ? Nijdam (talk) 07:09, 21 December 2011 (UTC)

There probably is some kind of difference between every one of your statements and the statement of Whitaker. This is to be expected as Whitaker was not an expert on the subject and did not make clear exactly what he wanted to know. You statements, on the other hand, are clearer and in the given context I have a much better idea exactly what question you require me to answer. Martin Hogbin (talk) 10:03, 21 December 2011 (UTC)

Tell me the difference, pointwise. Nijdam (talk) 23:27, 22 December 2011 (UTC)

Nijdam, you are missing my point. Even if the two descriptions were exactly the same I might formalise them differently according to their contexts. Martin Hogbin (talk) 23:31, 22 December 2011 (UTC)

Okay, tell me which points have the same formulation and what the difference is in the context.Nijdam (talk) 10:56, 23 December 2011 (UTC)

If the problem was an exam question (in particular one written by you) in which door numbers chosen and opened were given, I would assume that the door numbers were intended to be significant and therefore to be used in the mathematical formulation of the problem. I would therefore use the 18-element sample space.

If, on the other hand, door numbers were not given, or were given just to clarify how the game worked, and the context was that of a simple mathematical puzzle, I would assume that the door numbers were not significant and use the {Goat, Car} sample space. Martin Hogbin (talk) 11:53, 23 December 2011 (UTC)

That's not an answer. And what's more. it's meaningless what you say. You said such before, and couldn't explain to what the door numbers were not significant. That's why i go point for point through the experiment. So Martin, take point 1 for a start and tell me the difference, if there is one. Nijdam (talk) 21:18, 23 December 2011 (UTC)

Sorry Nijdam but I have got a little lost here.

This is the setup?

  1. There are three boxes, numbered 1,2 and 3.
  2. One of the boxes contains a price, the others are empty.
  3. The price is put at random (uniformly, to avoi any misunderstanding) in one of these three boxes.
  4. You are unknown with the position of the price and you may choose one of the boxes and have what's in it.
  5. After you choose one of the boxes but before you open it, I open one of the not chosen boxes, showing it is empty.
  6. This game may be repeated, and I will always open an empty box after you have made your initial choice.
  7. If There are two empty boxes for me to choose from, I open one of them at random (uniformly ...).

What question do you want me to answer? Martin Hogbin (talk) 22:35, 23 December 2011 (UTC)

I will answer this one:

If you did choose box No. 1 and I opened box No. 3 showing it is empty, what then is the probability the price is in box No. 1?

As an exam question set by Nijdam

I have just been studying sample spaces and have been advised to start with a sample space that includes all the facts given in the question. I assume that all that is required to give an answer is provided in the question.

1 As the boxes are numbered I shall use the box numbers to set up my sample space. 3 The probability that the prize is in box 1 is 1/3, and boxes 2 and 3. My sample space must include the event that the prize is in each box.
4 My sample space must include the event that I choose any of the boxes.
5 My sample space must include the event that you open an unchosen empty box.
7 I can assign equal probability to all the above events.

I therefore set up a sample space with 18 elements and condition it appropriately, greatly pleasing by teacher.

As a simple puzzle

Maybe some of the data given is irrelevant. Perhaps there are some valid shortcuts that make the puzzle easier to explain. Symmetries perhaps, they are very powerful and intuitive.

1 There are three boxes. Do the numbers matter. Let us see.
3 The probability that the prize is in any given box is 1/3
4 So far the problem is symmetrical with respect to box number. There is nothing to distinguish any box from any other thus the box numbers are thus far irrelevant.
5 Nothing still distinguishes the boxes. The only important thing is what is in the box.
7 Box number is confirmed to be unimportant.

I therefore set up a simple sample space containing everything that matters, what is in the box I initially pick. It contains only two elements {Prize -1/3, Nothing - 2/3} Martin Hogbin (talk) 15:14, 24 December 2011 (UTC)

As the original question

Please concentrate on the question. Let me spell it out again. I take point 1: 1 There are three boxes.

Does this make a difference whether you consider this an exam question or as the simple MHP puzzle?? that's the simple question I like you to answer. Is it that difficult?Nijdam (talk) 19:16, 24 December 2011 (UTC)

Clarity of the term 'vacuum' in electromagnetism

Transcription from Talk:Metre:

To Martin: I agree that BIPM and NIST have deliberately not "qualified" their use of the term 'vacuum', I suspect because the historical usage of 'vacuum' in the context of SI units will define it by implication. One might, however, inquire whether a general reader of WP will understand that very specific context without a little guidance. Although it is possibly uncharitable of myself, I do subscribe to the (possibly erroneous) idea that you are resisting changes here less because of the espoused reasons than because you fundamentally wish to avoid any indication that vacuum in the context of SI units refers to Vacuum#In electromagnetism. If I am mistaken about this, please tell me so. Brews ohare (talk) 17:00, 11 December 2011 (UTC)

If there were any ambiguity in NIST's and BIPM's use of the unqualified term 'vacuum', their use of the term is made crystal clear with their web-site postings of c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0. Do you agree? Brews ohare (talk) 17:12, 11 December 2011 (UTC)

However clear you may think it is that BIPM meant something different from that which they wrote I think we must use the actual words they used in the standard.
My reasons for this are not so much that I actually disagree with you but that I think you are jumping to conclusions. I am happy to discuss this on my talk page. Martin Hogbin (talk) 21:18, 11 December 2011 (UTC)

Continued discussion: Martin: Can you spell out for me (i) the statements I've made that you do agree with me about, and also (ii) my statements that amount to "jumping to conclusions"? Brews ohare (talk) 21:59, 11 December 2011 (UTC)

We seem to agree that BIPM and NIST have deliberately not "qualified" their use of the term 'vacuum'.
OK Brews ohare (talk) 00:58, 12 December 2011 (UTC)
I disagree with you that BIPM intend to specify any particular type of vacuum. The reason for this is that the expected difference between the SOL in different vacua is insignificant compared to current experimental uncertainties. They do not need to address what might turn out to be the very difficult question of which kind of vacuum to use because they simply do not need to.
I agree that they do not have to specify any real vacuum, but that is because they chose a different path. They have defined, instead of a realizable vacuum, an ideal reference state, not realizable even in principle, to which measurement of any real vacuum's properties may be referred by measuring its refractive index (for example) or its relative permeability (for example). Consequently, no matter what is discovered about the vacuum of electrodynamics or QCD vacuum, those discoveries can be related to the reference vacuum, which remains aloof from all that. All this is exactly the same as for a medium like air, and is done the same way. You will recall the NIST calculator for refractive index of air, which takes as input the vacuum wavelength, c0/f. Brews ohare (talk) 00:58, 12 December 2011 (UTC)
The EM values you mention are based on the SOL in vacuum so your reference to a vacuum in which these constants have particular values is a tautology not physics or metrology. Experimentally c and the EM constants are the same in all vacua.
I agree that the speed of light in Vacuum#In electromagnetism is not physics. It refers to a idealized reference state, unobtainable in the laboratory, even in principle. For one thing, Vacuum#In electromagnetism has defined properties, not measured ones. For another, Vacuum#In electromagnetism fails to agree with some observed phenomena, for example, all realizable vacua must allow for vacuum fluctuations and virtual particles. These behaviors have verified experimental consequences, like the Lamb shift, and lead to electromagnetic properties of all real vacuums (whatever else may be discovered about them) that differ from the defined ones of Vacuum#In electromagnetism. These differences are experimentally small, but the point isn't their size but that they exist in real vacua, while even in principle they play no role in Vacuum#In electromagnetism.
I do not agree that metrology plays no role, and would say in contrast that metrology is the primary concern of BIPM and NIST and so is the underlying objective in their choice of definition for a reference state. That is, this reference state is intended to make measurements simpler and easier to compare with one another by basing them upon the departure of the measured system from the universally recognized reference state of 'vacuum'.
The final statement that "c and the EM constants are the same in all vacua" is incorrect, and really beyond debate. For example, you may recall that the speed of light in a medium is c/n, and n is not identically one for all wavelengths and all frequencies and all field strengths and all polarizations and all directions in any medium other than the reference state of Vacuum#In electromagnetism. If you need references on this subject, check the ones out I have provided elsewhere.
When QM starts to become significant in the realisation of the metre the standards authorities will have to use the best theory of the day to decide exactly which vacuum they wish to use. This is how things work in metrology. Until it matters it does not matter. Martin Hogbin (talk) 22:34, 11 December 2011 (UTC)
This idea is not properly framed. The definition as the distance light travels in 'vacuum' in 1/299,792,458 seconds is independent, of quantum mechanics, of course. The realization of the second may change with advances in atomic physics, like frozen atoms and so forth. That changes the metre, but not the reference vacuum which remains a medium with defined electromagnetic properties that have no relation to measurement.

To avoid meandering about between the many topics above, I propose a series of sub-headers where individual discussions can be segregated. Here is an example starting point: Brews ohare (talk) 16:58, 12 December 2011 (UTC)

To begin

I'd suggest that we can lay to rest quite quickly the statement "c and the EM constants are the same in all vacua", so that might be a good starting point. The relative permeability and permittivity of field-theoretic vacuums is described in Kurt Gottfried, Victor Frederick Weisskopf (1986). Concepts of particle physics, Volume 2. Oxford University Press. p. 389. ISBN 0195033930. and more recently in John F. Donoghue, Eugene Golowich, Barry R. Holstein (1994). Dynamics of the standard model. Cambridge University Press. p. 47. ISBN 0521476526.{{cite book}}: CS1 maint: multiple names: authors list (link) and also R. Keith Ellis, W. J. Stirling, B. R. Webber (2003). QCD and collider physics. Cambridge University Press. pp. 27–29. ISBN 0521545897. Returning to the vacuum of a relativistic field theory, we find that both paramagnetic and diamagnetic contributions are present.{{cite book}}: CS1 maint: multiple names: authors list (link) Brews ohare (talk) 16:44, 12 December 2011 (UTC)

My conclusion is simply that the beginning statement is incorrect. If you wish, it can be revised to say "c and the EM constants are nearly the same in all vacua". Brews ohare (talk) 17:01, 12 December 2011 (UTC)

Please note that I said, ' Experimentally c and the EM constants are the same in all vacua'. In other words there are no experiments that can determine the difference in c and the EM constants over a distance of a metre. By how much do you think c varies between different vacua, measured over a distance of a metre? Martin Hogbin (talk) 22:30, 12 December 2011 (UTC)
Fine, Martin. This point is settled. Let's go on with the next one. Brews ohare (talk) 23:12, 12 December 2011 (UTC)
You have not answered my question. Martin Hogbin (talk) 09:25, 13 December 2011 (UTC)

BIPM does not need to specify a particular vacuum because of small experimental differences between them

This is an interesting discussion. In my mind, experiment plays no role, whether the differences are large or small. Now vacuum#In electromagnetism very clearly uses defined electromagnetic properties, which clearly indicates experiment is not a factor. Quite simply, although definitions can be changed under practical duress, so long as a definition is in place, the properties are determined to infinite decimal points, unlike any measured quantity. So the above header appears to suggest that, in your view, the 'vacuum' of BIPM used to define the metre is in fact not vacuum#In electromagnetism, although, in your view, their 'vacuum' applies to any vacuum that is compatible with vacuum#In electromagnetism until better experiments come along. Is that your view? Brews ohare (talk) 17:12, 12 December 2011 (UTC)

A somewhat different version is that 'vacuum' in the BIPM definition of the metre is a stand-in that is expected to survive any future developments, but for the moment it is provisionally selected to be a defined entity vacuum#In electromagnetism, but without any unyielding commitment to this form of vacuum. Thus, although today there is no difference between 'vacuum' and vacuum#In electromagnetism, the term 'vacuum' is used in a more inclusive sense that envisions the possibility that one day 'vacuum' will not be vacuum#In electromagnetism. Possibly a change will be motivated by advances in measurement techniques, but of course, metrology has to weigh a variety of practical considerations (for example, reproducibility and availability of the standard), so experimental advances are not the only thing entering a decision to change. How's that? Brews ohare (talk) 17:20, 12 December 2011 (UTC)

The first formulation has to address what is the role of the postings of c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0. The second formulation has no problem with these postings, as they are the 'vacuum' of the moment. Brews ohare (talk) 17:27, 12 December 2011 (UTC)

The problem is that there may be an infinite number of things that might affect the speed of light most of them, as yet, undiscovered so if your objection is the purely philosophical one that we have defined an exact speed in an inexact medium then this is a problem that we will always be stuck with. Nevertheless, if you do agree to have an exact speed in an inexact medium, nothing bad happens.
When experimental uncertainties become comparable to the theoretical differences in c between different vacua over a distance of a meter then we will need to define our vacuum more precisely. Until that time, I take the BIPM wording to mean 'any kind of vacuum you like'. This is how standards have always worked. For example, when the temperature of the caesium atoms used to define the second started to become a significant aspect of measurement uncertainty, a defined temperature was added to the standard. In fact it was a theoretical and unobtainable temperature.
It is the same with the vacuum. When we can measure the difference over a distance of a metre, we will need to specify exactly what kind of vacuum we mean. It might be one of several types of unobtainable theoretical vacuum depending on what is considered to be the most fundamental according to the theory of the day. Martin Hogbin (talk) 22:50, 12 December 2011 (UTC)
Martin, this discussion of measurement seems to suggest you are adopting the first of the two choices I outlined above. Is that correct? If so, why not the second one, which seems to me the simpler of the two because it avoids all this discussion entirely? Brews ohare (talk) 23:15, 12 December 2011 (UTC)
You seemingly will not address the question of what role is played by the postings of the exact values c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0. It seems to me that if BIPM wished to suggest that any vacuum with properties within ±Δ of these properties is 'vacuum' within their definition, then they would have done that. What is your explanation for using definitions in these postings: how are they supposed to be used in practice? Brews ohare (talk) 03:48, 13 December 2011 (UTC)
I assume, like you, that the exact values of the EM constants might be considered to apply to the same notional type of vacuum as the exact value of c but BIPM have not specified what type of vacuum this is. We could do our own OR and by looking at other references make our own decision as to what type of vacuum this ought to be but that, in my opinion would be wrong.
When we are close to being able to measure a difference between different vacua over a distance of a metre the standards authorities, along with theoretical experts, will have to decide on the best treatment for the whole subject. We cannot predict what this may be. It could be, for example, that Maxwell's equations need to change. Maybe an additional constant will be added, breaking the strict relation between the current EM constants and c or maybe a more complex change will be required.
At the moment we have some exact constants defined in inexact circumstance. Let us leave it like that until the experts decide it needs to change. Martin Hogbin (talk) 11:57, 13 December 2011 (UTC)
You have said that one might "assume" that the posted μ0 and the subsidiary derived properties with the exact values ε0 and Z0 apply to "the same notional type of vacuum" as that in which the defined value c, c0 applies. Is that an assumption you make yourself, or do you feel there is some alternative, equally likely assumption? If so, what is it? Brews ohare (talk) 15:35, 13 December 2011 (UTC)
You have said that " BIPM have not specified what type of vacuum this is", which I would take as meaning that they haven't spelled out just how one might prepare this "notional vacuum" in a laboratory. My stance on this, which is not OR at all, is that any laboratory vacuum will exhibit such experimentally established observations as the Lamb shift and spontaneous emission. Such phenomena result from interactions with virtual particles and vacuum fluctuations that cannot occur in the "notional vacuum" because its specified parameters ε0 and μ0 are incompatible in theory with these phenomena. These experiments cannot be explained using the "notional vacuum". In other words, the "notional vacuum" is fictitious, or differently described, ideal. Brews ohare (talk) 15:43, 13 December 2011 (UTC)
The "notional vacuum" is not useless, even though it is not realizable in a laboratory, because it can be used as a reference state. So, for example, one can calculate a "vacuum wavelength" as c0/f using a measured frequency, even though this vacuum wavelength is in principle not observable in a laboratory. Then one can calculate an observable wavelength by putting in the properties of a realizable vacuum, in the manner of the NIST calculator. Brews ohare (talk) 15:58, 13 December 2011 (UTC)
The electromagnetic properties of a medium, such as air, are related to the reference state using various formulas. These in turn are established by an extended process of verification, for example, the polarizability and permeability of molecules, which are verified by a host of experiments, such as optical absorption spectra, chemical behavior, etc. This smorgasbord of experiments establishing a formula is not restricted to measurements of the speed of light. Brews ohare (talk) 16:55, 13 December 2011 (UTC)
I have not traced the steps behind the formulas in the NIST calculator, although references are provided. In any event, any theoretical attempt to arrive at these formulas for comparison of theory and experiment will introduce the reference state because all the theoretical models are based upon Maxwell's equations with source terms that are taken to indicate the departure from the reference state. Brews ohare (talk) 17:05, 13 December 2011 (UTC)
There seems to me to be the possibility that there exist two reference states: (i) the "experimental" reference state underlying the experimental procedure establishing that (for example) the formulas in the NIST calculator indeed describe a departure from the reference state (maybe, for example, an assumption that the reference vacuum corresponds to zero partial pressure), and (ii) the "theoretical" reference state established by the theoretical procedure employing Maxwell's equations with sources that cause departure from their reference state. Maybe that concern is what you are worried about? So far as I can tell, this possible incompatibility of "experimental" and "theoretical" versions of the reference vacuum is not a concern of physicists. Brews ohare (talk) 17:26, 13 December 2011 (UTC)
In any event, for the moment the vacuum#In electromagnetism is taken as the reference state for experiment and for theory. Maybe one day, as you may already have said, a theoretical connection between vacuum#In electromagnetism and the experimental procedure for establishing departures from this state will be elaborated and will introduce corrections to a simple assumption that the reference state corresponds experimentally to (for example) zero partial pressure. Brews ohare (talk) 18:12, 13 December 2011 (UTC)
Martin, please excuse me if the above reads like a lecture. I have difficulty in conveying a conversational tone on a Talk page. In any event, I do appreciate your willingness to engage in such an extensive back and forth. Brews ohare (talk) 18:27, 13 December 2011 (UTC)

You seem to go in to a lot of pointless detail. The fact is that we have exact values in an inexact medium. The reason the medium is not defined any more exactly is because it makes no measurable difference. Martin Hogbin (talk) 23:29, 13 December 2011 (UTC)

Disappointed that you find the above pointless, Martin, possibly a simple consequence of not reading it carefully? I made a concerted effort to explain matters from what I see as an accurate standpoint. That is, the "fact" is that we have an exactly defined but not physically realizable reference vacuum with defined electromagnetic properties c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0. Measurement of these properties is not a consideration, inasmuch as they are defined. I think your dismissal of the points described above is cursory, and unfortunate. Brews ohare (talk) 01:56, 14 December 2011 (UTC)
Your notion that "we will need to define our vacuum more precisely" on the occasion "when we can measure the difference [between different choices for vacuum] over a distance of a metre" is not correct. Improvements in measurements might establish that (for example) QCD vacuum was a good model for a vacuum prepared in some fashion, but that development would not force a change from the present reference vacuum, because any advance in the measured properties of real vacuum is easily expressed as an improved accuracy in measuring its departure from the reference vacuum. Theoretical descriptions of QCD vacuum and other field-theoretic vacuums already are expressed in this fashion, relative to the reference vacuum. Brews ohare (talk) 15:16, 14 December 2011 (UTC)
To elaborate with an example, reference vacuum is independent of frequency and wavelength. Thus, any measurement of a real vacuum that demonstrates such a dependence of that vacuum can be described as a frequency-and-wavelength dependent departure from the featureless reference vacuum. The use of a departure from reference vacuum is exactly how media that are not vacuums are described today, both theoretically and experimentally. Such theoretical and experimental departures are the basis of relative permittivity and relative permeability, which are identically unity by definition in the reference vacuum. Brews ohare (talk) 15:43, 14 December 2011 (UTC)


If you are saying that the vacuum in which c and the EM constants have their defined exact values could be defined as 'that vacuum in which c and the EM constant have their defined and exact values' then I would be hard pushed to disagree with you but this is not physics, it is just a tautology. Exactly what else this should be taken to imply about the vacuum is not clear.

A similar case has already occurred in relation to the definition of the meter. How gravitation should be taken into account was not clear at one time. This was clarified not by specifying a type of vacuum but by consideration of the meaning of the word length' itself. As it happens the necessary theory and terminology was already in existence and the issue was resolved by defining the metre to be a unit of proper length.

In the case of QM effects we do not yet have the necessary theory to fully resolve this issue but luckily the practical effects are well below current measurement uncertainties. My point is that you are preempting decisions that may need to be made at some time in the future but which have not, in fact, yet been made. Martin Hogbin (talk) 10:34, 15 December 2011 (UTC)

Martin: These remarks of yours are not pertinent to anything I have said above. My remarks seem to have passed you by in their entirety. Your commentary about measurements and their progress are not relevant to the use of vacuum#In electromagnetism as the reference vacuum for electromagnetic properties of all media, including whatever may be learned about particular vacuum states and the effects of gravity upon them. And speculations about what future choices of reference state might be made by the BIPM or its related agencies are simply speculations: what is in hand is the understanding of what is done today. Brews ohare (talk) 16:38, 15 December 2011 (UTC)
To make a crude analogy about the role of a reference state, the Ten Commandments can be used as a basis for judging human behavior, regardless of one's particular beliefs about their status or origin in the scheme of things. All that is necessary is to be able to judge whether a particular behavior is consonant with the standard or not, or, in the case of the reference medium, to be able to make an estimate of how far and in what way any particular medium departs from the standard. Brews ohare (talk) 16:49, 15 December 2011 (UTC)

Brews, I cannot really understand what you are getting at. All I can say is that there is disagreement (at least between you and me) about the meaning of the term 'vacuum' in the BIPM definition of the meter so we are best to stick closely to the wording used in the definition rather that to try to interpret it however obvious you may think your interpretation is. Martin Hogbin (talk) 10:06, 16 December 2011 (UTC)

I wouldn't characterize your incomprehension of my remarks as disagreement exactly; maybe as disconcerting. Brews ohare (talk) 16:06, 16 December 2011 (UTC)
In any event, a WP reader is very likely to find reference on WP to the posted c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0 and can draw the obvious conclusion that this is what BIPM and NIST think of as the properties of what they term 'vacuum'. If they have doubts about this connection they have multiple secondary sources, for example, that draw this conclusion for them. Brews ohare (talk) 16:37, 16 December 2011 (UTC).
My only clue to your reservations in this matter seems to be a belief of yours that the vacuum in which the standard metre is measured has to be real (which is undoubtedly true) and that vacuum has to be the 'vacuum' referred to by the BIPM (which is not true, this 'vacuum' is a reference state and is not a real vacuum at all). The standard meter can be measured as wavelengths in air, if one wishes, and the departure from the reference state is then calculated using the NIST refractive index calculator. Or, it could be measured in partial vacuum and the corresponding corrections for that medium could be used. (These corrections will evolve, of course, as measurement techniques improve.) In any event there is no need at all for the standard metre to be measured in the reference vacuum and, in fact, that is an impossibility as the reference vacuum is a fictitious unrealizable ideal reference. Brews ohare (talk) 17:14, 16 December 2011 (UTC)
Another example of an unrealizable reference datum is absolute zero which is the starting point for the Kelvin scale and shares with reference vacuum the similarity that it is unobtainable. Brews ohare (talk) 17:25, 16 December 2011 (UTC)

Brews, I have stated my opinion as clearly as I can and you do not seem to understand what I am getting at. You have stated your opinion as clearly as you can and I do not understand what you are getting at. I do not see any point in continuing. Martin Hogbin (talk) 17:50, 16 December 2011 (UTC)

Am I to understand that my assessment that you wish to have the metre measured in 'vacuum' is incorrect? I understand that the "standard metre" as established in some medium is found only to within a measurement error, described by NIST as dominated by errors in finding partial pressures, and not to errors in the theoretical formulas used. So, if you like, the "standard metre" is itself unrealizable, because any attempts to reproduce it are subject to measurement errors in characterizing the medium, and to various errors of interferometry. Brews ohare (talk) 17:55, 16 December 2011 (UTC)
In any event, if you feel we cannot identify what the differences between us actually are, so be it. Brews ohare (talk) 18:55, 16 December 2011 (UTC)

My thoughts and opinions

You have said here on Talk:Metre that "I am not going to get involved in an edit war but it looks to me as though Brews is adding his own personal thoughts and opinions to this article (despite the citations)."

You may recognize that this comment lacks any specifics about exactly what constitute my "personal thoughts and opinions" that I have added to this article, "despite the citations".

Frankly, I find this kind of oblique approach to criticism in the vaguest of terms to be reprehensible. It is accusation from the shadows without exposure to fact. If you have something to say, say it clearly and specifically! Brews ohare (talk) 17:29, 17 December 2011 (UTC)

Brews, I do not want to have this discussion all over again on the Metre article talk page. You seem to want to make some point about the speed of light definition of the metre but I cannot understand what it is. I can see no encyclopedic purpose in the material that you have added but rather than us arguing endlessly about it let us see what others think. Martin Hogbin (talk) 18:19, 17 December 2011 (UTC)
Martin, how can we expect others to express their ideas about what you think when you have not said what it is you object to, but have only registered an nonspecific vague reservation? Basically, you are asking others to come up with and crystallize objections that you cannot formulate yourself. Or, maybe you are rallying the troops?
In contrast to your remarks immediately above, I am making no comment whatsoever about "some point about the speed of light definition of the metre". In my opinion, this notion of yours is a fixation that causes any comment to be refracted through this prism: it is an idée fixe of yours. Brews ohare (talk) 19:39, 17 December 2011 (UTC)
What has been said about the metre is simply that its definition in BIPM vacuum can be related to laboratory approximations of the metre that employ real media, interferometers, and frequency measurements. All these real world impediments introduce measurement uncertainties in any laboratory attempt to fashion the standard metre, and thus an exact version of the ideal standard metre of the definition cannot be verified in a lab. I'd say that the description of these practical uncertainties is in no way a commentary upon the nature, value, or advisability of the definition of the metre, despite your reaction to the contrary. Brews ohare (talk) 19:56, 17 December 2011 (UTC)
Realisations (or delineations) of the metre will always imperfect and subject to uncertainties and if that is what you want to say that is fine with me but why not use the standard metrological terminology rather that inventing your own terms like the 'standard metre' and 'laboratory approximations of the metre'. Martin Hogbin (talk) 21:52, 18 December 2011 (UTC)
Martin: I am glad we are now on the same page. I made some changes to play down the term "standard metre" as you have suggested. Brews ohare (talk) 17:41, 19 December 2011 (UTC)
I have made some further changes to simplify and used more standard terminology. Martin Hogbin (talk) 11:54, 23 December 2011 (UTC)
Hi Martin. These changes seem OK. Now you have entertained the notion of realization of the metre, which involves use of a real medium. I believe that concept involves recognition that any realization of the medium with defined electromagnetic properties c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0 will involve uncertainties, which involve among other problems the impossibility of avoiding vacuum fluctuations and virtual particles. I'd guess you are OK with that too? Brews ohare (talk) 14:16, 23 December 2011 (UTC)
A concomitant of the uncertainties in realizing this idealized medium with defined properties is that no realization of the defined speed of light c0 is free of uncertainty either. Brews ohare (talk) 14:28, 23 December 2011 (UTC)
I am sorry but I do not understand what you are getting at. If you are happy with what I have don so far I will make some more changes, mainly to avoid repetition if the same facts. Let me know what you think. Martin Hogbin (talk) 17:28, 23 December 2011 (UTC)

outdent Hi Martin: I was wondering whether the notion of "realization" might illuminate our discussion of the medium with defined electromagnetic properties c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0? That is, we might discuss this medium from the stance of trying to effect its realization.

In particular, the specification of this medium is a matter separate from its realization. Hence, the discussion of a reference state with defined properties is released from some of our digressions, which in this light now appear to be a discussion of the realization of this state, and not a discussion of the reference state itself. Brews ohare (talk) 17:20, 24 December 2011 (UTC)

To pick one point in particular, the defined speed of light c0 is one characteristic of this ideal medium. Then realizations of the metre, or observations of the measured speed of light in some real medium, can be discussed in terms of that real medium being an approximate realization of this ideal medium. Brews ohare (talk) 17:30, 24 December 2011 (UTC)

You are back to your tautology again. The medium in which the light has the defined exact speed in metres per second is indeed a medium in which light has a defined exact speed on metres per second. I cannot see what other point you are making. Martin Hogbin (talk) 22:14, 24 December 2011 (UTC)
Martin: try to look at things from a different angle, the angle that the properties of any real medium can be related to that of an idealized medium with defined c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0. This comparison is effected through the relative permeability μ ⁄ μ0 and relative permittivity ε ⁄ ε0 of that medium. That can be done for partial vacuum, for example, a medium with c=c0/n, n the refractive index. If one is deliberately trying to reduce n to n=1 by pumping down the partial vacuum, one may view the sequence of partial vacuums at any particular stage of pumping down as better and better realizations of an ideal medium, an ideal medium where it is actually the case that c=c0. I take this as a pretty conventional viewpoint.
What may appear less conventional to you is that this viewpoint sets up the idealized medium with defined c, c0, μ0 and the subsidiary derived properties with the exact values ε0 and Z0 as a "reference" medium, and removes any requirement that this medium be "real": it is just a reference point, like absolute zero, perhaps. Consequently, our previous discussion in which you speculated that BIPM chose this standard only because experiment was too inaccurate to tell the difference between this medium and, say, QCD vacuum becomes unnecessary: this reference medium need not be closely related to any real vacuum, any more than, say, it is related to a ferroelectric, because the real vacuum and the ferroelectric both can be characterized relative to the standard medium via their particular values for the ratios μ ⁄ μ0 and ε ⁄ ε0. Improvements in measurement do not affect the standard's μ0 and ε0. Rather, the more accurate numbers show up as better values of μ ⁄ μ0 and ε ⁄ ε0. Brews ohare (talk) 23:08, 24 December 2011 (UTC)