Wikipedia talk:Manual of Style/Dates and numbers

Archives
1, 2, 3, 4, 5, 5a, 6, 7, 8, 9, 10, 11, 12, 13, 14, 14a, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48 49, 49a

See also:

• Wikipedia talk:Naming conventions (calendar dates)
• Wikipedia talk:Timeline standards
• Wikipedia:Measurements Debate

---

Note on Archives:

The recent discussion on linking of dates from 9 March 2006 to 13 April 2006 is in archives 42 through 46, plus 48. 42, 43, 44, 45, 46, 48

---

Numbers from zero to ten

I propose two changes to the numbers section. First, move the general comments about numbers to the top, before the comments on large numbers - seems more logical that way. Second, update the general comments section as follows. The comment that "some users prefer to spell out one to ten" is not helpful in a Manual of Style. We should be bold and have a standard. I suggest the following which is common in many manuals of style.

"In the body of an article, numbers from zero to ten should be spelt out as one, two, three, etc. Numbers greater than ten should use numerals. An exception is for multiple groups of numbered items in the same sentence - use the same format for like items, e.g. "They were seated in chairs ten and eleven in rows 5 through 15". In tables or other lists, numerals would generally be used for all numbers. It is considered awkward for a numeral to be the first word of a sentence: either recast the sentence or spell the number out. A consistent approach is required within each article. " Rillian 13:07, 26 May 2006 (UTC)

I have a feeling that it's like that because there isn't agreement over this internationally, just like English spelling and AD/CE, etc. If no one objects, however, then a standard is fine with me. Also just to clarify, under these rules, would it be acceptable to start a sentence with the word "Eleven" or not? --Spangineer^[es] (háblame) 13:15, 26 May 2006 (UTC)

Eleven would be fine, it's numerals that you want to avoid at the start of a sentence, e.g. "21 cats walked past the store" versus "Twenty-one cats..." Rillian 14:15, 26 May 2006 (UTC)

I believe this is consistant with MLA style. I don't really know what APA has to say about it, though. -- uberpenguin @ 2006-05-26 14:13Z

According to my MLA handbook, their perference is that any number that can be spelled out in one or two words should be but to use numerals for others. So MLA would say "ten, twenty-one, one hundred, thirty million", but "3.75, 875, 25 thousand, 421 million". If we adopted that standard, it would be a broader statement than just zero to ten, and 11 and greater. Rillian 14:19, 26 May 2006 (UTC)

don't agree. I think it's excessive bond for editors, which really does not help. -- tasc ^talk_deeds 14:20, 26 May 2006 (UTC)

Here are the relevant sections from the APA manual

• Spell out common fractions and common expressions (one-half, Fourth of July).
• Spell out large numbers beginning sentences (Thirty days hath September . . .).
• Spell out numbers which are inexact, or below 10 and not grouped with numbers over 10 (one-tailed t test, eight items, nine pages, three-way interaction, five trials).
• Use numerals for numbers 10 and above, or lower numbers grouped with numbers
• Use combinations of written and Arabic numerals for back-to-back modifiers (five 4-point scales).
• Use combinations of numerals and written numbers for large sums (over 3 million people).

Rillian 14:39, 26 May 2006 (UTC)

Here are some notes from the The OAH Magazine of History Style Sheet

1. Whole numbers. Whole numbers from one through ninety-nine are spelled out in ordinary text, as well as any of these numbers followed by hundred, thousand, million, etc. The same general principle is applied to ordinal as well as cardinal numbers. 2. Consistency. Numbers applicable to the same category should be treated alike within the same context, whether paragraph or series of paragraphs. Do not use figures for some and spell out others. (E.g., "The population of Gary, Indiana, grew from 10,000 to 175,000 in only thirty years.) Rillian 14:41, 26 May 2006 (UTC)

I generally approve of the recent addition that was made in relation to this matter. Small numbers are often like crass little blops in a sentence, and WP has far too many of them. (Except that it might be one to nine spelt out, not one to ten.) Tony 16:05, 26 May 2006 (UTC)

This proposal is absolutely correct and should be implemented. - Centrx 22:26, 26 May 2006 (UTC)

To be added?:

Whole numbers from zero to ten are spelled out as words. Numbers above ten may be written as numerals or spelled out in words, where appropriate. Within a context or a list, style should be consistent. (Example: There were 5 cats, 12 dogs, and 30 birds. or There were five cats, twelve dogs, and thirty birds.) It is considered awkward for a numeral to be the first word of a sentence: either recast the sentence or spell the number out. Within each article, style should be consistent.

-- Centrx 18:41, 2 June 2006 (UTC)

I really like to "two words" guideline. Tasc said of one of the proposals "it's [an] excessive bond for editors" - not really, no-one has to follow the MoS when they're writing, MoS is a. guidelines and b. something that sub-editors (in the old sense) would apply, although if writers follow it so much the better. Rich Farmbrough 08:00 9 June 2006 (UTC).

So that would be one through twenty and intervals of ten on up to a hundred? I think there is a problem having a policy where 91 is written as a numeral and ninety is written out as a word, when the only difference is a hyphen and a generally one-syllable word. -- Stupid, blinko question added by me, Centrx on some day past

No, "any number that can be spelled out in one or two words". Rich Farmbrough 12:56 29 June 2006 (GMT).

Since the above paragraph is the result of agreement in the rest of this section and no one has objected to it, I will add it. It is still closer to what you propose than the current styleguide. -- Centrx 04:41, 12 June 2006 (UTC)

It might be worthwhile to also note that you'd almost never want to spell out numbers when used as a value in conjunction with a unit. -- uberpenguin @ 2006-06-12 05:19Z

I see no problem with "twelve kilometres", "nine kilograms", etc. —Centrx→talk • 22:39, 27 June 2006 (UTC)

Samples disappeared

Would someone point me to the discussion about removing the following two formats from the page:

[[1958]]-[[02-17]] : 1958-02-17

[[1958-02-17]]: 1958-02-17

Otherwise, I suppose we are to restore. -- User:Docu

The discussion is archived at Wikipedia talk: Manual of Style_(dates_and_numbers)/archive48. However there was a doubtful interpretation of what consensus means. A sizable minority pleaded for keeping these formats, but were denied. Thanks for the support here and I agree they should be restored. −Woodstone 18:09, 29 May 2006 (UTC)

Why would these formats be appropriate for prose? If the argument is that they are unambiguous, then they are still unambiguous if the alphabetic name of the month is used, and without hyphens, that is: 1958 February 17 -- Centrx 21:02, 29 May 2006 (UTC)

Thank you for pointing me to the archive. I tend to agree with the general idea, but would formulate the result a bit differently. As this wasn't specifically adressed, I suggest we readd the sample and remove the invention that they shouldn't be wikilinked (#ISO date formats). For tables, lists, DOB/DOD, the format is useful. -- User:Docu

They should definitely be wikilinked where they occur. However the first format seems an obtuse variation of the second and could be omitted, and I would like to see some encouragement to use ISO dates only in specific contexts. They still show up as full dates for most of us. Rich Farmbrough 08:05 9 June 2006 (UTC).

Dates of birth/death with additional spelling of name

At the beginning of biographies, there are several variations to add a second spelling, e.g. (Russian: Влади́мир Влади́мирович Пу́тин) to Vladimir Putin. From some of the solutions used:

(1) with semicolon

Vladimir Vladimirovich Putin (Russian: Влади́мир Влади́мирович Пу́тин; born October 7, 1952) is a Russian politician.

(2) with comma

Vladimir Vladimirovich Putin (Russian: Влади́мир Влади́мирович Пу́тин, born October 7, 1952) is a Russian politician.

(3) with parentheses

Vladimir Vladimirovich Putin (Russian: Влади́мир Влади́мирович Пу́тин) (born October 7, 1952) is a Russian politician.

Which one is the optimal or preferred one? Which other ones are in use? -- User:Docu

I have absolutely no idea, and judging by the fact that I'm the first response in (I think) two days, few others do.

If you ask me, I like the semicolon. Makes most sense. To be honest, don't know much on the matter. Neonumbers 07:42, 2 June 2006 (UTC)

I agree, the semicolon makes the most sense. A comma, certainly, is not the right punctuation for distinguishing such data, especially when a comma may already be used in the date, so a comma is not appropriate. Having multiple parentheses seems excessive and not clean formatting. -- Centrx 17:32, 2 June 2006 (UTC)

Thank you for your reponses. Semicolon looks fine to me as well. I will go with that for now and list more variants as I come across them. -- User:Docu

Non-breaking space before non-abbreviated unit

I’m certain that a non-breaking space is required between the number value and the unit if the unit is abbreviated, but is it required when the unit is not abbreviated? For example, is the   in 50 centimetres necessary? --HeteroZellous 19:37, 1 June 2006 (UTC)

If the number is written as a numeral ("50") then there should be a non-breaking space so it is not left hanging at the end of a sentence. If the number is written out as a word ("fifty") then there need not be a non-breaking space. -- Centrx 00:14, 2 June 2006 (UTC)

These should be handled by the software, not written in manually... — Omegatron 00:29, 2 June 2006 (UTC)

I agree that the software should handle this but I have just fixed a problem of the word "million" falling to the start of the next line leaving "£7" hanging precariously at the end of the preceding line. I had to use   to prevent a premature devaluation. -- Alias Flood 02:07, 2 June 2006 (UTC)

I agree too that it should be handled by the software (or, anyway, in a single point of maintenance). As a general remark (meta-guideline?), I would advise against complex guidelines such as "if it's abbreviated do this, if it's spelled out do this instead". In my experience on wikipedia they highly increase the error rate from both newcomers and expert editors, and consequently the amount of article fixes and edit history noise, for a very little gain. —Gennaro Prota^•Talk 16:17, 2 June 2006 (UTC)

Understood. Thanks everyone. --HeteroZellous 10:10, 2 June 2006 (UTC)

Re Centrix: That's not the impression I get, and it's not the way that the wiki code is currently (full disclosure—it was inconsistent a few weeks ago, and I changed it to its current state). Either way though, not a big deal; if that's the way people want it, let's make it more clear on the actual page. --Spangineer^[es] (háblame) 12:27, 2 June 2006 (UTC)

There are many formatting policies not implemented or enforced in the wiki code, that doesn't mean they aren't part of the recommended style. By the premise of the initial question, the nbsp in an abbreviated unit is not in the wiki code either. In software, this would be implemented by having a list of unit names, which could not be exhaustive? -- Centrx 17:29, 2 June 2006 (UTC)

One would hope that the WP:MoS would at least be the one thing that complies with the WP:MoS. If you click the section edit link at WP:DATE#Units of measurement, you'll see things like "the Moon is 380,000 kilometres (240,000 mi) from Earth". If policy is truly that   always follows numbers, that should be explicitly stated and the examples here be fixed. --Spangineer^[es] (háblame) 13:13, 3 June 2006 (UTC)

Just to be clear: the SI defining standard ISO-31 does not require a non-breaking space between the number and the unit symbol. It specifies a normal blank. −Woodstone 14:16, 2 June 2006 (UTC)

This applies to the abbreviated units too? Also, does the standard simply ignore it, or does it explicitly state that there should be a regular space? -- Centrx 17:29, 2 June 2006 (UTC)

The standard states "a space is always used to separate the unit from the number". This is true both for symbols (they are never called abbreviations) and written out names. There is no explicit mention of non-breaking. In their own documentation, I found by peeking at the source, they are not really consistent, but often use the html "nobr" tag, like in <nobr>123 km</nobr> to keep number and unit together. However this does not seem to work on wikipedia. −Woodstone 21:36, 2 June 2006 (UTC)

The <nobr> tag is no longer standard HTML. It has been superseded by the non-breaking space entity. — Gulliver ✉ 00:20, 5 June 2006 (UTC)

Uh? Has it ever been standard? —Gennaro Prota^•Talk 00:55, 5 June 2006 (UTC)

Good point. It is a proprietory Netscape tag. It is no longer necessary. The two modern solutions are   and <span style="white-space: nowrap;">...</span>. — Gulliver ✉ 12:59, 7 June 2006 (UTC)

"Do not append an s for plurals of unit abbreviations."

At least in the case of "lb" versus "lbs", it seems to be an AE/BE issue (I haven't yet found an unequivocal statement yet, but googling various combinations, and especially comparing US[1] and UK[2] dictionaries seems to reflect a BE/AE diff), so saying 'only "lb" is correct', and people routinely making that 'correction', doesn't seem compatible with 'all forms of English are welcome on Wikipedia' and related guidelines. 24.18.215.132 23:35, 4 June 2006 (UTC)

Is this an example where spelling out of the unit would be appropriate? Rich Farmbrough 08:38

"Whit Monday" and "Super Tuesday"

As these days always occur on Monday and Tuesday respectively, the articles Whit Monday and Super Tuesday link to Monday and Tuesday. I suggest that we specify here that repeat occurrences of days of the week or months should generally link. -- User:Docu

I don't think it needs an extra mention, the MoS already states that they should be linked if there is a specific reason to do so. -- Centrx 17:35, 5 June 2006 (UTC)

I think it's worth mentioning, just in case someone would want to see samples of specific (or "particular") reasons. -- User:Docu

What do you mean "repeat occurrences"? Rich Farmbrough 08:10 9 June 2006 (UTC).

Those happening generally on the same day/month, such as Whit Monday. -- User:Docu

Ah, I see what you mean now. Actually, this criteria does not directly correspond to exactly what is appropriate for linking. An article could use the word "Monday" multiple times, yet be only referring to a special chronology of events not appropriate for linking, whereas a holiday may very well only have one use of the word and should still be linked. —Centrx→talk 06:53, 15 June 2006 (UTC)

Repeat links to the same word is a different issue, not specifically addressed here. The event would reoccurr on Mondays, not the word "Monday". -- User:Docu

I still don't understand. Can you give some specific example of what you mean. Also I do not know why the guidance has been changed when nobody else seems to agree or understand. bobblewik 18:37, 15 June 2006 (UTC)

A nice sample referred to above is "Super Tuesday commonly refers to a Tuesday in early March.". I changed it as Centrx seemed to agree on the principle to the point that he suggested that it's not worth mentioninng as the page already suggests to link if there is a specific reason. -- User:Docu

Centrx, how should we formulate it to avoid a possible confusion with "repeat occurences of the word itself"? -- User:Docu

I don't know, maybe "Articles that document events that always occur on a specific weekday should link to that weekday" would be accurate, but it seems long. I still think that we should probably just keep the principle that it be linked if it is relevant. —Centrx→talk 08:14, 19 June 2006 (UTC)

As it happens that these are delinked, it's probably preferable to mention it. Similar to your suggested wording, "Weekdays and month names refering to events that generally occur on a specific weekday/month should link." would do. -- User:Docu

The following might describe it even better: "Month names when used to describe that an event/fact regularly occurs in a specific month should link." and "Names of the days of the week when used to describe that an event/fact regularly occurs on a specific day of the week should link." -- User:Docu

Example sentence in "Units of measurement" section

One of the example sentences was changed by Hardern for standardization purposes. The original sentence starts off in the metric system (English system) than switches to English (metric). I believe that this was originally done to show the look in both systems. The problem was is that one would probably never see a sentence structured in such a way. It wasn't the best example, so Hardern and later Woodstone changed the sentence so that it read: Use digits and unit symbols for values in parentheses and for measurements in tables. For example, "a pipe 100 millimetres (4 in) in diameter and 16 kilometres (10 mi) long". For clarity's sake and to show the reverse, I added: "or a pipe 4 inches (100 mm) in diameter and 10 miles (16 km) long", at the end of the example sentence.—MJCdetroit 03:22, 9 June 2006 (UTC)

It was me that created the example. As you suggest, it was intended to show both systems. However, I also intended it to show that source data should go first. You said that you think such a sentence is unlikely, but it is entirely plausible in some parts of the world (e.g. the UK). Furthermore, such 'source-data-first' sequences occur right here in Wikipedia (you will find plenty of examples in car and aircraft articles). But it was just an example and I don't mind it being changed.

See the above section titled Small deregulation of 'units of measurement' section.. It contains a proposal to remove this bullet entirely because it fails my 'guidance-about-guidance' (described in that section). Removal of the bullet would end the problem that you describe. bobblewik 15:26, 10 June 2006 (UTC)

Symbols for square miles

We have a variety of symbols for squares and cubes of units. Metric units are easy. Where it says 'sq km', I either expand it to 'square kilometre' or 'square kilometer' or I make it a symbol 'km²'.

However, non-metric units are less easy. I don't usually worry but mixing styles is inconsistent. For example the US state infobox e.g. Montana pairs 'km²' with 'sq mi'. I would not like to resolve the inconsistency by going to the language dependent form 'sq km'. In other articles we see 'mi²' and 'mile²'. What do other people prefer? bobblewik 21:47, 12 June 2006 (UTC)

We have here a contrast between scientific and traditional units. I see no objection in extending that to scientific (power 2) and traditional (square) surface measure. So a pair "km²" with "sq mi" should be ok. Another alternative is of course to convert to acres instead, avoiding the square altogether. −Woodstone 22:08, 12 June 2006 (UTC)

I prefer 'sq mi'. Here's my reasons why:

• The NIST gives two abbreviations for square mile—'mi²' (following after the metric pattern) and 'sq mi' (the traditional pattern). I like 'sq mi' over 'mi²' because it sets it appart from 'km²' and when quickly reading it won't be confused for square metres (m²).
• The U.S. state department uses 'sq mi' for square miles Example.
• The Encarta and Britannica encyclopedias uses 'sq km' and 'sq mi'. While the World Book spells out every thing.

In common usage you are more likely (in my experience) to see 'sq mi' than 'mi²' and I have never seen 'mile²' before. I have seen 'sq miles' used before and that would probably be fine for an abbrivated conversion in the text of an article but it probably would be too long for tables and infoboxes. That's why I believe 'sq mi' is the better abbreviation for square miles. Those are my thoughts. —MJCdetroit 00:41, 13 June 2006 (UTC)

That US DoS link you give actually uses the same ill convention as the encyclopædias you name, but with dots. The only thing worth learning from these is consistency. As we certainly won’t change “km²” to “sq km” this would mean we had to use “mi²” (although I consider abbreviations of non-metric unit not as symbols and thus would append a dot). We then would also have to use “mi/h” instead of “mph” etc. I’d like that, but you won’t get consensus on that as long as everyday experience in the USA differs from this. This English legacy is just a complete mess. Christoph Päper 07:41, 13 June 2006 (UTC)

Why not change "km²" to "sq km"? —Centrx→talk 08:56, 13 June 2006 (UTC)

Because "sq km" is not a valid SI unit. −Woodstone 17:58, 13 June 2006 (UTC)

The abbreviation 'km²' seems to be well established in Wikipedia. Just because we use 'km²' for square kilometres does not mean that we have to use 'mi²' for square miles and 'mi/h' for miles per hour (as suggested by Christoph Päper). We have the flexability to use what ever we see fit; hense the discussion. I believe that we should use the scientific/metric pattern for SI/metric units (i.e. 'km²' not 'sq km') and use the traditional pattern for customary units (i.e. 'sq mi' not 'mi²') as I stated above. —MJCdetroit 18:28, 13 June 2006 (UTC)

The square kilometre is indeed a valid SI unit, and "sq km" is an abbreviation of "square kilometre". Just because it doesn't use the "km²" notation does not mean it is not a valid SI unit. —Centrx→talk 22:43, 13 June 2006 (UTC)

I don't see what's wrong with using "km²" and "sq mi" throughout. Both notations are far more common than their counterparts, and IMO the added benefit is that they visually pretty distinct between themselves. The "consistency" of using "km²"+"mi²" or "sq km"+"sq mi" would be quite artificially imposed. Duja 09:41, 16 June 2006 (UTC)

Fully agreed. Anyone who who is concerned that using both km² and sq mi is "inconsistent" is advised to heed the words of Emerson. 165.189.91.148 20:03, 22 June 2006 (UTC)

There are quite a few different ways in which people write 'km²' (e.g. 'sq.km', 'sq. km.', 'sqkm', 'sq kms' etc). They can all be made consistent using a monobook tool. Anyone that wants to do this, see the Wikipedia_talk:Manual_of_Style_(dates_and_numbers)#Unit_format_tool section on this page. Once it is set up, you only need to click once on a 'units' tab. It deals with a lot of unit inconsistencies but it does not currently deal with the many different styles for square miles (e.g. 'sq.mi', 'sq. mi.', 'sq mile', 'sq m', etc). It also has a 'dates' tab to deal with solitary days, months, years. Hope that helps. bobblewik 11:20, 23 June 2006 (UTC)

As a matter of closure, can we conclude that:

• km² & sq mi are the preferred abbreviations and
• sq km & mi² are acceptable but are not preferred and
• that the following abbreviations maybe acceptable for use within the text but are not acceptable for tables and infoboxes: sq miles & mile²?

Also, 'sq m' should never be used for square miles (or square metres).—MJCdetroit 12:40, 23 June 2006 (UTC)

Personally, I'd discourage sq km & mi² as much as possible. "sq" versions are likely "Americanisms", while "²" versions are likely "Europeanisms"; normally, an American would be interested in the sq mi version and an European (and most of the rest of the world) in the km² version. Thus, the symbols should be presented in the most common notation. Duja 13:55, 23 June 2006 (UTC)

What is wrong with "sq km"? Why would that be only an "Americanism"? —Centrx→talk • 06:40, 24 June 2006 (UTC)

Maybe it's just my reflex from my native tongue and engineering education, but it looks fairly odd to me. As someone pointed out above, it's not a valid SI notation either. I'd rather let an e.g. Briton comment further... ["sq km" site:*.uk Google search] seems to contradict my opinion... Duja 08:08, 24 June 2006 (UTC)

I don't agree that it is an Americanism vs Europeanism (just type sq km into yahoo.fr). It is simply the traditional vs scientific way of abbreviation. There is nothing wrong with 'sq km', but what is most common? I agree with Duja that km² and sq mi are the most common abbreviations for square kilometres and miles. Therefore, we sould prefer/incourage the use of those two. —MJCdetroit 14:10, 24 June 2006 (UTC)

There is no reason given why "sq km" is not acceptable in SI. The name of the derived SI unit "square kilometre". "sq km" is an abbreviation of that name. While it is not the official symbol, it is an abbreviation of two words and there is no prohibition of it. "km²" is more common in scientific use, but on Google search, we find ~9.5 million hits for — "sq km" -wikipedia -"km²" — and ~10 .1 million hits for — km² -wikipedia -"sq km" —. This does undercount "km²", because for example "km2" has relevant hits but can't be used in the search because most of its hits are not the unit, but at the very least "sq km" is common. —Centrx→talk • 20:37, 24 June 2006 (UTC)

—Centrx→talk • 20:37, 24 June 2006 (UTC)

Quote from the official SI brochure (p 38):

It is not permissible to use abbreviations for unit symbols or unit names, such as sec (for either s or second), sq. mm (for either mm2 or square millimetre), cc (for either cm3 or cubic centimetre), or mps (for either m/s or metre per second).

So it is either km² or square kilometre (note: "km" is not an abbreviation, it is a symbol). −Woodstone 22:01, 24 June 2006 (UTC)

Okay. Still, "sq km" is common, and what is the ambiguity in "sq km", which is the justification for that rule? —Centrx→talk • 01:20, 25 June 2006 (UTC)

Proposal: Mention about spelling out numbers up to 100

While we currently have the manual say spelling out numbers up to 10, I propose that we at least in one sentence explain the standard formal English usage of spelling out numbers up to 100. I was asked by an editor "why spell out "fifty" and not "five hundred and sixty-two" and responded that it was formal English practice because it was simpler and less complex verbiage for numbers up to 100. We could also do this implicitly by separating numbers above 100 from those between 10 and 100, because we really don't want people spelling out (as they "may") 14583, however this would be still more complex because it should still be acceptable to spell out "thirty million". A suggestion above from the MLA handbook about spelling numbers out that can be put into one or two words may be the way, but this does seem unusual and it may be difficult to word it properly? —Centrx→Talk 23:54, 14 June 2006 (UTC)

Currency

In the examples, why does GBP have a space before the digits but the other currencies have no space? -- SGBailey 12:19, 17 June 2006 (UTC)

Not only GBP, but also CNR has a space. This gives away the key: the $ sign is customarily written attached to the numbers. This is retained if the $ is preceded by a precision which dolar is meant. The all-letter abbreviations like BGP and CNR are followed by a space. −Woodstone 14:52, 17 June 2006 (UTC)

Unit format tool

For those interested in making unit formats consistent, please feel free to use my 'unitformatter.js' tool.

To get the tool, copy the entire contents of User:Bobblewik/monobook.js to your own monobook. Then follow the instructions in your monobook to clear the cache (i.e. press Ctrl-Shift-R in Firefox, or Ctrl-F5 in IE) before it will work. To make the tool work, simply click on the 'units' tab in edit mode.

Regards. bobblewik 12:57, 18 June 2006 (UTC)

Clarification. After you activate the "edit this page" tab, the tool places a few tabs to the right at the very top of the page. When you enter either of those tabs, the tool makes a number of substitutions and shows them in "differences" mode. It is then up to you to cancel, save or continue editing. −Woodstone 17:14, 18 June 2006 (UTC)

What is the big deal with non-breaking spaces?

The guidance says: Use &nbsp; for the space (25 kg) to ensure that it does not break lines..

However, this guidance appears to be misinterpreted. Somebody recently said: the Manual of Style states that non-breaking spaces should be used with units. I don't see changing "20mm" to "20 mm" with a normal space as an improvement, since neither are in accordance with the manual of style.

I have always thought that the nbsp is vastly over-sold. Can we do something about this misinterpretation? bobblewik 19:26, 21 June 2006 (UTC)

Firstly, the SI rules prescribe a space, without specifying if it should be non-breaking. Secondly, since an average line contains say 120 characters, and a unit is mostly 2 characters, the chance that non-breaking has any impact is about 2%. So the tool gives at least a 98% improvement in compliance. −Woodstone 21:06, 21 June 2006 (UTC)

Could you please clarify the above? What do you mean by "98% improvement in compliance" and how do your derive it? I presume you mean that non-breaking HTML entitites should always be used. I think they are cumbersome. --Cedders^tk 09:21, 22 June 2006 (UTC)

The tool changes 20mm to 20 mm. So a non-standard format is changed into a standard SI format in all cases. However when the space happens to be at the end of a line, it is not according to the WP style guide. This happens when the end of the line is reached within the two characters "mm". So about 2/120 chance (assuming a linesize of 120). My personal preference would be to use a normal space and adapt the manual of style accordingly. −Woodstone 11:01, 22 June 2006 (UTC)

Why? What is the problem with nonbreaking spaces? Did one pick on you in high school or something? ;) Chuck 12:54, 22 June 2006 (UTC)

It really isn't that hard, and adding nbsp improves readability. Also, we can't always assume a linesize of 120 (since screen size varies), and for some articles (such as ones on tropical storms), there are alot of units used, increasing the likelihood of a problem. --Spangineer^[es] (háblame) 15:06, 23 June 2006 (UTC)

While I'm totally behind following SI style recommendations, I agree that using &nbsp; is very cumbersome. I see the utility of having a non-breaking space between quantities and units, but I don't think it's worth burdening the author with having to type a bunch of character entities. Someone suggested before that the non-breaking space might be better handled by the Wikimedia software, and I tend to agree. It wouldn't be terribly difficult to add the capability to add &nbsp; characters appropriately where known units are detected. -- uberpenguin @ 2006-06-23 15:17Z

That would be great, but the developers are overworked, so it might not get done right away. However, remember that Wikipedia is a work in progress—the number of authors who actually care about nbsp might be rather negligible, and that's OK. The idea of this style guide isn't so that we can delete people's badly formatted entries, but so that we can agree on the "ideal" style. If no one cares enough to implemenet it, fine, but I think there are enough sticklers like me roaming WP:FAC to at least keep our best articles looking great. 100% compliance with all guidelines isn't mandatory, it's the goal, and it's not a problem if some people focus on a few particular ones and ignore the rest. If enough people care, it'll all get done. --Spangineer^[es] (háblame) 15:24, 23 June 2006 (UTC)

I'd hack together the code for this myself if I knew anything about Wikimedia's development cycle and thought that it might actually make it into Wikipedia eventually. -- uberpenguin @ 2006-06-23 15:32Z

I have changed it so that the recommendation is less demanding (must/should), now it is a subordinate part of the item about using a space, and says "Preferably". —Centrx→talk • 06:36, 24 June 2006 (UTC)

Sorry for posting to an aging discussion, but I am the user quoted above; I said I don't see changing 20mm to 20 mm (with a normal space) to be a worthwhile improvement. I think that the context is important here: this was in a discussion about Bobblewik's unit formatter monobook tool. I feel that if you're going to use a tool that automatically formats units, that tool should format them with nbsp. Doing so results in a slightly higher false positive rate. I think it's silly to say that it's not worth it to put up with that and you'd rather just auto-add regular spaces.

I don't think that there's anything wrong with writing articles using regular spaces between units. It's a lot more effort to insert nbsp, and it's more important to spend your time writing a better article than worrying about the nitpicky details of unit formatting. However, I think if you're going to make it your mission to correct unit formatting, you should correct it all the way, to use nbsp. TomTheHand 17:27, 30 June 2006 (UTC)

Several people have asked me to add that to the scope of my tool. I understand the frustration that they have. I have not spent the time and effort to do what they want and to eliminate the additional false positives. Remember though that I am moving Wikipedia towards the solution you want, not away from it.

If the community wants to jointly develop these two 'units' and 'dates' tools, I would be delighted to share the code with the community and advise where I can. It is all 'open source'. Indeed, there is actually nothing to stop people copying it right now. bobblewik 17:55, 30 June 2006 (UTC)

You've definitely been a huge help. It's just that when I read the above quote I thought I sounded pretty irrational, and I wanted to step in and explain the way I feel in detail. TomTheHand 18:00, 30 June 2006 (UTC)

OK. My mantra is that there is no such thing as user error. All good-faith editors are rational almost by definition. I was attributing a deficiency to the guidance, not you. I apologise if it could be interpreted the other way round. bobblewik 18:20, 30 June 2006 (UTC)

Percentages

The manual currently states "The format of the numeric and percentage terms should match. Thus pair 7 with % and seven with percent." Yet, I found that when using Bobblewik's new tool numerous articles pair a numeral together with percent, as "45 percent". This seems reasonable to me, and it is acceptable in the Chicago Manual of Style. It is also very common elsewhere on the Internet and in printed text. What is the justification for requiring that the word percent only be together with a spelled out number? Is this a possible mis-application of the principle that numerals always be paired with numbers in a list or in the same context (see Wikipedia:Manual of Style (dates and numbers)#Numbers in words? —Centrx→talk • 22:34, 21 June 2006 (UTC)

Personally, I thing "45 percent" looks bad. if you're spelling it out, then spell it out. If you're using numerals, then use symbols. This is not particularly different than for currency. We don't use "50 dollars"; instead it's "$50". Chuck 12:57, 22 June 2006 (UTC)

The manual of style does not state that you must always pair numerals with the dollar sign, in "$50" rather than "50 dollars". It has an example which could be interpreted that way, but that would be over-reaching: the manual just ignores the matter for currency. If "45 percent" looks bad, why doesn't "45 people"? If it is because the number is so small, then that is a different matter not specific to percentages or currencies, whether numbers ought to be spelled out. Why is "percent" or "dollar" any different from the thousands of other words or countable nouns? —Centrx→talk • 00:05, 23 June 2006 (UTC)

From the article: "use US$100 or one hundred United States dollars'"

Your example using "people" is a little silly because there is no symbol for people. There is a symbol for percent. To answer your general question, "percent" and "dollar" are different because there are broadly recognized symbols that are generally paired with numeral representations. For example, it is sloppy to write "100 million dollars". It's either "$100 million" or "one hundred million dollars". This is the same reason you would not write "forty-five%". By your logic, this would be a perfectly reasonable alternative. Chuck 02:35, 23 June 2006 (UTC)

I know what the MOS says. That is an example. The examples are not exhaustive, a mere absence of an example does not mean a style is not acceptable; there are, throughout the MOS, perfectly acceptable formats for which there is not an example. There is no statement that pairing "100" with "dollars" is not acceptable.

The example of "people" was to elucidate how exactly you thought it was ugly when numerals are paired with words everywhere else.

Numerals paired with words are common and appropriate style. If you open up the New York Times, if you look at several books, you will find that it is rather normal. If you want to recommend against written English usage, you need to provide some better justification than simply saying it is "sloppy" or "ugly".

This does not entail using "forty-five%". It is just using symbols instead of words, for no reason, in what is supposed to be formal prose. The default in all prose is to use the spelled out word, except when doing so would be substantially less clear to the reader or where space is precious. In such exceptions, the word in question is changed to make it more clear. This is the reason for writing "1,393,040" rather than "One million three hundred and ninety-three thousand and forty", but it is not a reason for writing "%" yet forgoing writing the more complex "forty-five" as "45". "Percent" is a word, and using it just like any other word is perfectly acceptable. —Centrx→talk • 07:41, 23 June 2006 (UTC)

Ah, but "percent" is not just a word. To me writing "45 percent" is the equivalent of writing "45 per hundred", which would be wrong per the MoS. You would either write "45 per 100" or "forty-five per hundred". Since you'd be using the numerical symbol for "forty-five" (i.e "45"), you should use the numerical symbol for "percent" (i.e. "%"). Chuck 16:53, 23 June 2006 (UTC)

Percent is a special case of ratio, not any old "25/41". For rounded, commensurate ratios, such as "per hundred", "per thousand", or "per million", it is normal to to find numerals alongside them. For percent, I found in the OED that even when the ratio is more explicit, as "per Cent.", numerals were still used, like:

• 1568 T. GRESHAM Let. 29 Aug. in H. Ellis Orig. Lett. Eng. Hist. (1827) 2nd Ser. II. 314 Th' interest of xij. per cent by the yeare.
• a1687 W. PETTY Polit. Arithm. (1691) vi. 99 The Interest thereof was within this fifty years, at 10l. per Cent. forty years ago, at 8l. and now at 6l.
• 1804 G. ROSE Diaries (1860) II. 136 The Funds rose 1 per cent. on the news.
• 1843 J. A. SMITH Product. Farming 153 The ash of the turnip bulb contains 16½ per cent. of soda.
• 1870 Nature 23 June 140/1 Neglect of it may result in an error up to 190 per cent.
• 1973 Times 21 Dec. 14/6 This trouble's cut my social life by about 35 per cent.
• 1990 Daily Mail 10 Mar. 10/2 A boy who was set on fire because he would not buy crack spent his 12th birthday yesterday recovering from 55 per cent burns.

These are in fact, more common (7 of 10), in these quotations, which are carefully chosen to be representative and canonical examples by the OED.

So, pairing the numeral with the word is common both now and in the literature, and is appropriate style in formal English. Why should the MOS recommend against it? —Centrx→talk • 06:31, 24 June 2006 (UTC)

I have now searched through all the archives, and it appears that this style recommendation, edit [3], was not made according to consensus and was made against valid arguments. The discussion that preceded the change, one month before it, [4], has three regular style editors stating positions that do not support the change, and three supporting it, one tentatively, one the implementing editor and another editor whose only contribution to the discussion was their vote of support. This is hardly consensus, some of the last comments in the discussion were about leaving it up to individual editor decision. Long after, [5], we find two regular style editors not involved with the other discussion recommending "percent" as standard style. Other slightly relevant discussions: [6]: Not especially relevant, but recommending against too much instruction creep. [7]: Discussion unrelated, but has U.S. Government Printing Office recommending acceptable "15 to 25 percent". Not found in this particulary discussion, but note also that this is Chicago style and AP style, among others. —Centrx→talk • 22:18, 27 June 2006 (UTC)

Centrx, what wording do you propose? bobblewik 22:58, 27 June 2006 (UTC)

Questioning space before unit

I would like to question the insistence on a space between a quantity and a unit, especially where the measurement is used as an adjective. I know that such spacing is in accordance with the SI standard, yet the style guide does permit separating thousands by commas, which is clearly not SI-compliant. In everyday usage, it would be far more common to see "35mm film" rather than "35 mm film". There is probably a good reason for it too: "75cl bottles" avoids any possible ambiguity. I would suggest restricting this guideline to scientific contexts. -Cedders^tk 09:18, 22 June 2006 (UTC)

"35mm film", "70mm film", etc. have become names rather than descriptions of the film size (IMHO), so spacing is not relevant. (It's not a question of adjectival vs. nominal usage.) — Arthur Rubin | (talk) 16:42, 22 June 2006 (UTC)

The official SI standard states:

Even when the value of a quantity is used as an adjective, a space is left between the numerical value and the unit symbol. Only when the name of the unit is spelled out would the ordinary rules of grammar apply, so that in English a hyphen would be used to separate the number from the unit.

It gives as example "a 35-millimetre film". −Woodstone 20:46, 22 June 2006 (UTC)

Any examples of adjectival use that are not in the body of an article? If it is in the body of an article, then it should not use the abbreviation anyway and a hyphen is appropriate. Wouldn't most infobox items not need it? For example, specifying 35 mm film might be an item in a camera infobox, yet the heading of that item would be "Film", so we would have "Film: 35 mm", not "Film: 35 mm film", same for the bottles example if that is used anywhere. —Centrx→talk • 23:19, 22 June 2006 (UTC)

Abbreviated month names

I reverted a link to Nov 16 in the DJ Screw article. Is this style always ok, or only in a quote? I can't find the answer in the MOS. Tim Ivorson 2006-06-22

In the body of an article, month names should always be spelled out, except, as you say, in quotes. It may be appropriate to abbreviate in an infobox or other place where space is short, but the infobox in question on DJ Screw does not appear to be such a situation. The MOS only implicitly indicates this, by examples of fully spelled months. This is normal English writing, but perhaps it should be explicitly stated in the MOS. —Centrx→talk • 00:10, 23 June 2006 (UTC)

America's Army has two good examples where abbreviated months are appropriate. —Centrx→talk • 00:12, 23 June 2006 (UTC)

Thanks for the info, Centrx. I think it'd be helpful to make the MOS explicit, but I think I should leave that to people who know what they're doing. Tim Ivorson 2006-06-23

Month-date ranges

I found nothing in the current article about what to do with yearless month-date ranges like "August 16–20" (i.e., "16–20 August"). There are many occurrences in Wikipedia articles of this type of date range, but there seems to be no policy statement on whether or how to format them, even if to say there is no policy. Should we:

• leave them unformatted (breaking consistency with other, formatted dates)?
• format part of them (e.g., August 16–20), which won't look right for half the users whichever way they're formatted?
• format the whole thing (e.g., August 16–20), again looking bad for half our editors?
• some other system not obvious to me?

At the very least, our policy page should acknowledge the problem. I didn't see anything mentioned in Wikipedia talk:Naming conventions (calendar dates). I didn't sift through the forty-eight archives of this talk page, but if there's some discussion buried in them, its resolution (or lack thereof) should long ago have been made a part of the policy page. Thanks for any assistance. ~ Jeff Q (talk) 07:41, 25 June 2006 (UTC)

Since it seems that formatting like "August 16–20" would with date preference applied show "16 August–20", which is not proper, in such cases I do not think the dates should be linked. A possible solution is to write it out as "from August 16 to August 20", which preserves date preferences and is otherwise perfectly acceptable. —Centrx→talk • 02:32, 26 June 2006 (UTC)

A horrible example is: [[November 12]]–[[15]], [[1942]] in Naval Battle of Guadalcanal (at time of writing). I propose that we add the guidance:

• Date ranges: Generally do not link..

bobblewik 20:11, 28 June 2006 (UTC)

This isn't guidance already? I'm surprised, since it's what I've always been doing or changing ranged dates to. It's so obvious I'd just always assumed it was in WP:DATE. TheGrappler 01:17, 1 July 2006 (UTC)

I strongly object to this proposal. Are you people trying to get your way simply by tiring everyone else out and hoping that you'll slip this back into policy when no one is looking - again? Rebecca 02:31, 1 July 2006 (UTC)

What do you propose instead? If you are referring to date linking, this "proposal" is not about date linking in general; it is a very specific case that, as far as I can tell, was not discussed at all in the recent date linking discussion in Archives 42–46,48. —Centrx→talk • 07:01, 1 July 2006 (UTC)

November 12-November 15, November 12-15 and November 12-15 are all fine by me per se, but I do think there's no reason to break the date preferences setting. — Preceding unsigned comment added by Rebecca (talk • contribs) [8]

Ok for me too. Linking has the additional advantage that it's clear what is meant. -- User:Docu

I agree that there is nothing wrong with [[November 12]]-[[November 15]]. But that format is rare.

Many articles have broken dates because some editors don't understand the difference between 'hyperlink' and 'preference mechanism'. Editors do things like: [[November 12]]–[[15]], [[1942]]. It is probably worse than all the other errors because the page looks broken when you see:

• 12 November-15.

Set your date preference to the non-US format and look at the the first sentence of: Naval Battle of Guadalcanal.

The proposal addresses that common problem. bobblewik 10:36, 1 July 2006 (UTC)

Furthermore, the example given by Rebecca presents exactly this problem.

Quote: ...[[November 12]]-[[November 15|15]] ... are all fine by me

That is broken and becomes 12 November-15. bobblewik 10:46, 1 July 2006 (UTC)

Date ranges & linking: in general no difference from other dates, I'd say. The only remark is that editors should be sensible to the implications, and make an appropriate choice:

• unlinked:
  • May 12-15 gives de facto precedence to US notation, so could only be used w.r.t. topics that are clearly under a variety of English that uses this notation as a standard (See WP:MoS#National varieties of English)
  • 12-15 May gives de facto precedence to European notation, so could only be used w.r.t. topics that are clearly under a variety of English that uses this notation as a standard (See WP:MoS#National varieties of English)
• linked:
  • Month + day at both sides of the hyphen/dash/"to", without pipes would give a consistent layout whatever the transformation by preferences setting of the user: [[May 12]]–[[May 15]] or [[May 12]] to [[May 15]], etc. (but yeah, this is not so "fluent", and probably not very good style)
  • Both with pipe, resulting in US notation (same remark as above for US notation, "unlinked"): [[12 May|May 12]]-[[15 May|15]] or [[May 12|May 12]]-[[May 15|15]]
  • Both with pipe, resulting in European notation (same remark as above for European notation, "unlinked"): [[12 May|12]]-[[15 May|15 May]] or [[May 12|12]]-[[May 15|15 May]]
• half-linked as above, e.g.
  • [[12 May|May 12]]-15 (US notation)
  • 12-[[May 15|15 May]] (European notation), but only linking the second date doesn't look too well I suppose. Although I could imagine something like "After a short campaign (12-15 May) the elections were held" wouldn't be too silly. --Francis Schonken 12:00, 1 July 2006 (UTC)

Do you have an alternative wording for the proposal? bobblewik 12:31, 1 July 2006 (UTC)

The proposal is to change nothing. Something that is evident from existing guidance, should not be used as an excuse to change prior agreed upon provisions in the guidelines. In other words: there is no problem (apart from maybe the problem that you're on a date de-linking crusade again, as remarked above by Rebecca, and on your user talk page). There's no reason to change linking/delinking of dates guidance, and I go from the assumption you know that. The so-called linking of date ranges "problem" you elaborated as a result of Jeff's question above is an artificial construct, ignoring current guidance. I think I demonstrated that. --Francis Schonken 12:50, 1 July 2006 (UTC)

Why not add "Incorrect: February 14–15 and 15–14 February" to Wikipedia:Manual of Style (dates and numbers)#Incorrect date formats? —Centrx→talk • 22:39, 1 July 2006 (UTC)

Proposal for section Numbers in words

It is common English style to have numbers up to 100 spelled out, and under the MOS, acceptable. However, the current version lumps numbers 10 through 100 together with numbers 100 and up. While it is in formal use appropriate to have "eighty-seven" spelled out, it is rare to have "five hundred and sixty-seven" and totally inappropriate to have "one million, four hundred and fifty-six thousand, seven hundred...", which the MOS read literally currently accepts. Therefore, numbers 10 through 100 must be distinguished from those above 100. Here is probably the simplest solution:

• Numbers between ten (10) and one hundred (100) may be written as numerals or spelled out in words. Within each article, style should be consistent.
• Numbers above one hundred (100) must be written as numerals, except at the beginning of a sentence.

Note that one problem with this is it forbids "five hundred" or "ten million", which with a little wording could be corrected.

There are two other possibilities. First, we can explicitly recommend that all numbers between 0 and 100 be spelled out as words in the body/prose/text of an article. Second, per suggestion above, at Numbers from zero to ten which states it is MLA style, we can recommend that numbers that can be written out in two words or less (e.g. "eighteen", "ninety-nine", "five hundred", "five million", but "101", "499", "5,100,000" or "5.1 million"). This is a convenient solution, and corrects the problem with the above, simple solution. See also Wikipedia talk:Manual of Style (dates and numbers)#Numbers from zero to ten and Proposal: Mention about spelling out numbers up to 100. —Centrx→talk • 22:58, 27 June 2006 (UTC)

I like your second possibility ("written out if in two words or less"). It sums it up nicely, I never thought of it that way. Neonumbers 06:03, 6 July 2006 (UTC)

• Support "if in no more than two hyphenated words". Regards, David Kernow 09:13, 7 July 2006 (UTC)

Done, see [9]. —Centrx→talk • 20:59, 9 July 2006 (UTC)

Proposal for section Percentages

The MOS currently recommends that numerals should always be paired with the percent symbol (%) and spelled-out numbers paired with the word "percent". This forbids the style of "55 percent", which is common on Wikipedia and is standard style in English prose, recommended by the Chicago MOS and AP style guide, common in the canonical quotations of the OED, s.v., and found in books[10][11], journals[12], and styleguides[13] everywhere. See also Wikipedia talk:Manual of Style (dates and numbers)#Percentages, [14], and [15].

Note, however, that while it is common and standard for "percent" to be written out in 'humanistic' copy, we find in above-mentioned discussions and in the Chicago MOS that in scientific or statistical use, the symbol % is more common or appropriate. However, such recommendations are for scientific journals or books, not for general-purpose encyclopedias that are covering scientific matters among many others. Note, for example, that the Encyclopedia Britannica uses "percent" uniformly (e.g.: "They consist of (by dry weight) roughly 50 percent to 75 percent crude protein, 4 percent to 18 percent fats,...". Therefore, the Percentages section ought to be changed to:

• Use the word percent or per cent for a percentage in the body of an article. Examples: "twelve percent", "12 percent".
• Use the symbol % with a numeral for a percentage in a table, infobox, or the like. Example: "12%"

—Centrx→talk • 23:27, 27 June 2006 (UTC)

I agree with Centrx's proposal with two modifications. First, the use of numerals versus numbers should be based on the Numeral guideline in the MOS (zero to ten, 11+) and we should recommend one spelling of percent, namely "percent". Rillian 12:30, 28 June 2006 (UTC)

Apparently, "per cent" is common in Britain and "percent" in America, so I was just following Wikipedia:Manual of Style (spelling)#Preferred variants in allowing both. —Centrx→talk • 22:19, 28 June 2006 (UTC)

For numeral use, the example is in the range in which it is acceptable to have either "12" or "twelve". Or do you mean that the section should explicitly state that percentages should follow the Numeral guideline? —Centrx→talk • 23:49, 3 July 2006 (UTC)

Some pro-symbol references:

Use the sign % instead of per cent (The Economist)

% in headlines and copy (The Guardian)

Symbols for numbers and units versus spelled-out names of numbers and units (NIST)

Unlike most publications, Wikipedia is international. Therefore we should take this into account when choosing between a language-specific words and a language independent symbols.

The original complaint from Centrx was that '55 percent' is forbidden. I like the current guidance but I am sympathetic to the desire for deregulation. However, the proposed new wording is not liberalisation, it is inversion. What was forbidden would be mandatory and what was mandatory would be forbidden. Could we agree on deregulation instead? bobblewik 18:54, 28 June 2006 (UTC)

Nothing is "forbidden" by the MoS - Rich Farmbrough 10:44 5 July 2006 (GMT).

Yes, this is not the most accurate word, but it is understandable within the advisory context of the MoS, and the fact remains that one is liable to find some editors changing article style to conform to the MoS, sometimes by script/semi-bot, and find that changing it back is met with calls to the authority of the MoS. While the MoS is not exactly binding in the free flow of the wiki, its substantial effect can be beyond mere recommendation. —Centrx→talk • 02:19, 6 July 2006 (UTC)

Deregulation would not require much of a proposal and could be implemented by removing the section. Note, however, that following the standard in spelling out units in text[16], rather than using symbols, would indicate that "percent" should be spelled out as well. Note also that "percent" spelled out is far more common on Wikipedia, and the former policy was against common practice. Also, the current proposal is not exactly an inversion, formerly any editor who wanted could use "fifty-five percent".

The NIST recommendation is irrelevant, it is for scientific and technical papers, which Wikipedia is not. Regarding the Economist and the Guardian, note that the BBC styleguide[17] and The Times of London styleguide[18] both have "per cent" written out (In the U.S. also, aside from the AP styleguide it is found in the NY Times, the Wash. Post, and Time Magazine), and furthermore that Wikipedia need not consider issues of space like newspapers and magazines, and that books, non-technical scholarly journals, and encyclopedias overwhelmingly use the word spelled out, while it is still much more common for newspapers and magazines anyway (see also Google News).

I see no reason why the Manual of Style should not state that "percent" is written out when:

• It is far more common style on Wikipedia.
• It is uniformly common and standard style in professional works, books, encyclopedias, etc.
• It is far more common style in magazines and newspapers
• Only scientific and technical journals and books, which Wikipedia is not, and a minority of newspapers and magazines, which Wikipedia is not like, use % instead of percent

—Centrx→talk • 22:19, 28 June 2006 (UTC)

Whichever way is chosen, it should be required (based on SI specification) to have a space between the number and the percent sign. The example should read:

• Use the symbol % with a numeral for a percentage in a table, infobox, or the like. Example: "12 %"

−Woodstone 19:06, 28 June 2006 (UTC)

This is very rare in general practice, and does not really apply to many uses of percent, which is not an SI unit. The MOS should not push such a minor usage against the vastly more common use on Wikipedia and elsewhere. Also, this is a separate issue for which there was a lot of debate in the archives, and so to keep this discussion clean we could have both in the example, though it might be best to just resolve it now. —Centrx→talk • 22:19, 28 June 2006 (UTC)

• Prefer "percent" per Encyclopaedia Britannica, but probably best left optional...?  Regards, David Kernow 09:15, 7 July 2006 (UTC)

Dates (years BC)

The entry mentions, "Articles for the year 500 BC and earlier should be redirected to the relevant decade. Articles for the year 1700 BC and earlier should be redirected to the relevant century. Articles for the year 4000 BC and earlier should be redirected to the relevant millennium."

Shouldn't this read, "Articles for the year 500 BC and later should be redirected to the relevant decade. Articles between the year 1700 BC and 500 BC should be redirected to the relevant century. Articles prior to 4000 BC should be redirected to the relevant millennium." The current wording is ambiguous and includes overlap. —The preceding unsigned comment was added by Dr1819 (talk • contribs) 19:42, 30 June 2006 (UTC)

It appears to mean:

• After 500BC: Link to year.
• 1700BC–500BC: Link to decade.
• 4000BC–1700BC: Link to century.
• Before 4000BC: Link to millennium.

—Centrx→talk • 07:09, 1 July 2006 (UTC)

It's the page itself that redirects, e.g. 502 BC (which is 500 years before/prior/earlier than year 1) redirects to 500s BC. One should still link to 502 BC. -- User:Docu

Different calendars

In the section on different calendars it says to use the Julian calendar for dates before 1582-10-15.

Dates before the adoption of the Gregorian calendar on 1582-10-15 should usually be given in the Julian calendar. The Julian day and month should not be converted to the Gregorian calendar, but the start of the Julian year should be taken to be 1 January (see below for more details).

I'm not totally clear on what this means. For example in citing history it may not be clear whether a date is Julian or Gregorian. On the other hand in using simulations (such as astronomical software) the dates are often presented as Gregorian dates no matter how far back the simulation goes. In addition, the Julian calendar is ambiguous for the starting year (see Julian calendar), so I assume this means use a Julian calendar with the AD/BC epoch. However, it's not entirely clear. Perhaps should simply say Gregorian calendar as the default unless there's some clear reason to use the Julian calendar. Is this common in the field of history to explicitly use Julian calendar dates before the first adoption of the Gregorian calendar? My assumption when I read Julius Cesear adopted the Julian calendar in 46 BC is that this is a Gregorian date: that this is 2051 years ago (with no 0 year in either calendar). Can anyone clarify this? Thanks in advance. --Cplot 06:31, 2 July 2006 (UTC)

After posing the question above I did find a useful tool that perhaps could be linked to in the manual of style to help make conversions. It's a | calendar conversion calculator for many different calendars (including Gregorian and Julian). It doesn't really answer my question, but I thought others would find it useful. It may also help in thinking through this issue. --Cplot 06:51, 2 July 2006 (UTC)

The Julian calendar doesn't say anything about numbering of years. In Julius Caesar's Rome years weren't even numbered, a year was known by the names of the two consuls of that year. The AUC ("Ab Urbe Condita") year number wasn't used as a means to number years in civil life. Livy wrote a historical work named "Ab Urbe Condita", but didn't "number" years in that historical work. Varro listed consuls, and only much later a numbering of AUC years was derived from a combination of historico-mythological works about the founding of Rome and the known lists of consuls. Numbering of years from the birth of Christ (as far as known at the time) originated in the early middle ages, so was completely unknown by Caesar, and, on the other hand, completely established by the time of pope Gregorius - so he didn't "change" that numbering of years, he only modified the start of months, and how many days months would have; these months were basicly still the same twelve months as in Caesar's day.

Another difference was at which month a year started. In old Rome a year started when the two new consuls came to power. Which was May in the oldest known times, later became the first day of March, and had more or less standardised on the first day of January by Julius Caesar's time (in fact since AUC 601 = 153 BC, see blue vertical line on this page: [19]). Also here "Julian Calendar" doesn't say anything about whether a year starts 1st of January, or 1st of March (or whatever other month). Gregorius did away with the differences (that apparently still lived on till his day), and in his calendar years always started first of January. For some countries that didn't yet move to Gregorian calendar at the time, a year still started at first of March (or whatever other day usual in the local calendar). That is the "old style" of some countries. So, moving to Gregorian calendar had two aspects: (1) having the year start at the first day of January, (2) adopting Gregorius' system of leap years, number of days for the month of February, etc. The first of these changes could be done independently of moving to the full Gregorian calendar. Anyway, for "old style" calendars, dates in the month of January and February could have a one year difference with calendars that started a new year on the first of January.

To answer your question "Is this common in the field of history to explicitly use Julian calendar dates before the first adoption of the Gregorian calendar?", the answer is yes, but you have to know that Julian calendar, by itself, says nothing about numbering of years. Doing otherwise (i.e. not using the Julian calendar for dates before the Gregorian reform) is called using a "proleptic" Gregorian calendar, which is thoroughly uncommon in historiography. As for the numbering of years, in current (Occidental) historiography the numbering of years starts at 1 AD/CE (and counting back from 1 BC/BCE for older dates). And this counting of years is influenced by which day of which month is seen as the first day of a year (which, for instance, appears in old style/new style differences for days in the two first, or last, months of a year). --Francis Schonken 09:08, 2 July 2006 (UTC)

Thanks for the reply that clear things up for me. I think some abridged version of this explanation could be included in the manual of style. Perhaps I take a stab at it. --Cplot 17:51, 2 July 2006 (UTC)

Inclusion of conversions

Under Wikipedia:Manual of Style (dates and numbers)#Units of measurement as of the time of this edit, the first point states that “Conversions should generally be included and not be removed.”  I hope that I’m not missing anything, but this seems to be a loophole to me.  I could add an astronomical amount of unnecessary unit conversions from the list of strange units of measurement but, according to policy, “Conversions should generally be included and not be removed.”  I don’t believe that the removal of these inessential conversions should be against policy.  Would anybody agree with me that these conversions should be limited to standard systems of measurement, such as metric and imperial conversions only? —Rofl 09:22, 7 July 2006 (UTC)

Makes sense. I suspect there is an intent to avoid "lawyering up" with too much specificity, leaving it to common sense and the WP principle (pillar #5) of not getting bogged down in rules. The general idea, I think, is to permit the use of both SI and the most common and relevant non-SI measurement (usually imperial or U.S., of the most comparable magnitude) for each quantity. ~ Jeff Q (talk) 09:32, 7 July 2006 (UTC)

Yeah, I see what you're getting at. (I was concerned about this when we redid that section, but no-one else seemed to consider it relevant). Jeff's right, the general idea is to allow conversions where necessary; but then you get disputes over what's necessary and stuff like that, I guess... the loophole needs to be fixed.

I think conversions should be limited to:

• all units typically used for that context (e.g. in astronomy, light years or AU)
• metric units
• U.S. units (or whatever they call them.

But knowing the nature of this, I bet there'll be some exception somewhere. It's just a tricky call... it's not so much related to pillar five (or at least, not in my view... we just try to give good, solid, reasonable guidance.) Neonumbers 12:15, 7 July 2006 (UTC)

There will be certain exceptions like in the Jersey article, but for the most part I agree with Rofl. I think that common sense will prevail if someone tries to input a conversion in some odd unit like cubits to describe the CN tower's height. Someone will revert it. —MJCdetroit 19:20, 7 July 2006 (UTC)

How about "Conversions useful to many English readers should generally be included and not be removed."? —Centrx→talk • 03:19, 9 July 2006 (UTC)

Discussion on the interpretation of Eras section.

Hello. There is an ongoing discussion on the interpretation of the Eras section of this MOS in Talk:Montanism. The passages used for the two different positions are:

1. Both the BCE/CE era names and the BC/AD era names are acceptable [..] When either of two styles are acceptable it is inappropriate for a Wikipedia editor to change from one style to another unless there is some substantial reason for the change. For example, with respect to English spelling as opposed to American spelling it would be acceptable to change from American spelling to English spelling if the article concerned an English subject. Revert warring over optional styles is unacceptable; if the article is colour rather than color, it would be wrong to switch simply to change styles as both are acceptable.
2. Normally you should use plain numbers for years in the Anno Domini/Common Era

The matter of the dispute is:

• in an article in which all of the years and centuries are AD/CE, is it possible/mandatory/prohibited to add "AD" to the first year/century nominated?

Please, help us to settle this matter.--Panairjdde 23:54, 8 July 2006 (UTC)

Which midnight?

Someone added to the guidline:

So for consistency, always use 12 midnight to indicate the start of a given day: whenever the day is relevant. For example: “The winner crossed the finish line at precisely 12 midnight 17 March 2001.” indicates the beginning of the day 17 March (the end of the day 16 March).

Is this really the common way of doing it? A more usual occurrence would be a due date, like in:

• submissions have to be handed in before midnight July 31

Would that not commonly be interpreted as midnight at the end of that day? −Woodstone 19:59, 9 July 2006 (UTC)

Yes, I don't think this is most common at all. Midnight implies the night of the given date, not the morning of it. For clarity, it might even be best to explicitly state "the night of" or "the morning of". I have reverted the changes of that line. —Centrx→talk • 20:44, 9 July 2006 (UTC)