Wikipedia talk:Manual of Style/Dates and numbers/Archive 73

minus signs
This page says:


 * The minus sign may be represented by a hyphen ("-") or by  ("&minus;").

The hyphen is far too short for that purpose. No respectable typesetter would use it. Compare:


 * 5 - 3


 * 5 &minus; 3

Within TeX, of course one uses a hyphen, and the reader sees this:


 * $$ 5 - 3 \, $$

Michael Hardy 22:48, 27 June 2007 (UTC)


 * WP:DASH seems to agree with you. Stephen Turner (Talk) 08:55, 28 June 2007 (UTC)


 * Wikipedia editors are not, in the main, typesetters, respectable or otherwise. Andy Mabbett 11:27, 28 June 2007 (UTC)


 * See also: Wikipedia talk:Manual of Style Jɪmp 15:41, 28 June 2007 (UTC)


 * Hyphens may not be used to represent minus signs. See MOS. Tony 11:41, 12 July 2007 (UTC)

Roman numerals
The Transactions of the Linnean Society of London express their volume numbers in Roman numerals. Should I follow their lead and write: ""... and was published in Volume X of Transactions of the Linnean Society of London."" or should I convert to Hindu-Arabic: ""... and was published in Volume 10 of Transactions of the Linnean Society of London."" ? Hesperian 04:42, 29 June 2007 (UTC)

thereby making it clear that "Volume X" is their usage? Andy Mabbett 09:07, 29 June 2007 (UTC)
 * Why not ""... and was published in Transactions of the Linnean Society of London, Volume X.""


 * Nothing wrong with Roman numerals in my book. Go ahead.  I like Andy's suggestion: not only would it be clear to readers that this is their usage but also to any editor who might otherwise feel like changing it to Hindu-Arabic. Jɪmp 23:55, 2 July 2007 (UTC)


 * Thanks folks. Hesperian 12:41, 9 July 2007 (UTC)

Metric/SI only
Why is their any allowance for non metric systems of measurement? Ancient units I can see, simply for historical purposes, but only the US still uses a system other than metric today; even they are starting to move over. Should not, due to the fact this is a global encyclopaedia, only use the metric system? Spacedwarv 07:19, 7 July 2007 (UTC)


 * Because the articles should be accessible to everyone. So it's appropriate, especially in articles about the U.S., to use units which Americans are familiar with. Of course, a conversion to metric should always be provided. Stephen Turner (Talk) 09:55, 7 July 2007 (UTC)
 * Metric is everyone’s system. US units are – well – US-only, Wikipedia is not. (Mandatory) dual units lessen readability for everyone, without relevant gain in universal understandability. Christoph Päper
 * It's not up to you or Wikipedia to change the way Americans think. The concept is obnoxious. Chris the speller 01:52, 23 August 2007 (UTC)


 * And another reason: it's usually correct to quote the units which the source used. Stephen Turner (Talk) 10:23, 7 July 2007 (UTC)
 * That actually only applies if the source was constituting something using non-metric units, not if it was reporting measurings. Christoph Päper


 * Also, many people from British commonwealth countries still don't "think" in terms of metric measurements, despite some of their governments being officially metric for years. To use only metric units would actually make the encyclopedia less universal. &mdash;MJCdetroit 19:52, 7 July 2007 (UTC)
 * How should they learn to think metric if they’re nannied with old or dual labels all the time? Christoph Päper
 * It is not compliant with Wikipedia's mission to take a side in measurement politics. Atropos 01:12, 24 August 2007 (UTC)


 * Any chance of a requirement to list all articles in SI instead of a simple recommendation? I can see the logic in that people do not think in it yet, but the acceptance of metric is global and in the few strongholds otherwise, it is inevitable. Spacedwarv 21:58, 11 July 2007 (UTC)
 * No! Current system is the right compromise, I think. Tony 11:38, 12 July 2007 (UTC)
 * Yes, too many people do feel or believe that way, not much thinking involved. (Actually I was just checking whether that had changed lately. I’m gone again now.) Christoph Päper

I find the current system just awful when it comes to readability. Most pages you read have both a unit and a conversion. Only 4% of the population are non-metric. That's technically a minority. I'd bet that higher than 4% of the population have "," instead of "." when expressing a decimal number, but "." has been made mandatory. I'd rather see the style for technical and scientific articles adopted for all articles: only SI, except for specific historical reasons. I can see the argument that for example an article on a town in the US, that it is reasonable to have both units. Wikipedia is international and non-SI units are way way in the minority. I'm sure all would agree that eventually the US will use SI - so why not lead the way? Jim77742 13:51, 22 August 2007 (UTC)
 * Please see my reply to Christoph Päper above. The US does not need you to lead it. Chris the speller 01:52, 23 August 2007 (UTC)


 * OK Jim77742, please try removing the use of "yards" from the article on American Football. ;) Fnagaton 14:18, 22 August 2007 (UTC)


 * I think I indicated above it would be appropriate to use yards in those types of cases. But it would need to explain what a yard is. Your US yard is defined as 914.4mm (exact). Jim77742 01:25, 23 August 2007 (UTC)

I'd strongly support a dual system in which US-related articles are US units (main) and metric (converted), and all other articles are metric alone. Tony 15:54, 22 August 2007 (UTC)


 * This is exactly what I'm saying.Jim77742 01:25, 23 August 2007 (UTC)


 * Me too. In the meantime, I think that the unit section can be simplified and improved. For example, replace the 'Which system to use' and 'Conversions' sections with:


 * ''*Editors are encouraged to use the metric system. Non-metric units may be primary at the consent of editors in the following cases:
 * ''**The article is US-related and non-scientific.
 * ''**The article is UK-related and discusses quantities that are not currently metricated (e.g. road speed in mph).
 * ''**There are compelling historical or pragmatic reasons why metric units should not be the main units.
 * ''**If editors cannot agree on the sequence of units, put the source value first and the converted value second.
 * ''**American English spells metric units with final -er (e.g. 'kilometer'). All other varieties of English, including Canadian English, use -re (e.g. 'kilometre').


 * ''* In general, editors should not remove conversions.
 * ''* Conversions adjacent to units can sometimes make common expressions (“The four-minute mile”) or quotes awkward. If this is an issue, consider a non-adjacent conversion (e.g. in a subsection).
 * ''* In general, units in text should be spelled out. Editors may agree that it is appropriate for text to contain symbolic forms after the first occurrence.
 * ''* Units in parentheses should be in symbolic form. For example, "a pipe 100 millimetres (4 in) in diameter and 16 kilometres (10 mi) long” or “a pipe 4 inches (100 mm) in diameter and 10 miles (16 km) long".
 * ''* Precision of source and converted values should match. For example, "the Moon is 380,000 kilometres (240,000 mi) from Earth", not "(236,121 mi)".
 * ''* Editors may wish to take advantage of conversion templates such as undefined undefined.
 * Actually, I would eliminate the last five of those bullets. That would simplify it even further. Guidelines should only exist if they solve a significant problem that would not be solved by the wiki. Lightmouse 18:08, 22 August 2007 (UTC)
 * Agreed. I'd even remove the UK exception. Miles are used on the roads, but that will be gone by the time of the 2012 Olympics. Jim77742 01:25, 23 August 2007 (UTC)

Can I propose the following words (stolen and modified from above :-)


 * Editors are highly encouraged to use the metric system alone without conversions. Conversions make articles less readable.
 * Non-metric units may only be primary (at the consent of editors) in the following cases:
 * The article is specifically US-related and non-scientific.
 * There are compelling historical or pragmatic reasons why metric units should not be the main units.
 * When the non-metric unit is used as primary, a conversion should nearly always be provided. Keep readability in mind though, for example in an article on American Football it would be appropriate to explain that a yard is 914.4 mm exactly the first time a yard is mentioned, but not thereafter.
 * Note that American English spells metric units with final -er (e.g. 'kilometer'). All other varieties of English, including Canadian English, use -re (e.g. 'kilometre') and so this can be simplified by generally using the formal abbreviations "km".
 * In general, editors should not remove conversions from these US-based articles
 * Precision of source and converted values should generally match. For example, "American football is played on a field 120 yards (109.728 m) long" is incorrect. Use "(110 m)" instead. Note that defining one yard (as exactly 914.4 mm) may be approriate.

Jim77742 01:25, 23 August 2007 (UTC)
 * In a direct quotation, if the text contains an obscure use of units (e.g., five million board feet of lumber), annotate it with a footnote that provides standard modern units rather than changing the text of the quotation
 * In scientific articles, SI units are the only units of measure, unless there are compelling historical or pragmatic reasons not to use them (for example, Hubble’s constant should be quoted in its most common unit of (km/s)/Mpc rather than its SI unit of s−1)
 * Where footnoting or citing sources for values and units, identify both the source and the original units.


 * This strikes me as blatantly and staggeringly pro-UK and anti-US. Mountains in England will be measured in metres, conversion to feet being forbidden, while mountains in California will be measured in feet, with a mandatory conversion to meters? Perhaps you think American readers of Wikipedia are an extremely small minority? Chris the speller 01:44, 23 August 2007 (UTC)


 * It's not pro-UK or anti-US it is pro-World. Three countries and 4% of the world's population use imperial units. And even in the US, science, medicine and the armed forces all use metric exclusively. The number of people who do not know what a litre is would have to be very much under 4% of the world's population. This is the internet. Wikipedia is not a US thing, it is a world-wide encyclopaedia. I've never read an encyclopaedia that has unit conversions on every measurement. It read's terribly! And while we're on mountains I'd even go as far to say that US mountains should have the primary measurement in metres with a conversion to feet. After all there is a template for all mountains on wikipedia. By the way, your surveyors measured your mountain in metres and then converted to feet for local publication. Jim77742 02:16, 23 August 2007 (UTC)
 * This is the English Wikipedia, correct? According to Wiki's own article on the English language, more than 2/3 of all native speakers are American.  That being said, your 4% statistic is worthless.  Americans do not only read articles regarding American subjects. So to have US-centric subjects have conversions and all others metric only is unfair if not completely retarded and counter-productive.  Your proposal makes wikipedia less universal instead of more universal. The MOS as written is very fair to everyone. &mdash;MJCdetroit 02:32, 23 August 2007 (UTC)
 * The English language article says that 1.8 billion people speak the language and that "Modern English is sometimes described as the global lingua franca". So when you want to speak globally on this planet - you use English. When you want to measure you use metric - that's what the world does. A lot of non-Americans from all over the world read the English Wikipedia - and the constant barrage of things like "the 10 km (6.2 mi) long and 100 mm (4 in) wide pipe branches into a 3 km (1.8 mi) long and 70 mm (2 3/4 in) wide pipe and a 7 km (4.3 mi) long and 40mm (1 1/2 in) wide pipe." makes readability so bad. (I know this is an exaggeration but a lot of articles look like this.) On the other side of the coin, the scientific articles with only SI read so much better. Jim77742 03:07, 23 August 2007 (UTC)


 * Jim77742 misses the point about the mountains example. Having the height of a mountain to the nearest millimeter is not important, it's presenting the information so the readers can understand and grasp it that's important. Using his logic, the majority of native English speakers write "color" instead of "colour", so all of WP should use "color", except in UK and Australia, where we will now write "The national colours (colors) of the United Kingdom are red, white and blue." I am not proposing that, since I am not insensitive to the traditions and feelings of the many readers who still cling to the archaic spelling even though they can understand "color". I am kidding, of course! Chris the speller 03:12, 23 August 2007 (UTC)


 * Don't like the conversions in parentheses; then skip over them when you read the article. That's why they are in the parentheses. Unless of course it is your system of measurements that is in the parentheses.  &mdash;MJCdetroit 03:31, 23 August 2007 (UTC)


 * Why should U.S. related articles be forced to have metric conversions, yet non-U.S. related articles should not be forced to have imperial conversions? All that does is prevent Americans from fully understanding an article they are reading. ---CWY2190TC 06:12, 24 August 2007 (UTC)


 * I strongly agree with the push to reduce unreadable clutter in non-US articles, and I have little time for people who can't or won't take on the world-standard measurement. (I also have no time for US obstruction of and refusal to join such institutions as the World Court; it's outrageous.) If this proposal is to move forward, it will have to be raised at MOS as well. Tony 03:44, 23 August 2007 (UTC)


 * Are such things, then, as the UN Convention on the Law of the Sea an example of "people who can't or won't take on the world-standard measurement"? Whether one likes it or not, it specifies territorial limits and Exclusive Economic Zones in nautical miles.
 * While professional articles in medicine are indeed almost universally metric, the daily practice of clinical medicine in countries that use some Imperial uses the units most familiar to the patient. There is enough problem in counseling an epidemic of obesity than demanding the patient know how many kilograms he or she is expected to lose.
 * A conversation with Canadian friends can be comical. They tend to mix systems, speaking of childrens' heights in Imperial but weight in metric. Some forget that I'm completely comfortable in Celsius, so they try to be courteous and convert, while I don't think it's just hot, but intolerable when they convert and tell me it's 98 Fahrenheit -- and I'm thinking Celsius.
 * Even with pure metric measurements in medicine, there is confusion because of changes in different metric systems. I started thinking of blood glucose in mg/dl nearly fifty years ago. When the values changed from the multiple sugars in the Fehling or Benedict reaction to the more accurate glucose oxidase tests, I kept up with the value. While I do think metric in these tests, it is not intuitively obvious that the switch from weight to molar concentrations is more clear. Howard C. Berkowitz 19:50, 9 September 2007 (UTC)


 * This push makes Wikipedia LESS universal. Tony you seem to be letting your political feelings regarding the US government navigate you.  It is not unreadable clutter to a third grade student in Ohio who needs to read about New Zealand.  Is it fair to that 10-year-old that the article would be presented in measurements that are unfamiliar to them? What is clutter to you, is valuable information to others.  You should ask yourself if this makes wikipedia more universal or less universal?  The MOS as written is very fair to all readers and does not alienate anyone. &mdash;MJCdetroit 04:10, 23 August 2007 (UTC)
 * Tony, your political leanings have no place in consideration of MOS issues. Wikipedia aims to be accessible to all readers, that means using both sets of units in all articles.  Americans make up a large part of the authorship and readership of the English Wikipedia.  It would be a grave service to this large group of people if we were to make our articles less comprehensible to them. Johntex\talk 04:19, 23 August 2007 (UTC)
 * Your argument about making articles more universal means, for example, the article on Ohio that a New Zealand child is reading should have all monetary units mentioned having other country’s values in brackets. (“Per capita personal income in 2003 was US$30,129 (EUR 22,205, GBP 15,090, JPY 3,492,704, AUD 36,977…), 25th in the nation.”) Of course that would be ridiculous, but even if you mentioned only a couple of currencies it would still be silly. Readability is the issue here. And your Ohio child when she reads that Aoraki is the highest mountain in New Zealand at 3754 m has learned something. And if she doesn’t know what a metre is then she can click on the “m” and learn some more. Jim77742 04:46, 23 August 2007 (UTC)

Please look again at the first proposal. I think it is achievable. The only conceptual change is an encouragement for use of metric. This is absent from the current text. It is largely a rewrite of the existing text (plus a suggested removal of five bullets that seem to me to be unnecessary rule creep). Unfortunately, MOS wording makes little difference to the prevalence of metric in Wikipedia, particularly if the Featured Article process is giving FA status to unmetricated articles (see my new section below). Incidentally, Jim77742 please give your source for the metrication of UK roads by 2012 Olympics. Lightmouse 11:08, 23 August 2007 (UTC)
 * http://www.timesonline.co.uk/tol/news/uk/article733885.ece Jim77742 12:27, 23 August 2007 (UTC)

My "political leanings", as they're referred to above, concern my view that it's undesirable to pander to the xenophobia of some Americans (some) as expressed in their insistence on retaining a system of measurements that has been discarded by the rest of the world.

Please consider this:

CONVERSIONS
 * Editors are strongly encouraged to use the metric system alone without conversions; conversions make articles less readable.
 * Non-metric units may be used as the main units, and converted to metric units, where:
 * the article is both specifically US-related and non-scientific; or
 * there are compelling historical or pragmatic reasons why metric units should not be the main units.

Where non-metric units are the main units
 * Spell out the main units and use unit symbols or abbreviations for conversions in parentheses; for example, “a pipe 4 inches (100 mm) in diameter and 10 miles (16 km) long”. The exception is that where there is consensus to do so, the main units may also be abbreviated in the main text after the first occurrence.
 * Converted values should use a level of precision similar to that of the source value; for example, “the highway is 32 miles (51 km) long”, not “(51.2 km) long”. The exception is small numbers, which may need to be converted to a greater level of precision where rounding would be a significant distortion; for example, one mile (1.6 km), not one mile (2 km).
 * Category:Conversion templates can be used to convert and format many common units, including, which includes non-breaking spaces.

Other cases
 * In a direct quotation, if the text contains an obscure use of units (e.g., five million board feet of lumber), annotate it with a footnote that provides standard modern units rather than changing the text of the quotation.
 * Where footnoting or citing sources for values and units, identify both the source and the original units.


 * So you propose to abandon the long standing consensus of using metric and imperial/U.S. customary measurements after an editor with only 144 edits rehashes an old discussion because he does not like seeing converted values within articles and because this would be a way to force us xenophobic Americans to start conforming to world standards. Is it possible that there are some shenanigans going on here?  Relating information so that all readers can easily and quickly understand it is not xenophobic; it's common sense.  However, proposing the removal of measurements that are mostly used by Americans does show a clear underlying bias against Americans. None of the other major online encyclopedias such as encyclopedia.com, Encyclopaedia Britannica, Encarta, Worldbook, and infoplease Encyclopaedia are metric only.  They provide the information that the reader is most likely to relate to, like both systems of measurements.  The proposal is ridiculous if not just outright anti-American.  &mdash;MJCdetroit 17:16, 23 August 2007 (UTC)

This is a silly discussion. There is no way that removing English units entirely from articles would ever have consensus, and even if you managed to engineer a consensus on this page it would not reflect the will of editors generally, a majority of whom are probably American. That kind of move would cause a wikiriot.

If an article or section is about something in America, or chiefly of interest to Americans, it should be in the English system because that is what we use. There should be mandatory conversion to metric units. If and when America goes metric we can revisit the issue. This is not the place to push agendas. If it relates to science or some other field where metric is the norm (e.g. some sports events), then use metric and convert if necessary to English. Articles that are chiefly about non-American subjects should use metric with no conversion required. If it's of universal interest that's a tougher call. Clearly there should be metric units, either primary or in conversion. It's a good question whether metric should be preferred and whether conversion to English should be mandatory.

There's no reasonable argument that the unit conversions make an article less readable. It's perfectly readable when you get used to it. That argument is as hollow as the argument by Americans that the metric system is too complicated for them. The short answer is "get over it." The whole debate over units, and the fervor that people bring to these debates on both sides, is ridiculous. Units are arbitrary. They are also a simple matter of arithmetic. If you were able to graduate from grade school you can handle units and conversions. Wikidemo 00:37, 24 August 2007 (UTC)


 * I do agree with Jim that conversions create a lot of clutter and make the text less readable. I wonder whether those on the oldspeak side might be willing to entertain a relatively minor change: that scientific articles be allowed to use metric without conversions, where there's consensus for this among the contributors? It makes sense, doesn't it? A fifth-grader in Ohio should be able to understand the distance to the moon in kilometres, yes? And if she doesn't, well, there's an opportunity for knowledge acquisition. Don't US schools acquaint all students with the metric system? That's what I've been led to believe. Tony 06:31, 24 August 2007 (UTC)
 * No, they really won't understand it. And neither will there parents, etc. May as will write the article in Greek - then they can be forced to learn an new language and a new alphabet. We simply don't have metric in everyday use to get instinctual understanding of it. Rmhermen 06:38, 24 August 2007 (UTC)


 * Um ... but they're intelligent enough to consult a scientific article? Doesn't add up. You just insist on a link for the first occurrence of each metric unit, simple as that. Tony 11:31, 24 August 2007 (UTC)
 * It depends what you mean by scientific. If it's truly a technical scientific article, sure.  no point talking about Chemistry or physics in pounds and inches.  If you want to learn about the subject you have to learn some terminology.  On the other hand if it's a natural history article about how large different subspecies of barn owls are or a meteorology article talking about the pressure caused by a wind shear, better mention both pounds and kilograms.  Reference on a specific subject comes from Wikipedia; basic technical literacy has to come from the schools, not us.  We have to address the audience.  We can't be so elitist we refuse to use terms people understand on the objection that they ought to know better.  Wikidemo 07:11, 24 August 2007 (UTC)
 * "Scientific" is already used as a classifier ("In scientific articles, SI units are the main units of measure, unless there are compelling historical or pragmatic reasons not to use them"). Are you raising a larger question here? Are you saying that the boundaries between scientific and non-scientific need to be further defined? No one has raised this issue about the existing point. Tony 11:27, 24 August 2007 (UTC) To this, I should add that if it ever came to the issue of whether an article is in the "scientific" category, the consensus of contributors would be required in any case. Doesn't that solve the issue. My proposal is a very conservative one. See new section at the bottom. Tony 14:42, 24 August 2007 (UTC)

It strikes me that the current system of "use what the source uses" is most appropriate and most accurate. To avoid WP:BIAS, it's not really appropriate to base it on any assumption of the audience. Of course, in areas where there is an established standard (SI for science), the use of sources to determine the unit should lead to the automatic use of SI in such articles. Outside of things like SI-for-science, I don't think that there's any overwhelming tendency anywhere in the english-speaking world towards particular measurement systems, let alone across all-of-the-ESW-except-the-US. Here in the UK, Imperial measurements are the norm culturally, and still legally in a few areas (such as road distances). SamBC(talk) 17:58, 24 August 2007 (UTC)

I'm all for reducing the clutter. Most of the world is used to the metric system; those who aren't shouldn't find it too hard to master.

One argument raised above is that WP shouldn't push people -- more particularly, people in the US -- to do this or that. All things being equal, I'd certainly agree. However, WP has already done this in certain ways. A clear example is WP's use of UTF-8 encoding for its articles and indeed for everything else. If you use a really old browser for en:WP, "foreign" characters are likely to be garbled. This isn't much of an issue for users of en:WP, but consider Japanese people using old browsers for ja:WP: all the characters will be garbled. UTF-8 is still unusual for Japanese-language websites, and ja:WP that's partly responsible for the disappearance of UTF-8-incapable browsers and thus the newfound ability of other people to use UTF-8 for Japanese. (I suspect it's a similar story for language versions with other scripts.) &para; Incidentally, many of the mentions of "US" above, should I think instead read "US, Burma and Liberia" (in one order or another).

How en:WP might start to move away from near-automatic conversions is a matter needing further thought. -- Hoary 09:13, 27 August 2007 (UTC)


 * The current policy should not be changed. Having imperial alternative units helps make articles more accessible at very little cost to readability. In addition to countries currently using the US Customary units there are many people living in countries that converted within peoples lifetime who still think in Imperial units. This is why the BBC News for example provides imperial unit conversions.


 * Wikipedia should not favor one system of units over another. Cultural considerations are as important as technical ones and favoring one system is pushing a point of view about such considerations. -- PatLeahy (talk) 15:34, 27 August 2007 (UTC)

Remember that such an enforcement would have far reaching implications. A selection of units we would no longer be able to use are light year, electron volt, parsec, atomic mass unit, Plank units, solar mass, speed of light (as in 0.1c), monolayers, and of course degrees Celcius. All of those units have important and legitimate uses, and there are many more like them. This is a poorly thought-out suggestion. Modest Genius talk 19:53, 3 September 2007 (UTC)

Contradictions II
Here's what I wanted to talk about - someone else moved my heading under "centuries" and then wondered why I want to talk about other contradictions under the wrong heading. If we don't exclude measurements, then it contradicts several examples in the Measurements section - 4 inches, 10 miles, 5° C, and 5 ft. Why do contradictions matter, whether or not they are about centuries? Spelling out numbers zero throught ten is the most frequently cited item on the page in my experience. If we don't tell people to do opposite things at the same time - as defined by the rules Wikipedia expects us to follow, not by the Chicago manual - then people would be more likely to obey, cite, and otherwise respect this guideline, and all the other headings on this talk page would be more likely to matter. Art LaPella 20:36, 10 July 2007 (UTC)
 * That someone else was me. I did so because it sure looked as if you were specifically referring to centuries and that's the way the discussion had been progressing.  Sorry to cause you strife.  I would say that a distinction can be made between contradictions and exceptions.  If we have this spell-numbers-out-from-zero-to-ten-but-not-from-11-up rule, then I'll be signing this post 11 July 2007 in the usual fashion but yours would have to have been something along the lines (the) ten(th) July 2007 or July (the) ten(th), 2007.  Would we want this?  I'd suggest not.  So exceptions should be allowed for. Jɪmp 00:17, 11 July 2007 (UTC)

Does that mean you endorse "excluding dates, times and measurements" or some variation that meets my goal of self-consistency? Or does it mean there should be a general disclaimer that there are exceptions to everything? Art LaPella 15:53, 11 July 2007 (UTC)
 * I would say that, in general, measurements and times should be excluded. The problem with such a blanket statement is that it is not always true. In a scientific article that is peppered with measurements, numerals make it easier to read, so they should be used. But in other kinds of prose those numerals (e.g. "The book, when originally published in 1778, was four inches square".) might be out of place. Context matters in measurement, I'm afraid. (By the way, I agree with the "exception" theory articulated above.) Awadewit | talk  16:06, 11 July 2007 (UTC)

When someone writes "WP:MOSNUM" in an edit summary, I believe (but I can't think of a good way to prove it) that they usually mean "Whole numbers from zero to ten should be spelled out as words in the body of an article." So I don't think this is a Ignore all rules situation; I think it should be fixed. How about "excluding dates, times and some measurements"? Art LaPella 18:57, 11 July 2007 (UTC)

Since you both like the word "exceptions", how about "Exceptions include dates, times and some measurements"? Art LaPella 19:09, 11 July 2007 (UTC)

Proposed revision of "Numbers in words"
I don't understand the title. Please provide feedback on this new version. Don't you all think it's time to bite the bullet on "billion"? (Except that it doesn't quite belong under the new title.) Should there be guidance on hyphenating spelled out fractions? I've added something, and I'm unsure about it. And what about hyphens in numbers such as "twenty-four"—mandatory or optional?

There's a problem with "*Fractions standing alone should be spelled out unless they occur in a percentage. If fractions are mixed with whole numbers, use numerals." that persists in my proposed amendment ("Fractions should be spelled out unless they occur in a percentage or with an abbreviated unit ("3.5 mm") or are mixed with whole numbers. Fractions are hyphenated ("seven-eighths")."). Fractions seem to be mixed up with decimal points.

EXISTING

Numbers in words


 * Whole numbers from zero to ten should be spelled out as words in the body of an article. Use numerals in tables and infoboxes.


 * Numbers above ten may be written out if they are expressed in two or fewer words, except in tables and infoboxes. Example: "sixteen", "eighty-four", "two hundred", "twenty million" but "3.75", "544", "21 million".


 * Proper names and formal numerical designations should instead comply with common usage. Example: Chanel No. 5, 4 Main Street, 1-Naphthylamine, channel 6.


 * Within a context or a list, style should be consistent. Example: "There were 5 cats, 12 dogs, and 32 birds" or "There were five cats, twelve dogs, and thirty-two birds", not "There were five cats, twelve dogs and 32 birds".


 * It is considered awkward for a numeral to be the first word of a sentence: recast the sentence or spell the number out.


 * Fractions standing alone should be spelled out unless they occur in a percentage. If fractions are mixed with whole numbers, use numerals.


 * Billion is understood as 109 in the United States, English Canada and most of the rest of the English-speaking world. (In most Germanic and Romance languages, it means 1012.) See the detailed discussion in English-language numerals, long and short scales. International monetary institutions generally recognize a billion as equaling 109 when it relates to expenditure, GDP, deficits.


 * Ordinal numbers should be spelled out in words using the same rules as for cardinal numbers. If digits are used, the ordinal suffix (e.g., "th") should not be superscripted. For example, "fourth", "twenty-third" or "23rd", "496th".

NEW

Spelling out numbers

General rule
 * In the body of an article, single-digit whole numbers (from zero to nine) are spelled out; numbers of more than one digit may be spelled out only if they are expressed in one or two words ("sixteen", "eighty-four", "two hundred", "twenty million" but "3.75", "544", "21 million"). Many people prefer all multi-digit numbers to be expressed generally as numerals, within the bounds of the exceptions below.

Exceptions
 * Dates and times.
 * Numerals that open a sentence are spelled out; alternatively, the sentence can be recast so that the number is not in first position.
 * In tables and infoboxes, all numbers are expressed as numerals.
 * Within a context or a list, style should be consistent (either “There were 5 cats, 12 dogs and 32 birds” or “There were five cats, twelve dogs and thirty-two birds”, not “There were five cats, twelve dogs and 32 birds”).
 * Fractions are spelled out unless they occur in a percentage or with an abbreviated unit ("3.5 mm", “⅛ mm”) or are mixed with whole numbers.
 * Ordinal numbers are spelled out in words using the same rules as for cardinal numbers. If digits are used, the ordinal suffix (e.g., "th") is not superscripted (“23rd” and “496th”, not “23rd” and “496th”).
 * Proper names and formal numerical designations comply with common usage (Chanel No. 5, 4 Main Street, 1-Naphthylamine, Channel 6). This is the case even where it causes a numeral to open a sentence, although this is usually avoided by rewording.

Hyphenation
 * Numbers from 21 to 99 not ending in zero are hyphenated (“fifty-six”), as are fractions ("seven-eighths"). Do not hyphenate other multi-word numbers (“five hundred”, not “five-hundred”).

Large numbers
 * Where values in the millions occur a number of times through an article, upper-case "M" may be used for "million", unspaced, after spelling out the first occurrence. This is particularly useful for expressing large amounts of money ("She bequeathed her fortune of £100 million unequally: her eldest daughter received £70M, her husband £18M, and her three sons each just £4M each").
 * Billion is understood as 109. After the first occurrence in an article, it may be abbreviated to unspaced “bn” (“$35bn”).

Tony 04:47, 12 July 2007 (UTC)


 * I have no objection other than the one you probably read under the previous heading . Nobody is arguing that 0-10 should always be expressed as words, so that rule should mention that there are exceptions. Art LaPella 05:03, 12 July 2007 (UTC)
 * I didn't read that previous entry, actually. Um ... can you be explicit as to your objection? I don't think this new version changes much WRT the ten/11 boundary for spelling out numerals, does it? If I could have it my way, I'd insist on numerals above nine as the norm, but I think too many people would object. Am I right?
 * So we need to know what change you'd require for this new version to have your total support. Tony 09:11, 12 July 2007 (UTC)
 * Since I had just called for "Exceptions include dates, times and some measurements" to be added to the "Whole numbers..." sentence, and your version omitted that suggestion without comment, I took that as unspoken opposition. But if you're not taking a position on my issue, your paragraph is otherwise fine with me. Art LaPella 21:33, 12 July 2007 (UTC)


 * Looks good to me. Sandy Georgia  (Talk) 14:50, 12 July 2007 (UTC)


 * If it were up to me, I'd have billion, trillion, etc. thrown out in favour of scientific notation.
 * Also, I'm not sure about your purpose behind the clause "many editors prefer all numbers above ten to be expressed generally as numerals". What does it matter what editors prefer (we're writing for readers)?  What do I do about this tid-bit of information? ... so many other editors prefer this ... great, me too ... and ... If it were up to me, I'd have the two-word rule replace the zero-to-ten rule altogether.  Who's hat was ten pulled out of?
 * Besides these points, I think this is an improvement. Here's food for thought, though.  Consider the first two points.


 * Whole numbers from zero to ten are spelled out as words in the body of an article, but are expressed as numerals in tables and infoboxes.


 * Except in tables and infoboxes, numbers above ten may be spelled out only if they are expressed in one or two words ("sixteen", "eighty-four", "two hundred", "twenty million" but "3.75", "544", "21 million"); many editors prefer all numbers above ten to be expressed generally as numerals. Within the bounds of the guidelines below, usage should be consistent within an article.


 * The first discusses numbers one to ten. The second discusses number above ten.  Both note the difference between prose & tables/info boxes.  Would it not be more straightforward to reorganise this so that one point discusses tables/info boxes and another discusses prose?  When to spell the number out need not then be mentioned in the former of these points.  Something along the following lines might do the trick.


 * Numbers are expressed as numerals in tables and infoboxes.


 * In the body of an article whole numbers from zero to ten are spelled out as words; numbers above ten may be spelled out only if they are expressed in one or two words ("sixteen", "eighty-four", "two hundred", "twenty million" but "3.75", "544", "21 million"); many editors prefer all numbers above ten to be expressed generally as numerals. Within the bounds of the guidelines below, usage should be consistent within an article.


 * Jɪmp 17:07, 13 July 2007 (UTC)

(1) I've stuck my neck out by changing the ten/11 boundary to nine/10, because it's so simple to conceive the spell-out-single-digits rule. Happy to hear objections on that. (2) "What does it matter what editors prefer"? It's a polite way of strongly recommending a usage without making it mandatory. (3) I suppose there's no support for The Guardian's "bn" as an abbreviation for "billion", is there? (4) I'm uncormfortable encouraging the writing out of "twenty million" yet the use of numbers for "21 million"; I'd rather "20 million". Tony 02:09, 14 July 2007 (UTC)
 * Jimp, I like your suggestions; I've tweaked and incorporated them. Please note:


 * Whilst I'd rather see the boundary pushed up higher (perhaps to twenty/21 ... being where we begin having to use two words to name positive intergers), I do like the non-arbitariness of nine/10 over ten/11.
 * I get what your saying about what editors prefer but can't help but feel as if the implication is that we're writing for editors rather than readers. Perhaps reword it so as not to mention who's doing the preferring.
 * bn for "billion" seems fine to me: no more ambiguous than billion itself but, as I've mentioned, I reckon we'd be better banning billion than biting bullets on it. I'd rather scientific notation for 108 and up.
 * Yeah, all words for 20 000 000 but numberals for 21 000 000 would be problematic but if we want 20 million we'll have to draw the line somewhere: we surely don't want 1 million ... I presume the boundary would be nine/10 million.
 * Jɪmp 16:42, 15 July 2007 (UTC)

Tony's responses:
 * OK, so we'll stay with nine/10, I guess. It's very common.
 * Changed "editors" to "people": is that what you meant?
 * I'll insert "bn" as an optional abbreviation, pending objections from others.
 * I suppose "eighteen million" and "18 million" are both allowed under both old and new versions. Guess the bit about keeping it consistent within a context or list will stop people mixing them awkwardly.

"Two- and one-digit numbers are hyphenated when spelled out ('fifty-six')..." But how would you hyphenate "nine"? There are no one-digit numbers like "fifty-six", so I don't understand why the words "and one-digit" were included. Art LaPella 03:45, 16 July 2007 (UTC)
 * Thanks for seeing that. Would this solve the problem? "Conjoined two- and one-digit numbers are hyphenated ...". Tony 05:56, 16 July 2007 (UTC)

I understood it, but it took me several seconds, perhaps because in mathematics, fifty-six is one number. An alternative is "Numbers from 21 to 99 not ending in zero are hyphenated...". Art LaPella 06:30, 16 July 2007 (UTC)
 * Clever man: my brain couldn't arrive at that. I'll insert it. Tony 15:39, 16 July 2007 (UTC)


 * 1. Let's consider three options;
 * a) ten/11—a seemably arbitary choice
 * b) nine/10—one digit when written as a numeral
 * c) twenty/21—one word when spelt out
 * I prefer b) to a) because this gives us a sensible reason but wouldn't you agree that c)'s reasoning is just as good (& in a sense better ... more positive—"avoid single digits by spelling the numbers out" vs "when a number can be written as a single word, go ahead and do so") but mostly I would just rather see the boundary pushed up rather than down—æsthetically speaking, words tend to look better than numerals in prose (to me atleast).
 * 2. People is good but how about recasting the clause into the passive?
 * 3. bn for "billion" is fine by me, the problem I have is with billion for "thousand million" (a somewhat different issue) but perhaps I swim against the current here—alas the current flows against logic ... but this is the English language, what's new?
 * 4. I think the consistancy rule solves a lot of these types of problem
 * Jɪmp 17:56, 16 July 2007 (UTC)

ten/11 is not at all arbitrary: ten is the greatest number childrens may count on their fingers. After "ten" begins the abstraction of the number. pom 12:24, 27 August 2007 (UTC)

m for million - with or without preceding space?
An editor has added a space before various occurrences of "m" for million, both in numbers (populations) and currency sums: 1.2m to 1.2 m and £3.4m to £3.4 m.  I find it clearer with no space (as do others: it wasn't my writing that was being edited). Is there any guideline or consensus on this? PamD 07:29, 12 July 2007 (UTC)
 * I don't like the space one bit. Tony 09:03, 12 July 2007 (UTC)
 * Curiously, I note that the editor who put the spaces in has now removed them! But is there a rule, anywhere? (Or is use of "m" for million deprecated anyway?) PamD 09:57, 12 July 2007 (UTC)
 * I don't mind them where values in the millions are used at least a few times in an article. It's easier to read—highly identifiable by the eyes. As to whether both upper and lower cases are acceptable ... I'm unsure. And I think people might balk at "k" for thousand, although I use it in writing grant applications and budgets. I do like "bn", the Manchester Guardian's abbreviation for "billion", as in "$3.2bn". In fact, "m" and "bn" are much more readily distinguishable than "million" and "billion". Tony 11:25, 12 July 2007 (UTC)
 * Space them, for the same reason we space everything else. BTW, I disagree strongly with the idea that "$3.2bn" is more distinguishable; it comes across to me as lazy abbreviationism for no particular purpose, and it is unfamiliar to most readers. My first thought was "huh? 3.2 buns? 3.2 in binomial notation? WTF does this mean?" Granted it only took about half a second to get it, but WP is written for people far below my level of education as well as us smarty editors; why make them get headaches for no reason? —  SMcCandlish  &#91;talk&#93; &#91;cont&#93;  ‹(-¿-)› 16:05, 13 July 2007 (UTC)
 * Where did you get this idea that such abbreviations are "lazy"? My only concerns are readability, smoothness and concision. We need to compare sample texts for a bird's-eye view of spaced and unspaced. Tony 16:09, 13 July 2007 (UTC)
 * Frankly, I find I always tend to first think "meters/metres" whenever I see lower-case "m". An alternative I often see, at least in American financial writings, is to capitalize the initial letter: $4.5M or $4.5B. (Usually there is no space between numbers and letter.) Askari Mark (Talk) 19:47, 13 July 2007 (UTC)
 * I agree entirely. Unspaced cap is best. Tony 01:56, 14 July 2007 (UTC)
 * Lower-case m for million is an abomination. Any scientifically or technically inclined reader will see this symbol as "metre" or "milli", depending on context. It is cognitively wrenching otherwise: your lecture comes to a dead stop, you think furiously for a few milliseconds, and then you say to yourself « Oh ! It must mean million ! ». Urhixidur 15:28, 16 July 2007 (UTC)
 * So? Do you mind the upper-case “M”, and for that matter, “B” or “bn” for “billion”? Tony 15:36, 16 July 2007 (UTC)
 * I agree that it's an abomination, and 1.2m (as a population figure) should really not exist. As for currencies, I do get the point of £3.4m as spelling out "million" would require a space, making for "£3.4 million", which is strangely grouped, looking more like "(3.4 pounds) millions" than "(3.4 million) pounds". Of course, 3.4 million GBP and 3.2 billion USD, or 3.4 MGBP and 3.2 GUSD using an SI-style prefix, is another way to go. I don't know what our policy on this is, but I don't really like these unqualified $s and £s in a multinational encyclopedia. Of course, most of the time it's clear from context whether the dollars are American or Australian or Canadian, but we can really never be too clear. Anyway, as I said, avoiding the space is pretty much the only point of using m and bn, so no, no space. -- Jao 15:39, 16 July 2007 (UTC)

Space after "c."? Examples mixed in MOSNUM
Have just looked at MOSNUM (having earlier been deterred by the "Out of date" tag). There seems to be an inconsistency:
 * Dionysius Exiguus (c.470 – c. 540)
 * Rameses III (reigned c.1180 BCE – c. 1150 BCE)

Is there really supposed to be a space after the c. for deaths but not births? I saw Dionysius, thought it was a typo, then saw Rameses. Does this need tidying up? (As a newcomer to the MOSNUM pages I hesitate to tread on what's probably hotly-disputed territory) PamD 10:02, 12 July 2007 (UTC)
 * I'm afraid that this is the result of my impulsive change earlier today (adding the spaces, since c.1850, to my eye, looks awkward); someone came along later and reverted most of them. Sorry for my part in this. Do people really prefer this without spaces? Tony 11:27, 12 July 2007 (UTC)
 * Yes. Johnbod 14:53, 12 July 2007 (UTC)

No, I feel that a space is called for. More strongly, I think it should be either all-spaced (c. 470 – c. 540), or without any spaces (c.470–c.540), according to WP:MOSDASH "All disjunctive en dashes are unspaced, except when there is a space within either or both of the items". So the question is simply whether a space always follows "c."

I will add spaces around the en dash for Charles Darwin's and Genghis Khan's examples, which are currently flat wrong.

I will leave alone the (fl. 760–772), because "fl." pertains to both years of the date range, and so the space is between "fl." and the date range, and is not within an item of the range. Compare with (c. 640 – 687), which indicates an approximate year of birth with a known year of death (a common situation for folks killed in battle). Does (c.640–687) muddy the distinction? I think it might. Chris the speller 16:24, 12 July 2007 (UTC)
 * Hat taken off to you, Chris. Tony 16:29, 12 July 2007 (UTC)


 * Aren't you contradicting yourself in saying you prefer "(fl. 760–772), because "fl." pertains to both years of the date range"; should you not then also prefer (c.640–687), where only the first date is approximate? I just find a sentence like  "Rameses III was ruler of Egypt c. 1180 BCE – c. 1150 BCE and was succeeded by his ...." too sprawling for easy reading. Johnbod 16:40, 12 July 2007 (UTC)


 * This inconsistency was introduced in the past couple of months, and circa does not belong with a space. —Centrx→talk &bull; 01:41, 13 July 2007 (UTC)
 * Why not? Tony 11:49, 13 July 2007 (UTC)
 * Indeed, why not? "c." is an abbreviation of "circa", so "circa 1250" should logically become "c. 1250" (it's not an abbreviation of "circa1250", after all), and there has been no expanation for the disappearing space. I suspect that some editors have simply fallen into the habit of skipping the space, and it now feels right to them. Chris the speller 12:54, 13 July 2007 (UTC)
 * I think this would most usefully be addressed by considering the style used by other encyclopedias, as has proved so convincing (to me anyway) on the spelling out of centuries above. Johnbod 13:48, 13 July 2007 (UTC)


 * OED seems to use no space, no stop: "c1234". Times style guide specifies space, no stop: "c 1234". Britannica seems to use stop and space: "c. 1234". ODNB seems to use no space, stop: "c.1234". (Wonderful specimen there is "Smith, Theodore (fl. c.1765–c.1810x23)" - I've no idea what the "x23" means!)   So there is precedent for all 4 variations: stop or not times space or not! Does that help?  PamD 15:16, 13 July 2007 (UTC)


 * This shouldn't even be a debate. Of course it should be spaced, for the same reasons we don't write "fl.209-256CE". —  SMcCandlish  &#91;talk&#93; &#91;cont&#93;  ‹(-¿-)› 14:36, 13 July 2007 (UTC)


 * I agree with SMc. And my personal preference would be for the dot. I don't like dots normally, but without it, this abbreviation seems just not as distinguishable as you'd like. We're so used to letter/number combinations, and c 1823 could be a train-engine number. BTW, I see ranges cropping up here with hyphens. MOS insists on en dashes. Tony 15:24, 13 July 2007 (UTC)


 * I also concur. Whether “c” or “c.”, it’s an abbreviation of a separate word, so there should be a space (although I’d ignore it in a table with space limitations).  Since the standard is to use a period/stop to denote such an abbreviation, as with “fl.” for “flourished”, the same should apply here.  I’m surprised that any style guide offers something as atrocious as “fl. c.1765–c.1810”; “flourished” itself means it’s an approximation, and is used when neither birth nor death information is available.  Askari Mark (Talk) 19:58, 13 July 2007 (UTC)