Wikipedia talk:Manual of Style/Dates and numbers/Archive 86

Archive 80

←

Archive 84

Archive 85

Archive 86

Archive 87

Archive 88

→

Archive 90

Suggestion for addition to the unit section: artillery calibres

Resolved

 – Three points in the guideline already address this, as noted below.

I have stumbled on cases where French battleships were said to have N-inch guns; while the measure might be more or less accurate, the French do not use imperial units, and the guns are named "M-mm gun" (just like it'd be accurate but ridiculous to measure the calibre of the guns of a US battleship in furlongs).

I think that the same is true of German, Italian, Russian and Japanese ships.

We might want to consider setting a rule that calibres must be given in the unit featured in the official name of the gun — hence the Jean Bart had 380 mm (15 inch) guns, the USS Massachusetts had 16 inch (406 mm), the Yamato had 460 mm (18.1 inch) guns, the Vanguard had 15 in /42 (380 mm) guns, the Tirpitz had 380 mm (15 in) guns, and so forth. Rama 10:25, 20 September 2007 (UTC)

We do not need a special guideline for that. Just be bold and do it. In the unlikely event of an objection, just quote the existing guidelines:

• For US-related articles, the main units are US units
• For other country-related articles, the main units are metric
• ... put the source value first ...

I would suggest that millimeters can remain unconverted in military applications, since they appear to be universal. Lightmouse 10:34, 20 September 2007 (UTC)

Yep; and it's probably good to spell out "mm" as the main unit (see MOSNUM on this). Tony (talk) 10:43, 20 September 2007 (UTC)

I would like to note a little nitpick: the official names of WWI and WWII German guns were generally in centimeters, not millimeters, so Tirpitz carried 38 cm guns. TomTheHand 14:28, 20 September 2007 (UTC)

Modern artillery is indeed universally metric, but this is definitely not the historical case. Even into WWII, the British often described artillery by projectile weight (e.g., 25-pounder on a tank), and this is nearly universal historical practice from the mid-19th century back. Calibers did appear often, but not universally, during the American Civil War. Again, I come back to where you said For US-related articles, the main units are US units, and ask again if you are thinking US versus everyone else. The Battle of Trafalgar, for example, was fought by French and British ships, all of which described their guns and carronades in projectile weight, not metric and not Imperial caliber. At the battle of the Spanish Armada (actually involving multiple countries and principalities), the flagships Ark Royal and São Martinho measured their batteries in pounds. Much of the Dreadnought revolution, of which British Admiral "Jacky" Fisher was the enfant terrible, used calibers of inches to refer to generations of capital ship batteries.

My fundamental point is that as much as you might like a universal rule, you aren't going to get one and be historically accurate. Please leave some of this to the judgment of writers. Bluntly, simply knowing metric diameters of shells doesn't give solid information alone: a US shell referred to as 16"/45 (406mm), the latter being the length of the barrel in diameters, was indeed a 406mm, but had more penetrating power than a Japanese 460mm (18.1 inch). The US has a series of guns that are 5"/127mm, but there is a significant difference in performance based on barrel length: 5"/38, 5"/54, 5"/62, and 5"/74. I've never seen the barrel lengths metricated, although I suppose they could be -- it's more of a dimensionless constant.

Note that I put the units but In WWII, while the Soviets were mostly metricated, they made a point (attributed to Stalin himself) that each caliber in millimeters would be different, to avoid the Anglo-American problem of ordering a shipment of 75mm ammunition and getting naval shells instead of light artillery shells. Eventually, there was a technical reason why both a howitzer and rocket needed to be 122mm, so the rockets were always called Katyushkas (and GRAD in their successor), never a caliber, to avoid confusion Howard C. Berkowitz 15:41, 20 September 2007 (UTC)

Points raised by others (cm, inches, pounds, non-US) seem reasonable. However, please note that my comment was confined to values expressed in *mm* within the source data. Lightmouse 17:34, 20 September 2007 (UTC)

Well... I'm not sure that I agree that a caliber in mm doesn't require an English/Imperial/U.S. conversion, for two reasons. First, while even the U.S. Army uses gun calibers that are defined in metric, many average Americans will better understand the size of the shell if they see it in inches as well. Second, English units will very often help to explain where the heck such odd calibers came from. For example, 7.62 mm? Why, that's 0.30 inches. 152 mm artillery? Six-inch guns. TomTheHand 18:27, 20 September 2007 (UTC)

Seconded. It's no burden to convert each value, once or twice, in article, in either direction. Metric is currently almost universal for this, but articles should stick primarily with source information, be it in pounds, inches, or centimetres. Sacxpert 09:37, 1 October 2007 (UTC)

Create list of suggested currency abbreviations?

[Discussion moved over from WT:MOS as off-topic there.]

I was reading MOS to see how to format Canadian currency, for an article on tips. WP:$ currently has two examples of dollars, in passing; Australian as AU$1 and U.S. as US$1. If interpreted to recommend two capital letters followed by a dollar sign, then Canadian would be CA$, but the Canadian dollar article says the Canadian government and IMF favor C$, while Editing Canadian English advises $Can or $CDN, and its only mention of CA$ is in software packages. My best guess would be to use C$, but whatever is chosen, it seems that a list of recommended currency abbreviations would be helpful in the MOS. If the XX$ format is recommended for dollars, New Zealand NZ$ and Jamaican JA$ are reasonably accepted, but Bahamas or Barbados (both beginning with "ba") also use dollars. ISO's two-letter country codes (that link also includes ISO's standard three-letter currency names) suggest BS and BB as country codes, so BS$ and BB$ (even if never used before) would be a way to follow the two-letter pattern, but B$ and Bds$ are suggested in Bahamian dollar and Barbadian dollar. Three-character ISO currency codes (BSD and BBD) could somehow be used, but that's inconsistent with using US$ and AU$. MOS suggests linking the symbol to a page about an obscure currency (₮146), which helps clarify the intent in any case. Prospect News' style guide has a list of their own recommended currency abbreviations, as a point of comparison. -Agyle 20:56, 19 September 2007 (UTC)

So why not be bold and create the list? Then we have a place to begin discussion of specific currencies. --JodyB _{yak, yak, yak} 11:58, 20 September 2007 (UTC)

I'd prefer using the ISO 3 character currency code (CAD) as being the definitive unambiguous version. If use of symbols is desired, I would combine with either 2 or 3 letter country codes as $CA or $CAN. $CDN doesn't lead to any extensible scheme to name world currencies. Mayalld 13:35, 20 September 2007 (UTC)

Jody, point taken. :-) I'll try to do that. Any guidance on approach (e.g. Mayalld's opinion) would be helpful. -Agyle 17:32, 20 September 2007 (UTC)

I like the ISO codes too. Inasmuch as that is used by the banking community, or so says the list, it would be an appropriate guide. --JodyB _{yak, yak, yak} 18:39, 20 September 2007 (UTC)

Some comments: First, "$US" isn't used by much of anyone, and should be advised against. "US$" is more readily understood by most readers than "USD", which many readers will not even recognize as a currency indicator if they do not work in the financial industry. Just because someone somewhere for some context has come up with a standard does not mean that the standard should be applied in all situations (really clear case in point: we write "Juana Ortiz was born in Mexico" not "...in MX", despite MX being the ISO 2-letter code for Mexico, and that system of 2-letter codes is a broadly accepted international standard. — SMcCandlish [talk] [cont] ‹(-¿-)› 01:15, 21 September 2007 (UTC)

I find it immensly odd that the wikipedia "standard" is to make up an abbreviation that you think the reader will understand, and only if one is not readily to hand or may be ambiguous, then fall back to the international standard three-letter codes. Many people will already have seen those codes anytime they book a flight, or do currency conversion. The ISO lists are readily available and WP should use that standard. Let the standards bodies manitain the list at their expense too. anon 2007-09-21 01:30 UTC

Huh? That's what I'm arguing against ("$US", "$C", "£UK", etc. are neologistic silliness). If you are referring to "US$", "C$", "GB£", etc., these are not Wikipedia inventions but common parlance; as far as I can tell so far there are used everywhere except in the banking industry and in financial journalism, where USD, CAD, GBP, etc. are preferred, but which few outside of finance recognize. Please do not confuse the existence of one organization's published standard for evidence that all other systems of notation are ergo necessarily random gibberish. It's just isn't so. — SMcCandlish [talk] [cont] ‹(-¿-)› 12:00, 21 September 2007 (UTC)

Candlish, your Mexico example doesn’t apply, because the discussion was about the choice of which abbreviation or symbol to use, not about whether to use one. In large tables with many country entries, MX would be an accepted substitute for Mexico.

I didn't get the point across clearly then (on a re-read, I'm certain that I didn't, as I was not very clear); it is: "Wikipedia does not apply a nomenclature standard simply because it exists; an actual rationale for using it (especially versus any competing conventions) must be presented, tested/argued, and agreed upon by consensus." — SMcCandlish [talk] [cont] ‹(-¿-)› 23:39, 21 September 2007 (UTC)

I’m absolutely fine with ISO abbreviations to avoid any country-centrism, but as a compromise I suggest to allow the replacemnt of the letter standing for the currency name, not the country (or similar issuing body, like the EU), by an established and readily available symbol, thus USD and US$, GBP and GB£ would be fine.

Anyhow, I find it more distracting that currencies, unlike all other units, are often put before the value (and also unspaced), not after it. — Christoph Päper 18:16, 21 September 2007 (UTC)

Um, but that's simply how it is done (universally as far as I can tell; Spanish speakers do not write "5,000–USD"; this is not an "Americanism" issue at all, if that is your implication. — SMcCandlish [talk] [cont] ‹(-¿-)› 23:39, 21 September 2007 (UTC)

Related articles: List of circulating currencies, ISO 4217 and Currency sign. --Gadget850 ( Ed) 09:27, 21 September 2007 (UTC)

This entire discussion is about 90% a waste of time, engery and brain cycles, since as WP:MOSNUM advises, the first occurrence of any currency term should be wikilinked to the article on that currency to begin with, thus obviating any confusion: "2 million dollars", "USD2,000,000", "US$2,000,000". —Preceding unsigned comment added by SMcCandlish (talk • contribs) 12:04, 21 September 2007 (UTC)

"2 million dollars" is no good: the wikilink won't show in print (such as Wikipedia 1.0). The others are fine, though (if by "USD2,000,000" you mean "2,000,000 USD"), but a little standardization wouldn't be bad. Personally, I like the ISO approach (where spelling the whole thing out is not suitable). -- Jao 18:39, 21 September 2007 (UTC)

I'm going to say something I've said a few times in the past: context matters. To this tune, SMcCandlish made a good point when he observed that the ISO three-letter codes are barely recognised outside finance. For this reason, I would not support using ISO in all contexts. Most newspapers would us "US$" when talking about US dollars outside their business section. The ISO standard is only appropriate in specialist finance articles, or where there are too many currencies in the same article for US$, NZ$, C$ etc. to be viable.

To address the original issue, it might be worth having some standard abbreviations (i.e. is Australia A$ or AU$? In New Zealand, we use A$.)—worth a try, anyway, I don't know how workable it'll be though. The principle is clear, though, so don't forget about it in the midst of details: the standard symbol and standard country specifier should be used, be put in the correct place (US$100 not 100 US$) and ISO codes may be used only in specialist finance articles. Neonumbers 01:26, 22 September 2007 (UTC)

In Australia, it's mostly "A$". Tony (talk) 10:42, 30 September 2007 (UTC)

Neonumbers said: "To address the original issue, it might be worth having some standard abbreviations". Agreed. So why not just use those in ISO 4217 rather than inventing the wheel, yet again, and making yet another new standard. The idea of standards is that once you have seen something in one place, it will be the same in other places. WP already has a page on ISO 4217 and hundreds of other sites also list the codes. 2007-10-01 11:45 UTC 85.210.15.78 —Preceding unsigned comment added by 85.210.15.78 (talk)

Abbreviation for nautical mile (M, NM, nm or nmi) and knot (kn, kt or kts)

Moved from Talk:B-2 Spirit (begin)
I don't agree that nm is a "standard" abbreviation for nautical mile. For maritime use, the only internationally accepted abbreviations I know of are nmi and M. Do aviation users have a different standard? Thunderbird2 11:23, 19 September 2007 (UTC)

Don't you think that the fact you're going around to LOTS of articles and having to change this is a hint? These are airplanes and it's a long-standing abbreviation used for aircraft. Read the article on the Nautical mile and you'll see that this is an accepted abbreviation and the standard abbreviation when used for aircraft.--Asams10 11:39, 19 September 2007 (UTC)

The nautical mile article states that nm is in widespread use, but not that it is a standard in any sense. I do not see anything in it specific to aircraft though, so perhaps I am missing something. I accept the possibility that aviation authorities may use different standards to the maritime ones referred to there, but I would like to see evidence for this. Do you have a verifiable source for your statement that nm is "the standard abbreviation when used for aircraft"? Thunderbird2 12:02, 19 September 2007 (UTC)

Perhaps I can ask you to provide a verifiable source that says that NMi should be used for Aircraft?--Asams10 12:57, 19 September 2007 (UTC)

The IEEE lists nmi as its preferred abbreviation, regardless of application. I think the American Institute of Physics does so to but I would need to check. I know of no international authority that recommends nm as an abbreviation for nautical mile. Thunderbird2 14:01, 19 September 2007 (UTC)

Even wikipedia conversion template uses the NM default. Template:NM_to_km --Asams10 14:11, 19 September 2007 (UTC)

The Wikipedia template is not an international authority on aviation terminology. Here are the IEEE guidelines. I will look up the equivalent AIP advice. Thunderbird2 14:21, 19 September 2007 (UTC)

Last I heard, IEEE didn't dictate distance measurement standards. Try the FAA. They use NM. --Asams10 14:40, 19 September 2007 (UTC)

So the abbreviation/acronym probably should be NM or at least nmi or M. I think the first 2 are clear to most, while M is probably not. If the first use is linked, there's no room for confusion anyway. -Fnlayson 15:03, 19 September 2007 (UTC)

The FAA is good enough for me. Can we agree on NM then (linked on first use)? Thunderbird2 15:09, 19 September 2007 (UTC)

I'm a little hesitant to use the above FAA document, because every entry in the glossary is capitalized. It may be that the FAA thinks NM should be capitalized, but it may also be that as a formatting issue every entry was capitalized. I think the FAA would be an excellent thing to rely on as long as we could see that the FAA uses NM in contexts in which everything isn't capitalized. *Tom does some Googling* This search seems to show that the FAA uses both "nm" and "NM" frequently. Sorry, I'm not trying to just make trouble; I don't have a definite answer. TomTheHand 17:19, 19 September 2007 (UTC)

When I see nm, I think nanometer, not nautical mile! 17:55, 19 September 2007 (UTC) —Preceding

unsigned comment added by Rangek (talk • contribs)

I'm an aerospace worker. I see NM I think Nautical Mile. I doubt very much the two would be confused in use. If the B-2 really had a range measured in nanometers, we'd probably not have bought them.--Asams10 18:04, 19 September 2007 (UTC)

I see no conflict between nm for nanometre and NM for nautical mile. TomTheHand has a good point though about a possible ambiguity in my interpretation of the FAA guidelines. Can anyone confirm that upper case letters are intended throughout (and not just used by default)? Thunderbird2 18:14, 19 September 2007 (UTC)

The FAA jurisdiction is just US aviation. There is also CASA, CAA, JAA, CAAC etc. By international agreement (Chicago Convention) they all use ICAO standard units or notify deviations from it. So ICAO would be a good place to look. The nautical mile is also used in shipping worldwide. So Wikipedia must address the domains of shipping+aviation and have a worldwide scope. Given the wide scope of this unit of measurement, I propose moving this discussion to Wikipedia talk:Manual of Style (dates and numbers) unless anyone objects within 24 hours. Lightmouse 09:55, 20 September 2007 (UTC)

A discussion on maritime abbreviations was held on the SHIPS talk page. The (cautious) consensus there was for using nmi, but the discussion was archived recently without having reached a clear cut conclusion. I support Lightmouse's proposal to move the discussion to MOS, but then I would ask for its scope to be broadened to include kn vs kt for the knot (see knot talk page). Thunderbird2 10:14, 20 September 2007 (UTC)

Following Lightmouse's tip, I've been looking for an authoritative ICAO publication. The best I can find is this Australian document that compares their national regulations with ICAO guidelines. Although not the final word, it seems to give the nod (for aviation use) to NM for nautical mile and kt for knot. Am I reading it correctly? Thunderbird2 14:29, 24 September 2007 (UTC)

I'm given to understand that the capitalized NM is used and that the lower case nm is used to denote nanometers. Hmmm, I thought kt denoted kilotons of yield for a nuclear warhead? --Asams10 14:55, 24 September 2007 (UTC)

Yes, it seems that NM is the preferred abbreviation for aviation, which solves the problem with the nanometre. For the knot both IEEE and AIP prefer kn, presumably for the reason you state. Can we agree on NM at least (for articles related to aviation)? Thunderbird2 17:09, 24 September 2007 (UTC)

I would support the idea of using NM for aviation articles, certainly all the official aviation documents I have seen use either NM or nm and never nmi. Perhaps if contentious this should be brough up on either Wikipedia talk:WikiProject Aircraft or Wikipedia talk:WikiProject Aviation rather than in just one aircraft article. MilborneOne 17:34, 24 September 2007 (UTC)
Moved from Talk:B-2 Spirit (end)
Lightmouse 19:20, 24 September 2007 (UTC)

@Rangek: funny, I thought the same thing. It was a long journey: 500 nanometer. Shinobu 16:23, 26 September 2007 (UTC)

The SI authority has a table of old non-SI units. It says:

• nautical mile: "the symbols M, NM, Nm, and nmi are all used; in the table the symbol M is used"
• knot: "There is no internationally agreed symbol, but the symbol kn is commonly used."

I am sure the SI authority wanted to avoid using 'nm' and 'kt' because they are already used. Several people here also mentioned that issue. Note that a lot of articles use [[Category:Conversion templates|convert templates]], so these should match any decision here. Lightmouse 11:01, 29 September 2007 (UTC)

I wish to summarise the facts and make a proposal. As I understand them, the facts, first for the nautical mile:

• The abbreviation NM is preferred by the ICAO for nautical mile
• The abbreviation nmi is preferred by the IEEE for nautical mile
• The symbol nm is reserved by SI for the nanometre

and for the knot:

• The abbreviation kt is preferred by the ICAO for knot
• The abbreviation kn is preferred by the IEEE and the AIP for knot
• The symbol kt is reserved by SI for the kilotonne

---

I cannot imagine everyone agreeing to use the same abbreviation, so my proposal is to offer a choice. For the nautical mile:

• permit use of NM and nmi for nautical mile
• discourage use of nm for other than nanometre

and for the knot:

• permit use of kn and kt for knot
• recommend use of kn

---

Comments anyone? Thunderbird2 23:18, 29 September 2007 (UTC)

I would like to go one step further and declare one preference for each unit, and strongly discourage any overlap with SI units:

• preferred usage is nmi for nautical mile and kn for knot
• do not use nm for nautical mile and kt for knot, because they conflict with SI units nm (nanometer) and kt (kiloton)

An exception for NM as alternative to nmi could be made.−Woodstone 09:59, 30 September 2007 (UTC)

Sounds like a good plan. I'd prefer Woodstone's version, but getting consensus for Thunderbird's "preference" system would probably be easier. Sacxpert 09:52, 1 October 2007 (UTC)

One issue I'd like to bring up is that English-speaking navies overwhelmingly use nm for nautical miles and kt for knots. I understand the issue of overlap with SI, and I understand that it's important to try to use units with less ambiguity, but I think it's also important to point out that when you're saying that we should not use nm and kt, you're saying we shouldn't do what perhaps the most important professional organizations to use these units do.

I'm sorry, Thunderbird, I know you and I have discussed this for a long time, and I don't want you to think that I'm suddenly withdrawing from my willingness to compromise. I'm just sometimes bothered by the way we talk about these issues, as I think actual usage is as important as SI's recommendations, if not more so. TomTheHand 14:47, 1 October 2007 (UTC)

I too would prefer a single recommendation, if this can be realised. Putting aside the knot for a moment, is there any mileage in following ICAO by suggesting NM as the preferred abbreviation for the nautical mile? I don't agree that following navy practice is wise if it conflicts with both aviation authorities and SI. Thunderbird2 10:46, 6 October 2007 (UTC)

Following the deafening silence in support of NM, I return to my previous quest for consensus, and suggest the following specific text to go in 4.3 Unit Symbols and Abbreviations:

---

• Avoid use of unit abbreviations that have conflicting meanings in common units systems such as SI and US customary units.
  • Use NM or nmi to abbreviate nautical mile rather than nm;
  • Use kn to abbreviate knot rather than kt or kts;
  • Link such units to their definitions on first use.

---

I imagine that the list of potential clashes is not limited only to the knot and nautical mile, and the proposed text is deliberately written in such a way as to facilitate additional entries. Comments and counter-suggestions are invited. Thunderbird2 12:00, 13 October 2007 (UTC)

I would still prefer to suggest a single preference. Proposal (in line with "mi" for mile):

• Use nmi (or NM) to abbreviate nautical mile rather than nm;

−Woodstone 12:37, 13 October 2007 (UTC)

I've been trying to be consistent with nmi, on the weird chance someone would confuse nautical miles and nanometers. Knots remain a challenge, as I can't think of anything -- perhaps "knots" -- that couldn't be ambiguous, if only pedantically such as kilotons and knots. While I recognize there are formal conventions on when to capitalize a prefix and not, they are so little followed that I'd rather not depend on them.

I did find that someone had inserted "(370.4 km)", with one decimal point, after I citied the international treaty language of a 200-nautical mile Exclusive Economic Zone. May I note that that was an treaty developed by many nations, including those using SI, and, while they found it appropriate to disambiguate miles versus nautical miles, they did not. Seeing the 0.4 following an integer authoritative measurement, to me, just feels wrong. In such cases, I would recommend the usual rounding practice: truncate if the first decimal point is less than 0.5, and round up the integer if it is 0.5 and above. This is said in the interest of consensus; I really don't think that any other systems should go into treaty or other legal descriptions, except those that were used by the legal drafters. Howard C. Berkowitz 15:48, 13 October 2007 (UTC)

You are demonstrating a serious deficiency in your understanding of "precision" and significant digits and the like. When somebody talks about a 200 mile EEZ, there is absolutely nothing about the precision of that number that would prevent us from expressing it as 1,215,223.097 ft or with even more precision. That is not a measured quantity; it is a defined quantity. It has infinite precision. It is only practical considerations of what could reasonably be measured that constrain us, and in most cases this distance could in fact be measured to within a tenth of a kilometer. ~!Gene Nygaard —Preceding signed but undated comment was added at 18:56, 13 October 2007 (UTC)

I just messed up entering the tildes above. But in this particular case, expressing it as 370 km is also problematic. That doesn't show you that in fact it is actually accurate to the nearest kilometer. Rather, it looks like it might only be much rougher, only accurate to the nearest 10 kilometers. Gene Nygaard 19:03, 13 October 2007 (UTC)

Imperial vs Metric: Keeping measured values straight in scientific and technical articles

Many of the sources I use for scientific articles give values in Imperial units. So for those data, I use the cited Imperial units as the primary value in scientific articles I contribute to and give a calculated approximate value for SI/metric following it in parenthesis. But then you have people who just hate Imperial units (I don't much care for those units either, but I keep that to myself when accuracy is concerned) and either switch the calculated approximate value for the cited one or they sometimes delete the Imperial completely. Then somebody else will often come by later and do another conversion and sometimes a switch in priority; now we a significantly incorrect datum.

I find this intolerable because it is replaces a cited value with something that is not in the cite and that can easily lead to something that is less accurate. So while I generally agree that scientific/technical articles should prefer SI/metric, I don't think that should be done to the detriment of article accuracy. So here is what I propose:

1. Never remove a cited value simply because it is in an Imperial unit. Instead, add the cited value to the reference while giving properly calculated SI/metric units preference in running text. For example "The tree grows to an average height of 4.6 m (15 ft) during its life cycle.<ref>Given as 15 ft in Doe, John (1995). ''Trees'', page 456</ref>

This would make the article text consistent by always having SI/metric units first, have Imperial in parens as appropriate (for accessibility purposes), while making sure that the original cited (often measured!) value is preserved. --mav 01:39, 26 September 2007 (UTC)

The MOSNUM clearly states that in scientific articles where there is consensus among the contributors not to convert the metric units...

That does not say you now have license to remove any and all non-SI values from articles that you deem scientific. Mass removal of conversions is what I stated would happen (here) if we entered this "scientific" clause. Any conversions that were removed should be replaced.

As for having SI units stated first in scientific articles despite the fact that Imperial units are the cited units, yes we had that discussion before (here) and we were all ok with that too. —MJCdetroit —Preceding signed but undated comment was added at 03:40, 26 September 2007 (UTC)

We are not all okay with that. And I am not okay with the vagueness of calling something a "scientific article" either. Gene Nygaard 19:45, 9 October 2007 (UTC)

I never did get round to tossing my tupence-worth into that discussion. Scientific context or otherwise, I'm happy with metric units' being stated first if and only if it is noted using <ref></ref> wherever these are not the units given by the source. The case of conversions from metric where context calls for preference's being given to non-metric should be treated similarly. In general, I'd have it a requirement that any conversions either be given as secondary (e.g. in brackets) or duely noted as such. Seems only reasonable to me. Jɪmp 04:26, 26 September 2007 (UTC) ... Yesterday's unfinished sentence finished today in brown. Jɪmp 00:35, 27 September 2007 (UTC)

Sorry to appear dumb, but I'm confused. Mav, is your quote about the tree a direct quote or a paraphrase? For direct quotes, an original Imperial unit must be retained, of course, with the conversion in square brackets rather than parentheses. That much, I hope, doesn't need to be spelt out in the guideline.

Assuming that it's a scientific article, if it's a paraphrase, doesn't it depend on whether consensus has allowed metrics only (in which case, use the metric equivalent unconverted), or not (in which case, use metrics as main with Imperial the converted unit/value, even if the original you're paraphrasing used Imperial)? Please correct me if I've got it wrong. Tony (talk) 04:40, 26 September 2007 (UTC)

The example is a paraphrase, not a direct quote. I concede (and agree) that priority should be given to SI values. What I would like, is an addition to the guideline that covers situations where the cited (often measured) data are in Imperial units. My simple suggestion was to require that the actual value and units be given in the citation. Simply removing the Imperial units or even giving them lower priority when the cite uses them is not acceptable from an accuracy and maintenance standpoint. --mav 14:08, 26 September 2007 (UTC)

Comment: I suspect that this issue affects only a minority of articles. Most scientists use SI. Shinobu 16:19, 26 September 2007 (UTC)

Scientists? That is beside that point for an encyclopedia (which is primarily a tertiary source constructed from secondary sources). This issue affects a high percentage of any articles that rely heavily on secondary sources produced in the United States. --mav —Preceding signed but undated comment was added at 17:29, 26 September 2007 (UTC)

I'm a scientist and I don't use SI exclusively. In fact, in my field only small "bench batches" (not more than 1,000 g) are weighed in the metric system everything else is done in U.S. Customary.

I agree with what Mav was saying, these articles are meant to be read by anyone and not just us scientists. None of the other online encyclopedias like Britannica or Encarta present [scientific] information from a metric-only POV. —MJCdetroit 18:12, 26 September 2007 (UTC)

I rather like Mav’s recommendation. Ask the average person which appears more correct, and most will say “4.6 m”, not “15 ft”. It’s an illusion of accuracy generated by the decimalization of the conversion. It’s also not uncommon for these conversions to go back and forth a few times as they go from one source to the next. While most editors won’t mess with clearly quoted measurements, it is not at all uncommon for someone to switch “15 ft (4.6 m)” to “4.6 m (15 ft)”, if only for consistency in the article, and thereby creating a false impression of which is the more accurate datum. If the original measurement was in US or imperial units, but the article (scientific or technical or not) chiefly gives preference to SI, then asking for some footnote to “ground” the original and most accurate measurement seems to me to be a very good idea. Askari Mark (Talk) 18:02, 26 September 2007 (UTC)

Data should be cited as given, with the possible exception of obviously approximate figures (when the source says "roughly 10 miles away", this is sufficient justification for saying "about 15 kilometers" and conversely). If a footnote, or a note on the source ("Data from this book in Imperial units") will help us to do so, then we should. Septentrionalis PMAnderson 18:16, 26 September 2007 (UTC)

People, you are misreading what I typed. This discussion is about the use of imperial units from scientific sources. Read that again, read the OP's statement and then read my original comment again, please. I also again assert that most scientists use SI. The one or two that don't are what people call "the exceptions that prove the rule", much like the popular story about the grandmother who smoked five packs a day and lives to become 98 years old, still going strong. Shinobu 18:41, 26 September 2007 (UTC)

I say let's not simply ask for some footnote but insist upon it wherever the source units are not given as the primary units (of course, with the exception noted by Anderson wherein the source is obviously a rough measure). Moreover, wherever those source units are not given in the text at all, require that they be given in the said footnote. This seems only reasonable & therefore propose to have words to this effect added onto the page. Jɪmp 00:32, 27 September 2007 (UTC)

Eh... I was under the impression that we already had a guideline like that. I didn't actually check, though. *checks* "Where footnoting or citing sources for values and units, identify both the source and the original units." That comes awefully close. Considering all values already have to be source per WP:V, doesn't this essentially insist on what you are asking the styleguide to insist on? Are we by now talking about fixing a problem that does not, actually, exist? Shinobu 15:34, 27 September 2007 (UTC)

But the problem we have is that some editors are using parts of this guideline to remove or subordinate cited values just because they are in Imperial units. I would like this policy to require them to place those values in a cite. --mav 13:54, 28 September 2007 (UTC)

Somewhat changing wiewpoint: When referencing standards, please use first the messaurements as originally defined. Case in point "Rail gauge" article where Decauville gauge is given as 598 mm, although it was originally defined as 600 mm. Reason repeated back and forth conversion. Seniorsag 15:58, 4 October 2007 (UTC)

The guideline currently says "Where footnoting or citing sources for values and units, identify both the source and the original units." If the original units were included in a <ref></ref> they should not be removed.

The amended guideline says that US units can be removed "where there is consensus among the contributors not to convert the metric units". In the cases you have seen mav did the person who removed the value gain consensus to do so? If not the conversion should not have been removed.

If the example you give about the rate of growth of trees is a real example of where this occurred then the standard for considering an article "scientific" is much broader than the examples people give when proposing removing such conversions from scientific articles.

In this example the imperial measure is not only true to the source but actually improves the article by making it easer to read by people who think in imperial units with negligible cost to readability. You can see in the history I had second thoughts about discussing this again but I decided to say this because the arguments made for articles about drugs or astronomical measurements don't apply here. This is a good example of why this is a mistaken policy.

-- PatLeahy (talk) 17:16, 4 October 2007 (UTC)

-- PatLeahy (talk) 16:56, 4 October 2007 (UTC)

Shinobu,

I guess you might say that I am "talking about fixing a problem that does not, actually, exist". Note, though, that you had to check ... and down pretty far in the list and that you had to rely on another piece of policy on another page (albeit a relatively well known and accepted one). So, yeah, you're right, the insistance I was calling for is already made but perhaps there still is a problem to be fixed. I'd call this a pretty important issue. The way it's put now, it's not to hard to overlook or misinterpret. I say give this issue the prominance it deserves: spell it out clearly leaving no room for interpretation.

Jɪmp 18:43, 9 October 2007 (UTC)

Can anybody point me to specific scientific articles "where there is consensus among the contributors not to convert the metric units". Gene Nygaard 19:54, 9 October 2007 (UTC)

Three questions

1. Why are the numbers ten to twenty not in the standard list of exceptions? I think most people would consider that more natural, and it also goes together well with English grammar/vocabulary and our own hyphenation rule. Made-up example: "The boy band consisted of eleven members." Notice how odd it looks when you substitute 11.
2. Why are subscripted radices not used in computer-related articles? Most computer literate people (at least those who know what 0x means) know (or could easily guess) what ₁₆ means too. There are computer languages that use other pre- or postfixes (e.g. &H07C0FFEE, DEADBEEFh) or none at all. Why tie ourselves to the syntax of a specific programming language when there's a perfectly usable, and standard, mathematical notation? Why change it just because it's about computers?
3. What to do with things like #1? Do we change them to "number one", or do leave them as is? Example: "Their album Termunterziel flew high in the charts and soon reached #1."

Lots of thanks in advance. Bye, Shinobu 13:54, 26 September 2007 (UTC)

• (1) Doess "11" look odd? Most style guides insist on it, with contextual exceptions. (2) Dunno. (3) IMO, # looks crass; better to use "No." or "Number". Tony (talk) 14:42, 26 September 2007 (UTC)
  • (3) "Number one" seems better for prose. In album chart tables we drop the "#" and just use the number, per WP:CHART. --PEJL 15:45, 26 September 2007 (UTC)

(1) Are you sure style guides universally recommend against it? The Guardian's may, but Chicago doesn't. There is of course the possibility that this is a cultural thing; I've looked at a few styleguides, and I think I know where I got the "up to twenty" rule of thumb ;-) (3) Okay, "number one" it is then, at least in prose.

Should we perhaps add some clarifying comments (if necessary) after we let this discussion run its course? Shinobu 16:14, 26 September 2007 (UTC)

1. Contextual exceptions aside, I'd rather have seen the cap on spelling numbers out bumped up from ten to twenty instead of down to nine.
2. I'd much rather we used standard mathematical notation for bases throughout. Making this exception for these articles and that exception for those makes the encyclopædia that much less readable.
3. I'd have the word "number" spelt out but allow "No." where the number is written as a numeral.

Jɪmp 00:18, 27 September 2007 (UTC)

I doubt you really meant "No.", rather than "no.", but just in case: It should be "no." if abbreviated, unless it's a proper noun in the context, like part of a title of something. — SMcCandlish [talk] [cont] ‹(-¿-)› 00:36, 10 October 2007 (UTC)

Response to Jim: (1) Why? You'd rather see "seventeenth–nineteenth centuries" than 17th–19th centuries"? The latter is much easier to read, takes less space, and in articles that contain lots of references to centuries, more easily identifiable. It is optional, all the same. (3) I think I agree in most contexts. Tony (talk) 10:40, 30 September 2007 (UTC)

I don't think I can give any better reason than "I like it." whether to stop at 9 or twenty is more or less a matter of taste. I kind of like the idea that was floating about a few months ago to spell out whatever you can in one or two words. Centuries, I reckon, we can consider seperately (though I'm still a speller-outer there). Jɪmp 03:49, 9 October 2007 (UTC)

Centuries are already handled separately, so the general recommendation wouldn't affect it, I think. I too would prefer to see it be "write out one through twenty". — SMcCandlish [talk] [cont] ‹(-¿-)› 00:36, 10 October 2007 (UTC)

I would prefer to handle centuries separately, like the style guide currently recommends. Anyway, is there anyone who strongly opposes spelling out ten to twenty? Also, on the 0x question, are there any people elsewhere, on other styleguides for example, that we might wish to contact to get clarity? Bye, Shinobu 18:00, 4 November 2007 (UTC)

The date format in ~~~~

Why don't I use ~~~~ ?

Because the resulting date format is non-compliant with ISO 8601.

When standard date formats are available, I find it highly improper to impose any ugly, illogical, non-compliant format on editors. Personally, I wouldn't bother to complain about the differences between 2007 September 26 and 2007-09-26. The latter is arguably superior, but the former has a point or two in its favor. (In particular, "September" often reminds me of the septentriones, one of the more pleasant parts of the neighborhood). I think both formats incomparably better than the internally inconsistent 26 September 2007. Iccchh.

The Wikipedia:Manual_of_Style_(dates_and_numbers) claim here that "ISO 8601 dates (1976-05-12) are uncommon in English prose, and are generally not used in Wikipedia" is based on at least 2 false premises.

1) There is no such thing as an "ISO date". There are ISO 8601 date formats, yes; but dates are dates. There is no alien species which can be outlawed as inappropriate or unnecessary.

2) Dates are part of language because they are numerical expressions and thus part of mathematics, which is the language of science and technology. Being part of mathematics, dates are thus not subject to grammatical conventions, rules or prohibitions based on "natural linguistic" principles. "English prose" does not intersect dates. Dates are numerical expressions; even when imbedded within prose, where they may be thought of as "foreign language" expressions.

Why should virtually all dates used within Wiki comply with ISO 8601? There are a number of important reasons.

a) It is an international standard. The benefits of consistent usage with authoritative meaning are many and widely recognized. Many very intelligent people have put a lot of thought and effort into crafting the most useful standard possible. It is a reasonable assumption that standard usages will be better understood by more people for a longer time than non-standard usages. Unless you are begging for trouble, it is almost always a good idea to comply with all applicable standards.

b) Ambiguity kills data. Non-compliant formats may be misunderstood. In the case of dates, this is exquisite. As of today, how am I supposed to know whether 10-07-07 is in the past or in the future? Years from now, what date is that?

c) The year must have precedence. If you had a pedometer or odometer that worked like the internally inconsistent formats, it would drive you crazy. Would you mind if your employer used a non-standard format for the amount of your paycheck? Those formats are thus obsolete.

d) It is essentially irrelevant how widely adopted the standard is. In fact, since ISO 8601 provides free sorting and sharply reduces errors, it is already widely used in IT and in astronomy.

The US Congress passed the Metric Act in 1866. There has been a lot of resistance to a few of the metric measures. Even the total loss of the $125,000,000 Mars Climate Orbiter because a NASA contractor used an Imperial unit didn't seem to wake much of anyone up.

It's still a good idea to use international standards. And to be grateful when they exist.

Some points editors may wish to consider before repeating previous comments in objection to these assertions here:

Some Christians object to spelling out the names of months in Western languages because they are, or are based on, the names of what they have the gall to call "pagan" gods. Thus, in records of "monthly meetings" one sees the abbreviation Mo, e.g., 7 Mo or 9 Mo for September (as opposed to mo, which means "married out" (of the church)).

I consider it generally improper, because inconsiderate, to provide a non-compliant value without also providing a standard value.

Because older dates in secondary sources have often become corrupted, the most original date is often the best choice. Footnotes can help immensely by providing details. If necessary, a section titled "Details on Dates" is a good alternative to promulagation of inaccurate or easily misunderstood "information". In an encyclopedia, it's a safe bet most readers won't care about a 5-year error in the placement of a date 100 or 500 years before the present time, and that a few will care deeply about every day, or even every hour.

Standard usages are better understood, not less. One of the goals of standardization is to eliminate repetitive thought about meaning. Confusion arises when people head off on non-compliant excursions. When people use standards consistently, text is more concise and clearer. This has cost implications, including cost implications in education. When parents and taxpayers pay for education about non-compliant practices, they are not receiving good value for their money. Also, people, whose mental ability for detail would marginally qualify them for jobs where they might shine in other ways, may be disqualified when a plethora of non-standard usages becomes too much to comprehend and process.

Brains should freed up for useful or artistic innovation, not tied up processing meaningless variation.

Because ours is an increasingly interconnected world, with more and more linkages between formerly disconnected fiefdoms, there is a crying need for standardization.

There is virtue in being an exemplar.

No amount of carping about dates is likely ever to stop men from embracing ambiguity in telling women, "I'll call you on Tuesday".

Walter Nissen 2007-09-27 01:31

Forgive Sinebot, he's only a bot. Dates not part of the English language? Anyhow, what do you propose? That all dates on WP be given only in ISO format ... good luck. Jɪmp 02:55, 27 September 2007 (UTC)

Well, almost all dates, yes. And not limited only to Wiki. In the alternative, what would you offer? That this persistent mess be conserved to throttle our grandchildren?

Walter Nissen 2007-09-28 00:14

Just FYI, complaints or suggestions about the date format in signatures cannot be addressed here. Any problem with that will have to be fixed by the developers, so it should be raised at WP:VPT. — SMcCandlish [talk] [cont] ‹(-¿-)› 00:39, 10 October 2007 (UTC)