Talk:International System of Units/Archives/05/2020

Who has the ultimate authority over the SI?

A previous version of the text stated that The ultimate authority over the SI rests with the General Conference on Weights and Measures (CGPM). In a revision of 22:49, 22 April 2020, this has been modified to read just Authority over the SI rests with the General Conference on Weights and Measures…, with a note that "ultimate" is a strong word; the CGPM reports to higher authorities (collective governments).

I am certainly not an expert, but I do believe the note is incorrect. The CGPM does not report to anyone or anything, not in the sense in which e.g. the CIPM reports to the CGPM. In particular, the decisions of the CGPM, at least as far as the SI, are both binding and final, by which I mean that they can only be changed by another action of the CGPM. No Member State, or even a collection of them, or even all of them together, can override CGPM decisions except by going through the CGPM itself: in other words, except by convincing other Member States to instruct their delegates (to the CGPM) to pass another resolution (of the CGPM), and it is that new resolution (of the CGPM) which overrides the original decision (by the CGPM).

^{It is of course up to the individual Member States to implement the SI domestically (or not). On the other hand, it seems to me, based on the text of the Metre Convention (in particular Art. 7 of the Annex), that probably no Member State can even unilaterally pass legislation that e.g. redefines the metre on its territory to something other than what the latest relevant CGPM resolution decreed. In order to be able to do that, a Member State would first have to leave the Metre Convention, i.e. cease to be a Member State.} --Reuqr (talk) 17:51, 28 April 2020 (UTC)

How many soldiers does the CGPM have? Jc3s5h (talk) 18:30, 28 April 2020 (UTC)

My note was not correct to say "the CGPM reports to higher authorities (collective governments)", but it is subject to the collective vote of the delegates of the member states. The definition of a delegate is pretty much someone who is tasked to act as a representative, and hence to vote in accordance with a state's instructions. Like "unique", "ultimate" is a word often abused and becomes very slippery. Let me pose another example: Is the government the ultimate authority of a country? If so, it is not subject to the wishes or control of anyone, yet it would be a perversion of democracy to suggest that a non-autocratic government, once constituted, is the ultimate authority over the country, even though the voters fit the (adapted) description No [voter], or even a collection of them, or even all of them together, can override government's decisions – at least within the constitution. For example, the CGPM can be retired/made redundant through a collective decision to abandon the Metre Convention made externally to the CGPM, and for argument's sake the SI placed under the control of another body. English words have slippery semantics, and when these are marginal, these words are best avoided. Perhaps "highest central authority"? —Quondum 18:59, 28 April 2020 (UTC)

To be sure, the counterexample with the government in a democracy is well taken. But I am inclined to bite the bullet here, and say that a democratically elected government from that example is indeed the ultimate authority with respect to legal matters. Consider the following two examples. 1. On the UK Parliament's website, we read that the Parliament is the supreme legal authority in the UK. It would seem that the word supreme would be subject to the same objections as the word ultimate. 2. Similarly, on the US Supreme Court's website, we read that the Court is the final arbiter of the law. In this case, too, one could problematize that proposition, given that the US Constitution can be amended. And yet, I think that neither of those statements is misleading. And I think that the statement that the CGPM is the ultimate (or supreme, or final) authority over the SI isn't misleading, either; and I suspect that the reasons why these three statements aren't misleading are probably very similar. I am not sure I will be able to pinpoint these reasons on the first try, but let me make an attempt. I suspect that the issues of ultimate authority are normally—setting theological contexts aside—evaluated relative to a system/framework of decision-making. I think it is normally understood that the system itself could conceivably be changed. So, I would say that, relative to the present system of decision-making as far as the SI, the CGPM is indeed the ultimate authority. I'm not sure how best to phrase that for the purposes of the article, but I don't think that central quite captures it. --Reuqr (talk) 20:08, 28 April 2020 (UTC)

Maybe it is just a personal thing, but for me, "supreme" or "final" work better than "ultimate". I'm one of those people who cringe at phrases like "very unique". Even better would be to use a source for phrasing this claim, lest we be at risk of editorializing. —Quondum 21:29, 28 April 2020 (UTC)

Of course, that would be ideal. I'll look around. --Reuqr (talk) 22:58, 28 April 2020 (UTC)

I've changed the phrasing to that used in 73 FR 28432 (‘Interpretation of the International System of Units (the Metric System of Measurement) for the United States’):

The SI was established and is maintained by the General Conference on Weights and Measures.

--Reuqr (talk) 17:43, 14 May 2020 (UTC)

Looks good – it is informative and sounds right. —Quondum 20:56, 14 May 2020 (UTC)

QES metric system

I noticed that one of the references (Carron, pp. 16–19) had quite a bit of information on the QES system, and suggested that it played a role in the development of the SI. This would mean that not all metric systems had the metre as a decimal multiple, since this one had it as a decimal submultiple. —Quondum 22:39, 20 May 2020 (UTC)

The text now says The metre is recognised in all of them, either as the base unit of length or as a decimal multiple or submultiple of the base unit of length. --Reuqr (talk) 22:24, 23 May 2020 (UTC)

Cool. I should have saved you the trouble and made that edit myself; I guess my thinking was that we should capture this source and its information about QES somewhere. I'll take a look at doing so at Metric system. —Quondum 23:00, 23 May 2020 (UTC)

Subtleties about quantities in different unit systems

"The problem is that, in general, one and the same electromagnetic physical quantity will have not only different unis in the CGS-ESU, CGS-EMU, and SI systems, but even different dimensions." This wording unfortunately perpetuates a confused way of thinking about unit systems that seems to permeate articles on unit systems in WP, but the modern architects of the SI do not seem to fall into this trap. The International System of Quantities (ISQ), notwithstanding its incompletely defined nature, gives unambiguous relationships between several quantities and is the basis for SI. If I say a circle has a dimension-x equal to 1 m in system A and 2π m in system B, I am effectively talking nonsense. You might recognize that if I am talking about the radius in the first case, then I am likely talking about the circumference in the second case, and thus that I am talking about distinct quantities and just confusing things by giving them the same name. Just because I use the same name to refer to an equally good way of sizing a circle does not mean that it is, in any rigorous sense, the same quantity. This is exactly what happens with the term "electric field strength" in SI and say the Gaussian system: the same name is being applied to distinct quantities, albeit related by a constant of proportionality. This insight may help in choosing better words. —Quondum 00:49, 17 May 2020 (UTC)

My instinct is to agree, but I guess this depends on the definitions of "quantity" and "dimension". My view is that any sensible definition of these terms ought to lead to the dimensions of a given quantity being independent of the selected unit system. Dondervogel 2 (talk) 05:53, 17 May 2020 (UTC)

My view too; this is why I'm always baffled when I see the three different ways electromagnetic quantities used to be expressed with different dimensions in the CGS system, depending on whether, IIRC, 4πε₀, or μ₀, or neither, were posited equal to 1 and thus dimensionless. I never thought the CGS electromagnetic units were "sensible", but when I first heard about them I was a mere high school pupil and the people using them were respected physicists so I didn't argue. Happily my physics teacher used MKSA, as it was called in the Sixties. — Tonymec (talk) 06:12, 17 May 2020 (UTC)

"the modern architects of the SI do not seem to fall into this trap" – I take that back. I was trying to find the papers (maybe by I M Mills?) that had an excellent analysis of quantities and units and made a lot of sense. However, I see recent papers by metrologists generally and Peter Mohr in particular that seem to be as confused about this as is the WP community, or at least contain as much disagreement. —Quondum 00:28, 18 May 2020 (UTC)

I've changed that note so that it now reads as follows:

The problem is that, in general, the physical quantities that go by the same name and play the same role in the CGS-ESU, CGS-EMU, and SI systems—e.g. ‘electric charge’, ‘electric field strength’, etc.—do not merely have different unis in the three systems; technically speaking, they are actually different physical quantities. Consider ‘electric charge’, which in each of the three systems can be identified as the quantity two instances of which enter in the numerator of Coulomb's law (as that law is written in each system). This identification produces three different physical quantities: the ‘CGS-ESU charge’, the ‘CGS-EMU charge’, and the ‘SI charge’. They even have different dimensions when expressed in terms of the base dimensions: mass^1/2 × length^3/2 × time⁻¹ for the CGS-ESU charge, mass^1/2 × length^1/2 for the CGS-EMU charge, and current × time for the SI charge (where, in the SI, the dimension of current is independent of those of mass, length, and time). On the other hand, these three quantities are clearly quantifying the same underlying physical phenomenon. Thus, we say not that ‘one abcoulomb equals ten coulomb’, but rather that ‘one abcoulomb corresponds to ten coulomb’, written as 1 abC ≘ 10 C. By that we mean, ‘if the CGS-EMU electric charge is measured to have the magnitude of 1 abC, then the SI electric charge will have the magnitude of 10 C’.

How is that? --Reuqr (talk) 21:53, 18 May 2020 (UTC)

Nicely written, and captures the spirit of my thesis. I'm not sure I have been able to get the same core argument (different physical quantities) from the references though. For example, IEC 80000-6:2008 says nothing aside from using the symbol '≘' Unicode U+2258 CORRESPONDS TO (without apparently defining it), which is merely a hint that the quantities are not the same, and that we should not say 1 G = 10⁻⁴ T. "Babel of Units" has a long explanation, but a scan didn't quite get a direct mention of "different physical quantities", though I probably just missed it. Nevertheless, I like it as it stands. —Quondum 23:20, 18 May 2020 (UTC)

Yes, the revised text is well worded. Great job! Dondervogel 2 (talk) 23:28, 18 May 2020 (UTC)

Having just said "we should not say 1 G = 10⁻⁴ T", I see we have in the article "(e.g. 1 gauss = 1×10⁻⁴ tesla)". —Quondum 23:38, 18 May 2020 (UTC)

To the following, I added notes explaining that '=' should really be '≘' : "1 abampere = 1 decaampere," "1 abhenry = 1 nanohenry," and "1 gauss = 10⁻⁴ tesla."

As far as references supporting the claim that we have three different physical quantities, explicit statements to that effect appear in the Page reference. Thus, I added two quotes to the references to that paper:

1. “Since the corresponding equations in different systems are not identical, the symbols in them must represent (slightly) different quantities, i.e. different mathematical models of invariant physical phenomena.”
2. “The precise statement involved in converting from emu to SI (the modern International System) is as follows. A magnetic field described by H = 1 Oe is also described by H = 1000/4π A/m. These statements describe the same phenomenon in terms of different quantities.”

These quotes appear as tooltips when one hovers the mouse pointer over the page number, like so:

…, they are actually different physical quantities.

Unfortunately, in tooltips, I don't know how to achieve italics other than by resorting to Unicode characters; here is what happens if one tries to do it using markup: , , .

While I was at it, I also added this quote:

3. “Similarly, it is incorrect to state ‘one abampere equals ten amperes’, ‘one tesla is ten kilogauss’, etc.; the correct verb is corresponds to.”

Here is how it appears as a tooltip: ‘… ten coulomb’, …
The symbol ‘≘’ is indeed never defined in IEC 80000-6:2008. Carron ("Babel of Units") says this:

… the IEC has used⁷² the notational symbol ≘ to mean ‘corresponds to’…
_{⁷²For example, IEC Report 80000-6 (2008); email from Joanna Goodwin of the IEC, April 16, 2014. The importance of “corresponds to” rather than “equals” has been emphasized by Page [Pa70a].}

Here [Pa70a] is Ref. .

Your resourcefulness in getting italics into the tooltips is impressive. Since we seem to have no better way, I like that approach. I wonder whether a character-substitution template would work.

I would suggest that we should do as we say. For example: "1 abampere ≘ 1 decaampere". —Quondum 00:17, 20 May 2020 (UTC)

Done; all those now use the ‘≘’ sign, with an explanatory note. --Reuqr (talk) 21:26, 20 May 2020 (UTC)

Apropos nothing much, I'm appalled at the 8th SI Brochure. From the 1st to the 7th Brochure, the maxwell vs. weber and gauss vs. tesla are considered to have different dimensions ("This unit is part of the so-called “electromagnetic” three-dimensional CGS system and cannot strictly be compared with the corresponding unit of the International System, which has four dimensions when only mechanical and electric quantities are considered. For this reason, this unit is linked to the SI unit using the mathematical symbol for “corresponds to” (^)."), but in the 8th they just use an equality sign (while keeping the correspondence only of the oersted with the ampere per metre, and that only because of a rationalization issue: "These units are part of the so-called “electromagnetic” three-dimensional CGS system based on unrationalized quantity equations, and must be compared with care to the corresponding unit of the International System which is based on rationalized equations involving four dimensions and four quantities for electromagnetic theory. The magnetic flux, Φ, and the magnetic flux density, B, are defined by similar equations in the CGS system and the SI, so that the corresponding units can be related as in the table. However, the unrationalized magnetic field, H (unrationalized) = 4π × H (rationalized). The equivalence symbol ≙ is used to indicate that when H (unrationalized) = 1 Oe, H (rationalized) = (10³/4π) A m⁻¹."). —Quondum 23:32, 29 May 2020 (UTC)

References