Wikipedia talk:Manual of Style (dates and numbers)/Archive 38

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 35 Archive 36 Archive 37 Archive 38 Archive 39 Archive 40 Archive 45

ISO 8601

What advantages does the ISO format have over spelling out the month? If the month is written out, the date is shown exactly as pronounced, albeit perhaps in a different order. There can be no confusion. Using a number for the month can be harder to read for someone not used to it, and some particularly clueless non-Americans (hah! I get to refer to a subset of non-Americans as being stupid for once!) may mistake the month as the day. What advantages does it offer? All I can see is that it's very handy for computer storage, since alphabetizing will automatically put everything in the correct date order, but that's hardly relevant, is it? —Simetrical (talk • contribs) 03:35, 10 February 2006 (UTC)

Okay, they're also useful for people who don't know the English month names. Given that most other major languages have very similar month names anyway, I don't view that as an issue either. —Simetrical (talk • contribs) 03:39, 10 February 2006 (UTC)
Everyone that can read English can, or can be assumed to (i.e. should be able to), recognise the month names. This is, after all, an English encyclopedia, presumably for English speakers.
Because there is no such date format as YYYY-DD-MM, it is highly unlikely that anyone would mistake the day for the month. Except for America, the order is always in order of precision/magnitude/length (don't know how to describe it). Only the American system is out-of-order. (Sorry, I just had to retort that non-American comment :P)
But anyway, I personally see no advantages (in fact, the opposite) and am happy to see it go. Unless they use it in tables or somethign like that. I suspect there's a few who disagree. Neonumbers 11:29, 10 February 2006 (UTC)
I do not quite understand what is being asked. Is it:
1. Why does ISO 8601 exist?
2. Should the MoS refer to ISO 8601?
3. Should the MoS section on ISO 8601 be improved?
bobblewik 11:53, 10 February 2006 (UTC)
I think it's "should the editor be allowed to type ISO 8601 dates". The ISO format is quicker to type, and it's converted by date preferences so it shouldn't matter. However, I never use it because I'm told anonymous users and new users have the date preference "don't convert anything". I wish they had a default of either February 10 or 10 February — I don't care which. Maybe that way we wouldn't get so many edits "correcting" the date from one to the other. Stephen Turner (Talk) 11:55, 10 February 2006 (UTC)
Ah. If forbidding use of ISO 8601 is the question, I would not vote for that constraint. The section does not add much value but I worry that removing it might lead to editors claiming that format is not permitted. If my worry is unjustified, then it could be removed. Otherwise it could be improved. Wikipedia is almost unique (even on the web) in being a specifically international resource so internationalisation is a good thing for us. I definitely think ISO 8601 should be permitted, even if the option is hardly ever used. Like you, I am also a non-user of date preferences but don't tempt me into debating that whole issue here! bobblewik 12:45, 10 February 2006 (UTC)
The English Wikipedia should be maximally comprehensible to English speakers. ISO 8601 is rarely used in English-speaking countries except maybe in computing, and even if it's not ambiguous, it will strike our readers as odd and hard to read. The date preferences are there, but a vast majority of our readers are unregistered, so they'll still see whatever's typed exactly as-is. Let's adhere to normal English-language dating conventions, and if someone really wants ISO, they can always change their preferences. —Simetrical (talk • contribs) 06:37, 12 February 2006 (UTC)
Indeed, date preferences do not work for unregistered users. Many registered users don't use them either. I wish we had statistics on usage. The date preference mechanism is a cure that causes more problems than the disease. I would be happy to see it abolished. But I am still a little confused as to what is being suggested. Please propose the new text that you would like. bobblewik 11:35, 12 February 2006 (UTC)

I would replace the section

The ISO formats, e.g., 2004-02-17 are less ambiguous but also not as widely comprehended. Redirects for other ISO forms should always be created. The YYYY-MM-DD format currently only follows the style of ISO 8601, but not the proleptic Gregorian calendar.

to read

The ISO formats, e.g., 2004-02-17 are less ambiguous but also not as widely comprehended, and therefore other formats are normally preferred. If for some reason using ISO is necessary, redirects for other ISO forms should always be created. The YYYY-MM-DD format currently only follows the style of ISO 8601, but not the proleptic Gregorian calendar.

And under "Incorrect date formats", change

Do not use numbers to express a month, except in full ISO 8601 format, which always includes the year. Always express a month as a whole word (for example, "February") to avoid ambiguity. In the ISO 8601 format, a leading zero is always added to single-digit months and days.


Never use numbers to express a month, unless for some reason the ISO 8601 format (which always includes the year) is necessary. Always express a month as a whole word (for example, "February") to avoid ambiguity.

Again, after consideration, I can't find a single reason that ISO should be used. I know it's annoying for me, at least, to figure out what the correct month is for any month-numbered format, and it's even more annoying when the numbers are in an odd order (such as, say, with the year first). As far as I can tell, month-named formats are completely unambiguous to everyone and nobody has any difficulty reading them. Maybe people are using ISO because they prefer it stylistically, but I think the ease of reading for some is more important than the stylistic preference of anyone. —Simetrical (talk • contribs) 04:08, 13 February 2006 (UTC)

Working in an international company, we used to have a mix of date formats (MM/DD and DD/MM) in communications with our partners in other parts of the world. Replacing this by uniform usage of ISO 8601 this has eliminated uncertainty and misunderstandings. I never met any people that did not immediately understand this format, but it might take some time to get fluent with it. Putting the year first gives the clue that this format is used. The proposed changes above unduly deprecate the ISO standard. I propose to change the above mentioned sections as follows:

The ISO formats, e.g., 2004-02-17 are not ambiguous and are therefore preferred. The YYYY-MM-DD format currently only follows the style of ISO 8601, but not the proleptic Gregorian calendar.


If no year is present, do not use numbers to express a month. In such cases express a month as a whole word (for example, "February") to avoid ambiguity.

Woodstone 08:42, 13 February 2006 (UTC)
We normally assume that everything is allowed unless forbidden. However, in this case we are specifically stating that ISO 8601 is permitted, presumably we think it is necessary to prevent a future dispute. Giving permission seems to be all that is required. I have no idea how or why there is a reference to proleptic Gregorian calendar and I do not think it is necessary. I am not aware of any other clear and present problems with this issue so we could try something like:
  • ISO 8601 date formats such as YYYY-MM-DD (e.g. 2006-02-13) are permitted.
bobblewik 23:05, 13 February 2006 (UTC)
I think this is actually a technical comment, that the wiki-linked YYYY-MM-DD does not follow the proleptic calendar as the ISO 8601 format recommends. The ISO 8601 days count one day at a time straight back into negative years (where -0001 is 2 BC). The Gregorian calendar doesn't actually cover years before adoption in the 16th century, and previous Julian calendar dates disagree with Gregorian dates. The comment just says that when a YYYY-MM-DD day appears, it's a Gregorian date after 1582 and a Julian date before 1582. —Daelin @ 2006–02–15 05:55Z

I don't see how "14 February 2006" is ambiguous. "12-02" or "12/2" are ambiguous, of course, they are already banned. ISO 8601 is not English, at the very least, it is not accurate prose. While its use in tables, to assist alignment or to save space, might work, I would not expect to see, "On 2006-02-14, some guy was lonely".

Let's forget about those who have set preferences for a moment, and consider new or unregistered users. (Remember that a format doesn't need to actually need to be "allowed" as written for user preferences to use that format, technically speaking.) I'd bet anything most people probably wouldn't care about "February 14, 2006" vs. "14 February 2006", but "2006-02-14" would throw them off-balance a bit. This is an encyclopedia. It uses a formal, accessible tone. It is not the responsibility of the reader to become fluent in a not-often-used date format (I won't say "rarely", the point is, in this type of writing, we generally spell dates out.) Wikipedia is an international English encyclopedia, not an international multi-lingual encyclopedia. Neonumbers 06:06, 14 February 2006 (UTC)

This is why I wish new and unregistered users were given a default format; so that they wouldn't see 2006-02-14 even if someone wrote it. Stephen Turner (Talk) 09:29, 14 February 2006 (UTC)
However, given that that isn't the case, I think it is right to mandate against the use of ISO 8601 dates. I know that I did use them at one time, not realising that unregistered users would see them. I would propose:
  1. Drop the last two examples from #Date formatting.
  2. Rewrite #ISO date formats as follows:
    ISO 8601 dates, for example 1958-02-17, should not generally be used. Even though the software will convert such dates according to users' date preferences (for example, [[1958-02-17]] → 1958-02-17), new users and unregistered users do not have any date preferences set, and will therefore see the unconverted ISO 8601 date. This format is often not understood, and so should be avoided.
  3. Rewrite the first bullet point of #Incorrect date formats as follows:
    Do not use numbers to express a month; always express a month as a whole word (for example, "February") to avoid ambiguity.
Stephen Turner (Talk) 09:51, 14 February 2006 (UTC)
As my following comment indicates, I strongly disagree with this idea. We should not rule against what is sometimes a good practice because of a current niggling technical detail. Yes, ISO 8601 dates should not usually appear in prose. No, we should not prohibit the use of ISO 8601 dates. I don't think ISO 8601 dates should appear in place where they should be wikilinked. A simple style guideline of "write out months in prose" handles that. —Daelin @ 2006–02–15 06:00Z
ISO 8601 dates are independent of context, including language. That's the benefit. It's also much easier to programmatically interpret. As far as good English style is preferred, the month should be spelled out. ISO 8601 is specifically intended for labelling, not prose. It belongs in mail headers, on soda cans, in tables, in charts, and anywhere a date should appear outside of a complete sentence. Anywhere you'd concievably want to use a numerical date, use ISO 8601. When the context is obviously an American English paragraph, there's not much benefit.
Please do not discourage ISO 8601 use, but emphasis the rule that things should generally be spelled out in paragraph.
I'd also like to comment that when a series of dates appear in a paragraph, I'd personally appreciate it if the ISO 8601 date were presented parenthetically so that I can more easily compare the sequence of dates. I say ISO 8601 because it's brief, significant-digit ordered, unambiguous, linguistically neutral (even English/American), and easily understood by everyone.
Daelin @ 2006–02–15 05:45Z
Thanks for your reply, Daelin. I don't think our positions are actually that far apart: maybe I just need to phrase my proposal a bit better by judicious insertion of "in normal text" or something like that. Let me ask this: is there ever a situation in which it's a good idea to have a wikilinked ISO 8601 date? That's really what I'm trying to "forbid". Stephen Turner (Talk) 10:18, 15 February 2006 (UTC)
That's a little tricky. I don't think there's ever a situation in which it's a good idea to have Wikipedia do date translation from ISO 8601. It seems that's the primary purpose of linking dates. It's a date though, so all the pedestrian usefulness of linking dates applies. It's just that, generally, where an ISO 8601 date should appear (not in prose), it should not be translated. This is doubly true because ISO 8601 dates should be proleptic, but wiki dates are not.
Now, if you're interested in the events that happened in a particular month or day, you can always look it up in Wikipedia the normal way. That's true of all wiki links, though. However, I'd be willing to sacrifice ISO 8601 links for the reasons we both cited.
Daelin @ 2006–02–16 01:09Z

I skimmed this section, and I just wanted to mention that one reason that most English speakers do not recognize the ISO 8601 format is becuase they do NOT see it. If people saw it used, they would grow accustomed to it, but they do not becuase it's not used. I agree that it's uncyclopedic for something to be written "On 2006-01-29...," but just becaue it doesn't look great in formal prose does not mean it should not be used elsewhere. And somewhere way up there was a comment that US usage of dates is not in logical order, being m/d/yy or m/d/yyyy (yes, I do mean m and d rather than mm and dd). However, the same is true (to a greater extent, possibly) of the d/m/yy or d/m/yyyy formats--logically, if one is looking at something, seeing the day first gives one no perspective of the range of times or even the order of occurence of events--instead, seeing the month first at least helps, but seeing the year first makes the most sense, it's in largest to smallest order and can be "alphebetized" or ordered easily and logically while still being chronolical. //MrD9 08:46, 20 February 2006 (UTC)

I completely agree with you. In my opinion, the best would be to adhere to the ISO standards as much as possible. . . that's what they were created for. Wikipedia is not your run of the mill encyclopedia, and I suspect that is not easy to find an english language encyclopedia in most non-english speaking countries, but you can definitely find the internet. And, of course, English is a required language in most countries.
In Colombia we use 1.000.000,00, yet Wikipedia seems to only recommend 1,000,000.00. Again, to eliminate confusion, ISO dictates either 1000000,00, 1 000 000,00, 1000000.00, or 1 000 000.00. There are also other issues of a similar manner that I won't go into.
Finally, I would also advocate the use of parenthesis when confusion is bound to happen in any general setting. For example:
The pencil is 11 cm ( 4 3/8 in ) long.
The Real Madrid is one of Spain's premier football ( soccer ) teams.
On that year, a CEO made, on average, 21 500 000 USD ( 21,500,000 USD ).
Keep writin'! ( writtin'? Maybe we should also standardize English spelling. . . just kidding! ) --GusCaicedo 2006-02-21T13:59-05

GusCaicedo, this is the English Wikipedia, and therefore quite naturally we follow the conventions of English, not other languages. In English-speaking nations, 1.000,000 is never used; it's always 1,000.000 or, in some scientific contexts, 1 000.000. There's no ambiguity in English there. Granted, there are many non-native English speakers on Wikipedia, but it's primarily intended for English speakers; otherwise we'd ban the use of less-common words where commoner ones would work, for instance, and maybe use more "logical" spelling or grammar despite the increased burden this would put on English speakers.

Anyway, are most of us agreed that at least ISO shouldn't be used in prose? What about dates of publication in references? —Simetrical (talk • contribs) 03:01, 28 February 2006 (UTC)

Dates of birth and death.

I would like to add the fairly standard format abbreviation fl. (floruit or "florished") to the DOBAD section e.g.

  • (fl.1214 - 1244)
  • (fl.1244, 1316)
  • (1202, fl. 1622)

This is as unambiguous as I can make it without (re)introducing b. for born and d. for died. Comments? Rich Farmbrough. 18:21, 10 February 2006 (UTC)

Perhaps it should be a link fl. Rich Farmbrough. 18:33, 10 February 2006 (UTC)
It has its use, but I would use it only when necessary and always link it, because I think a lot of people wouldn't be familiar with it. Stephen Turner (Talk) 18:43, 10 February 2006 (UTC)
Strangely enough I have come across fl. used twice on wp since mentioning this. Rich Farmbrough. 23:20, 12 February 2006 (UTC)
In lists of people, by name (my own concern) or in Dabs, the fl. lk is uneeded and a bad idea. These navigational tools do their job best by lking, with stunningly rare exceptions, only to the articles (bios, where fl. is concerned) that are the respective bios of the people listed or the articles that might have had the title if there weren't competititors for the title. There is no reason to lk fl. in lists and dabs, in light of two facts:
  1. There's no navigational need for users to know more than that the date after fl. has something to do with the person they are seeking (which BTW is all that "fl." communicates to anyone).
  2. Those who wonder what it means have a logical place to find out: the lk that is separated from it by the preceding blank and open-paren, which leads to an article that (supposedly) has the dates in it and is a place far less burdened by the lk to floruit.
--Jerzyt 23:26, 15 February 2006 (UTC)
Dab pages should not have any links, apart from to the articles. That's already covered by Wikipedia:Manual_of_Style_(disambiguation_pages)#Individual_entries. I don't have a strong opinion about lists. Stephen Turner (Talk) 10:06, 16 February 2006 (UTC)
That isn't an absolute rule, but a qualified one, though I'd say it should apply in this case (the most obvious time when something else could be linked is when the dab link is for the time being a redlink to something that should eventually be written). In any case, that's what Jerzy's explanation implied wrt the issue at hand. Gene Nygaard 15:52, 16 February 2006 (UTC)

Village pump discussion titled: "Is it permissible to implement the guidance in the Manual of Style?"

There is a Village pump discussion (sparked by the date link dispute) titled: . "Is it permissible to implement the guidance in the Manual of Style?". Interested parties are invited to contribute.
bobblewik 15:43, 11 February 2006 (UTC)

Percent: space or no space

Before the current guidance on percent, we had a long discussion about:

  • Editors must not use a space i.e. '7%'
  • Editors must use a space i.e. '7 %'
  • No guidance i.e. both options are permitted.

I have not looked back at the archive but I seem to remember that we decided not to mandate. Can we discuss this again because User:Freakofnurture tells says it must be the first option of the three. If this is the case, the MoS needs to say so. bobblewik

It does not necessarily have to be the first option of the three, but "deciding not to mandate" means not running around with a bot changing them to your personal preference en masse. Are you noticing a pattern here? Ambi 14:57, 12 February 2006 (UTC)
No space, of course. I've never seen any usage where there is a space between a percent sign and the number. Do you have any links to websites where there is a space between the two characters? Talrias (t | e | c) 14:58, 12 February 2006 (UTC)
I wasn't in the previous discussion, but I think there shouldn't be a space, and that spaces should be removed when they occur. Does anyone have access to a major print manual of style which covers it? Stephen Turner (Talk) 16:37, 12 February 2006 (UTC)
I wasn't involved in the previous discussion either, but the space between the number and the percent just seems weird and uncommon. I definitely don't think we should be changing things to a space followed by a percent sign without citations from "real" manuals of style. JYolkowski // talk 17:12, 12 February 2006 (UTC)
Well, I think good practice usually dodges the issue. In science writing, for instance, "per cent" is usually discouraged, in favor of more exact descriptions of density, concentration, or ratios with zero-power units (eg, 4 m / 10 m = 0.4 m0 = 0.4). While the BIPM manual does not mention percent (except as "per cent" in an older reference to the purity of the kilogram standard), ISO 31-0 does, and NIST describes it as recommending a space.
  • This system is 53% efficient. (Conventional)
  • This system is 53 % efficient. (ISO 31-0)
  • This system is 53 percent efficient. (common)
  • This system is fifty-three percent efficient. (Chicago)
That said, I think most style manuals recommend spelling out "percent", and Chicago at least recommends spelling out numbers up to 100, which demands the unit for "0.01" be spelled out. At the very least, most manuals recommend spelling out numbers up to ten, so percent should probably be spelled out for consistency, considering how often percents fall below ten. —Daelin @ 2006–02–13 02:54Z
That doesn't cover things such as charts, however, where all style guides would agree that using numerals is appropriate. As for the issue at hand, no reputable English-language publication that I've heard of uses a space, ISO's bizarre and arcane standards be damned. The whole point of the MoS is to have us use the same conventional styles as reputable print sources so that we look polished, and a space is completely unheard-of in encyclopedias or other normal sources.

(It might be that science journals do things differently, though. If so, that would give me pause.) —Simetrical (talk • contribs) 04:13, 13 February 2006 (UTC)

Well, I'm a little disappointed. A cursory search reveals many science journals either don't use a space, or specifically note that no space appears.
The Journal of Animal Science uses no space. The Wisconsin Undergraduate Journal of Science notes a space does not appear before the percent (%) sign, but does appear before the degree-temperature (°C) sign. I don't have the APA style guide, but the examples of APA percent usage I see do not include a space. The Chicago Manual of Style indicates no space.
Many note that "percent" or "per cent" should appear in text, and the symbol (%) appear only in tables and figures (eg, New Mexico Journal of Science). This neatly sidesteps the issue because properly typeset tables place the units in the heading cells. —Daelin @ 2006–02–14 01:14Z
Daelin has a very good point: numbers up to ten should definitely be spelt out, numbers up to a hundred should arguably (possibly) be spelt out (in prose). It could be argued that seven percent is a bad example, and should be replaced with forty-nine percent (or 49%) or something like that.
In prose, I would always write out "percent", whether or not the number (more than ten) is "49 percent" or "forty-nine percent"; my reasoning for this is that the "49" isn't really a "numeral", just a commonly accepted "abbreviation" (or shortened version) for "forty-nine". (I just realised that doesn't match the manual's current wording! Any views on this?)
Where "49" is a "proper" numeral, maybe in science articles, I don't know where, I'd say no space (49% not 49 %).
My user page says I'm on a wiki-break, this is my rare post, I may not be back for a while. Neonumbers 05:24, 14 February 2006 (UTC)
I see this debate is drifting. Spelt out formats are dealt with separately. Can we restrict ourselves to simple issue of the symbolic format please? bobblewik 07:39, 14 February 2006 (UTC)
No, because they are not unrelated. We have no argument about the need for a space between the number and the spelled out word or words (then it's just a question of whether there is a single word or a space making it two words).
But I seem to remember someone going on a tear and eliminating many of the spelled-out words, and consequently making the question of what to do with the symbol. You did do that, didn't you, Bobblewik? So maybe if you go back and put back the spelled out words where they were, we can limit the discussion here to the question of what to do with spacing when it is the symbol involved. Gene Nygaard 03:34, 16 February 2006 (UTC)
I think it looks better without a space, but I think it looks better still with the word spelt out. I'd prefer it if the MoS said it should be spelt out. Also, to throw another option into the mix with an example from the world of science, MNRAS journal style says it should be 'per cent', not 'percent', so I'm in the habit of using that. Worldtraveller 12:24, 14 February 2006 (UTC)
Right you are, bobblewik. I think this is one of those areas where accepted standards contradict accepted practice. When it comes to liberal studies and even soft science, I think we should let them get away with the spaceless variant, since we let them get away with all manner of greater mistakes already. :)
However, I think that in tables the space should be used, so that the percent symbol will align with other unit symbols. The only other unitless symbol which does not require a space is the arc-degree (°). Though, generally, the units should appear in the table headers and not in the data cells. If this isn't possible, the table format probably needs significant improvement.
Daelin @ 2006–02–14 20:52Z
I don't understand your point about "tables" at all. If you have one column including the numbers (preferably right aligned) and another column containing the units (left aligned), then the space will automatically come in with the padding in the cells, etc. But otherwise, if you just have a column with a mixture of units, there is no way the numbers are going to align themselves if some of the units are "m" and some "kPa" and some "kJ/(mol·K)" in any case. Even just m and km keep the numbers from aligning if both the number and its units appear in the same table cell. Gene Nygaard 03:43, 16 February 2006 (UTC)

Two points:

  • space or no space. I can't think of the last time I saw a space. I think it should definitely be no space. (And I so agree with the comment earlier "ISOs bizarre and arcane standards be damned". Personally I think ISO should be banned from WP. Hardly anyone uses it. Few would understand it. It is IMHO pointless on WP.)
  • Re per cent, percent or %. IMHO we should always go with %. The reason is simple. Contrary to some stuff written earlier, per cent is used quite widely, but largely in some disciplines and in some forms of English. So using per cent or percent is another of those flashpoints (US versus UK spelling, BC/AD versus BCE/CE, etc) that we should avoid like the plague. % is used everywhere. As to using the written form before some numbers and % before others, that would be unwise. Closed edit sources can ensure a consistency. Open-edit sources like here cannot. So if we say use the word(s) before number 'x' and % after, we'll simply have edit wars erupting, or have to do constant explanations as to why we write 49 per cent/percent and 50%, or whatever the changeover number from one to the other is. Saying simply "always use %" dodges all the minefields. I doubt the space issue would arise because apart from ISOs (*make a spitting noise*) who else on the planet actually puts in a space? FearÉIREANNIreland-Capitals.PNG\(caint) 07:29, 19 February 2006 (UTC)

"first' vs "1st" etc

Is there a policy about "first' vs "1st" etc? does it go by "names of numbers" (spell out small ones? Bubba73 (talk), 04:31, 16 February 2006 (UTC)

Capitalization of unit names

Hi all,

Should we capitalize unit names? The problem came up when using amperes. I believe that we should only capitalize when refering to the scientist.

Tony 05:58, 19 February 2006 (UTC)

You are right, the unit names should be lowercase, even though unit symbols for units named after people start with a capital letter. One of the quirks of English is the capitalization of proper adjective identifiers of those units as in degrees Celsius and degrees Fahrenheit, however (in the old days it was degrees Kelvin, but now that kelvin is a noun, the unit name kelvins is lowercase). Gene Nygaard 09:53, 19 February 2006 (UTC)

Binary unit prefixes

a table in this section shows Kb for kilobyte; this should be KB, to match the rest of the table, to match common usage of KB and to match the rest of wikipedia. I just do not know how to find the source of the table to make this change. Help! please. Thanks Hmains 20:02, 19 February 2006 (UTC)

done. −Woodstone 20:28, 19 February 2006 (UTC)
Actually, common usage is KB = 1024 bytes, so using KiB = 1024 and kB = 1000 bytes is correct. — Omegatron 03:33, 20 February 2006 (UTC)