Wikipedia talk:Manual of Style/Dates and numbers

Unofficial anagram of the Manual of Style

Number ranges more generally

Meant to bring this up a long time ago, but forgot. The logic and resultant rule we're applying to date ranges – to give them in the form 2002–2011 not 2002–11 – applies to all numeric ranges (outside of directly quoted material) and we need to state that explicitly. I keep running into sporadic WP:WIKILAWYERing along lines which can be parodied as "I can use 'p. 2002–11' if I wanna because MOS:NUM#Ranges only technically applies to dates, and no matter how much WP:COMMONSENSE dictates that that the reasoning applies across the board, the rules don't quite say it, so ha ha ha." Let's just fix it.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  06:18, 3 October 2017 (UTC)

It seems to me that the preference for a single piece of guidance in Wikipedia:Village pump (policy)/Archive 129 #WP:DATERANGE ambiguity and stylistic concerns RfC debate would apply equally to the choice between full number ranges and abbreviated ones, in particular phrases like "pages 153–158" vs "pages 153–58" or "pages 153–8". It would be a step forward, I think, to encourage one style and deprecate the others. Reading the RfC, I believe that there is likely consensus for a new sub-section in Wikipedia:Manual of Style/Dates and numbers #Numbers titled "Number ranges" with guidance along the lines of "Number ranges, including page ranges, when given as ordinals, should state the full number both for the beginning and end of the range." I expect that the usual exceptions like "In tables and infoboxes where space is limited ..." and when quoting a source directly would need to be mentioned, along with a few examples.

You will either need to start a new RfC, or try a bold edit to this page and see if it sticks. Or perhaps there will be lots more debate here, whence we can determine consensus. Cheers --RexxS (talk) 18:56, 9 October 2017 (UTC)

Took the bold approach [1], since all the reasoning in the RfC applies to other date ranges. We can of course RfC it again if necessary.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  23:35, 10 October 2017 (UTC)

User:SMcCandish, I'd much prefer to retain the option. How do you like journal page ranges such as "43241–43248". Really? I'd do a service to readers: "43241–48". Tony (talk) 01:36, 11 October 2017 (UTC)

The problem I see is that for every case of "43241–8" (which is honestly hard to be certain is intended to be a range rather than some kind of obscure serial number or something, so yes I do think "43241–43248" is preferable) there's going to be about 1000 cases of "125–9" or "1901–05". The last kind of case is seriously problematic since it looks like a Y-M date; many fonts do not distinguish clearly between - and –, and many readers are not clear on the distinction even if they can see it. Maybe a caveat can be added that if the range begins with a number higher than 2100 the short form can be used. That should avoid any potential confusion with dates up to the near future. PS: We can't do something like "If it could be confused for a date, use the long form" because some of this data is auto-generated by templates and bots and such. We can't depend on every number range to be human edited. That'll be even more true the more WikiData stuff is implemented here.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  03:38, 11 October 2017 (UTC)

"125–9" or "1901–05"—Isn't that what Chicago demands? Tony (talk) 08:04, 11 October 2017 (UTC)

CMoS has a complicated four-tiered system for compressing ranges, geared to avoiding ambiguities (they say, though I don't see that it deals with date-confusion), and based on how large the starting number is. I strongly suspect their system is too fiddly for WP to accept (plus it's a manual thing, and can't reliably be done in an automated way). The CMoS system, and any others for compression, are concerned with paper publications where space-saving is a concern. CMoS also recognizes the don't-compress rule we implemented for dates, and the compress-maximally ("2934–8") approach some want to use here. Maybe more importantly, it expects citations in a consistent format with a particular order of authors, dates, pages, etc., and a more limited number of such parameters, where we have utter chaos. The potential for confusing dates and other things is much higher on WP, especially since people are empowered by WP:CITEVAR to invent their own citation "styles" (according to CITEVAR's regulars, anyway), and our templated citations have parameters for just about every kind of source ID there is, many of which are numeric. PS: Anyone should feel free to revert the WP:BOLD addition of the "Number ranges" section, of course. If it needs more discussion and perhaps more exemptions or whatever, we can work that out.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  17:01, 11 October 2017 (UTC)

Any further input on this?  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  05:36, 30 October 2017 (UTC)

On a related issue, Mac. Why is it that a range of general numbers e.g. 10–15 is OK to be written this way, but 1–9 should be spelled out as "one to nine"? At appears to be inconsistent to me. William Harris • (talk) • 09:45, 20 November 2017 (UTC)

Surely it all depends on context? Citing "pages 1–9" seems perfectly reasonable to me, as does "the boat could carry one to nine additional guns". Who would write those differently? I think the problem is in expecting us to be able to write prescriptive rules to cover every possible circumstance. The English language simply isn't that amenable to regulation. --RexxS (talk) 14:13, 20 November 2017 (UTC)

Right. In running prose, we'd normally write "one to nine" because we normally write "one" and "nine". But we wouldn't do that to things conventionally given in numerals, like page numbers, addresses, sports scores, vote tallies, measurements, etc.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  19:34, 20 November 2017 (UTC)

Many thanks for both of your replies. It was measurements that I had in mind as I contribute largely to biology-related subjects. I certainly would write 5-7 kg, or 2–7 km, but that is not what the guide advises. It would look strange stating that "the smaller subspecies weighs from five to seven kilograms and the larger subspecies 10-14kg." Following from your earlier lead, Mac, I have been WP:BOLD on the article page. William Harris • (talk) • 20:02, 20 November 2017 (UTC)

@William Harris: Where do you think the wording needs clarification? I'm not sure where your "not what the guide advises" is coming from. MOS:NUM#Number ranges doesn't suggest anything like "five to seven kg" [or "... kilograms" or "... kilogrammes"], and only illustrates ranges with figures. If anything, we might need to clarify that one should write "five to seven survey respondents out of ten", when the numbers are not something we normally always put in figures (scores, votes, measurements before units, etc.), since the section doesn't actually illustrate this usage.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  22:45, 20 November 2017 (UTC)

PS: I think I've IDed the problem, and have addressed it under separate cover at #Missing point in "Numbers as figures or words" section, below.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  22:54, 20 November 2017 (UTC)

Regarding: Numbers as figures or words, Generally, in article text: "Integers from zero to nine are spelled out in words." We now turn our attention to the meaning of this word "general" - (of a rule, principle, etc.) true for all or most cases. What I am referring to does not appear under the Notes and exceptions nor does it appear under Number ranges. Somewhere, it should. William Harris • (talk) • 00:33, 21 November 2017 (UTC)

 Fixed [2].  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  00:30, 22 November 2017 (UTC)

Jewish dates

Is it possible to add Jewish dates in templates (for example {{cite news|date="7 Tishrei, 5775 // October 1, 2014"}})? --Jonund (talk) 19:19, 6 November 2017 (UTC)

Not in {{cite news}} and other Citation Style 1 templates. Dates are checked for validity. Only Gregorian calendar months are supported, and years are limited to the current year + 1 (e.g. in 2017, years up to 2018 are supported). – Jonesey95 (talk) 14:43, 29 November 2017 (UTC)

Decades

The example given "1960s Boston" - since it is referring to "Boston of the 1960s" I believe it should be written with an apostrophe in this case! i.e., "1960's Boston." Be good! 238-Gdn (talk) 07:39, 7 November 2017 (UTC)

No. 1960s is a plural not a genitive. Your example "1960's Boston" would be refer to Boston in 1960, not for the ten years as a whole. By the way be careful of "decade". To be strict (which as an encyclopaedia we should) the decade runs from 1961 to 1970 inclusive, not from 1960 to 1969. See the discussion on the page. Martin of Sheffield (talk) 09:47, 7 November 2017 (UTC)

No, your second point is wrong. A decade, as the article says, is any period of ten years. There is absolutely a decade of the 1960s, and it runs from 1960 to 1969. There is also a "seventh decade of the twentieth century", which runs from 1961 to 1970. --Trovatore (talk) 16:50, 13 November 2017 (UTC)

Ah, you're right that it's a plural, but it is also genitive (or possessive). In that case, it should be 1960s', with the apostrophe after the s. By the way, it's not my example, but one of the examples offered in the text, which is not accurate, and is therefore confusing. I suggest changing the example and writing a new paragraph with examples in which use of the apostrophe would be required. Be good! 238-Gdn (talk) 10:11, 7 November 2017 (UTC)

[3] It's no more a possessive than are 1963 Boston or 19th-century Boston, though constructions such as from the nineteenth century's strictures to the 1960s' permissiveness are conceivable. EEng 10:42, 7 November 2017 (UTC)

• Don't use an apostrophe. It's not necessary, and it will just inspire people to insert the apostrophe all over the place ("in the 1960's"). The "1960s Boston" case is not a possessive unless you are engaging in farcical anthropomorphism. It's simply an adjective, in exactly the same form as "nineteenth-century Japan". It is possible to conceive of "the nineteenth century's Japan" but this is silly poetics, not encyclopedic writing. PS: The fact that one has a hyphen and the other doesn't clearly illustrates that the two constructions are different in nature, which was harder to detect with a single-word case like "1960[']s".  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  21:29, 13 November 2017 (UTC)

Separators other than hyphen

Neapolitan mother picking nits (EEng did not place this here, but he wishes he had.)

Let's stop arguing this through edit summaries.

It seems we have three separate concerns. Stanton doesn't like changing between singular and plural in the same sentence. EE notes that the other entries in the table are plural, and points out that you can say "letters other than X". My concern is that "hyphen" is not the name of the hyphen in the same sense that X is the name of the letter X.

Is that an accurate summary? If we agree on the problem, we might have a better chance of finding a solution. --Trovatore (talk) 16:40, 13 November 2017 (UTC)

Oh for crying out loud, this is a table of minimalist instructions:

• Do not abbreviate year
• No comma between month and year
• Comma required
• Do not abbreviate year (and, yes...)
• Do not use separators other than hyphen

It's not supposed to be full sentences with all the trimmings (Do not use any separator other than the hyphen) and the particular choices made depend in part on the examples involved, and the surrounding injunctions. (I'm the primary author of the table – with the help of various of our esteemed fellow editors of course – so call it ownership if you want.) But if it keeps you up at night, knock yourselves out. I can't guarantee, though, that years from now I won't forget and tinker it back. EEng 17:31, 13 November 2017 (UTC)

But it is supposed to be written in English, not Pidgen. There is a clear difference between "Do not use separators other than hyphen" and "Do not abbreviate year", since 'year' in that context is standing in place of an actual year: the latter is tantamount to saying "Do not abbreviate 1849 to '49"; while the former isn't an injunction against writing 'yr' or some other abbreviation of the word 'year'. A hyphen is itself, the name of the separator, and in that context requires an article when used in the singular, otherwise it reads jarringly. I'll just repeat here for completeness that other singular terms in the table like "a dot", "an abbreviated month", "a leading zero" are rendered complete with an article, and quite rightly. --RexxS (talk) 18:14, 13 November 2017 (UTC)

OK, whatever. A table like this is meant to give short, punchy injunctions that can be taken in at a glance; they're as far away from the formality of article text as you can get. Next you'll be turning Comma required between day and year into A comma is required between the day and the year (or maybe, I suppose, The comma is required between a day and a year, or maybe some other combination of definite and indefinite, according to some elaborate analysis). You guys are being silly.

And don't forget to change No comma between month and year --> The/A comma must not be used between the/a month and the/a year. 'Cause articles are important.

EEng 18:58, 13 November 2017 (UTC)

• The fact that you can say "letters other than X" is kind of irrelevant. I can say "I don't drive any cars but Toyotas" and that comes off better than with a singular "Toyota". Same issue here. I really don't care that much, other that by doing the "Do not use separators other than hyphens" plurality agreement, we also get rid of the "I don't like your telegraphic 'Do not use separators other than hyphen'" objection someone had; two birds, one stone. I have no idea why EEng thinks that "hyphens" is objectionable here when clearly it isn't. The fact that "hyphen" isn't linguistically impossible to use here doesn't make it the better choice, when it solves nothing and raises two (or is it three?) independently observed problems. But I really WP:DGAF. I was just responding to REVTALK minor editwarring triggering my watchlist, by providing a solution that both sides should find acceptable. If they won't, there's nothing I can do about that. :-)  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  21:19, 13 November 2017 (UTC)

Watch out, someone'll tell you you should write "I don't drive any cars but Toyota's". If we settle on Do not use separators other than hyphens I probably won't unthinkingly change it years from now. EEng 21:33, 13 November 2017 (UTC)

In het Nederlands zou je kunnen zeggen "Ik rijd geen auto's behalve Toyota's" :p Dondervogel 2 (talk) 21:50, 13 November 2017 (UTC)

The only people who would tell you to write that are all Grocer's. --RexxS (talk) 23:30, 13 November 2017 (UTC)

Um, "I don't drive any cars but Toyota's" is entirely correct (if a bit odd-sounding – which is why I was joking that this crowd might want to impose it). Compare "I don't like any fish but mama's". EEng 23:47, 13 November 2017 (UTC)

Seems a bit "if Mark Twain had a time machine".  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  09:04, 16 November 2017 (UTC)

User:EEng#Museum_of_If_Mark_Twain_Had_Been_a_Gynecologist. EEng 18:25, 16 November 2017 (UTC)

• Can we agree on Do not use separators other than hyphens? EEng 04:25, 16 November 2017 (UTC)

  Sure.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  09:04, 16 November 2017 (UTC)

  Of course. --RexxS (talk) 17:39, 16 November 2017 (UTC)

Done. EEng 14:08, 20 November 2017 (UTC)

Symbols for dot and micro

What symbols should {{convert}} display for units that need a product dot or a micro sign? I suspect that convert uses the wrong symbols. Examples:

• Unit kW.h has symbol kW·h (uses middot). Discussion at Template talk:Convert#Dots.
• Unit um has symbol µm (uses micro). Discussion at Template talk:Convert#Micro symbols.

Dot

• ⋅ = U+22C5 = multiplication dot = dot operator = &#8901; = &sdot;
• · = U+00B7 = interword separation = middle dot = &#183; = &middot;
• WP:MOSMATH says to use sdot for math products.
• WP:MOSNUM says to use middot for products of units, example m·s for metre-second.

Micro

• μ = U+03BC = Greek small letter mu
• µ = U+00B5 = micro sign (displayed by µ in at least two browsers)
• Micro-#Symbol encoding in character sets says to use mu, not micro.
• WP:MOSNUM says to use mu for μm.

I'm planning to change convert to use sdot (instead of middot) and mu (instead of micro). Using sdot contradicts the first WP:MOSNUM link above. Thoughts? Johnuniq (talk) 00:50, 18 November 2017 (UTC)

The mid-dot in unit combinations, is clearly a multiplication symbol and should use the same symbol as any other mathematical multiplication (I don't care which of the two dots is adopted, but it shold be the same dot for both).

The μ in μm is a Greek lower case mu - no other character should be used for this purpose IMO.

In other words, I agree with Johnuniq's proposal to harmonise these. Dondervogel 2 (talk) 09:21, 18 November 2017 (UTC)

• I support.

Wrt the dot, the two MOSes are contradicting and one should be changed (it follows, WP:MOSNUM should be changed). I note that per SI, the unit is an algebraic formula (allowing calculations), so in m·s (for metre-second) it is the multiplication symbol.

Wrt the micro/mu: as an SI-prefix it is a letter, not a math operator symbol, so the Greek letter is the obvious one. The confusion arose when Unicode created two characters for this one letter. (But maybe outside of SI-prefix, the "micro" does have a meaning?) -DePiep (talk) 12:36, 18 November 2017 (UTC)

• Support updating WP:MOSNUM to suggest to sdot, and to use mu and sdot in templates for units. I imagine that WP:MOSNUM may have been written without considering the alternative to middot. —Quondum 23:54, 19 November 2017 (UTC)
• Question: ping @SMcCandlish, RexxS, Tony1, and EEng: I'd like to hear from the regulars here, would be nice if the two MOSes and {{Convert}} all align. -DePiep (talk) 19:59, 20 November 2017 (UTC)
• Support updating WP:MOSNUM to use sdot for compound units, and to use mu and sdot in templates for units. I am used to the sdot (dot operator) indicating a product, but of course that might be old-fashioned! It wouldn't hurt to settle on Johnuniq's suggested scheme, and would only require bringing MOSNUM into line – I'm fairly certain readers won't be confused by seeing compound units like m⋅s and kW⋅h, as opposed to m·s and kW·h. Slightly tangentially, quantities like torque are vector products, rather than scalar products like work, but nobody is going to complain about us using N⋅m (I hope). --RexxS (talk) 20:35, 20 November 2017 (UTC)
• I'll go along with whatever you guys settle on. EEng 20:44, 20 November 2017 (UTC)
• Support John's proposal, and fix mosnum to say sdot instead of middot. Kendall-K1 (talk) 21:41, 20 November 2017 (UTC)
• Looks to me like we should be consistently using:
  • ⋅ – U+22C5 (dot operator a.k.a. multiplication dot), ⋅, ⋅, ⋅
  • µ – U+00B5 (micro sign), µ, µ, µ

The others are just approximations, and the names of these two unmistakably indicate they are the correct characters for these purposes. The other dot is a typesetting character, not a maths operator, and the mu is a Greek letter not a technical symbol (though the latter is obviously derived from the former). It's irritating that Unicode actually does this stuff with characters that are visually identical, but it does, and we're stuck with it. PS: I don't agree that "as an SI-prefix it is a letter, not a math operator symbol". "Micro sign" != "math operator". It's simply a symbol, and the correct one to use here, because μm is not Greek (i.e., we would not put it or the full version, micrometer/micrometre, inside {{lang|el|...}}). This is science and other fields repurposing what originated as a Greek letter, for symbolic purposes unrelated to rendering Greek text. Saying we should use the Greek letter μ is tantamount to saying we should not use the ™ symbol, but instead use <sup style="font-variant: small-caps;">TM</sup>.
 — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  21:43, 20 November 2017 (UTC)

Now that you mention it I do recall the special micro sign, distinct from mu. So yes we should use that. EEng 22:13, 20 November 2017 (UTC)

Please review the discussion at Template talk:Convert#Micro symbols where reasons for using mu not micro are given. Extracts: Micro-#Symbol encoding in character sets + WP:MOSNUM + "in every case that I checked, the encoding is U+03BC". Johnuniq (talk) 22:26, 20 November 2017 (UTC)

Skimming that, I don't find it compelling. I too was involved in Unicode mailing lists back when, and many decisions where typical, arbitrary (in the negative sense) "design by committee" gaffes, while many were not and were pointedly about separating symbols from natural-language characters that look like them. One person's anecdote about what they remember about what they thought might have been the intent back when, well, it just isn't very meaningful. What other people are using in their house style isn't very informative here, either, especially when the characters are essentially identical. (This is a bit like the dash-versus-hyphen-versus-minus and x-versus-×-in-math disputations; it really doesn't matter one whit that the average publisher just uses "-" and "x"; we use "−" and "×" in a maths context because they're the functionally correct characters to use, even if the average person neither notices nor cares.) There's a Unicode character specifically for micro- reduced to a symbol, it has not been deprecated, we have no indication it will ever be deprecated (any more than will IPA symbols that also coincide visually with natural-language ones in various writing systems), and it has no connection to rendering of Greek-language text, so we should obviously use the dedicated micro- symbol, as intended, when we mean micro- not mu in Greek-script quoted material. The fact that various off-WP writers have preferred the Greek letter is almost certainly because character-picker tools like Windows Character Map and PopChar (Mac OS) make it easier to find the Greek letter, in a code block for Greek right after the Latin alphabet, than to find the micro- symbol, mingled in confusingly with various other symbols. You have to know where (in your preferred tool) to look for it. That's a good reason to spell out what character it is and how to insert it as a decimal or hex HTML character entity, and to have templates auto-emit it.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  23:15, 20 November 2017 (UTC)

See my comment at the other thread. Micro is a compatibility decomposable character. It "would not have been encoded in the Unicode Standard except for compatibility and round-trip convertibility with other standards" (section 2.3) Kendall-K1 (talk) 00:29, 21 November 2017 (UTC)

But it has been, and it's there, so we should use it instead of misapplying a glyph from the Greek-language character subset for non-Greek purposes. I would love it if Unicode eliminated all duplicate and near-duplicate glyphs (especially since various characters that should be in 8-bit Unicode (UTF-8) are not due to lack of namespace room) but it's never going to happen.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  04:19, 21 November 2017 (UTC)

The Greek letter mu has been designated per SI to be the symbol for SI-prefix micro-. Once and where declared a symbol, it is not the "Greek language" any more (Greek script it is, but language =/= script). It is not up to Unicode to overrule that SI definition. Anecdotal background is not an argument, nor is Skimming that [micro/mu discussion], I don't find it compelling a strong argument. I point to the Kendall-K1 quote above @00:29: that the micro sign is a compatibility decomposable character, per a major Unicode policy. That is: added to provide compatability with legacy code systems (think old ASCII with limited number of code points so only selected non-Latin characters were imported into Latin set). Also, nowhere does Unicode claim or even suggest the micro sign should replace the letter mu anywhere. Also no need to go to the "confusion" area of the web, of Unicode, or of anyone's understanding. Both Unicode and SI are perfectly clear in this, and the web can handle this clarity.

And no, the micro/mu issue is not the same as the sdot/middot issue (as stated below), because the dot is a math operator, and the mu-prefix symbol is not. Math operators have different semantics from other dots, the prefix symbol has not: it is still the Greek letter. -DePiep (talk) 09:11, 21 November 2017 (UTC)

This is turning completely circular, and I decline to reiterate my position just because you're reiterating yours. We've both made our case. I don't want this to turn into one of those "SI mebi- and gibi- prefixes" types of near-interminable disputes. There are arguments in both directions, and consensus will settle on one or the other, probably not on the basis of technical assertions about which is more "correct" according to SI, or, well, the SI prefixes debate would not have dragged out for so long and would not have concluded the way it did.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  09:39, 21 November 2017 (UTC)

I added half a dozen of new arguments, especially responding to your posts. Skimming may be not enough. Also, your denigrating editsummary is not adding arguments. -DePiep (talk) 09:55, 21 November 2017 (UTC)

"Meh" isn't a denigration of anyone or anything; it's an expression of (usually growing) indifference and a feeling of tedium [4][5]. The arguments you added don't appear to be new, but recycling of those made in the earlier discussions already linked to. I don't think I have anything new to say about them, and I doubt anything new is forth coming in support of them, so I don't see the point. We have both made our case, and others can give their input.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  19:09, 21 November 2017 (UTC)

re SMcCandlish Both your "glancing" and your "meh" remarks are dismissive wrt the discussion proces and their contributors (see WP:IDIDNOTHEARTHAT). I note that I made several replies in my response to your posts. List of Newies, short: "not Greek language", "language =/= script", "different semantics", "Unicode policy", "Unicode does not claim ...", "compatability", .... None a repetition, also because this was my first reply to you. -DePiep (talk) 15:14, 22 November 2017 (UTC)

I'm not declining to hear anything, just to continue arguing about it. I've already indicated I WP:DGAF about the outcome of this. I've presented an argument for using the Unicode symbol that's intended specifically as a symbol for the kind of purpose at issue here, and you've presented a case to use the Greek letter, based primarily on some regret within the Unicode Consortium for having done things this way in the first place. I'll be okay with whatever consensus emerges on the matter. Neither view is crazy; one just favors using the tool as-is, the other favors using the tool as its creators have reimagined it (without them actually changing the Unicode spec itself; if they actually marked U+00B5 deprecated, I would take your side in this).  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  19:12, 22 November 2017 (UTC)

• mu. The use of mu as a metric symbol predates SI, let alone Unicode - for example, the 1911 Kaye and Laby uses it. SI is independent of Unicode and does not define symbols in terms of Unicode; it defines the micro- symbol as mu, not as any particular Unicode character. 92.19.24.9 (talk) 23:21, 20 November 2017 (UTC)

  We know. WP works in Unicode like the rest of the Web. Where Unicode provides two glyphs for the same thing (the character specified by SI), one a technical symbol, and one specifically for Greek text, we use the the former for the former purpose, we don't misuse the latter for it, or misuse the former for Greek material. Again, this is exactly the same issue as hyphen, dash, and minus characters, or x versus ×. It does not matter that some other publishers chose to use - for minus and x for multiplication because they don't care about Unicode distinctions; WP is not among those publishers. Another example is duplicate-looking glyphs in Greek and Latin, and in Greek, Cyrillic, etc. We don't use the Greek ones in Cyrillic or the Latin ones in Greek just because we feel like it or they're easier to type, or any other reason.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  00:26, 21 November 2017 (UTC)

  SMcCandlish – I see nothing in your argument other than that the micro should be used because it exists as a separate character with no other semantics (whereas mu also has this meaning). That they look the same is no accident: the one was supported by an early ad hoc extension to the ASCII character set without Greek characters predating Unicode; it is arguably the same character with an archaic encoding, and is evidently what led to your lament: that it looks the same. It is also clear that it is a legacy encoding that is effectively deprecated by Unicode (see quote by Kendall-K1). I have checked several SI standards PDF documents and web pages such as at NIST, they uniformly use mu, not micro. Unlike the argument for retaining the archaic inch in the US, adopting the Greek mu in this MoS would have no expected transition impact. —Quondum 13:47, 21 November 2017 (UTC)

  "should be used because it exists as a separate character with no other semantics" is exactly the reason we use the proper character for minus, for multiplication, and for many other things. I don't see a solid rationale for diverging from that pattern. The fact that some people (including some at Unicode) want[ed] to deprecate the one character in favor of the other is immaterial; it has not been deprecated. By way of analogy, many at W3C and WHATWG wanted to deprecate (for what they thought were good reasons) the non-semantic font-styling HTML elements like <b>, <i>, <u>, and <s> in HTML5 (in favor of using styled spans for these things when they do not have a semantic value conveyed, respectively, by <strong>, <em>, <ins>, and <del>) but this did not end up happening, so these elements are in fact still in use and legitimate to use. (They did deprecate <tt>, in favor of <code> and the related semantic tags, so <tt> should not be used here or elsewhere, even if browsers are generally not going to choke on it). "I wish this were deprecated" isn't good enough for me, especially when it runs into "use of micro- and the symbol for it in English has F'-all to do with rendering Greek-language material". I can't think of anything else to add that won't simply be rehash of what I already said. Bazillions of things are among those that some people want to deprecate and which do not actually get deprecated; life goes on.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  19:09, 21 November 2017 (UTC)

  SMcC, saying "we don't misuse" at best begs the question of whether it is misuse. Use of Greek letters is not only appropriate, it's prescribed. It's worth remembering that the U+00B5 coding is a bodge that predates Unicode; you'll find it in the 256-character set of the original IBM PC (later known as Code page 437) as one of the 14 upper- and lower-case Greek characters that were squeezed in to support various mathematical, scientific and technical uses. Unicode supported that bodge, of course, but it also supports the full Greek alphabet so there is no longer any need to use U+00B5. As Unicode Technical Report #25 puts it, "Implementations therefore need to be able to recognize the micro sign, even though U+03BC μ is the preferred character in a Unicode context." 92.19.24.9 (talk) 21:34, 21 November 2017 (UTC)

References

• Just in case my own position was not clear (shifting my position from neutral to use of sdot on the mid-dot question; unchanged on μ):
  • I support use of the multiplication symbol for all multiplication operations, including multiplication of units (seems self-evident)
  • I support use of the Greek letter mu for the μ in μm because μm is an SI symbol, and the SI symbol for μm uses a Greek letter mu (again, this seems self-evident to me, but I see that others are arguing the opposite - hence the need to reiterate the logic).

Dondervogel 2 (talk) 10:51, 21 November 2017 (UTC)

• This comment is a very incomplete substitute for a methodical summary of points raised which I don't have time for. The dot issue has been resolved and convert will use sdot. The micro/mu issue is a problem with many supporting mu based on reasons that seem convincing. The pushback supporting micro is strong but seems to be based on a belief that micro's existence is proof of its correctness. Countering that is the claim that micro exists only for certain technical reasons (see "compatibility and round-trip convertibility" above). What evidence is there to support micro? I was once a strong supporter of micro as can be seen at User talk:128.178.189.30 but my faith started wavering due to DVdm's comments there. I still support one thing I said, namely Unicode is a black hole from which useful information rarely escapes. Johnuniq (talk) 22:19, 21 November 2017 (UTC)

The quote I gave in the other thread seems pretty conclusive to me: "The purpose for the inclusion of compatibility characters like these is not to implement or emulate alternative text models, nor to encourage the use of plain text distinctions in characters which would otherwise be better represented by higher-level protocols or other mechanisms. Rather, the main function of compatibility characters is to simplify interoperability of Unicode-based systems with other data sources, and to ensure convertibiliity of data." In other words, it's not considered a separate character. It's more like the difference between the character "é" and the character "e" combined with a separate COMBINING ACUTE ACCENT. Kendall-K1 (talk) 23:02, 21 November 2017 (UTC)

Not a correct summary of my argument, but I'm beyond the WP:DGAF line at this point.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  00:00, 22 November 2017 (UTC)

My guess about the idea of "compatibility characters" is that some once commonly used 8-bit character sets define hex B5 as μ (mu). For example, see ISO/IEC 8859-1 (Latin alphabet No. 1). That presumably makes it more likely that Unicode U+00B5 (micro) would be displayed with a micro sign for the B5 on a system that did not use Unicode. Thanks Kendall-K1, that makes a convincing case that micro was for compatibility but mu is the preferred symbol. Convert moves slowly but it will be switched to use sdot and mu in due course. Thanks all! Johnuniq (talk) 23:51, 22 November 2017 (UTC)

I changed WP:MOSNUM to recommend ⋅ for the dot operator instead of ·. I gather we all agree with that. Johnuniq (talk) 00:28, 23 November 2017 (UTC)

Punctuation of sports scores and vote tallies

I notice that the page says "Sport scores and vote tallies should be given as figures, even if in the zero-to-nine range (a 25–7 victory; ...)". Can I interpret this as expressing a preference for an en-dash as the preferred punctuation mark in such instances – versus, for example "a 25-7 victory" or "a 25:7 victory" or "a 25 : 7 victory"? —BarrelProof (talk) 20:38, 20 November 2017 (UTC)

I also found "a 51–30 win;   a 22–17 majority vote;" at MOS:DASH, which seems to confirm that this is the preferred convention. —BarrelProof (talk) 21:12, 20 November 2017 (UTC)

• Yes; that's prescribed at MOS:DASH. The hyphenated "a 25-7 victory" is news style (in which WP is not written), and thus just a minor style error in encyclopedic/academic writing. "A 25:7 victory" is just wrong, and not a minor error but misleading/confusing, because that markup (with or without spaces) expresses a ratio not a count: "a 25:7 solution of [whatever in whatever]".  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  21:47, 20 November 2017 (UTC)

Missing point in "Numbers as figures or words" section

Something "obvious" that I'm pretty sure we used to have is missing from WP:Manual of Style/Dates and numbers#Numbers as figures or words: use figures before a unit. I.e., we do not want to see "seven kg" but "7 kg" (except "Seven ..." at the beginning of a sentence, if moving "7 kg" to elsewhere in the sentence doesn't work well for some reason). It seems very strange to me that this bit of very basic and central advice, which experienced editors all seem to be following, has somehow gone missing. The chart at WP:Manual of Style/Dates and numbers#General guidelines on unit names and symbols illustrates the figure style pretty consistently, while permitting some examples of spelled-out number words (presumably for sentence-initial cases).

At least one thread on this page (namely, part of the discussion at #Number ranges more generally) indicates some confusion about this, an inference that MoS is somehow expecting/demanding "five to seven kg", due to lack of an exception for measurements, so this probably should be resolved.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  22:53, 20 November 2017 (UTC)

Well, the table says Do not spell out numbers before unit symbols ... but words or figures may be used with unit names. There's a kind of division of labor under which Numbers as figures or words deals mostly with unitless stuff + non-scientific stuff like money and minutes/hours, and Unit names and symbols deals with hardcore units. I'm always torn about whether to duplicate advice like numbers-as-figures-or-words in two places. EEng 02:28, 21 November 2017 (UTC)

I've learned the hard way that it's best to do so, concisely and with a pointer to the more detailed section. People do not read MoS as a document from start to finish, but by shortcuts to sections to answer a specific question. If they don't find it where they think they should find it, they tend to just make an assumption (often incorrect) and stop looking.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  04:15, 21 November 2017 (UTC)

This ought to take care of it.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  04:45, 21 November 2017 (UTC)

Instruction creep to remove from "Numbers as figures or words" section

I propose removing this line from the "Notes and exceptions" subsection of WP:Manual of Style/Dates and numbers#Numbers as figures or words:

Personal ages are typically stated in figures (8-year-old child).

It simply isn't correct, and is patent WP:CREEP. It's entirely normal English, including in an encyclopedic register (perhaps especially in one) to write "eight-year-old child". Many of us do this, and the alleged exception is inconsistent with the general "Numbers as figures or words" rule, for no good reason. We gain nothing – for readers or for editors – in having this line-item. This is simply someone's personal preference, and I don't believe it represents consensus at all.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  22:59, 20 November 2017 (UTC)

Agree. I suspect this probably began in the newspaper style "John Smith, 8, was also injured." EEng 02:11, 21 November 2017 (UTC)

Yeah, and that's a style we wouldn't use here. We don't even do that in infoboxes; we use "Died: June 22, 2008 (aged 71)" style.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  04:46, 21 November 2017 (UTC)

Obviously. EEng 05:20, 21 November 2017 (UTC)

I agree with SMcCandlish and EEng.  – Corinne (talk) 16:04, 22 November 2017 (UTC)

I agree with Corinne. EEng 00:14, 23 November 2017 (UTC)

But not McCandlish?  – Corinne (talk) 00:21, 23 November 2017 (UTC)

I agreed with you in agreeing with me and SM. It's a lovefest of agreement. EEng 02:33, 23 November 2017 (UTC)

• Yes, please. There's another figure–word rule I hate: you can't start a sentence with a figure? But I can't imagine the moaning and wailing that would greet a proposal to remove that. Tony (talk) 02:48, 23 November 2017 (UTC)

I agree also. Except for the bit about numbers in figures not starting a sentence, which generally helps readability (except possibly when it's a list). MapReader (talk) 10:23, 3 December 2017 (UTC)

MOS Seasons

I note that no protocol is listed for the use of the terms fall and autumn - perhaps it's worth pointing out which countries use each term? (AFAIK, fall is used in all Western Hemisphere English-speaking countries, and autumn elsewhere, but I'd prefer some confirmation from Canadian, Jamaican, etc. editors) Grutness...wha? 08:24, 26 November 2017 (UTC)

You might have heard of that little island off the coast of Europe where the language first developed. Fall is considered dated, archaic or dialectal here, though it was in common use from 1550 to the end of the seventeenth century, when autumn took over as the usual term for this season. (Perhaps, by Western Hemisphere, you meant that big lump of newly-discovered land to the west of the Atlantic? ) Dbfirs 08:45, 26 November 2017 (UTC)

Pretty much most of the Commonwealth (ie which includes most English speaking countries outside of the US) says autumn, including my home country of Australia. That's far short of all Western Hemisphere English-speaking countries. But this is already covered by WP:ENGVAR. Just as long as we don't use either as a date because those of us south of the equator have to do mental gymnastics every time somebody says fall 2017 or autumn 2017 and mentally convert that into our spring 2017.  Stepho  talk  09:55, 26 November 2017 (UTC)

The difference in usage is also mentioned at Comparison of American and British_English#Frequency. I agree that it is much more important to avoid the use of a season with a year except in regional articles. Dbfirs 10:25, 26 November 2017 (UTC)

"You might have heard of that little island off the coast of Europe where the language first developed". Yes I have, @Dbfirs: I was born there and have lived in Commonwealth English speaking countries my whole life. It just seemed surprising to me that - after a lengthy discussion on the page about the order of numbered dates (e.g., "1/4/71") and the fact that the numbers are used differently in different parts of the world - no mention was made of this difference in usage, especially since it could cause confusion when the manual claims that "the same names are used for both northern hemisphere and southern hemisphere seasons", which implies that each season has only one name which is used worldwide at the appropriate time of the year. @Stepho-wrs: usage in Australia (and where I live in New Zealand) is irrelevant to its usage in the Western Hemisphere, as neither country is there. A lot of Canadian websites seem to use the term fall in place of autumn, as do some Jamaican and Bahamian sites. Grutness...wha? 00:49, 27 November 2017 (UTC)

I've made a tweak to MOS:SEASON as a result of this question--in case it wasn't obvious before, you should avoid the use of seasons entirely for dating. --Izno (talk) 14:54, 26 November 2017 (UTC)

Thanks. Grutness...wha? 00:50, 27 November 2017 (UTC)

The term Western Hemisphere was new to me and I confused it for Western Civilisation. Anyway, I think we're good.  Stepho  talk  11:30, 27 November 2017 (UTC)

Grutness...wha? 00:45, 28 November 2017 (UTC)

Some are born Grut, some achieve Grutness, and others have Grutness thrust upon them. EEng 02:25, 5 December 2017 (UTC)

MOS:NUMERAL

I was wondering why this says "Integers from zero to nine are spelled out in words", when "ten" is clearly more readable than "10". I would have thought changing "nine" for "ten" in the MOS would make sense? MapReader (talk) 10:05, 3 December 2017 (UTC)

It's just an arbitrary dividing line, so please can we just leave this alone? EEng 14:25, 3 December 2017 (UTC)

Far from arbitrary - there are no more single digit numbers available to use higher than "9". At higher bases we're forced to invoke letters, so "0-9" is a pretty bright (and immovable) line.   ~ Tom.Reding (talk ⋅dgaf)  14:32, 3 December 2017 (UTC)

Well, since this is a MOS discussion, we need to keep courteously correcting one another until one of us drops dead from exhaustion. It really is arbitrary. For example, apparently AP says up to 9/nine should be in words, but Chicago says up to one hundred/100 – not ninety-nine/99 – unless (Chicago says) you feel like using the 9/nine rule. This is may be the purest example of point A2(b) at User:EEng#mossy. EEng 14:50, 3 December 2017 (UTC)

Personally I think that any number that can be expressed with a single word (rather than a two-word combination) is more readable as a word than in figures, except in special circumstances like lists and tables. So this would put the dividing line at "twenty" - which is actually a pretty good place for it. Or, if you argue that words like fourteen are effectively four-teen, you could put the upper limit at "twelve". Either of these would be more logical and sensible than "nine", since the additional figure when you increment to "10" decreases readability, and the reduction of one letter from "nine" to "ten" increases it - so "ten" very clearly has a greater readability premium over "10" than "nine" does over "9". MapReader (talk) 19:34, 3 December 2017 (UTC) MapReader (talk) 19:34, 3 December 2017 (UTC)

Interesting ordinal date contradiction with WP:DATERET

On Erhard Maertens, the first (indeed, only) major contributor, scope creep, chose a date-number format that contravenes the Unacceptable date formats table, namely, Do not use ordinals (1st, 2nd, 3rd, etc.). What is the proper resolution here, and can WP:DATERET's wording be made clearer/stronger as to how these conflicts should be resolved in the future?   ~ Tom.Reding (talk ⋅dgaf)  13:22, 3 December 2017 (UTC)

I reverted to your version i.e. without the ordinals. Nothing needs changing in DATERET, because when it says, "If an article has evolved using predominantly one date format", that obviously means one acceptable format. EEng 14:09, 3 December 2017 (UTC)

The absence of "acceptable" was enough to dissuade scope creep to use unacceptable date formats, and enough ambiguity for me to bring it here. I'd say that's worth a 1-word 'bloat' addition. But I'll defer to consensus, of course.   ~ Tom.Reding (talk ⋅dgaf)  14:50, 3 December 2017 (UTC)

You've proposed [6] adding acceptable to the guideline i.e.

If an article has evolved using predominantly one acceptable date format, this format should be used throughout the article...

But when the negation of a proviso would make the guideline absurd, then the proviso is unnecessary. The negation of acceptable date format is date format (whether acceptable or unacceptable) i.e.

If an article has evolved using predominantly one date format (whether acceptable or unacceptable), this format should be used throughout the article...

which is absurd. While I realize you were momentarily flummoxed by the situation, I really think that was an anomaly. Under your logic we'd have to add acceptable in a bunch of places, and if we miss one then next thing you know someone will be saying that in that one case the omission of acceptable means the unacceptable is, in fact, acceptable, leading to angry talk page discussions, personal attacks, and indefinite blocks. EEng 15:48, 3 December 2017 (UTC)

MOS:DATERET appears to be a catch-all section, so I don't think it's absurd for someone to infer (whether acceptable or unacceptable), and indeed someone did. Also, when you said that obviously means one acceptable format, it's more accurate to replace "means" with "implies". For a catch-all section, explicitness is better than implicitness, so "acceptable" wouldn't have to be placed everywhere, I think. I would not want the scenario as you describe it to play out either, though.

I should caveat all this by saying I don't frequent MOS:Talk at all, and only now added it to my watchlist, so I'm unsure of how frequent a problem this is. While it might appear to me to be a non-trivial problem, it may, in practice, be an infrequent occurrence. If it's indeed rare, then it's probably not worth further discussion, so I'd like to know more about its frequency before conceding.   ~ Tom.Reding (talk ⋅dgaf)  16:35, 3 December 2017 (UTC)

Take it from us – unwatch while there's still time! – EEng

To my recollection it's never come up that someone's suggested that DATERET implies that an "unacceptable" date format should be retained. EEng 17:20, 3 December 2017 (UTC)

@EEng: don't you need to get back to the important things? Like updating the counter? --Izno (talk) 17:40, 3 December 2017 (UTC)

I thought about that, but I'm not really sure this is a dispute about date formats. It's more a dispute about a dispute about date formats. EEng 17:43, 3 December 2017 (UTC)

See post just below. Now I'm resetting the counter. EEng 17:54, 3 December 2017 (UTC)

There's a another circumstance where we have language like this: MOS:STYLERET. The "acceptable" wording came from ArbCom (in three cases, cited in WP:MOS's footnotes). Since DATERET is just STYLERET narrowed to dates, the "acceptable" proviso applies to it. It's true that "acceptable" seems redundant in the context of an entire MoS page, but virtually no one reads these pages top-to-bottom, just a section at a time, arrived at by shortcut. So, it's probably better to err on the side of clarity than brevity in this case. If someone just reads that a date [or other] style established in an article should not be changed without good reason, and they do not continue on to what styles are even acceptable at all, they'll naturally conclude that "July the first in the year of our Lord nineteen hundred and thirty-six" is a style they can revert to defend in an article of which they're the primary post-stub author. Not a desirable outcome. Even if they would not carry the day in such an argument, the time wasted on the argument is undesirable, as are poorly-watched articles with divergent style in them in the interim.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  01:27, 4 December 2017 (UTC)

Well, my prediction is that if acceptable is in there, then this hypothetical person will just say to himself, "Well, 'July the first in the year of our Lord nineteen hundred and thirty-six' is acceptable to me" and we're back where we started. So I guess we'll have to say, "acceptable under WP:DATEFORMAT together with all the provisos here and there about which and wherefore the various formats are designated to be used, and in what combinations". EEng 02:46, 4 December 2017 (UTC)

I've only seen this happen one time, and the WP:GAMING attempt did not WP:WIN. It was, however, why STYLERET was reworded: 'On some questions of style, the MoS provides more than one acceptable answer; on other questions it gives no guidance. The Arbitration Committee has expressed the principle that "When either of two styles are acceptable ...' – pre-loading this with what "acceptable" means in the context. We can do something similar but much more concise here, e.g. If an article has evolved using predominantly one [[#Dates and years|acceptable date format]] .... All it takes is a word and a wikilink in this case.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  12:30, 4 December 2017 (UTC)

I relent. EEng 12:34, 4 December 2017 (UTC)

Hi Tom.Reding, Hi EEng, I think there needs to be an awareness that the date formats used in most European countries and to a certain extent, those countries that European countries, conquered, e.g. Britain conquering India, means that India uses the date format 10th May 2019, and other countries. I propose an extension is made to the format to enable this format. For most British, reading a date like 10th May 2019 is natural. Not 10 may 2019. That is read here as ten May 2019, which doesn't sound right. It should be 10th as in tenth of May 2019. As regards the WP:DATERET I have been using it fend folk off who have changed to American Date formats. I don't know what your thinking it, but the American format dates subverts British English articles, and reduces them. It is not fair really. I propose an extension or whatever the procedure to extend the MOS in this specific instance. Thanks.scope_creep (talk) 17:45, 3 December 2017 (UTC)

SMcCandlish. can you handle this one? EEng 18:04, 3 December 2017 (UTC)

Undercover Wikipedian spot checking date formats in various Commonwealth newspapers

This is perennial rehash. We go over this several times per year and consensus never changes because the facts never change. The style "10th May 2019" isn't the one used in Commonwealth English, it's one of many styles ("10 May 2019", "10th of May 2019", "2019-05-10", even "May 10[th], 2019" in some publications). And "10th May 2019" style isn't limited to ComEng; it's also sometimes used by (mostly older) North Americans. The dominant style in ComEng publishing (we don't care how people write letters) is "10 May 2019". This is easy but tedious to verify by going down lists of notable, extant British, New Zealand, Hong Kong, Irish, Australian, etc., news publishers and seeing how they do dates. I just spot-checked about 30 of them, and all use "10 May 2019" format consistently, except: one used the weird "10 May, 2019" [almost certainly a software problem]; one used a mixture of those two; four used a mixture of "10 May 2019" under headlines and "10th May 2019" in running text; four used "May 10, 2019" (three were from the same publishing company); and two used "10 May 2019" in articles but "2019-05-10" style under headlines. So, a) "10 May 2019" clearly dominates, and b) it is conclusively proved that "10th May 2019" is not normative in Commonwealth English, just attested. WP doesn't need to use it because the th is redundant, like adding "of"; "10 May 2019" is perfectly understood by all English-language readers, including Americans (many of whom prefer it, especially in technical and academic writing).  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  01:27, 4 December 2017 (UTC)

Maybe this page should have an FAQ. EEng 02:46, 4 December 2017 (UTC)

Not a bad idea. The main MoS page does (and we should add a couple of things to it, but not many).  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  12:30, 4 December 2017 (UTC)

And an edit notice that you have to digitally sign by typing in the edit notice's content. EEng 12:34, 4 December 2017 (UTC)

Why don't you explain the rationale behind proscribing ordinals for dates in the section of the MoS that proscribes them? FactotEm (talk) 20:24, 4 December 2017 (UTC)

'Coz there ain't no rationale. It's just a choice, as good or as bad as any other consistently applied choice. (The alternative - not to make a choice - would be chaos). Leastways that's how I see it. Dondervogel 2 (talk) 20:59, 4 December 2017 (UTC)

Exactly. EEng 21:13, 4 December 2017 (UTC)

But there is a rationale, based on established style guides in the world of publishing and common usage, or have I misunderstood something? FactotEm (talk) 15:45, 5 December 2017 (UTC)

There's no one "established style guides in the world of publishing and common usage". There actually are publications using ordinal dates e.g. check out this Edinburgh bus schedule [7]. But each publication needs a house style or everything looks like a mess, and no-ordinals is ours. It's actually amazing we're able to operate effectively with the choice of MDY and DMY – the last thing we need is to turn that difficult marriage into a love triangle. EEng 15:54, 5 December 2017 (UTC)

An Edinburgh bus love triangle sounds positively 1st class (although those double-deckers can be a bit bumpy). Martinevans123 (talk) 16:48, 5 December 2017 (UTC)

To be clear, other than finding it a little odd at first, I do not have, nor have I ever had a problem with this standard. Apparently a lot of people do, though. The MoS is effectively banning by diktat a convention that many people routinely use in spoken English. The replies to my original post indicate the decision to adopt this standard was purely arbitrary. However, SMcCandlish's post above, and points made in previous discussions, seem to indicate that the preponderance of the cardinal format in modern usage, reflected in many reputable style guides, influenced this decision. This seems a reasonable justification to me. I just don't understand why it isn't stated somewhere more prominent than the talk pages. FactotEm (talk) 11:28, 6 December 2017 (UTC)

When I was at school I was taught to write "10 May 1969" and read this out loud as "the tenth of May, nineteen sixty-nine". Sure, written and spoken forms are different, but Wikipedia is (for the most part) a written encyclopaedia, not a spoken one. (Question: Is there an equivalent of MOSNUM for "spoken" articles? The point being raised would be more relevant there, especially as there at least two different ways of reading (the year) 2019) Dondervogel 2 (talk) 16:03, 6 December 2017 (UTC)

We would need new technological support for a "voice MoS" to be feasible or even meaningful, e.g. a way to force a pronunciation like "the tenth of May" when "10 May" is followed by a year and read aloud. Assuming such a thing were a good idea; there are people who do not read that construction that way at all, but say "ten May", exactly as written.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  08:09, 7 December 2017 (UTC)

The source rationale for the rule isn't stated in MoS because (see a zillion prior "source the MoS" discussions) MoS is not an article, but a guideline and we do not cite sources in guidelines, we provided the advice that consensus has concluded to provide, which is based on internal consensus discussion, often informed by sources, but not entirely derived from them; consensus is also derived from collective editorial common sense and experience of what works well here and what does not. Further, if we sourced this then people would want to source everything, and then MoS would be enormously longer. MoS is not part of the encyclopedia content, is not a statement of facts about the world, and is not providing advice to anyone otehr that WP editors for how to write WP content. In the very small number of places where MoS has an external source, it's for when we're deferring to an off-site standards document like HTML5 or something from IETF, WCAG, ISO, etc.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  08:09, 7 December 2017 (UTC)

Dear Wikipedians!!, You WP:MOS folk, I think, I suspect that like most folk, you are resistant to change. But we live in a changing world. SMcCandlish most newspapers, designed in the UK, and other English speaking nations, are designed to be read by transatlantic audiencies, and as such use the simplest of date format as possible, the American date format. SMcCandlish I see you have never written a very large, complex article, perhaps involving 100's, or 1000's of hours of research, then to write it, in British English, and realize you can't use British date formats, because at some point, somebody is going run a bot against the article and remove the ordinal dates. It is disheartining and not really fair. Your rationale is therefore incorrect. I will round up some evidence. Out of the 2.2million books published in the UK every year, some of them must use it. I will find the research. What is the formal method to raise a change in the WP:MOS. scope_creep (talk) 00:54, 5 December 2017 (UTC)

• Let me save you, and everyone else, some time. This has been chewed over many, many times, and something seismic will have to happen – like Jesus, Mohammed, and Yahweh appearing on TV as a Three Tenors tribute group and ordaining it in song – before it's going to change. If you still want to waste your time on this, start by reviewing the innumerable prior discussions listed here [8], and if you have something new to say, then get back to us. Finding examples of ordinals in the wild has zero value, because we already know that. Every publication has its house style, and that doesn't mean allowing everything that someone somewhere has exemplified.
• While we're here, let me give you another tip. The surest and fastest way to make a fool of youself on Wikipedia, and particularly here at MOS, is to present yourself as oh-so-more experienced than your fellow editors. The proportion of people here who do serious literary, governmental, commercial, or academic research and writing is very high, and to be quite blunt your post above is barely literate. You're not fooling anyone.

EEng 02:18, 5 December 2017 (UTC)

• User:EEng That is a bit harsh. Your purpose here seems to be push back, without supplying a cogent argument. What are you talking about? You seem to have the attitude that the MOS is inviolate, cast in stone, like the commandments, unable to change. I was making a point, that you have choosen to ignore. I will look through your list, find the research, analyze it, and if the research bears it up, the MOS will be changed. scope_creep (talk) 08:55, 5 December 2017 (UTC)

Given that you wrote I see you have never written a very large, complex article, perhaps involving 100's, or 1000's of hours of research, you should consider yourself to have got off easy. And there's a reason there's a box near the top of this page: It has been X days since the outbreak of the latest dispute over date formats. EEng 09:21, 5 December 2017 (UTC)

What about "10,000's of hours of research"? Martinevans123 (talk) 16:51, 5 December 2017 (UTC)

Too bad Carl Sagan's not with us any longer, or we could aspire to billions and billions of hours of research.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  08:11, 7 December 2017 (UTC)