Wiktionary talk:Spelling Variants in Entry Names - Draft Policy/BP April 2006

This stuff brought across from Beer Parlour April 2006.--Richardb 09:26, 15 April 2006 (UTC)Reply

First quarter 2006 US vs. UK flamewar

Latest comment: 18 years ago77 comments13 people in discussion

initial discussion

As is the regularly recurring cycle on en.wiktionary, questions have been cropping up again recently regarding American vs. Commonwealth spellings of words.

Since my comments in the past may have been unclear, I'd like to say that as an American, I think any respectable dictionary should list the UK spellings only as errors, perhaps used in a nonce fashion for comedic or Shakesperian effect. But they would be better off deleted.

In the interest of NPOV however, I have made numerous enormous concessions in my behavior regarding the incorrect UK spellings. I do not intend to change those compromises; that is, I'm not about to start deleting the UK spellings, even though I know in my heart it is wrong to include them here, especially indicated as valid spellings.

The purpose of this thread is to revive the older discussions so that those who are new, or have otherwise missed salient points of the conversation can adapt to the current practices.

My understanding of the current prectices (that I disagree with) are:

Separate entries must be created for UK/CW spellings.
Separate entries must be created for US spellings.
Each must indicate the existance of the other in the ===Alternate spellings=== heading, before the definitions.
Translations should not be duplicated, but rather limited to the "older" UK/CW spellings.

It is my hope that this quarter's discussion of the topic will not immediately revert to ad-hominem attacks, nor other such flamboyant nonsense.

Confer:

--Connel MacKenzie T C 17:09, 28 March 2006 (UTC)Reply

I firmly support discussion of this important matter. However, it is unclear to me how you intend to bring this up as earnest dialogue, pretending to "hope" it will not result into flamboyant nonsense, while starting off with exactly such palaver about "erroneous" UK spellings. Or is this your humour again? — Vildricianus 18:43, 28 March 2006 (UTC)Reply

In all seriousness, what is the current policy? Is Wiktionary:Policy Think Tank on American or British Spelling an accurate synopsis of where the issue stands? —Scs 20:37, 28 March 2006 (UTC)Reply

There is no firm guidline nor policy. That makes the issue resurface regularly. --Connel MacKenzie T C 07:08, 29 March 2006 (UTC)Reply

Palaver? Not at all. Commonwealth spellings are erroneous here in the US. Why would you think they are not? Sorry for generally trying to lighten the mood (especially on IRC) but I am quite serious here. --Connel MacKenzie T C 07:08, 29 March 2006 (UTC)Reply
- Palaver I say. — Vildricianus 09:48, 29 March 2006 (UTC)Reply
- You say "here in the US", but "here" for me, and indeed for many others, is not the US. I was not aware that en.wiktionary had suddenly been moved to en-us.wiktionary. Oh, wait…it hasn't. The lack of smileys in your text makes it difficult for me to tell whether you are being humorous or inflammatory. Please elucidate. HTH HAND —Phil | Talk 09:21, 29 March 2006 (UTC)Reply

translations

I think the translations not the definitions is big problem since as people have pointed out they might theoretically mean slightly different things depending on spelling. Since we strictly speaking should have at least three independent quotes for each sense of each spelling we really need entries for all variants. If nothing else to show the quotes somewhere. A minor issue is what to do with misspellings that are so common that it is even possible meet the criteria for inclusion.

Now back to translations, while of course the different spellings might possibly mean slightly different things such nuances are unlikely to effect translations except in rare cases. So the big question is where the translations should be. For example colorize, colorise, colourize and colourise all translate to färglägga in Swedish and I can't imagine any language that translates them differently depending on the spelling. Now I suppose you could argue that 2 of them are misspellings but even so they seem to be widely used none the less. Still, it leaves two of them. The point is that one of the spellings really must be the main entry. Note main entry, not right entry. --Patrik Stridvall 14:21, 30 March 2006 (UTC)Reply

I don't see any misspelling there. — Vildricianus 15:57, 30 March 2006 (UTC)Reply
- What region of the English speaking world would construct colorise or colourize? --Connel MacKenzie T C 20:57, 3 April 2006 (UTC)Reply

Well, it's a theoretical problem, as you say, Patrik, so it's unlikely there will be many cases like this. We'll just deal with them as they arise.

Here's an example that already exists. "Program(me)" is spelled "programme" in UK English and "program" in US English. However, the only spelling in the computing sense is "program" in all varieties of English. So this requires treatment at program that will be absent from programme. We just need to ensure that we provide that treatment. I don't think it is such a big problem. — Paul G 14:58, 30 March 2006 (UTC)Reply

Actually, I don't agree at all with the first-gets-it-all principle. Yes, that would be the easiest for us, contributors, but we're making this for the user, aren't we? I don't think anyone looking up "colorize" should be referred to "colourize" or vica versa for any information whatsoever, not even translations. Beside the fact that it's not fair whatever way you turn it, it will make the user wonder. "Do we prefer UK spellings?" - "No we don't, but the UK version got here first." Pardon my language, but this sounds like rubbish. Each entry needs the relevant information there where it belongs, not in a variant which happened to be created first. Yes, there will be duplication. But until the software can handle this properly, we'll need to balance out these entries in order to respect the N in NPOV. For the user, it doesn't help having this first-gets-it-all rule. — Vildricianus 15:57, 30 March 2006 (UTC)Reply

more discussion

In a comment on the proposal below, Paul G wrote:

...we are aiming at doing more that that here. We want to eliminate the unintentional bias that is introduced by having a page for (say) "aeroplane" that gives full treatment of the word, and another for "airplane" that just says "See aeroplane" as if "airplane" is a mere variant rather than the US spelling of the word.

I'd like to consider the possibility of not worrying about this, after all. Educated people understand that color and colour are two variant spellings of essentially the same word. Educated people understand that that no one spelling is universally "right" or "wrong" or "good" or "bad"; they're just different, that's all. Educated people understand that it's useful to have one page (not two) on which all the central, relevant information about a unique word is to be found. Educated people understand that (for the moment, at least) Mediawiki requires pages to have exactly one name. So, I assert, educated people do not see any bias when color redirects to colour, or vice versa; all they see is the unavoidable logistical repercussion of the simple facts that spelling variants exist, and that Mediawiki is the way it is.

(Now, it's true, uneducated people might perceive the "unintentional bias". But -- and I'm not trying to be elitist or anything; this is a plain fact -- uneducated people don't use dictionaries, so let's not worry about them so much.)

Some day, perhaps, Mediawiki will have a way for one article to have two (or more) different names, with absolutely no way of telling which is the "main" or "preferred" name and which are the variants. (~~Wiktionary~~ Wikipedia could of course use such a mechanism, too.) That's really the only solution to the "bias" problem. Until then, I suggest we agree that the superficial appearance of "bias" is one that can best be solved by user education.

Up above, Connel MacKenzie wrote:

The tremendous British prevalence here on en.wiktionary has taken a toll, leaving en.wiktionary looking less like a dictionary and more like a joke.

I'd like to challenge this assertion, too, because I just don't see it that way at all. En.wiktionary is a dictionary that is rolling up its sleeves and getting down to the business of defining words; it is wisely and maturely not getting bogged down (well, present company excluded :-) ) in internecine, relatively unimportant, utterly unwinnable arguments about The One True Official English Spelling. Guess what? The English language does not have "one true official spelling", if for no other reason that there is no one, single body to officiate it.

I also don't see any "Comonwealth cabal", nor do I understand what Connel is referring to when he talks about "arrogantly, Britishly dismiss the notion". Sure, there are a lot of Commonwealth spellings on Wiktionary. So what? The English invented the language, after all (or, at least, their cabal conspired to get it named after them :-) ), so I really don't see the problem if their spellings get used, even in a reference work that gets read in America. There are plenty of American spellings on Wiktionary, too. Again, so what? (Emerson's quote on consistency springs to mind.)

—Scs 16:13, 2 April 2006 (UTC)Reply

Well-reasoned of you, and I'd perhaps agree, if it hadn't been for such an obvious section title. Your solution will evoke further disagreement in the future, if it could allay the current one at all. — Vildricianus 12:45, 3 April 2006 (UTC)Reply

Eh? I didn't think I'd proposed a "solution", nor did I use a section title. Was it someone else you meant to reply to? —Scs 17:18, 3 April 2006 (UTC)Reply

I fail to see how User:Scs' ad hominem attack is well-reasoned. I never said that en-us is "The One True Official English" and my comments very clearly, repeatedly expressed that I had no interest in trying to make such an assertion. As for his idiotic statement that there is no cabal, one needs only to look at the statistics. Again, as the only American contributor in the top 10 contributors on en.wiktionary, (starting out at en.wiktionary a year or more later than several of the others) I have perhaps made a dent. But I've been stymied several times with specious arguments such as "a UK spelling already exists" == "The One True Official English is en-uk." Or arguments such as "terms shouldn't be entered as anything other than redirects."

I still feel it is irresponsible of Wiktionary to list UK spellings such as colour, flavour or parlour without identifying them as spelling errors in American English. It is not as if we don't know they are errors. But the commonwealth cabal in place refuses to let them correctly be listed as such. I find that odd, as I'd assume those same parties would want the American spellings likewise identified as incorrect in commonwealth English.

Until we have some kind of official policy indicating that both need to be entered, and both need to be identified properly, we'll continue to have periodic flamewars on the topic. Any other approach is certain to offend one side of the pond or the other (as has been demonstrated several times in the past now.) --Connel MacKenzie T C 17:36, 3 April 2006 (UTC)Reply

Whoah, Connel, calm down! I'm sorry you percieved an ad hominem attack, but truly, none was intended! I didn't say you said "One True Official English", and if you want to tar that statement of mine with anything, call it "hyperbolic" or a strawman, please. But at any rate, it's an objective fact: there is no one true official English, or else we wouldn't be having this debate.

Since you brought it up again, though, I'd like to ask why you're apparently so worried about asserting that words like colour are "spelling errors in American English". Would it not suffice to say that "color" is the accepted spelling in American English, and "colour" the accepted spelling in Commonwealth English, and leave it at that (i.e. and not brand either of them as "errors")? —Scs 18:40, 3 April 2006 (UTC)Reply

Perhaps I misread your statement. I did not consider the notion that you were suggesting "user education" as this is a wiki and therefore such a thing is impossible. I'll try to remember you are not making a personal comment. :-)

I was probably wrong to bring up the erroneous tagging again. As I said before, if I enter any British/commonwealth (is Commonwealth a proper noun?) spellings in, say, Microsoft Word, I'll be prompted to correct it. Those with the default "auto-correct" feature turned on may not even see the replacement. Now, since we aim to be correct in what we say about words, it does not make sense to lead someone on incorrectly. A visitor that arrives here and looks up the single word colour would not have any indication that what they entered is not a word in American English. Simply having a note somewhere that says it is British does not convey quite enough information to be useful. --Connel MacKenzie T C 20:49, 3 April 2006 (UTC)Reply

In terms of the hypothetical visitor, see my new proposal below, which ends up addressing this (at least, if we can agree on having a single page for colo(u)r rather than two).

As far as Microsoft Word is concerned, and please don't take this as any kind of personal attack, I really, really don't care what it does. This is not because I'm a Microsoft basher, but simply because (as I've said before) there is no one authority on "correct" English spelling, and even if there were, it certainly would not be a software company in Redmond.

Here's a thought experiment. You're editing a manuscript (using your U.S. copy of Microsoft Word), and you happen to be including an excerpt from the Guardian Unlimited:

God, the archangel says, is also disturbed by Mr Blair's remark that while religious beliefs might colour his politics, "it's best not to take it too far".

(There's nothing special about this quote; it was just one of the first hits I got in a Google search for "Blair colour".) Now, when you type or paste in this quote, Microsoft Word is likely to give you the dread squiggly red underline for "colour". Is this a problem? Me, I don't think so. There's nothing intrinsically wrong with stringing together the six letters c o l o u r, even within the shores of the revolutionary colonies. Whatever it is that the squiggly red underline means, it is not, "Thou shalt not use this spelling; it is wrong; correct or remove it at once".

—Scs 22:43, 3 April 2006 (UTC)Reply

Well, I'm not running MS Word right now, but on Linux, I use the command "spell." When I type ctrl-D after pasting in your text, it (correctly) informs me that colour is not a word. --Connel MacKenzie T C 01:15, 4 April 2006 (UTC)Reply

If you think that the Linux "spell" command carries any more (or less) weight here than Microsoft word, you haven't understood my point at all. At any rate, there is no dispute that "colour" is not the preferred American spelling. But that doesn't mean that it's "wrong", and it certainly doesn't mean that it's "not a word" (just as, of course, "color" is not wrong, either). —Scs 16:12, 4 April 2006 (UTC)Reply

This is probably a naive comment of mine, but I fail to see how on earth this can be such a big deal. En.wiktionary includes both American and British English, as well as Australian, New-Zealand, South-African, Irish, Indian, Canadian and any other regional variant of English whatsoever; therefore, all words and spellings of either variant should be included, treated and valued with the same esteem, regardless of any personal affiliation or custom, in other words, with a neutral view. The problem arising out of identical translations for two different spellings must not influence our stance on this general principle, as it is clearly secondary to it. Even if we were to have our main or sole purpose to be a translating dictionary, we should respect English in all its variants, no matter what the result is for our layout, format, translations etc. I wonder whom of the main contributors refuses to honour/honor this principle. — Vildricianus 18:07, 3 April 2006 (UTC)Reply

Rather than naming more names, I'd rather work towards solidifying a policy acceptable to all. Then stating it explicitly so that it no longer resurfaces (as it does now on a regular basis.) This is perhaps the poster-child of why we should have something like Wiktionary:Votes. --Connel MacKenzie T C 20:49, 3 April 2006 (UTC)Reply

Proposal

A header for editability. — Vildricianus

The "first-gets-it-all" principle is not ideal, but I'm not sure we have anything better right now. Avoiding duplication is important because full pages for both/all spellings quickly get out of synch (with information being added or corrected on one page only).

I proposed the following solution before, but I don't remember what became of it. I'll use "color"/"colour" as an example.

Have a single page called "color/colour" or "color, colour", or something similar; the name of the page lists the variations in alphabetical order, so there can be no claims of language bias (although, for most UK/US variations, this favours the US spelling).
Move the entire contents of color and colour into this page and format it appropriately so that meanings for a particular spelling are distinguished.
Turn color and colour into redirects to this page (but see the next bullet point).
As "color" is also a word in Latin and Spanish, color would actually be more than a redirect: the English entry would say "See color/colour", and the Latin and Spanish entries would remain on that page.

I think this is a simple solution that would clear up all the issues around spelling variations once and for all. — Paul G 16:25, 30 March 2006 (UTC)Reply

There is still the possibility to include the page color, colour (or whatever) as were it a template: {{:color, colour}}. I don't know if it is feasible, but at least theoretically there is a possibility that one could put the common info on that page and include it in both spellings... \Mike 16:43, 30 March 2006 (UTC)Reply

Templates would make it too difficult to change content. I read elsewhere about Paul's idea, and couldn't understand why this didn't make a resolution at the time. Sounds reasonable. — Vildricianus 16:55, 30 March 2006 (UTC)Reply

Also, there should be a better way to name these things. Commas imply some phrase; slashes are part of the url syntax. Perhaps color;colour? Note that there's no space. — Vildricianus 09:18, 31 March 2006 (UTC)Reply

Agreed - this was one of the things I was unhappy about with my proposal. The phrase "there, there" could be wrongly interpreted as a page giving two identical spellings of "there", for example. — Paul G 10:50, 31 March 2006 (UTC)Reply

People here seem to believe that words which have different spellings in American and British English often have slightly different meanings because they are spelt differently. This is clearly not the case. Nuances in meaning exist due to cultural and geographical factors. And this is the case for all English words, not only those that are spelt differently. Simply tagging words with the appropriate templates (Template:US, Template:UK, etc.), as we have always done, does the job. It is obvious that a meaning tagged as AmE will be spelt the American way, one tagged as BrE the British way, and one tagged as both can be spelt either way (according to where one comes from, or wants to have one's work published). Ncik 18:12, 30 March 2006 (UTC)Reply

Also, if what we're worried about is translations, or specifically, the problem that might arise if the Commonwealth "colour" ought to have a different translation into, say, Klingon than does the U.S "color", it's already a much bigger problem that there might not be a single perfect translation at all, i.e. neither for colour nor color. It's often the case that a single word in one language will be translated to different words in some other language, depending on the sense in which the word is used. Any translation scheme must accomodate (or at least acknowledge) this possibility, and having done so, if it happens that some of the distinctions between translated-from senses end up being correlated with distinctions between translated-from spellings, then not only is this no problem, but it makes it even easier to document (for a particular translated-to word) which sense is being translated from.

(I suppose there's also the question of which spelling to use when translating to English. Does de.wiktionary say that Farbe is translated to color or colour? [Answer: de.wiktionary.org has Farbe → colour, but here in en.wiktionary.org we've got Farbe → color.])

—Scs 20:37, 2 April 2006 (UTC)Reply

This is true; however, we are aiming at doing more that that here. We want to eliminate the unintentional bias that is introduced by having a page for (say) "aeroplane" that gives full treatment of the word, and another for "airplane" that just says "See aeroplane" as if "airplane" is a mere variant rather than the US spelling of the word. We also want to eliminate the duplication of effort and inevitable inconsistencies that would arise if we had two (or more) pages, one for each spelling. The page "aeroplane, airplane" (or separated by whatever punctuation mark is chosen) would indicate where "aeroplane" is the correct spelling (UK, where else?) and where "airplane" is the correct spelling (US, where else?), and then give all of the definitions as they currently stand. In a few cases (such as "program"), there are senses that have only one spelling, and in this case, the senses themselves would be marked accordingly; so (excuse my concise definitions):

program, programme

A series of planned events
A leaflet outlining such events
A TV show
(always spelled program in all varieties of English) A computer program

or something like that. — Paul G 10:50, 31 March 2006 (UTC)Reply

This bothers me. What about historical spellings of words? More to the point, what about the fact that many US spellings are in fact archaic UK spellings? We can't list all alternative spellings on the page title. In my opinion the best solution is to have duplicated information under all relevant headings – the problem of pages becoming out of synch with each other seems to me the lesser evil. Widsith 15:03, 1 April 2006 (UTC)Reply

I don't like Paul's proposal either (for the same reason). But isn't our page layout flexible enough to incorporate all information on historical spellings on one page? This could be done under the "Etymology" header (alternatively a "Word history" header if we ever decide to create one), and by means of various usage notes and an expanded "Alternative spellings" section. Ncik 15:19, 1 April 2006 (UTC)Reply

Why don't we just wait for the Indians to get more net active. Then clearly the commonwealth spellings will overwhelm the American spelling in numbers of users ! :-) --Richardb 00:34, 2 April 2006 (UTC)Reply

I think the historical spellings, etc, could all be catered for in a section (maybe called "Spelling" towards the top of the page):

program, programme (as the page title)

==Spelling==

Program is the usual spelling in the US (and wherever else). The spelling programme is archaic (or historical, or whatever)
Programme is the correct spelling in the UK (and wherever else). However, in the computing sense, program is the correct UK spelling.

We could put any other information we like in there about who uses which spelling, which spellings are historical, archaic, etc, and then leave the rest of the article to give the meanings, etc.

Widsith, Ncik, do you have an alternative proposal that we could consider? — Paul G 10:05, 3 April 2006 (UTC)Reply

Well I'm not really convinced of the need for any new proposal....as I said above I don't think it's too infeasible to keep a separate page for every current spelling, i.e. color and colour would both exist with very similar information on them (though each would have different citations reflecting the different spellings). Historical forms are a bit different, personally I'm coming round to the Word History idea but we don't need to worry about that till it's become an issue. Widsith 11:49, 3 April 2006 (UTC)Reply

further discussion

Personally, after reconsidering this entire issue, I think it's perhaps the best solution to deal with it as we're doing now. That is, two pages for colour and color, and trying to keep them as much in sync as possible. I agree that this is not very constructive, and even though I partly like Paul's proposal, I'm not sure whether this would help our project a lot. Perhaps we first need to experiment a bit with a low-profile entry (not color/colour). — Vildricianus 12:45, 3 April 2006 (UTC)Reply

Please also note that Paul's suggestion was tried (by user:Dmh?) with color, colour & color coloured and deleted by Ec as nonsense over a year ago. --Connel MacKenzie T C 17:10, 3 April 2006 (UTC)Reply

Hm, why was that? — Paul G 09:44, 4 April 2006 (UTC)Reply

As I recall, it was deleted as an abandoned/failed experiment. --Connel MacKenzie T C 16:51, 4 April 2006 (UTC)Reply

another proposal

Here's what I would propose, for now at least. This isn't just a policy statement; it also touches on goals and explanations. But what it says isn't really very different from the status quo, as I understand it. (In other words, I'm not proposing any new policy here, mostly just restating the current one.)

That is a wildy false statement. What you propose is even worse than Paul's "partial redirect by section" proposal. What follows on here (based on your incorrect assumption that redirects are acceptable by anyone) is very well formatted, but totally unacceptable. Your choice of torch is interesting; do you realize that in America, a torch is only a wooden stick with oily rags on one end? If you tried to say "torch" in America to refer to a flashlight, you would not be understood; you'd probably be suspected of being mentally retarded. --Connel MacKenzie T C 22:50, 3 April 2006 (UTC)Reply

I've clarified the "torch" example. (It wasn't hypothetical; it's pasted directly from torch.)

As far as the assumption that redirects are acceptable: I'm prepared to be proved wrong, but I had gotten the impression that plenty of people do find them acceptable. (Not perfect, but a decent compromise under the given constraints.) —Scs 22:56, 3 April 2006 (UTC)Reply

Where did you get that assumption? The only person that suggests it these days is Paul; each time the topic comes up he re-suggests it innocently (sometimes suggesting that is has consensus) as if he'd never suggested it before, nor ever heard the arguments against it. It gets tiresome. Consensus has never been to use those redirects here on en.wiktionary. Some experiments with them have been made, but AFAIK, each has been undone. --Connel MacKenzie T C 01:05, 4 April 2006 (UTC)Reply

Wow. Assuming I'm the Paul referred to, I find it extraordinary at how I am being misrepresented here ("each time"; "innoncently", "suggesting that [it] has consensus", "as if he'd never suggested it before nor ever heard the arguments against it". Is that really how you see me, Connel, or are you playing it up a bit here? — Paul G 09:51, 4 April 2006 (UTC)Reply

Paul, on this topic, it seems to me that your normal rational self takes a vacation. --Connel MacKenzie T C 16:54, 4 April 2006 (UTC)Reply

Paul doesn't need me to defend him, but he's so polite he might not say anything at all, so let me point out that denying his rationality here is uncalled for, and could be seen as offensive. Paul's been utterly rational on this topic, it's just that he's arguing from different premises and opinions than you are. You would do well to remember that many of your opinions are just that, also.

People have been bending over backwards to assume rationality and good faith on your part in spite of the wildly provocative way (yes, it really did look that way) you opened this thread. You might think about returning the favor. —Scs 01:43, 5 April 2006 (UTC)Reply

I do not mean to offend; I was stating my opinion as a matter of fact (that is, from my perspective, his actions and statements on this topic do not coincide with his normal behavior, polite bearing and refreshingly clever intuition.)

I opened this thread in a calm manner, compared to the two inquiries (on my and other's talk pages) that immediately preceded re-opening the topic. I have been clear from the outset that I am not ignoring the CW perspective, merely stating the inverse of it: an American perspective. In doing do, I have gotten a far too healthy dose of negative responses. I do wonder why. Perhaps it is too much to swallow when coming from the (flawed) perspective that UK English is The One True English. The only thing even mildly incitful I did, was putting "flame war" in the topic heading (but even that has proven to be partly accurate.)

It is curious that the prevailing mindset here is still not one of openness to allowing all words in all languages (as is the Wiki Way.) Isn't it clear by now, that whenever an attested alternate spelling exists, there must be two separate entries to be accurate?

--Connel MacKenzie T C 03:24, 5 April 2006 (UTC)Reply

I guess there are different kinds of "openness". From where I sit, statements like "any respectable dictionary should list the UK spellings only as errors" and "colour is not a word" don't look very open. In what way do the arguments I've been making look not open?

With respect to having separate entries, no, it's not at all clear that "there must be two separate entries to be accurate". For example, we have one page on bald, even through it's also a completely different and unrelated word in German. (And of course this is not an isolated example; it's just the first one I thought of.)

Finally, with respect to your suggestion of a "perspective that UK English is The One True English": that would indeed be flawed, but, again, I just don't see it. I certainly don't see it in this thread, and in particular not in any attempt to (say) unify the color and colour pages. —Scs 17:00, 5 April 2006 (UTC)Reply

I have made one little change in the described use of the "Alternative spellings" section. I've also alluded to the possibilities of combined titles (e.g. "colour/color"), and of separate entries with guaranteed-identical, transcluded content, but I don't get the impression there's consensus around those yet so I'm leaving them as ideas still under discussion.

Suggestions, criticisms, rewordings welcome. (In particular, there's probably a better taxonomic nomenclature than "variants".)

En.wiktionary is a dictionary of the English language, embracing several distinct variants such as British (or "Commonwealth") English, American English, Australian / New Zealand English, Indian English, etc.

In some cases, of course, these variants involve different words for the same idea, different meanings for the same word, different spellings for the same word, and words unique to a particular variant. (If there were no such differences, they wouldn't be variants!) Wiktionary entries must therefore be careful when defining these mixed-use words to indicate how the words are used in each variant.

When two variants have different words for the same idea, those entries should be tagged with their variant:

torch

1. a stick with a flame on one end used as a light source

2. (British, Aust) a portable source of electric light

Synonyms

* flashlight (US)

----------

flashlight

1. (US) An electrical hand-held lightsource.

Synonym

* torch (UK, Aust)

Cross-references between the other-English "translations" can be in the form of synonym lists (as in the examples above), or directly in the definition (e.g. "flashlight: (US) An electrical hand-held lightsource (a British torch).").

When two variants have different meanings for the same word, again, each sense in that word's definition should be appropriately tagged:

subway

1. (North American) underground railway.

4. (British) underground walkway, tunnel for pedestrians.

When a word is specific to one variant, it should obviously be so tagged:

godown

1. (Indian English) A building for the storage of goods; a warehouse.

The situation is trickiest when two variants have different spellings for the same word. In this case, it is preferable to collect the word's etymology, definitions, and other information in a single entry, to avoid duplication of effort, and so that translations can be consistently listed. The alternative spellings are listed in the "Alternative spellings" listing:

colour

Alternative spellings

* colour (Commonwealth English)

* color (US)

All alternatives (including that of the nominal headword) should be listed in the "Alternative spellings" section, as shown.

An unavoidable technical limitation is that any entry must have one title, which will perforce use a particular spelling. The preferred solution is to list the entire entry once under one spelling, and to use redirects to that entry from the other spelling(s). The choice of which spelling "gets" the entry, and which spellings are redirects, is almost accidental; the current practice is simply that the spelling first used when an entry is created stays with the entry, and that the later-added spellings are redirects. (This approach, though unabashedly empirical, does have a certain Solomonlike appeal to it.)

This issue has aroused considerable debate, but the contention can be minimized by observing that there is no claim or assertion of primacy or "correctness" attached to the choice of spelling of the main entry, versus the redirects. All spellings listed in the "Alternative spellings" section are equally valid in the context of their respective variants. The fact that one spelling happens to be listed in an entry's title is an artifact of Wiktionary's database architecture; it is not a value judgment.

To further reduce any appearance of bias, it has been suggested that the title of such an entry be something like "colour/color" or "colour,color" or "colo(u)r", with all individual spellings as redirects to it. This proposal is under discussion but has not reached consensus.

To reduce the appearance of bias, it would also be possible to retain one arbitrary spelling as an entry's formal title, but to list the variants in the entry's various sections:

colour

Alternative spellings

* colour (Commonwealth English)

* color (US)

Noun

colour/color

1. The spectral composition of visible light.

So that visitors unfamiliar with these issues will not perceive any unintended bias, it might be appropriate to include a templateized disclaimer at the top of multi-spelling entries:

Due to technical limitations, this entry's title uses a particular spelling, and is redirected to from other spellings. No value judgment is intended by these choices. See the "Alternative spellings" section for the list of all spellings of this word and their status.

This disclaimer is inspired by the {lowercase} template which Wikipedia uses for words which are supposed to start with a lower-case letter (see e.g. Wikipedia:zsync).

Finally, it is worth asking whether an enhancement to the Wikimedia software could be pursued which would enable a single entry to exist under multiple names, to completely eliminate the implication that the "main" entry uses a "preferred" spelling, or that there is anything inferior about spellings that use redirects.

(It has also been suggested that much the same effect could be achieved without any software changes, by having two or more distinct pages, each containing identical content transcluded from some central place via a template or other mechanism. This idea poses difficulties for those editing the content, and is still being discussed.)

—Scs 22:26, 3 April 2006 (UTC)Reply

Whomever corrected my error with the heading levels here in this BP section, thank you. --Connel MacKenzie T C 22:50, 3 April 2006 (UTC)Reply

You're welcome. :-)

Thank you Scs. --Connel MacKenzie T C 01:05, 4 April 2006 (UTC)Reply

I do not see how software could overcome the bias issue. No matter what technical solution is proposed, the result is still sub-optimal. If a user looks up the word color would we then have lots of {{PAGENAME}} tags within the page to display only "color" and not "colour"? That would then raise accusations of bias from the Commonwealth proponents, would it not? (Also, given inflected forms, I don't think such a solution is possible anyhow.)

Simply applying the wiki default policies here would be a monstrous improvement to this (Scs') proposal...that is, as Vild said, just have two entries.

yet another proposal

[Another proposal made at this point in the thread, at 04:54 on 4 April 2006 by Connel MacKenzie, is now archived at Wiktionary talk:Project - Keeping Translations Common and Synchronised Across Different Spellings/BP April2006. —Scs 13:52, 15 April 2006 (UTC)]Reply

yet more discussion

[Having] two separate pages... still strikes me -- please don't anyone take this personally -- as retarded. It's a "share the misery" approach; it's the compromise that succeeds not because it's any good, but because it's the least unpalatable.

All we're centralizing so far is the translations, and while those are significant, they're not the only or even the most important part of the entry that are problematically duplicated, that are at risk of diverging. The definitions, examples, and etymologies are still redundant. (Although someone mentioned that an earlier discussion deemed etymologies to be necessarily distinct, for reasons that escape me.) —Scs 03:04, 5 April 2006 (UTC)Reply

The NPOV issue arises again if the definitions are shared. Furthermore, when looking at colored vs. coloured I think you'll see dramatic differences in the definitions (etc.)

The etymology of the American terms must be different; they are, after all, bastard children of the commonwealth spellings (courtesy of Noah Webster.)

--Connel MacKenzie T C 03:28, 5 April 2006 (UTC)Reply

See comment below. —Scs 22:08, 5 April 2006 (UTC)Reply

The pages look very similar now because of the previous iterations of this flame war, not because they should look the same. As they are given the opportunity to diverge, they should gradually become more accurate. --Connel MacKenzie T C 03:31, 5 April 2006 (UTC)Reply

In what way(s) are they currently inaccurate? —Scs 22:08, 5 April 2006 (UTC)Reply

I disagree with you, Scs. Actually, the only things that are shared by these pages are the translations. Etymology: different; pronunciation: different; derived/related terms: different; and so forth. Certainly the definitions; when first looking at Connel's proposal I also thought about expanding this system to all corresponding sections, including defs, until I realized the definitions are even the main point of difference between these entries. Also, American articles should be written in American English, right? Even more differences.
Connel, American spellings on -or are historically as correct as Commonwealth -our. See Paradise Lost, there are only words on -or (IIRC). This is comparable to -ize/-ise, where the former predates (and is etymologically correcter than) the latter.
Up till now, I haven't got any complaints about this proposal. — Vildricianus 10:25, 5 April 2006 (UTC)Reply

Right. So as the "-*our" pages are corrected, they will reflect that they are valid archaic/obsolete spellings (everywhere, not CW nor US) while the "-*or" pages obviously will not contain those same definition lines. --Connel MacKenzie T C 14:18, 5 April 2006 (UTC)Reply

On reflection, some of these arguments are really not so convincing. It's been asserted that all sorts of things have to be different on the separate and differently-spelled U.S. vs. Commonwealth pages, but I'm not seeing those differences. (Perhaps this is because, as Connel suggested above, they've been artificially synchronized, altough at first glance they don't look the worse for that, if so.) In particular:

Definitions and examples. The definitions on the color and colour pages are, as mentioned above, virtually identical. The definitions on the colored and coloured pages (which Connel suggested I look at for "dramatic differences") are virtually identical. Besides the spelling and the irrelevant divergence in senses 6 and 7, the only difference that I can see is that the colour page includes tags for countable vs. uncountable. (Am I missing something?)

Spelling and other usage within definitions. Now, it's certainly true that a U.S.-slanted entry for "color" is likely to use "color" and other U.S. spellings in its definitions and examples, while a U.K.-slanted definition is likely to use "colour". But this is potentially true of every single definition of any word on Wiktionary. If this is a problem, then we need separate definitions and examples for every word, or in other words, we need a separate en_uk.wiktionary.org and en_us.wiktionary.org. But if we don't need to make that split, if we can tolerate U.K. spelling and usage in a definition that might be read by an American, or American usage in a definition that might be read by the rest of the world, then I don't see why we can't tolerate such quirks on a hypothetical unified color/colour page.

Pronunciation. Obviously, the pronunciation of many or most words differs between Commonwealth and U.S. English. If we can capture those differences adequately on a page like father, then we don't necessarily need two separate pages to capture differences in U.S./U.K. pronunciation for color/colour.

Etymologies. A couple of people have asserted that the etymologies for colour and color are different. But let's look:

colour: From Old French coulour, from Latin color. In American spelling the 'u' was dropped from colour to simplify the spelling. In British spelling the 'u' remains.

color: From Latin "color" via Old French "coulour"; in U.S. spelling the 'u' was dropped from colour to conform to the word's Latin origin. In the rest of the English-speaking world the 'u' remains.

So the only significant difference is that the two pages give different reasons for why the spelling is different in the U.S.

Now, in terms of hypothetically consolidated pages, I grant that there are additional complications for the color page (which has to contend with the so-spelled Latin and Spanish words), and the tyre page (which has to contend with the city). I grant that, for some people (though I don't know how many), the transpondian spelling variations loom large, and that the appearance of stigma associated with redirects is a real issue. But for all of these other alleged reasons why colour and color have to stay on separate pages, the reasons either aren't compelling (it's demonstrably possible to minimize the differences, as color/colour shows), or the reasons extend past color/colour to suggest that we end up needing (to butcher a phrase) two separate dictionaries separated by a common language. —Scs 22:08, 5 April 2006 (UTC)Reply

Please clarify what you are trying to say there at the end. I, for a long time, advocated using redirects (particularly for inflected forms of English words!) Many reasons exist for not using redirects here. The primary reason is that other language entries might share the same spelling. The secondary reason is the havoc caused to interwiki links. A third reason is the certain spellings are not NPOV. (I'm sure many more arguments were ofered a year ago, when I started entering redirects for inflected forms of English words. I don't feel like looking them up right now.)

If en.wiktionary.org has the practice of not using redirects, then is is a very British POV to quash American English spellings. To shoe-horn multiple entries into single entries (as has been done and I repeat: has not yet been undone on these entries) is obviously not a neutral point of view. I have not suggested having separate en-us and en-uk Wiktionaries (you, from your POV have.)

The only component of the entries that is unlikely to diverge (and diverge by a lot they will!) is the translation sections. Artificially merging the entries has always been and will always be, POV. --Connel MacKenzie T C 03:09, 7 April 2006 (UTC)Reply

I do wonder what you think my "POV" is. I think the only strong opinions I have here are that (1) having separate pages for e.g. color and colour is lame, and (2) redirects do not necessariloy connote second-class status.

I haven't seen anyone trying to "quash American English spellings", and I'm certainly not trying to. (If there were to be a combined color/colour page, and if it were not called "color/colour", it would likely be under color, seeming if anything to "quash" the British spelling, since it would make more sense to share the same page with the Latin and Spanish spellings.)

When I mentioned the possibility of separate en-us and en-uk wiktionaries I was not proposing or advocating it! What I was trying to show was that many of those who insist that color and colour must remain distinct should also, by logical extension of their own arguments, find themselves requiring such a wholesale split.

And that is what I was trying to say at the end there, which I shall try to clarify. The question is, are color and colour so closely related that they deserve to be discussed on the same page, or are they so different that they require two separate pages? And if they require two separate pages, are there other pairs of related-but-not-identical words that should similarly be split?

Let's look at some other cases:

sewer pronunciation 1 (a system of pipes) versus sewer pronunciation 2 (one who sews). Different etymologies, different pronunciations, totally different meanings, yet they share the same page.

periodic etymology 1 (repeating) versus periodic etymology 2 (chemistry, per + iodic). Again, completely different etymologies, pronunciations, and meanings, yet they share the same page.

father. Different pronunciations in the U.K. versus the U.S., but one page.

subway sense 1 (underground railway) versus subway sense 4 (tunnel for pedestrians). Different meanings in the U.S. versus the U.K., but one page.

Here we have pairs of words with completely different meanings and etymologies, or significantly different pronunciations, sharing the same page. Yet color and colour, which are clearly the same word but with a minor regional spelling difference, are consigned to separate pages. Why should spelling be the difference that trumps all others?

It has been repeatedly argued here that color and colour ought to be on separate pages because, aside from their spelling, they have or ought to have significant differences in the way their pronunciations, etymologies, definitions, or examples are listed. But if there ought to be separate color and colour pages for those reasons, then by the very same arguments, there should be separate pages for U.K. vs. U.S. father, and U.S. vs. U.K. subway, and chemistry vs. common usage periodic, and the two very different senses of sewer. If U.K. versus U.S. spelling and usage matter in definitions and examples, then every word (even if it's spelled and defined the same) potentially needs separate U.K. and U.S. definitions — hence, the hypothetical, reductio ad absurdum en.uk versus en.us split.

Real dictionaries do have separate entries for different words that happen to be spelled the same, such as sewer and sewer or periodic and periodic. Real dictionaries don't tend to have separate, redundant entries for color and colour — if they list both as separate headwords, one is invariably a "see" (i.e. a redirect) to the other.

—Scs 02:12, 11 April 2006 (UTC)Reply

Scs, entries on the English Wiktionary are distinguished by spelling - why we don't use Wikipedia-style disambiguation, I don't know. I wish we did, but we don't. With that premise in place, to not make the spelling distinction when the spelling distinction exists is POV. The "real" dictionaries you refer to are not multilingual dictionaries, but rather one or the other - the American Heritage Dictionary has an obvious bias towards the American spellings, while the Oxford English Dictionary has an obvious bias towards the British spelling. It is wrong for any Wikimedia project to adopt one bias or the other. The is the heart of neutral point of view! --Connel MacKenzie T C 02:39, 11 April 2006 (UTC)Reply

Ligatures

What about fetus, foetus and fœtus? Or is the ligature extinct in modern English? Jonathan Webley 06:46, 31 March 2006 (UTC)Reply

Briefly: US English drops the "o" and UK English retains it; the spelling with the ligature is archaic in UK English and obsolete (or archaic? or erroneous?) in US English. — Paul G 10:50, 31 March 2006 (UTC)Reply