Wikipedia talk:Requests for mediation/Romanization of the Russian language

Hi, I shall be the mediator for you case. It is important that all answers and question on this page are free of hosility so that we can have a rational discussion, although I must say that so far both parties have been good natured. Let me start by saying that I have no knowledge of Russian, and so approach this from a very netural point of view. I have read through all the evidence submitted in the Mediation request and on the basis of that have a few initial questions.

1. It seems that User:Kuban kazak, you have a problem with the current transliteration convention for Russian on wikipedia. What steps have you taken to resolve this matter?
2. Do you (User:Kuban kazak) feel you have more of a problem with the system of changing the convention than the actual convention itself?
3. Do you feel that a vote may be the best solution to this problem; by having a properly conducted poll on Wikipedia:Naming conventions (Cyrillic), with arguements on both sides presented? Would this resolve the problem effectively for both parties, and for the wider interested Wikipedia community generally?

If you could leave your respnses below, and please feel free to leave questions for each other, or for me. If for any other reason you need to contact me, and do not wish to leave a message on Wikipedia then e-mail me, Special:Emailuser/Wisden17, which I check a number of times a day. --Wisden17 16:43, 5 April 2006 (UTC)

Ezhiki's reply and comments to question 3 above

The whole purpose of Wikipedia:Naming conventions (Cyrillic) (shortcut WP:CYR) is to establish a new convention of romanizing Russian; one that would hopefully be free of of the current system's deficiencies. So far the progress has been slow and active participants far and few between. Still, there is progress.

My problem (one that I am hoping to resolve with this mediation), however, is in regards to the interim period, when WP:RUS is still in effect, but WP:CYR is still being developed. It is my firm belief that until WP:CYR project reaches consensus all Wikipedians must adhere to current policies (which, in this case, is WP:RUS). Starting making allowances now would lead to nothing good—if everyone starts to apply their own standards to Russian transliteration based on the fact that "it's gonna change anyway" it can introduce a lot of mess if the WP:CYR keeps its slow pace. Furthermore, while there are several critics of the current WP:RUS policy, they all have slightly different view on what should work. Allowing them to use their personal guidelines that work best for them, citing "lack of consensus on current policy" will further deteriorate the Russian romanization practices, which are not in their best condition as is. Lack of consensus on existing policy is not a good enough reason to ignore it. Obviously, when that policy was adopted, consensus existed. Using an analogy, if we get lots of users together who are discontent with existing 3RR policy, it does not matter that they can and are in the process of developing a hypethetical 10RR policy—until that new policy is in effect they still must observe 3RR.

I hope my points are clear enough. I'll be glad to answer any questions as they arise.—Ëzhiki (ërinacëus amurënsis) 17:03, 5 April 2006 (UTC)

Cossack's reply to all questions

Although the questions were asked to be treated separately, I'll do them in one. First of all the problem is not on whether or not one system is better The problem is in the way the English readers are presented with articles that have the most simple yet accurate version of the system. ALL of the systems have their flaws that make them unsuitable for WP. The best way is to obviously start from scratch and make a flexible system, where appropriate - borrowing practices from existing systems and where appropriate - making new ones. However this is where wikipedian policy - No orginal research in effect makes it a closed loop. To answer question one - yes, I did more than once suggested for changes and given the reply I got, that is the answer to question two as well.

As for the rest, then one has always remember, more often than not, there are common exceptions for Russian words that will not fit any system, but will be more widespread. In such cases I think some consistency has to exist not for the whole articles, but for those that are split into groups of articles. If anything then with Moscow Metro stations it is fully acceptable to bypass the conventions rules in such a case. In other words, treat translit systems as American and British spellings, if a group of articles has predominately more popular (by more popular I mean Googling) results with one system, then for consistency move them all to that system, which I did and which Ezhiki tried to prevent me from doing.

So in fact we arrive at a contradiction. There is a system which in some cases has and in some cases does not have more popular usage (I'd say about 50:50 to be honest). Trying to augment it and make some concrete changes is received with a boot - No original research. Trying to overstep the system to reflect on more popular usage is also received with a boot. After my brief time on the WP:CYR, I got the impression that the whole programme is nothing more than a hoax...--Kuban Cossack 18:54, 5 April 2006 (UTC)

Responses to my initial questions

Right, let me say that I can see both of your points. On the one hand we have the situation where we have a policy(WP:RUS), albeit not a brilliant one, and a proposal to change it to WP:CYR. Kuban Cossack, you seem to suggest that WP:CYR is of little use, and that a better system would be to treat it as we treat American and English differences in spelling.

So, my question would be: have you made this proposal before, about scrapping WP:CYR (and can we please try not to refer to it as a 'hoax' as statements like that do not help discussions) and instead using a system akin to that with American//English spellings and usages?

I must admit my immediate reaction is that such a scheme would be impractical for several reasons:

• A policy of American/English was needed due to the vast numbers of editors who use two very clear differences in spellings. In the system you would propose there would be a limited number of editors, and be more than two options for spelling.
• Why overcomplicate matters. Would a better solution not be to create a specific page with words which are the source of conflict in both WP:RUS and WP:CYR, and for a discussion then to take place regarding those words?
• Thirdly I think that there seems to be enough problems creating a new policy that by trying to create such a drastically different policy would only lead to more confusion.

If you have not proposed your idea for having one system per page then you ought to and gauge peoples reactions and views to it (as I personally cannot comment on how easy a system such a thing would be to use, as I do not have experience of Russian). What seems to be key is that you get support for any policy idea you have, as otherwise such a policy is worthless: you need people to follow a set policy (you must 'all be singing from the same hymn-sheet').

So I would ultimately suggest, Kuban Cossack, that you formally propose your idea of the system akin to the American/English system, and leave a message regarding your new page on the talk pages of WP:CYR and WP:RUS, and on the Community Portal. This way you can effectively gauge if such a policy would have the support you think it deserves, but ultimately if it doesn't you must stick to the currently agreed policy, and keep lobbying to change it. If you can produce a strong enough argument, and if there is still problems regarding WP:CYR then ultimately more editors may come to share your view. But it seems until that time you must follow the accepted policy.

I would welcome your comments on this advice, and the questions I have posed. --Wisden17 19:25, 5 April 2006 (UTC)

I really wouldn't like having a policy similar to one regulating British/American spelling, because it ultimately boils down to having to review each word on case-by-case basis. One of the main reasons we even have WP:RUS is to avoid just that—having to look at each case individually. To me, using one system for all cases, even when occasionally it produces unusual spellings (and I do not mean words that have hundreds of thousands search engine hits such as Moscow—those are covered by the "use common English name" policy), is a far better solution than having no system at all. However, until we have a definition of what "common English name" is (at least within WP:CYR or similar proposals), there will always be arguments as to what is "common" and what is not really. Still, that is better than having to argue about every single word separately, and I don't see how any of the Moscow Metro station names would fall under "common English use", even though some of the spellings produce more google hits than others.

Having said all this, I have no objections to Cossack's submitting (a) different policy proposal(s) for review if he chooses to follow that route. Ultimately, I trust the community will select the best policy of those proposed, but in the meanwhile it is my belief that those policies which are already in effect should be upheld.—Ëzhiki (ërinacëus amurënsis) 19:43, 5 April 2006 (UTC)

See that is the exact problem, kind off saying we were here before you, we made those rules what makes your original research better than the "accepted" policy and why should we chage it for you? That is exactly the case here, by trying to augment the system to reflect on a few changes for example Dostoevsky vs Dostoyevsky or Medeleev vs Medeleyev. In almost all cases by googling the first result would give much more hits. So I logically proposed for the rule of having the ye combination to represent the -ye sound be dropped from the system, but that would mean original research. Not because one word outdoes the other on a scale of 6:1 or 3:1 or in case of Kiev vs Kiyev 1000:1; but because there is a systematic correlation between the different rules that are used for any real word. Even googling the same Metro station like Belyaevo vs Belyayevo gives a 140 x advantage. That can be said for nearely all Metro stations. That was mearely an example and there many others.

Of course it is possible to bypass this problem by simply making the translit system a guideline less of a rule. Most of the versions (in many respects) are more or less similar and interchangeble. So for not-so-well known places, like small towns that no-one will ever know of it is possible to apply the guideline in a firm mode. However, if there is a case that a word gets more hits because it does not fit the transliteration, I think it is suitable that in a case when Novogireevo (55 700) vs Novogireyevo (361) has a clear lead, then WP has to reflect that case without any reference to the transliteration policy or guideline. --Kuban Cossack 20:17, 5 April 2006 (UTC)

From my viewpoint you make a very strong point [User:Kuban kazak|Kuban Cossack]] and you clearly have a lot of data to support your argument. But it does not matter what I think, it matters what the Wikipedia community thinks. I really think you ought to create a name for you policy, create the article page, with all the data and examples you have. Post links from the WP;RUS and WP:CYR pages, and from the Community Portal. I would then, as a member of the community, read your proposals and the arguments for keeping either of the other two systems. You could then try and change peoples' views that way. What you must understand though is that none of this will happen overnight, and you appear to be rather annoyed that whilst you appear to have such a strong argument that no=one else appears to be listening or agreeing with you. Like I said I know little about Russian, so what you must try and do is persuade all of those who actively edit Russian related pages to accept your policy proposal. It may not be easy to do, but, if as you clearly believe it is a better policy then in the Long-Run it will benefit Wikipedia enormously. You could make a huge contribution to Wikipedia by creating a unifying policy, but to that you will need to convince people of its validity and need (although that seems to be apparent, as there are numerous flaws with both other systems).

I think this is your best course of action, and I'm sure that you could find other users to help you, indeed it may be an idea for Ëzhiki (ërinacëus amurënsis) to help you set a page for your policy, if you would like help with that. The key here is that all interested parties should work together, for the good of Wikipedia, so that eventually there will be a consensus on what the best policy is and why it is the best policy. You appear to have lots of fact and evidence to support your case, and if you were to lay this out on a specific page it would have a lot of clout to it. By posting a link to it on the Community Portal you will get people interested in the debate who will bring a fresh pair of eyes to the debate (indeed it may help to have people who speak absolutely no Russian to air their views on the subject).

The overriding need is for a policy which is voted on, or agreed upon by a consensus. There is very little point in one user waging a war against the 'system'. Kuban Cossack, you clearly have a great deal to offer Wikipedia, and it would be fantastic if your main contribution was to create a unifying transliteration policy, or simply a unifying policy on how to deal with Russian names etc..

I welcome both of your views on such a suggestion. --Wisden17 20:47, 5 April 2006 (UTC)

I believe that Cossack is perfectly capable of submitting a new policy proposal himself, but should he need any assistance, I, of course, agree to help.—Ëzhiki (ërinacëus amurënsis) 21:16, 5 April 2006 (UTC)

Comment by Ezhiki to Cossack's passage immediately above

[after edit conflict] We discussed this point before as well. Wikipedia:Search engine test describes very well why google tests are not as reliable as people want them to be and why they should be used as a rough approximation only. To save our mediator's time, I'll reiterate the main points:

• the number of Google hits is not a reliable indicator of common English usage (e.g., the majority of those "Novogireevo" hits are either pages targeting foreigners in general, not necessarily Anglophones, or pages where a Russian name was romanized for internal technical purposes such as file names). It would be much more reliable to query fairly recent English-language publications (books, research papers, and media publications), but unfortunately there is no such tool available to general public.
• the number of google hits may change over time, which is important to remember with hit ratios of lower order.
• the number of hits returned by google is affected by google search algorithms. If one is to query for "novogireevo" and go to the last page of search results (page 36 as of this writing), one will see that the number of unique results is only 356. "Novogireyevo" produces eight pages of results with 71 unique results. Now, using Google Advanced Search, narrow down the results to English-language pages only, and the number of unique results drops to 161 and 52 correspondingly. One can further tweak this by limiting the domains to exclude Russian servers or invent a great number of other legitimately-looking ways to alter the results, maybe even reversing the trend. The problem, however, remains—for scales of less than roughly 1:1,000,000 google tests suck and for scales of less than 1:10,000 they are pretty much meaningless.

With all that in mind, having a unified transliteration system does no longer seem all that bad. It is also important to remember that while the article can have only one title, the number of redirects to it is not limited, and more or less common spelling variations which are out of compliance with existing translit policy can still be mentioned in the article body. A combination of redirects and alternate spellings covers the needs of readers very well, and a spelled-out romanization policy equally well covers the needs of the editors. When you title an article, would you rather apply a system that is used by scholars and governments and is common and comprehensible (and I don't mean WP:RUS in its current state), and be done with it, or would you rather prefer to query search engines straight, backwards, and sideways, and then backwards again, then decide which result is "more common" and on what basis, and whether one result is better than other, and then defend your variant from the people who do not agree with your judgement? To me, the answer is obvious. Furthermore, the existing rules are written out in such a way that they only allow for "common Engilsh names" and WP:RUS. If it's not "common", it's romanized, period. Search engine tests for borderline cases are not a part of any of these two policies. Again, a proposal can be made to include them, but it's the same story all over again—if you wish to propose a new rule, by all means do so but please abide by existing rules until yours goes through.

Finally, neither I, nor other editors who support my views on romanization practices, are snobs who "were here first", and hence are always right. Those who were here first created many policies, but none of these policies are set in stone. There is a procedure that can be followed to (try) changing any of Wikipedia's policies. Propose, discuss, review, vote, accept (or decline)—these are five essential steps that should not be skipped. Meanwhile, old policies are upheld. If one has no time or desire or patience to go through these steps, one should be content with what's already there. That's all I am trying to say.—Ëzhiki (ërinacëus amurënsis) 21:16, 5 April 2006 (UTC)

Cossack's Comment to Comment

The question is not wether to romanise a name, it is how to romanise that name. In almost all circumstances we try to use both. We do not transform Ivan into John (which would be the equivelent name), however in case of Alexander and Aleksandr we translate it. As for Google the case is even if you do run the filters on such non-common words that Ezhiki did, it still shows that Novogireevo outnumbers Novogireyevo. In the end it is about a district (as well as a metro station) in Moscow and hence most of the publications [English!] about that would be written by native Muscovites and Novogireans. That is a solid and undisputable fact. That is what ultimately influences the English languages' use for foreign terminology. That is what got Peking to be dropped in favour of Beijing. The way things currentely stand is that most of the Russians who romanise the Russian language do not apply the certain set of rules that I have pointed out which the current romanisation guideline is trying to enforce. Wikipedia, of all places, has to reflect on this, because I will not expect to see native Anglophones writing about such a far-edged place like Novogireevo.

However in case of such well known places and names like Dostoevsky which places like Amazon.com use. Moreover with Dostoevsky even if give Dostoyevsky, amazon will mirror the result, however if anyone zooms in on the cover of any book (as all are published by western press) that first y miraculously drops from the title. That is, IMHO, says something about which translit rules would the world society use. It is a case when there is a continuous, systematic disregard to the guideline that Ezhiki is trying to enforce. Full stop.

So either we accept that either we need to redesign the translit guideline to fit the needs of the English language and thus remove the no original research barrier. OR we augment how the guideline is applied to wikipedia.

BTW, what is Ezhiki's personal opinion on the matter of using the (-ye) after a vowel? (If anything I might as well start off with convincing my key opponent :)--Kuban Cossack 14:40, 6 April 2006 (UTC)

Again you appear to have a lot of evidence hear, and like you say you need to convince people. This is the point. Ezhiki has a problem with the way you've gone about proposing your idea, in that effectively you haven't. What you need to do is start an article which codifies your proposal, sets out why you think the other two are wrong (which you appear to have gone a long way doing above), states how yours would solve the problem, and then say why yours is better in other regards (for example you have the example of Amazon, so you could argue that there should be continuity between major Internet sites).

I'll be happy to help you set up a page if you require any help, and Ezhiki has also offered to help. I think the problem is that you are misunderstanding the nature of the problem. You need to get a consensus on the matter. You have gone a long way to convincing me that your proposal is a good idea, indeed I would most likely vote in favour of it, but that is the point, you must try and convince more people than just me. Sure you will not satisfy anyone, but as long as you can get a general consensus then your proposal will have been a success.

So, what I would propose you would do is this:

• Set up a Wikipedia article which formally sets out your proposal with evidence supporting it and why it would be better than WP:RUS and WP:CYR
• Start a discussion on the Talk page of this article.
• Post links to your article's talk page from Wikipedia_talk:Romanization_of_Russian and Wikipedia talk:Naming conventions (Cyrillic).
• Post a link to your article and start a discussion on Wikipedia:Village pump (policy)
• And finally, once you have got a good discussion going post a link on Wikipedia:Community Portal so that you can get the views of more of the Wikipedia Community.

Once you've undertaken all of these steps you may well find that you have a lot more support for your policy, or you may not. If you do, and your policy get adopted, then you've achieved a great deal. If ultimately this fails, initially, then you must work within the current policy, and as time goes on more and more people may become aware of the problems with the current policies and come around to your ideas. So don't give up trying, but whilst you are working towards your new policy you cannot simply force it upon pages that you happen to be involved with editing.

I would welcome both your views on this suggestion. --Wisden17 15:00, 6 April 2006 (UTC)

Ezhiki's answer to Cossack's e/ye question

[after an edit conflict, so some points are identical to Wisden's above]

About the "(y)e" after vowel: I honestly don't care; whatever the WP:CYR (or any other alternative policy that may appear in the future) leads to, I'll use. What I do care about is that:

• we abide by the current rules why they stand;
• we propose and discuss new rules, but do not implement them until they are agreed upon by all parties.

For me personally it is also important that we use a translit system which is documented and is not an empirical approximation of practices outside of Wikipedia (i.e., if we go with BGN/PCGN, then we use "ye", if it's ALA-LC, then we use "e"). This is why I am hesitant to use any simplified and customized systems, because simplification and customization points will always leave door wide open for new disputes. Current WP:RUS is a simplified and customized system already, and look at all the disputes around it. The less wiggle room we leave in the rules, the less time we will spend arguing about the fine points of romanization from the empirical point of view and the more time we'll have left to do actual editing.

But this, of course, is my personal view only. The only way to find out if that's what the community wants/does not want is to let them know. The draft of my proposal is already at WP:CYR and so is yours and Michael's; so let's refine them and put them out for a formal vote. Whatever variant the community accepts (even if it's ugly GOST :)), you can count on me to accept and enforce it from that point onward. As long as it's a communal decision, and not that of a single user, I am going to be fine with it.—Ëzhiki (ërinacëus amurënsis) 15:20, 6 April 2006 (UTC)

Cossacks reply

Well here is where we have our solid disagreement: use a translit system which is documented and is not an empirical approximation of practices outside of Wikipedia. The problem is that even in the way WP:RUS currentely operates, Original Research barrier is already overstepped, like no apostrophes in titles, I am just saying to take it one step further, accounting of the overwhelming popularity of the negligence to rules like ye/e. Ie the same Dostoevsky. All I am trying to say that if a policy is adopted without modifications there should not be a barrier on exceptions like the (ye/e); alternatively a new Wikipedian version for Romanisation of Russian be adopted and drafted by us, wikipedians.

So what to do know, well I would gladly except some of your suggestions and support, given that a garuntee that the Original research boundry be lifted. Until then, I shall try to take a more active part in WP:CYR, but since the project is still undergoing discussion I would like to propose that the conflicted Metro station articles stay as they are (to avoid unecessary extra page moves) until the policy be voted and cemented into Wikipedia. --Kuban Cossack 15:32, 7 April 2006 (UTC)

(Un)fortunately, neither I nor anyone else can guarantee that original research barrier will be lifted. That is not something one person or a group of people can decide; that's up to the community, which again reminds us of the importance of making proposals publicly available. On the other hand, you have the right to propose whatever you want and whatever you think will work the best for the community.

As for the "barrier" itself, it is one of the points covered by WP:CYR. True, current policy contains elements of original research (which I admit I myself introduced), but that is exactly why it is so controversial—no matter what elements of OR one introduces, there will always be someone willing to contend those points. Seeing that, I joined people who wish to develop something better.

Anyway, we are going in circles now. Our disagreement regarding whether or not the system should be customized is irrelevant to this mediation (the point of which is to confirm that current policies must be upheld even though consensus no longer exists). That disagreement can very well be discussed at WP:CYR. All I am trying to persuade you to do is to not ignore the rules you do not like but follow a proper procedure. If we have agreement on that, let's close this mediation and move to WP:CYR. As a compromise, I promise not to move/correct any of the existing metro articles as long as you create new/edit old articles in the spirit of WP:RUS. That's inconsistent, true, but it will serve us both as a good motivation to continue to work on WP:CYR. You won't get the impression that I am stalling and procrastinating to keep WP:RUS active for as long as possible, and I won't get the impression that you are trying to impose your views on romanization practices without having to go through proper procedures. So, how about that?—Ëzhiki (ërinacëus amurënsis) 17:15, 7 April 2006 (UTC)

That to me seems like a very good suggestion, and would be a very good result for this mediation. I hope that both of you have felt that you have been able to easily discuss these issues. I know a lot of these arguments have been used by you before, but it helps a lot to have everything contained on one page as opposed to being spread across lots of Talk Pages. As I said I'll be happy to help with this matter at a later stage if you won't another pair of non-Russian eyes to view things. I feel that a real compromise has been reached in a short space of time, and that whilst you'll still have your viewpoints you can now understand, respect, and bear in mind the other's. --Wisden17 00:51, 8 April 2006 (UTC)

as long as you create new/edit old articles in the spirit of WP:RUS. On the contrary, I will gladly revert other conflicting articles like Krasnye Vorota and Okhotny Ryad as long as there is no -ye in the Metro, that would be breaking the inconsistency, but in return I am ready to abandon my push for the -iy and -yi altogether, simply because Google or not, it is still much less popular than the simple -y for both cases. As for WP:CYR, I will need your support there for the no -ye rule. As the rest goes, well I think the dispute is settled... --Kuban Cossack 23:07, 8 April 2006 (UTC)

To me this seems like the perfect outcome in that both of you have compromised and Cossack you've started to look at whether your idea would be popular enought to gain enough support to get adopted, and this is indeed what you will have to look at when putting together any planned policy. As I said if you need any help, especially a pair of eyes untrained in all the complex philiogical theory behind the proposal, and simply want to know if it is easy to understand, then I'm very happy to. I'm glad that we've managed to settle this dispute and I wish both of you a really good time of Wikipedia, and we may well bump into each other in the future, as I'll now take a close watch on how the policies develop and how Cossack, you do, in trying to get some of your ideas into the community for rigouous debate and comment. --Wisden17 00:03, 9 April 2006 (UTC)

Thank you, Wisden, for your guidance and willingness to return to this topic later as an independent observer. I think the outcome was very good. Also, thank you, Cossack, for agreeing to this mediation and your willingness to continue working in this area. I am looking forward to working with you at WP:CYR. Regards, Ëzhiki (ërinacëus amurënsis) 15:01, 10 April 2006 (UTC)