Wikipedia talk:Manual of Style/Dates and numbers/Archive 56

This is an archive of past discussions about Wikipedia:Manual of Style. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 50

←

Archive 54

→

Proposed Amendment to Section 1.2

Section 1.2 of Manual of Style (dates and numbers), concerns the linkage of dates that contain the months and the day in that month. I would like to propose amending this section to include the correct English day/date format.

I would like to point out that there is some major bias towards the US in this Manual of Style (as with many articles and other guides). It uses many examples of how the date is formally written or spoken, that are very rarely used in many English speaking countries other than the United States.

For Example:

In the UK, Australia and New Zealand the majority of people say: The 5th of March, 2006. It is very rare to hear somone say 5 February 2006, as the later does not make any sense, it implies that there are 5 seperate Februarys. This also is the same for February 5, implying there was a February 4 and a February 3, etc, but this makes no sense either as there has been millions of Februarys since the begining of the Georgian callendar and there is no use in counting them all.

The use of these other formats make no sense, are inapropriate to use to determine a date in time, and are somthing that is used primarily in the US. The majority of the English speaking world can speak English in a way that hasnt diviated from the English language enough to result in ambiguity or inaccurateness, so thats the way it should be in the English Wikipedia, the correct way.

The aim of this proposal is to change the way that the day and the month are written in articles, especially biographical articles. (The article on the late Steve Irwin, has a date format that hardly anyone here in Australia uses: 22 February. I changed it to 22nd of February, and someone changed it back to 22 February. My brother pronounces the word' mirror' as though he is saying 'merra', he watches alot of americanised TV. I personally, feel that I have a responsibility to ensure that 'American English' stays in the country that it originated from, just as 'Australian English' must stay in this country. I don't want American English infiltrating one of the worlds largest encyclopedias, because it is the English Wikipedia not the Polish Wikipedia, not the Japanese Wikipedia, not the American English Wikipedia, not the Australian English Wikipedia.

Proposed Changes:

That 22 February and February 22 be changed to 22nd of February or the 22nd of February.

This implies that it is the 22nd day in the month of February. There is no ambiguity, it is more universally accepted amongst english speaking countries, and maintains the English language in the English Wikipedia.

It should be noted that date formats such as dd/mm/yy, mm/dd/yy, yyyy/mmm/dd, etc, are a seperate issue and not included in this proposed amendment.

Thanks for reading and lets hope we can set things straight and do things the right way. Nick carson 02:44, 5 September 2006 (UTC)

I do think this format should be allowed, as it is the way it is read and it is acceptable formal English. —Centrx→talk • 07:06, 5 September 2006 (UTC)

But of course I am a transplant from the 18th century. —Centrx→talk • 03:48, 6 September 2006 (UTC)

Oppose. It's acceptable only in formal (i.e., archaic) English in the UK, anyway. (We're tour guides for visitors from the UK and Germany at the moment, and that's their opinion.) If you can get the developers to accept that as a date format for wikilinking, and it really is commonly used in Australia, neutral. — Arthur Rubin | (talk) 13:47, 5 September 2006 (UTC)

Fair enough, but I shouldn't have to get anyone to accept the fact that 22 February and February 22 don't make any sense, its insane. Nick carson 08:41, 6 September 2006 (UTC)

Oppose. On the basis that the present system has the merit of brevity, that too much angst has already been caused over dates and introducing another is a bad idea. I also agree that it looks archaic. --Guinnog 13:51, 5 September 2006 (UTC)

My opinion is that the current format looks childish. Angst felt in the past is no excuse to dismiss an issue. If you actually read my proposal you would have realised that I didn't propose to add another format. Now go and read it. Nick carson 08:41, 6 September 2006 (UTC)

Oppose, for two reasons. First, I think you've failed to distinguish between spoken and written English; it's the normal spoken form, but not the normal written form. Secondly, such dates do not respond to users' date preferences, which is the whole point of this section. Stephen Turner (Talk) 14:11, 5 September 2006 (UTC)

Your right, I should have distinguished between written and spoken. But what is written, I read, Reading somthing that is not written is just making things up. Written and spoken are, therefore should, be the same thing. As for user preferences, there is no option to display the correct '3rd of June' format, that is why I am proposing this amendment, same as above, make sure you read before you oppose somthing, go and see for yourself in your preferences. Nick carson 08:41, 6 September 2006 (UTC)

Oppose. I've never heard such rubbish. Tony 14:50, 5 September 2006 (UTC)

Right back at you. Nick carson 08:41, 6 September 2006 (UTC)

Oppose. The proposal confuses spoken Australian English with written Australian English. Powers ^T 17:57, 5 September 2006 (UTC)

You don't know what your talking about. In Australia we write dates as dd/mm/yy, and we speak dates as '4th of September', it is inapropriate incorrect american english to write a date as '4 September' or 'September 4'. It is apropriate in an encyclopedia to write it the way it is spoken because it reads far far easier, neater, tidier and professional. It also makes sense. Nick carson 08:41, 6 September 2006 (UTC)

I think there's a difference between spoken and written English, which the proposal does not allow for. We say "the 22nd of February", but we normally write "22 February" (moreover, we (or rather, I) read "22 February" as "the 22nd of February". Neonumbers 05:43, 6 September 2006 (UTC)

If you or sombody you know writes '22 February' you or they have been influenced by American English. If you take time to think about it, the formats '4 May' and 'May 4' make no sense as they imply multiples of the month of May, not a day in the month of May. Nick carson 08:41, 6 September 2006 (UTC)

Stop voting. The reasons are what counts. There is no point in just repeating what others have said. Stephen Turner's single comment is enough to prevent this proposal from being implemented. If I had made a vote and were looking back on it now, I would remove it. I really find this disgusting. —Centrx→talk • 05:56, 6 September 2006 (UTC)

Pitty is not required. And you are incorrect, a single comment is not enough to prevent any proposal from being implemented. Nick carson 08:41, 6 September 2006 (UTC)

That one comment pointed out that at the moment your proposal is technically unworkable. It should be obvious that with this being the case, the proposal is pretty much dead until the technical barrier is overcome. Christopher Parham (talk) 21:10, 7 September 2006 (UTC)

I agree, and I apologise for my harsh comment. I, too, have been caught in the same way, where I started a vote (on the basis of misinformation), and it achieved nothing after the first comment. Tony 06:10, 6 September 2006 (UTC)

My apologies; I did not see it as a vote, but rather as a collection of reasoned views. I have therefore removed the word "Oppose" from my post. At least it seems like we're almost all in agreement. Neonumbers 06:18, 6 September 2006 (UTC)

People, you all must read through my proposal again, imploy sommon sense (if you can manage it) and do things the right way. Nick carson 08:41, 6 September 2006 (UTC)

First of all, it's rather conceited to refer to one's preferred style as "correct", especially when one clearly has no idea what correct English is. ("Inapropriate", "hasnt", "diviated", "inaccurateness", "thats", "word'", "imploy", and "sommon" are not English words. "Lets", "later", and "your" are, but are not used correctly. Apparently the irony of declaring "Thanks for reading and lets hope we can set things straight and do things the right way" was lost on Nick. Secondly, there is no "ambiguity" in what is meant. Everyone knows what "22 February" means. According to Nick's logic, the phrase "9:15" somehow implies that this it is the fifteenth nine. "22 February" means that this is the 22nd time that it's been February this year. "22 February 2006" means it's the 22nd time it was February 2006. If one imagines each day in February as being another instance of February, then "22 February 2006" does mean "the 22nd February 2006". Thirdly, it's rather hypocritical to claim that it's wrong to write less than one expects one's audience to say, and yet write "22nd". After all, if one expects one's audience to say "twenty-second", shouldn't one write "twenty-second? If someone reads "22nd" and says "twenty-second", isn't that "making things up"? And if one allows "twenty-second February" with the understanding that it should be interpreted as "twenty-second day of February", then it seems to me that one should be okay with just "22 February". If it's okay to leave one thing implicit, why is it not okay to leave the other? All language, to some extent or another, is metaphoric. There is one matter, however, on which we share a similar view. "Mirror" has two syllables, not one. Flarity 02:24, 25 September 2006 (UTC)

Date linking

I just noticed this in an article: June 9 (Old Style)/June 19 (New Style), 1619. Is it really necessary to link dates when they are in this complex sort or format? Rmhermen 02:54, 1 September 2006 (UTC)

Yes, it's necessary so that they appear as "June 9" for Americans and "9 June" for Brits etc. The rule is "If a date includes both a month and a day, then the date should normally be linked to allow readers' date preferences to work". Stephen Turner (Talk) 06:32, 1 September 2006 (UTC)

Does anyone ever say 9 June (Old Style)? Old style would not have written it that way. Rmhermen 15:07, 1 September 2006 (UTC)

I'm not sure how old style would have written it, but I would certainly write it that way. Stephen Turner (Talk) 15:28, 1 September 2006 (UTC)

Commas in dates that appear mid-sentence

When using an American-style date in the middle of a sentence, which would be correct:

"on May 10, 1815 at Fort McKay"

or "on May 10, 1815, at Fort McKay" ?

(The difference is the comma between the year and the subsequent word.) Omphaloscope talk 15:44, 1 September 2006 (UTC)

I believe commas go after dates regardless of whether it's day-month-year or month-day-year. Powers ^T 20:20, 1 September 2006 (UTC)

This seems to be a matter of common sense and context, rather than a hard and fast rule. Consider this sentence: Registering for our fitness programs before September 15, will save you thirty percent of the membership cost. This is an example from a style guide that labels its use here as incorrect. My own personal rule of thumb is that a comma indicates a pause and as such, putting a comma after every date is guaranteed to result in many awkward-sounding sentences. Many U.S. style guides make no mention of commas after dates at all, so it is obviously not a mandatory inclusion:

However, I find another guide (from something called the Capital Community College) mandates using a comma after a year in a full U.S. date, but not when the date is in international format. I note that some sentence constructions demand the use of a comma, regardless of date format. My own practice has been to get rid of non-essential commas, but lean heavily on the side of keeping them if there is any doubt. --Jumbo 17:50, 6 September 2006 (UTC)

It isn't even really a separate rule, it's just that for some reason, people don't seem to realize that dates, like other qualifying phrases, require commas. Take, for instance, the sentence "In 2006, John bowled a perfect game." Now replace "In 2006" with another phrase: "In Denver, John bowled a perfect game", "While wearing a clown suit, John bowled a perfect game", "Setting a new personal best, John bowled a perfect game", "On Tuesday, John bowled a perfect game". The need for a comma isn't really created by the date so much as the fact that it's a separate clause; any other separate clause would also need a comma.Flarity 02:42, 25 September 2006 (UTC)

Using the metric system in Wikipedia

The discussion below was moved from Village Pump
I notice that most articles in Wikipedia still make use of the emperial standards. An example would be Carol Yager. For any persons unfamiliar to emperial standards, the article is useless.

Arguably, this is because the USA still retains the emperial system, being one of only 3 countries not adopting the system. See Metrication

All other countries use the metric system and as Wikipedia is targeted at the whole world, I believe it is important that only metric standards be used, at least in addition to the emperial standards.

I did not find any reference to this important aspects.

The standard is to use both metric and Imperial measurements. See WP:MOSNUM. Kusma (討論) 13:21, 29 August 2006 (UTC)

WP:MOSNUM#Units of measurement User:Pedant 16:28, 29 August 2006 (UTC)

Won't someone help implement the metric system in our country :( 24.126.199.129 19:33, 29 August 2006 (UTC)

When I was in grade school, we were taught that the United States would be metric by the time we graduated high school. At the time, the president's name was Lyndon Johnson. Don't hold your breath. Fan-1967 20:29, 29 August 2006 (UTC)

A quite simple way of contenting both conventions lovers is the "hover" htlm feature. print metrics and let the hover be imperial. -- DLL ^{.. T} 18:44, 30 August 2006 (UTC)

"My car gets forty rods to a hog's head, and that's the way I likes it!" Kidding aside, I say simply remain consistent within a medium. If measurements for a subject are commonly given in imperial units, then that's what should be used. I like the "hover" idea, as well. – Someguy0830 (Talk | contribs) 18:55, 30 August 2006 (UTC)

The issue is that to USan and British editors, road measurements are 'consistently' given in miles. To Candian and Australian readers, road measurements are 'consistently' given in kilometers.

Another issue is the rounding problem. As an example a road distance in an article source given in miles may have been originaly rounded, for instance 25.4556 miles becomes 25 miles. Now when you do a simple translation of this to kilometers, it becomes 40.2336km, and is rounded again to 40km. The original figure would have been converted to 40.9668171 and rounded to 41km. --Barberio 19:24, 30 August 2006 (UTC)

Why not just treat this like we treat differences in British and American English. Either one is ok, and which one is used depends on who wrote the initial article, and on the "bias" of that author. --Frescard 20:11, 30 August 2006 (UTC)

I don't think that it could be a simple a solution as this really. For instance, if an US editor creates and article on a road that travels mostly in Canada but partialy in the US, should it remain as Miles even when the relevent sources all use km and a conversion would have to be made.

I'm actualy suprised we don't have a Wikipedia:Manual of Style (measures and metrics) which would cover this, since it's a hugely important issue to an encylopedia. --Barberio 21:07, 1 September 2006 (UTC)

Use both. In that specific case, I would say metric first makes the most sense, because that's the measurement that you say is the directly verifiable one. x km (y miles) --tjstrf 21:46, 1 September 2006 (UTC)

This debate has been underway for some time (see Wikipedia:Measurements Debate) reaching the solution Pedant pointed out above (see Wikipedia:Manual of Style (measures and metrics). And it matches tjstrf's recommendation. And yes, I did link Barberios Wikipedia:Manual of Style (measures and metrics) above.

Cheers - Williamborg (Bill) 19:39, 2 September 2006 (UTC)

After reading the various comments more carefully, I see we don't cover the cyclic rounding error issue in the MOS. Perhaps the solution to avoid this is to provide the reference in a footnote by using <ref>The original source was XYZ which used metric units.</ref>. That way we avoid cyclic rounding induced errors. I'll make (propose) that change in our policy. Williamborg (Bill) 19:57, 2 September 2006 (UTC)

The discussion above was moved from the Village pump

Actually, the issue is adressed quite nicely on the page:

"Conversions should generally be included and not be removed.
If editors cannot agree about the sequence of units, put the source value first and the converted value second."

In other words, cyclic rounding should never happen, because you use the sourced value as the primary. If the original was in metric, you will use the metric, with the imperial as a paranthetical converted unit. Reffing the measurements would be good, of course. --tjstrf 20:46, 2 September 2006 (UTC)

Always reffing the sources of the measurements should be the rule. Sometimes that is the only way to explain otherwise anomalous measurements. Some sources for lighthouses list the light ranges in miles. As these ranges are over water, nautical miles would normally be used instead of statute miles, but the abbreviation for nautical mile is nm, and the sources use m. So, which type of mile do you assume when you convert to km? A real doozy is this, where the area of cities is given in square meters, and the population density is given without units being specified, but which turn out to be persons per square mile. -- Donald Albury 23:29, 2 September 2006 (UTC)

My $.02: use both. What's so hard about that? jgp ^T_C 23:58, 2 September 2006 (UTC)

I don't thinkg that nm needs to be converted; I live in a metrics country, and nm is used. Tony 01:15, 3 September 2006 (UTC)

Nanometers are used in the U.S., too. It's angstroms we could do without, since they only involve a decimal point shift for someone who insists on using obsolete units. But if you are talking about the other "nm", including the conversions helps to identify what that ambiguous symbol is, without wondering why they are using nanometers in that particular context. That's only part of the reason for using them, of course. Gene Nygaard 03:59, 3 September 2006 (UTC)

Sorry, "nautical miles", as abbreviated above. Tony 06:13, 3 September 2006 (UTC)

The [SI brochure] says:

Nautical mile: As yet there is no internationally agreed symbol, but 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.

Information provided for interest only, not as a suggestion.

bobblewik 11:03, 3 September 2006 (UTC)

Looks like a good reason to always give the measurement and unit from the source. -- Donald Albury 17:57, 3 September 2006 (UTC)

I like Williamborg (Bill)'s changes in the units of measurement section asking for citation and refs of source units. This would allow the unit order to stay consistant throughout an article. The example that comes to mind is in the article on France where every measurement is written metric (imperial) except one which compares France to Yeman and the state of Texas. In that case in appears (but is unclear) that the square miles are the sourced data and sq km are converted. Citing the source would clarify that and allow the units to be switched to maintain consistency in the article. For the sake of example, I think there should be an example in the section on how citing a source unit would look. — MJCdetroit 13:37, 7 September 2006 (UTC)

In the case where one value comes from a source and other other is converted, the source value should be provided first since it is the most accurate. The converted value should be provided second, in parentheses, to indicate that it is not original information but is provided as a convenient aside to assist readers unfamiliar with the original units. Christopher Parham (talk) 21:14, 7 September 2006 (UTC)

Part of the problem in this discussion, MJCdetroit, is the ambiguity in the meaning of "source". It is only as a first approximation that the immediate source used by a Wikipedia editor to find the information matters, unless it is a direct quote. In fact, for most things, you can find some sources such as this which use SI units, some sources which use non-SI metric units, some sources which use English units, and some which use various combinations of units.

What really matters is the units in which a measurement was originally made, or more often the units in which something is designed with nice, rounded numbers. If it is a measurement has been made many time by different people in different units, then unless someone's particular measurement is relevant, it doesn't matter much which comes first.

Where it does matter, switching them doesn't improve clarity. Rather, it hides important information as to the actual precision of the measurement. It helps identify cases in which the conversion has been stated with way too much precision, or sometimes way too little precision.

In some long-archived discussion, I discussed how putting the original first might have helped discover a conversion error involving the use of gallons in Canada, where the value was given in a converted number of litres first, then expressed in gallons in parentheses--U.S. gallons, not Canadian gallons. As a result, people thought that the numbers were correct (as long as you realized that what looked like a conversion from litres went to U.S. gallons), but in fact the number of litres was wrong, because it was actually from an original in Canadian gallons but converted as if those gallons had been U.S. liquid gallons. Gene Nygaard 08:59, 8 September 2006 (UTC)

I'm unaware that “the source value should be provided first since it is the most accurate” is a common convention. I see several reasons I’d prefer not to use it:

It is unclear that this convention is common in general publications; most references generally select one (perhaps Metric) as the first and the other (perhaps English) as second. Wikipedia shouldn’t create new conventions.
Using that convention leaves ambiguity and internal confusion. For example in a recent edit to Svalbard#Geography two references were cited: Encyclopædia Britannica and Web publication of Statistics Norway. The first provided English units but did not cover all islands in the archipelago. The latter used only Metric, but covered all islands. I chose to go with Statistics Norway for metric since it was more recent and comprehensive, but followed Encyclopædia Britannica when leaving the English units added by a previous editor. Both were citable—so which should go first?
If one used that convention, then I could wind up with English cited first at one point in a paragraph and metric another. For example one could use Encyclopædia Britannica for island areas, but Statistics Norway for glacier field areas, with the unusual format of an island area of 500 mi² (1280 km²) which has an glacier field area of 1020 km² (400 mi²).
All numbers should be referenced. Wikipedia:Verifiability, which is official policy, indicates “The burden of evidence lies with the editors who have made an edit or wish an edit to remain. Editors should therefore provide references. If an article topic has no reputable, reliable, third-party sources, Wikipedia should not have an article on that topic.” This most certainly applies to numbers. So capturing the original units is no big deal.
A footnote advising which are the derived numbers is unambiguous. Footnotes are easy using the formatting of Wikipedia:Citing sources.

Skål - Williamborg (Bill) 03:57, 8 September 2006 (UTC)

I'm not sure why you say capturing the original unit is no big deal; the original unit is the most accurate value we have. When you convert you are degrading the accuracy of the information in order to present it in another, possibly more convenient way. The reason we should present the converted value in parentheses is because it is worse information. Any editor who can comprehend the source value will want that more accurate figure. Generally, I would describe accuracy as the overriding concern. In the situation like you describe, for instance where you are consciously drawing a comparison between figures, they should be in the same unit. As you say, footnotes can be used to explain that the source value was converted to elucidate the comparison. Christopher Parham (talk) 04:51, 8 September 2006 (UTC)

Apologize for lack of clarity. Of course capturing the "original unit" is a "big deal". All I'm suggesting is the appropriate way to capture it is already established by Wikipedia. Use references to capture which unit was used by the source. Williamborg (Bill) 05:14, 9 September 2006 (UTC)

For some amusing and instructive examples of what can go wrong when converting units, see this. Scroll down to the image of the kid holding the giant ruler. -- Donald Albury 11:45, 8 September 2006 (UTC)

Those of us who learned engineering using sliderules have long been amused by the computer generations penchant to make conversions without being aware of the number of Significant figures. Donald Albury's joke above is a classic example of such nonsense... Thank for the laugh - Williamborg (Bill) 05:21, 9 September 2006 (UTC)

I cannot do much more than second what Gene Nygaard & Christopher Parham have written. Yes, "the source value should be provided first since it is the most accurate" if this leaves an article with inconsistancy as to which system appears first, so be it: this is the lesser of the two evils.

If this is a new convention, it won't be the first Wikipedia has created. For those of us who care about accuracy it would be a convention worth adopting or even creating.
I don't agree that "Using that convention leaves ambiguity and internal confusion." I don't see how Bill's Svalbard example shows ambiguity or confusion. This is just an instance where you have two sources using different units. In this case both systems are on equal footing. I'd suggest wherever we have more than one system on equal footing we choose the metric system unless we have good reason not to.
"If one used that convention," Bill points out "then I could wind up with English cited first at one point in a paragraph and metric another." Yes, this is true but it's a small price to pay for accuracy.
"All numbers should be referenced." Bill points out. Again, this is absolutely true. From this Bill concludes "So capturing the original units is no big deal." This conclusion is false. I don't happen to have a copy of the Encyclopædia Britannica on hand. I doubt that there even exists a copy within a 16.09344 kilometre radius. You can't expect readers to have easy access to any and every book quoted as a reference. External websites as references are not necessarily any better in this respect and are often worse seeing as they can change.
"A footnote advising which are the derived numbers is unambiguous." Yes, this would be a good solution. However, making a footnote does take a little more work.

Indeed I'd even suggest the second recommendation be made stronger. At the moment it reads as follows.

If editors cannot agree about the sequence of units, put the source value first and the converted value second.

I'd suggest it be strengthened to something more along the lines of the following.

Put the source value first and the converted value second unless there exists a compelling reason to choose another sequence of units.

Or perhaps something like this.

Put the source value first and the converted value second unless editors agree that there exists a compelling reason to choose another sequence of units.

Jimp 08:24, 26 September 2006 (UTC)

Of course, using footnotes negates the need for using order to show which is the source unit. We could make the recommendation that wherever appropriate the metric unit be put first and if this is not the source unit then that the source value be include in brackets with a footnote stating that this is the source value. Jimp 04:33, 3 October 2006 (UTC)