Wiktionary talk:Russian transliteration

Definition from Wiktionary, the free dictionary
Jump to: navigation, search

Attribution[edit]

Most of this table was copy-pasted verbatim from the wikitext at w:Romanization of Russian. Wikipedia's licence requires that the contributors to that article be credited here. —w:user:Mzajac

Actually we don’t need to have all of those transliteration schemes. We only needed them temporarily for comparison to decide what system we would use here. —Stephen 00:29, 16 October 2007 (UTC)
Stephen, I see black boxes in the 3rd row. Can I fix it? Mallerd 20:38, 3 November 2007 (UTC)
Do you mean that the background color for the English transliteration appears black on your computer? On mine it is a light grey. Yes, please fix it. I use a laptop which probably does not show the colors very accurately. —Stephen 12:46, 4 November 2007 (UTC)
Ok, I fixed so I can see it clearly, I can't speak for you I'm afraid. Mallerd 13:53, 4 November 2007 (UTC)

Contradictory and redundant[edit]

There is a separate romanization section in Wiktionary:About_Russian. It contradicts this chart in some details, and adds a bunch of rules which seem to be aimed at turning this into a phonetic transcription system rather than transliteration.

Phonetic transcription (in IPA and/or other systems) belongs in the "Pronunciation" section. A romanized representation of the Cyrillic orthography belongs with the headword, and in a "translations" section. (My opinion: keep this page and scrap the other, redundant section.)

This guideline should only remain in one place, to avoid this kind of forking.—Mzajac 05:21, 7 March 2008 (UTC)

Where are the contradictions? Just glancing at the two, it was not immediately apparent to me. This page is the main one, and the other was a work in progress that never got very close to being finished. Of course, there are some people who like to have IPA, but most Americans are not used to either IPA or SAMPA. But whenever someone wants to add IPA, the pronunciation section is definitely where it belongs. —Stephen 05:45, 7 March 2008 (UTC)
I've added a more detailed critique at Wiktionary talk:About Russian#Romanization critique. —Mzajac 23:36, 7 March 2008 (UTC)

Move to Wiktionary:Russian transliteration[edit]

Since this transliteration system is specific to Wiktionary, I'm proposing moving this page to Wiktionary:Russian transliteration, under Category:Wiktionary:Transliteration.

Please see discussion at Wiktionary:Beer parlour#Organization and Wiktionary:Beer parlour#Transliteration appendices. —Michael Z. 20:00, 12 April 2008 (UTC)

* Note: ё is transliterated as o following the consonants ж, ч, ш, or щ.[edit]

I fully support this, I would even suggest to change the romanisations - Gorbachov, Khrushchov, etc. However, no single official romanisation caters for this.

http://en.wikipedia.org/wiki/Romanization_of_Russian Anatoli 03:57, 24 March 2009 (UTC)

Alternatives[edit]

Similarly to ts/c, do we need to provide alternatives for ž, š, č and šč as zh, sh, ch and shch? I think it's worthwhile, since they are intuitive for English speakers and could be used if the diacritics are missing.

I always hesitate between ja/ya, je/ye, jo/yo, ju/yu, although "y" also stands for "ы", it may mislead English speakers into believing that "j" stands for /ʤ/, especially when Йоханнесбург (Johannesburg) is transliterated as Joxánnesburg.

For x we also need kh. Anatoli 00:31, 17 April 2009 (UTC)

This is a guide to Wiktionary's transliteration “standard,” not an encyclopedia article about transliteration. Please read Wiktionary:Transliteration and romanizationMichael Z. 2009-04-17 02:45 z
I understand this but who sets the standards in Wiktionary, aren't we doing it? If you look at the varieties of standards here: http://en.wikipedia.org/wiki/Romanization_of_Russian, a few of them include the suggested variants. Anatoli 03:18, 17 April 2009 (UTC)
The table in that encyclopedia article is documenting different variations in use. This table is setting the one standard which we use. This is a critical difference. No standard “suggests” variation, much less the free-for-all you are entering in here, or else it wouldn't be a standard, would it?
Don't you understand that the transliteration of a word is meaningless if you don't know exactly what the letters represent? Michael Z. 2009-04-17 14:42 z
The Oxford Guide to Style says “[w]herever possible, adhere to a single transliteration system throughout a single work. In texts using transliterated Russian, as well as Belorussian, Bulgarian, and Ukrainian, authors and editors should avoid mixing, for example, the usual British ya, yo, yu; the Library of Congress ia, io, iu; and the philological ja, jo, ju.” (Ritter 2002, p 350.)
For Ukrainian, we use the philological (a.k.a. scientific, academic, linguistic, international, scholarly) method of transliteration. For Bulgarian we do the same. For Russian we've been using a bastardization of the same. Let's just do the right thing, no? Michael Z. 2009-04-17 14:50 z
Support. Let's stick to one and only one romanization system to avoid ambiguity. I suggest we adopt w:Scientific transliteration as our sole standard and abstain from mentioning alternative transliterations altogether. We can make exception for letter ц, which can be transliterated either as c or ts. --Vahagn Petrosyan 15:10, 17 April 2009 (UTC)
If you make an exception, then you're not following the standard. How would you justify adding an optional English transliteration for only one letter, when the rest follow a century-old European standard? Let's just do what the big boys do. Michael Z. 2009-04-17 15:19 z
The translations are made by multiple users who may not know of our local standard. Transliterations ts, sh, zh, ch, shch, kh, ya, ye, yo, yu are standard, used in the media, official documents and are well understood by public - Russian and English speaking. They are not ambiguous either. Letters x and c could ambiguous and are a bad choice if we want to help people who have no idea how to read in Russian or this transliteration method. The "right thing" depends on what you think is right, isn't it, Michael? As there is more than one standard for the Russian transliteration, then you can't say this is right and this wrong. I agree that for Slavic (and other) people c, ž, š, č and šč are more intuitive but x is not. It is important to allow the easiness of typing transliterations for random editors, ž, š, č can't be typed on standard keyboards. Methods like BGN/PCGN are used to transliterate Russian passports, bank transactions, etc. Letters ž, š, č are never used in any Russian official document. Anatoli 22:53, 17 April 2009 (UTC)
I see that my arguments are ignored and my translations are "corrected". Anatoli 07:47, 18 April 2009 (UTC)
Yeah, well, the Russian letter ц was transliterated c thrice in one entry. You changed it to ts in one of them, leaving the other two. I corrected it.
You are proposing that we handle Russian transliteration differently from how anyone else in the world does, and differently from how we have handled all languages until now. This can't be helpful to anyone. I'm at a loss for words... Michael Z. 2009-04-18 23:33 z
I meant "федерация", not "птенец" (translations). What do you mean, anyone else in the world? Do you have any proof? A simple Google search will show that Russian names in the English like transcription are more common. As for dictionaries, they normally don't even use the transliteration but IPA at best. The link from WIkipedia also shows that there are more than one actively used standards. So "anyone else" is simply incorrect. If this is becoming so emotional, I won't insist but I am being objective and wish to help, not to upset. If my proposal has no support I won't push for it. Besides, I suggested an alternative, not a replacement. Anatoli 00:07, 19 April 2009 (UTC)
This is not an emotional issue, but I admit I am frustrated in my inability to communicate the simple practical realities of transliteration. Please be patient with me.
Every academic source in the world which uses transliteration chooses one method as its standard in any context, is what I mean. Every dictionary which cites foreign etymologies will use one standard for transliteration. For example, the OED mentions Russian and Ukrainian terms in etymologies transliterated according to the British Standard (it does not show their pronunciation). Every author who writes an article uses one standard to transliterate Slavic terms; typically a standard set down by the journal he submits the article to. Every author who writes a book uses one standard for Russian names; typically either the international method for terms in linguistics, or LOC for historical names or bibliographic titles in almost any other context.
No one – NO ONE – uses a random mix of standards in a single work. You demand proof? The proof is that there are no examples. Michael Z. 2009-04-19 00:50 z
Hello Michael. Consistency and standards are two different things. I asked for the proof that this transliteration method is used by everybody in the world, not that once you've chosen a system, you stick to it. I have been consistent and haven't broken the rules yet, before my proposal was approved, by the way. Variants c/ts were there before my last edit. My suggestion wasn't to mix both systems either but to allow random translators use the easy alternative, based on English spelling of Russian names or BGN/PCGN standard, without having to copy/paste characters that don't exist on standard English keyboards. The reality is, that most Russians are also better familiar with the English transliteration than Czech or Croatian alphabets. If you are opposing this proposal so vehemently, feel free to revert my edit (I may do it a bit later myself). Note that c/ts was there before my edit. I didn't mean to cause you any grief.
Wiktionary is not done just by one author, so some variations, IMHO may be still possible. Anatoli 01:23, 20 April 2009 (UTC)
Well, we have had a standard, such as it is was, and if editors enter different transliterations without familiarizing themselves with the guidelines, then we correct them. That's how all of the guidelines work. Of course this transliteration method is used by no one in the world, because Wiktionarians have made up some new rules which contradict the phonological basis of the linguistic transliteration system. Using something which looks like linguistic transliteration but has subtle and meaningless differences can only mislead readers.
Within the variation of the current last guideline, a transliteration that doesn't correspond to linguistic transliteration is inferior to one that does, so I will improve them when I see them.  Michael Z. 2009-04-20 17:08 z [updated for clarity: I was referring to the former near-linguistic transliteration]
At least this version clearly doesn't represent a standard. I won't endorse the previous version by reverting the page to a confusing near-standard. Your changes render the intro line obsolete. Would you please update it to indicate what this is intended for?
By the way, if we should accommodate all editors in the allowed transliterations, why not add в =w, е = ie, ё = e, ё = io, ж = j, й = i, й = ĭ, х = h, ч = tsch, ш = sch, щ = shh, щ = schtch, э = è, э = é, э = ė, ю = iu, я = ia? Michael Z. 2009-04-20 21:01 z
I've left 2 letters I am not happy about - х kh/x ц ts/c, until agreed. My preference is kh over x and ts over c. Anatoli 04:29, 22 April 2009 (UTC)
What are you doing? Please stop randomly adding and removing items in the table. Until we agree on the purpose of this page and a strategy to fulfil that, these unexplained changes are meaningless.
Due respect, but your or my preference for transliteration of particular letters doesn't concern us. We should choose a suitable transliteration system for this dictionary, not make up our own by aggregating Croatian and English phonemic expressions according to our whims. Michael Z. 2009-04-22 16:36 z
What am I doing? I changed the table back to what it was before my edit. I understand what you are saying and I didn't suggest all the possible varieties of transliterations for all languages but those that seemed more appropriate for the English Wiktionary. You don't have to be mean, Michael. Please stop complaining and change it to what you want. I won't touch this table. Anatoli 23:03, 22 April 2009 (UTC)
Well, as you just noted, you didn't remove the newly-added х = kh/x, so I had no indication that you'd intended to revert the page. I don't want to be mean, but I cannot comprehend the impulse to include a bunch of options as our “standard” and expect that readers will benefit from transliterated Russian when they can't tell what it represents.
A reason to publish transliterated Russian is that readers unfamiliar with Cyrillic can see a consistent representation of a word, deterministically derived from the Cyrillic. They can compare the word to other Russian words, and to transliterations in other publications which use the same standard. They can compare it to cognates in other languages, including Ukrainian and Belarusian transliterated by the same method, and to other Slavic languages, all of which have consistent transliteration methods or Roman spellings. They can understand unusual historical spellings and alternate orthographies found in etymologies.
Adding a single change to the system, or relaxing the standard by adding alternates, completely blows away its utility. There's no point in having any transliterations at all, when the reader can't tell whether Gorbačov is supposed to be Горбачов or Горбачёв, or whether tsar represents цар or тсар, etc.
It doesn't matter if readers aren't familiar with the particular standard, either. If they are familiar with the actual standard, then our current near-“standard” will mislead them. There will always be unfamiliar readers, no matter what we choose; if they are interested, they can look it up. If there is a demonstrated problem with looking it up, then we can make it easier. Michael Z. 2009-04-23 00:38 z
Anatoli, I am sorry to disagree with you and I used to think similarly to you, before mine interest in Serbo-Croatian emerged and before I had a landmark conversation with one Frenchman living in Bulgaria, who explained to me how pernicious this contemporary encroaching Anglicisation of the transliteration standards in Eastern Europe was for Bulgarian and how vital it is to adhere to the scientific transliteration (ж, ш, ч are pronounced similarly in Bulgarian and Russian, so perhaps you will find it convincing too). It may be true that sh, ch are a simplification for the Anglophone reader, but if you just imagine how would one German-speaking user with no Anglophone heritage or knowledge pronounce zh..., whereas ž, š, č are pronounced the same way by every (educated) speaker of languages using the Latin script. The uſer hight Bogorm converſation 16:53, 22 April 2009 (UTC)

As I said already, do what you wish to do, I won't touch the entry (the project page). I was under impression that kh/x was already there, so I didn't change it. c/ts was definitely not my addition. I've been using x anyway, in translation. Because you are still frustrated, I don't want to deal with you and read your angry condescending talk. I am trying to be nice to everyone and I don't like when I am treated as if I am an ignorant or a delinquent. I'll try to follow the standards for Russian translations. No need to reply, the project page is yours! Anatoli 00:54, 23 April 2009 (UTC)

A couple of aspects of the Russian transliteration[edit]

(moved from User talk:Stephen G. Brown)

Hi Stephen,

I just wanted to ask about your preferences with the Russian transliteration, namely for letters like я, ю, е, ё. Do you prefer to follow Wiktionary:Russian_transliteration? I noticed you sometimes add ' to make consonant soft, rather than using j. Not sure if you do it consistently. I initially always used "j" (пять - p'at') but then I changed to adapt to the convention here at Wiktionary, i.e. started using this method: pjat'. Now I am confused. I usually ignore the difference in the pronunciation of "е" in небо and тест, romanising both as simply e (nébo, test, although, strictly speaking it should be n'ébo). I do the same for letter "э" - simply using Roman "e", like the page above. Perhaps we should discuss and/or update the rules. --Anatoli 05:24, 29 March 2010 (UTC)

I only add ’ for ь, " for ъ. Otherwise, I never put it. For я, ю, always put ja, ju (pjat’). For ё, put jo, as njóbo (except after a sibilant: чё, шё = čo, šo). Whenever you see a ' used, it was added by one of the native Russian contributors who is not accustomed to our transliteration system. For Cyrillic е, use e after a consonant, but je at the beginning of a word or between vowels (jéli-jéli, délajet, tésto, nébo, v"jéxat’). Э is a rare letter, just write it as e (éto, étot). The only problem is when Cyrillic е is pronounced like э. In that case, I use ɛ.
For the special case of the reflexive particle -ся in the infinitive or third person, since it is not palatalized, I write ’ instead of j: učít’s’a, kážets’a, but učílsja. —Stephen 05:52, 29 March 2010 (UTC)
I see. I saw your edits (I think) where you didn't follow this, that's why I asked. Generally I use the same principle but including the cases with -ся in infinitive - just transliterating, not describing the pronunciation. Besides, this is consistent -ться and -тся are always pronounced -цца (reflected in the Belarusian spelling). I can employ the rule "е is pronounced like э" from now on. I think it's helpful for learners to know, which foreign words with е are pronounced with the palatalisation and which are not (there are many variants too), but there would be too many entries to be fixed, as I never used "ɛ" before. Then we need to update the transliteration page. "ъ" actually needs ”, not " (I think " may cause a problem). I was often lazy and used ' for "ь", not ’ --Anatoli 06:23, 29 March 2010 (UTC)
No, I always follow this. If I did not somewhere, then it was a typo. I think that those places where you thought I had not followed this were actually written by someone else. I often work on a specific part of an entry, such as the declension, and may not correct other parts such as transliteration. All of these rules, including -ся, are (or were) mandated in Wiktionary:Russian transliteration, I believe. While ъ needs ”, it is hard to type, so " substitutes for it. Most editors also put ' for ’, which seems to be fine. I have been using " for years and have never noticed a problem with it. However, I think it might be a good idea to allow э for ɛ, for ease of typing: tэst. I have not looked at the transliteration page for a long time. It used to explain the use of ɛ, but maybe it has been deleted or changed in recent years. —Stephen 06:41, 29 March 2010 (UTC)
Since you already used "ɛ", let's use it. (The Russian transliteration system seems more or less scientific, unlike the Arabic). I will copy to my file of useful characters. Anatoli 09:31, 29 March 2010 (UTC)

Excuse me for butting in, but if you start changing the transliteration system to represent pronunciation (for which we have a standard subheading and format), then it's no longer transliteration. The point is to represent the Cyrillic spelling, so it is more accessible to readers who don't know the alphabet.

Furthermore, if you do it on an ad hoc basis, ignoring the guideline that other editors are following, then we end up with an undocumented mix of two systems, and our transliterations become meaningless to the reader. Every instance of transliterated Russian in this dictionary is now suspect; it's poisoned, and we may never be able to clean it up completely.

Please let's not just make things up as we go along. Michael Z. 2010-03-29 16:15 z

When Conrad's transliterator is up and running, the transliteration will happen automatically, both in translation tables ({{t}}), inflection lines ({{infl}}, {{ru-noun}}) and elsewhere ({{term}}, {{l}}, etc.). The tool will be using the table here, which can be modified. The transliterations added up until now will be automatically retransliterated according to this chart, which means that various schemes used by various contributors will be unified. It also means that we can't transliterate е as "ɛ" in some places and as "e" in others: the tool can't guess it's tɛst in тест but testo in тесто. We will have to show the peculiarities of pronunciation under ===Pronunciation===. --Vahagn Petrosyan 16:50, 29 March 2010 (UTC)
That looks great. Michael Z. 2010-03-29 19:15 z
The transliterator is good but there's no substitute for manual work. It will still require the word stress. Humans not only can provide irregular pronunciation but also add the word stress. It's not clear how the tool is going to treat iotised letters in different positions and whether sibilants + iotised vowels will be transliterated differently (e.g. "чёрный"). Michael, rather than telling us off, please mention, which part of our discussion you don't agree on. Stephen has been contributor in Russian for years, it counts more than one instruction page, which can be adjusted if we agree on something. We could add special cases, like the treatment of "э" and "е", if it's not palatalising consonants. We can move the dsicussion to the transliteration talk page. Vahagn, please join. --Anatoli 21:56, 29 March 2010 (UTC)
I'm not telling you off, and I'm not debating the relative merit of the guideline at WT:Russian transliteration. Just pointing out that by ignoring the guideline, you are entering incorrect transliterations in a great many entries, and making the dictionary inconsistent and the reader's documentation meaningless. Who's going to clean up after you?
(By the way, the process of transliterating text has nothing to do with pronunciation, iotation, or sibilants. It's a mechanical exercise, transcribing letter-for-letter. If you don't do that, then you are fudging it up.) Michael Z. 2010-03-29 23:49 z
I asked these questions because I want to be consistent with other editors or agree on the methods if we have differences, so that we don't have to clean up anything afterwards. I have followed the transliteration rules in the last couple of years, even if I didn't agreed with them at first. What needs to be cleaned after me? I don't remember you being very active in Russian or cleaning up somebody's transliteration, Michael (sorry if I am wrong). Besides, you yourself described or agreed on cases where letters cannot be one-to-one, e.g. "ё" after sibilants and the Russian "е" after consonants vs after vowels or special letters (ь and ъ). If we transliterate mechanically both "ель" and "лента" as "jel’" and "ljenta" or "el’" and "lenta", we do a disservice to users. My preference and the current practice is "jel’" and "lénta". --Anatoli 03:57, 30 March 2010 (UTC)
Okay, I reread all of the above and realize that you are talking about variations in transliteration, but haven't been using them. Sorry for lecturing at you. But it would be a mistake to adjust transliterations for non-standard pronunciation or to add new rules. The transliteration must convey the original spelling (while the IPA represents pronunciation). So a linguist can explain that, e.g., “the spelling ego is pronounced /jevo/” to anyone who is familiar with the two century-old linguistic standards, w:scientific transliteration and w:IPA, but may not know the Cyrillic or Glagolitic alphabet.
Wiktionary's rule exceptions, like changing e to je at the beginning of a word, are not found in scientific transliteration. (But they could be handled correctly by a machine transliteration system, and specifically, I see that Conrad's transliterator supports matching word beginnings and ends.) Michael Z. 2010-03-30 06:08 z
I don't disagree on transliteration "jegó" but I usually supply the reading in brackets as (read: jevó). --Anatoli 09:38, 30 March 2010 (UTC)
Well, the entry его#Russian has a correct pronunciation /jɪˈvo/, but it has been copied in the mis-transliteration jevó. Unless you read Russian, you can't tell that the standard pronunciation of this word differs from its spelling ego (or jego, following the Wiktionary method).
Slavicists do sometimes use the same kind of notation to indicate pronunciation, as jevó. There's nothing preventing us from supplementing the IPA with such notation, just as we do with English respelling (our “enPR”), but that belongs in the “Pronunciation” section of an entry. Michael Z. 2010-03-30 16:27 z
Well, I have not changed any rules or introduced anything new. I have been doing it consistently the same way for years. I find the IPA pronunciation unreadable and unhelpful, and I think most users prefer to get their information from the Witionary transliterations, including the pronunciation considerations. When the native Russian add transliterations, they follow our system to a degree, but they also tend to mark soft consonants with ' instead of j (p'at'). I would rather have their assistance as they see fit to offer it, and I fix their translations as I find time. I don’t think we have any problems in our Russian transliterations, and if somebody wants to add IPA’s, they are certainly welcome, but I don’t use them because I find them simply unreadable. —Stephen 19:59, 30 March 2010 (UTC)
If you don't like IPA, then go ahead and add scholarly transcriptions to the #Pronunciation section. That would be a welcome improvement to the dictionary.
But please don't enter incorrect transliterations by “including the pronunciation considerations.” If you do that, then they are no longer transliterations. Don't you understand that their purpose is to represent the Cyrillic or Glagolitic orthography? Michael Z. 2010-03-31 02:17 z
The transliterations that I enter are the same ones that I was entering before the transcription page ever existed, and when the transcription page was finally written, it was a description of what I was entering. If you have in the meantime changed the rules so that what I am entering is no longer correct, then you should change it back the way it was. In any case, you are never going to get the transcriptions that you so badly want unless you do them all yourself. If you try to force your personal system on everyone, people will just stop offering transcriptions. If you like, I will stop adding or correcting transcriptions and will turn that job over to you. —Stephen 02:25, 31 March 2010 (UTC)
It's you who's championing your own personal system, which is exists only at WT:Russian transliteration. Wiktionary chiefly uses standard Slavistics transliteration for Cyrillic and Glagolitic in other languages. I think it's a mistake to invent a system when the rest of the world uses a century-old standard. But that's not what I'm writing about.
I'm concerned that you may ignore the transliteration rules we have. That's what you seem to be doing when you enter jevo for его (jego), or add undocumented innovations like -s’a for -ся (-sja), or ɛ for э (e). I'm not changing the rules. You wrote them, so how about you follow them? Michael Z. 2010-03-31 03:32 z
I do follow them. I have always been completely consistent. The only things that have changed are that х vacillated between kh, x, and χ, but finally settled at x; and ц began with ts, then either ts or c, and finally settled at c. Everything else is the same. The "undocumented innovations" were at one time documented. If they are undocumented now, it is because the documentation was removed. As well, jevó and xoróševo have been the standard since the very beginning, and were at one time documented as well. I recall discussing it with you years ago. —Stephen 03:47, 31 March 2010 (UTC)
I am sorry if my questions created a discussion that causes some grief. Gentlemen, do you mind if I move this discussion to Wiktionary_talk:Russian_transliteration? Note that my transliteration questions were related to translations, not entries. IPA would be irrelevant and cumbersome there, nobody puts IPA in translations and I will continue to use the established practice of transliterating in Slavistic methods with accent marks. My preference is to have the tone marks in the entries as well. IMO, exceptions in readings such as jegó/jevó should be close together in the entries and visible. --Anatoli 04:04, 31 March 2010 (UTC)

Automatic transliteration[edit]

Here I cooked up a script to transliterate Russian according to current rules:

function DoTranslit(src)
{
        var xlat = {'а': 'a', 'б': 'b', 'в': 'v', 'г': 'g', 'д': 'd', 'е': 'e', 'ё': 'jo', 'ж': 'ž',
                'з': 'z', 'и': 'i', 'й': 'j', 'к': 'k', 'л': 'l', 'м': 'm', 'н': 'n', 'о': 'o', 'п': 'p',
                'р': 'r', 'с': 's', 'т': 't', 'у': 'u', 'ф': 'f', 'х': 'x', 'ц': 'c', 'ч': 'č', 'ш': 'š',
                'щ': 'šč', 'ъ': '”', 'ы': 'y', 'ь': '’', 'э': 'e', 'ю': 'ju', 'я': 'ja'};
        var accents = {'a': 'á', 'e': 'é', 'o': 'ó', 'i': 'í', 'u': 'ú', 'y': 'ý'};
        src = src.toLowerCase();
        src = src.replace(/ё(?!́)/g, "ё́"); // make ё stressed if it isn't yet
        src = src.replace(/([жчшщ])ё/g, "$1о"); // ё becomes о afer ж, ч, ш, щ
        src = src.replace(/([аяэеоёуюыиъь]|\W|^)е/g, "$1йе"); // е becomes йе after vowels, ь, ъ аnd word-initially
        var result = ''; var ch; var ch2;
        for (var i = 0; i < src.length; i++)
        {
                ch = src.charAt(i);
                if ( ch == "́" )
                {
                        ch2 = accents[result.charAt(result.length-1)];
                        if ( ch2 != undefined )
                                result = result.substr(0, result.length-1);
                }
                else
                        ch2 = xlat[ch];
                result += ch2 == undefined ? ch : ch2;
        }
        return result;
};

Can it somehow be implemented here? —This unsigned comment was added by Panda34 (talkcontribs) at 06:48, 2 July 2010.

What is that language? It resembles C++, but it is definitely not C++. The uſer hight Bogorm converſation 12:19, 2 July 2010 (UTC)
Eh... I thought it was Javascript... Like, to run in the browser... But I'm no Javascript programmer, just made it up looking at MSDN documentation, so I may be wrong. It runs in my Internet Explorer, anyway. But javascript is disabled in wiki? This could be used, like, when entering Russian assisted translation to automatically fill 'transliteration' edit box. No big deal but it saves you from copy-pasting those acutes and wedges above letters which is a bother for me.--Panda34 13:35, 2 July 2010 (UTC)
You may want to read the discussion above: the automatic transliteration script has been written for a long time already, it just awaits approval. --Vahagn Petrosyan 14:58, 2 July 2010 (UTC)
Yes, it may be Javascript. I am only familiar with a couple of programming languages, among which C++ and this one appeared similar to it. The uſer hight Bogorm converſation 15:01, 2 July 2010 (UTC)
Acutes are no bother at all. Just choose Icelandic from the language templates from the combo box which is to be found below the save page and show preview buttons and there you have all you need for the vowels - á, é, í, ó, ú, ý, whereas ć, ś and ž are available in the Slavic Roman template. The uſer hight Bogorm converſation 15:05, 2 July 2010 (UTC)
Now I noticed that the acutes are available in the Slavic Roman template as well. Præsumably, because they are in use in Slovak. The uſer hight Bogorm converſation 15:09, 2 July 2010 (UTC)

Result[edit]

It's done now by Lua in Module:ru-translit. Ignatus (talk) 21:07, 6 March 2013 (UTC)

ISO 9[edit]

Could some explain to me, why you (we? they?) decided to invent your (our/their) own rules instead of using the ISO 9 standard, which sets out a uniform transliteration for all cyrillic alphabets? VZakharov 09:51, 14 March 2011 (UTC)

We had numerous calls for suggestions, comments and votes over the past eight years and this is what we wound up with. Some of the languages that use Cyrillic are also written in Roman, and we would prefer to use the standardized Roman spelling as the transliteration in those languages. In many languages, letters have a different phonetic value from Russian or Bulgarian. Anyway, numerous calls for opinions, suggestions, and preferences on this subject over the years and at some point we had to say that the way we have decided is the way we will do it. —Stephen (Talk) 02:20, 15 March 2011 (UTC)
That doesn't really offer a “why,” Stephen.
I believe we use standard linguistic transliteration systems for almost all languages. Using this mixed-up, made-up “system” for Russian is unhelpful for Wiktionary's readers. Please feel free to propose a new vote about this any time. Michael Z. 2011-04-03 01:05 z
Language specific policies are created and adhered to by contributors who actually work in those languages, Russian entries or translations into Russian, in this case. We only have a small number of regular contributors who work with Russian and we are happy with the current method. If you look at our Arabic transliteration - it's way off standard transliteration but it's created by Arabic contributors. If a vote is created, the contributors of Russian should have the weight. --Anatoli 08:43, 3 April 2011 (UTC)

ɛ[edit]

The addition of ɛ seems to be purely about pronunciation. This information is already conveyed in the pronunciation section of an entry. Isn't this merely complicating the transliteration system without furthering the aims of transliteration?

If we wanted the transliteration to also convey pronunciation, then we'd have to add about a dozen other rules and exceptions, and it would no longer be transliterating at all. Michael Z. 2011-04-15 19:35 z

We do have a number of exceptions in transliterations: что, чтобы, конечно use š, not č to show the actual pronunciation e.g. конечно (konéšno), if г is pronounced as /v/ or /x/, we show it: ничего (ničevó), лёгкий (ljóxkij). The rest can be derived from the basic knowledge of Cyrillic and the Russian pronunciation. It is meant to meant to be helpful to the average user, remaining consistent at the same time. One can also have: конечно (konéčno) (read: konéšno), which is also fine, as long as it helps users to learn the pronunciation. Just having "konéčno" is inaccurate. It's not unlike the transliteration of the Japanese particle , which normally is pronounced "ha" but when it is a particle, it's "wa".
As for ɛ, the transliteration when we transliterate "konéšno", it is assumed that /n/ is palatalised, as we don't use "kon'éšno" like in the vast majority of cases. интернет (intɛrnɛ́t) would make it clear that both /t/ and /n/ are not palatalised. The alternative would be harder - adding n'e, t'e, s'e, etc. where it's palatalised and ne, te, se, etc. where it's not, e.g. kon'éšno/internét.
Although Russian is a phonetic language, word stress and a few exceptions, like with the readings of ч, г and е are hurdles. Providing IPA for each word is cumbersome, not everyone can read it and we don't use IPA in translations. It's not a new addition, I only added what other contributors have been using for a long time, if you remember our discussion on Stephen Brown's talk page. --Anatoli 01:42, 16 April 2011 (UTC)
By that logic молоко should be transliterated malakó. I disagree with your and Stephen's opinion that we should show pronunciation in transliteration. That information belongs under the ===Pronunciation=== header. --Vahag 09:38, 16 April 2011 (UTC)
By the way, it's very common practice in Russian dictionaries to show pronunciation only in cases of notable exceptions, and only in parts where the exception takes place, like: неизвестный [-сн-], so current transliteration practice may be (at a little stretch) said to follow this scheme - non-exception parts are spelled orthographically, exception parts are spelled phonemically. See also another place, where this на колу мочало discussion took place:Strange romanization rules --Panda34 12:06, 16 April 2011 (UTC)
Vahagn, we don't transliterate malakó because it follows the standard rules of reading Russian. Only exceptions can be transliterated differently. --Anatoli 01:25, 17 April 2011 (UTC)
You mean exceptions in pronunciation, right? The nature of the exception is that the pronunciation doesn't match the spelling. But if you transcribe the pronunciation in the “transliteration” section, then you are obscuring the nature of the exception for anyone who doesn't read Cyrillic letters. You seem to be missing the point of transliteration altogether. Michael Z. 2012-03-05 07:37 z

Indeed, using a Greek letter for a Roman transliteration is just wrong and very confusing. Though Roman is the most common script for Cyrillic to be transliterated into, transliteration can be done into most any script, including Greek. Greek letters should be reserved for Greek transliteration, and Roman letters for Roman transliteration. This seems to be a transfer over from IPA phoneticization, which should be kept separate from transliteration. Nicole Sharp (talk) 15:53, 14 April 2013 (UTC)

Agreed, surely there has to be something better than ɛ, it's very confusing, what am I supposed to interpret it as? Mglovesfun (talk) 16:02, 14 April 2013 (UTC)
ɛ is not Greek, it is from IPA. The Greek letter you are thinking of is ε...we don’t use any Greek in the Russian transliteration. It is common practice to use ɛ for this purpose in Russian transliteration. It’s important to show that the preceding consonate is not being palatalized, and e would indicate palatalization. If you are a student of Russian, it is an important consideration and the use of ɛ is well recognized and understood by those who study Russian. —Stephen (Talk) 16:57, 14 April 2013 (UTC)
Greek letters used in the IPA are supposed to have their letter shapes harmonized with the Latin alphabet (I don’t know if every font designer got the memo).
But the Unicode latin small letter open e is not a normal Latin letter, nor is it used outside of IPA. In romanization, it’s bad for readers to use unfamiliar characters that they don’t know how to pronounce, it’s bad for editors to use difficult-to type characters, and it’s bad practice to add open-ended romanization rules that only very literate native-language readers can apply or interpret. It’s also bad to turn romanization into phonetic transcription, and to make a complicated, non-standard system with many exceptions and unexplained ambiguities. This is according to all the professionalsMichael Z. 2013-04-14 18:07 z
Also problematic to choose letters lacking capitalized forms, because then you can’t precisely transliterate citations in initial caps or all caps. Michael Z. 2013-04-14 18:12 z
As I commented below in #ISO 9 redux, writing phonetically and transliterating are two very different things. We already have IPA available to write pronunciations; transliterations are about scripts, not sounds. Someone reading a Romanization should already know what Roman letters correspond to what Cyrillic letters, and what those Cyrillic letters sound like. Having a transliteration makes it easier for people who are not native to the local alphabet to read (e.g. I can read Cyrillic, but about 2-3 times as slow as when reading Roman letters), but it should not be used as a replacement for how to pronounce the words either with IPA or other phonetic writing systems. The Wiktionary system seems to be trying to merge a phonetic writing system with a transliteration system, which is very problematic since they are two different things. Nicole Sharp (talk) 22:18, 14 April 2013 (UTC)
I’m 100% with you regarding transliteration vs. transcription. However, there are some pronunciation issues in romanization for a general, multilingual dictionary. 1. transliterations should reflect the source language’s phonology as well as possible. So Russian год is transliterated god, not plb or 239. 2. transliterations should be comprehendible and pronounceable in the target language. So год is not πνδ or ★♦✎. Also, 3. a multi-language dictionary should use the same principles in related languages, if not the same system.
An English-language reader should be able to compare the spelling and phonology of Russian горох = goroх vs Ukrainian горох = horox, Ukrainian ґанох = ganok vs. гаркати = harkaty vs. харкати = xarkaty. Michael Z. 2013-04-15 03:11 z
You can continue suggesting and criticising the page but the transliteration method we use - phonetic method showing exceptions is not going to change much. This is the method agreed by people who actually work and use Russian, also by Russians, over years. Nobody likes everything, you don't have to like the system. Transliterating letter-to-letter doesn't help anyone but showing how to pronounce words when a basic knowledge of the Russian phonology and alphabet doesn't help, is important. Russian transliterates English words phonetically or semi-phonetically - "трейдер" (trader), "таймер" (timer), "Брайтон" (Brighton). We use stress marks even if the words in Cyrillic do not have an acute accent - "слово" - slóvo. I have already used Japanese, Korean, Arabic, Hebrew, Hindi, Greek transliteration methods as examples, some have much less letter-to-letter correspondence than Russian - silent or unwritten letters, sounds that change their pronunciation depending on the position, exceptions - words pronounced differently if they have a different meaning. Other Slavic Cyrillic-based languages can't be used as examples, as they are more phonetic Russian and don't have exceptions but Mongolian could use phonetic transliteration when words are not pronounced the way they are written. --Anatoli (обсудить/вклад) 03:38, 15 April 2013 (UTC)
Getting back to the topic, where are all of the people who have been romanizing Russian with ɛ for all of these years? Michael Z. 2013-04-15 05:55 z
Some are still here. I can ask you again "where were you?". Specifically about "ɛ" - the common practice has become a standard. The alternative was transliterating all instances of Cyrillic "е" after consonants (except for letters ж, ш, щ, ч and ц) as "je" (like Belarusian) or "'e" and as "e" when consonants are not palatalised. We would end up with "tjeljegrámma"/"t'el'egrámma" (телеграмма), "pjerjestrójka"/"p'er'estrójka" (перестройка) but "brend" (бренд). Loanwords where consonants are not palatalised are not as common, so "telegrámma" (телеграмма), "perestrójka" (перестройка) but "brɛnd" (бренд) was adopted, so that palatalisation is not marked in any way but marking some words with "ɛ" where it doesn't occur. The transliteration is practical and is based on the needs and the common practice, not to cater for ISO's or people coming from outside and imposing their views. --Anatoli (обсудить/вклад) 23:06, 15 April 2013 (UTC)
Where was I when? I was the first to discuss this question right here, after you changed the system without comment.[1]
What is the standard you mentioned? Michael Z. 2013-04-15 23:14 z
I know you're watching closely this page. I meant, where were you when the symbol was used in translations and entries when it has become a common practice with editors working with Russian? The symbol wasn't my invention but I adopted the usage as well, I discussed the usage with Stephen Brown who was very active with Russian before adding it here. --Anatoli (обсудить/вклад) 23:31, 15 April 2013 (UTC)
So the use of the epsilon is a convention on en.Wiktionary, but not in any standards? Michael Z. 2013-04-16 03:08 z
The page doesn't claim to be a standard elsewhere. The discussion from 2010 about "ɛ" was copied to this talk page - I spotted it started using a while before it was discussed and added here. For demonstration, some pairs of words, which have palatalised and unpalatalised consonants in front of "е", all of them are loanwords and how to pronounce them is completely unpredictable. If even "ɛ" is confusing for someone per discussion, what's the use of IPA?
  1. текст (tekst), тембр (tɛmbr) (i.e. IPA(key): /tʲekst/ and IPA(key): /tɛmbr/)
  2. коррекция (korrékcija), тенденция (tɛndɛ́ncija)
  3. патент (patént), тент (tɛnt). --Anatoli (обсудить/вклад) 04:15, 16 April 2013 (UTC)
I have been doing ɛ for this purpose since I first started, which was before I even had registered a user name (about nine years ago). It was in common use in Russian-language study materials ever since I first began studying Russian in the early 1960s. It is used in Romanov’s Russian-English Dictionary...Smirnitsky’s Russian-English uses э for this purpose: тент (tэnt). It is an old and well-known standard. There is nothing confusing about it.
Anatoli, you have the patience of Job. I don’t know how you manage to keep answering this nonsense. I place Mzajac’s theories on Russian transliteration in the same bin as KYPark’s ideas about Korean/Indo-European genetic relationship and Korean and Indo-European etymologies. —Stephen (Talk) 11:18, 16 April 2013 (UTC)
Stephen, neither the Smirnitsky 1962 nor Romanov 1964 that I have seen have any transliteration at all. Perhaps you are mistaking their IPA pronunciations for transliterations? Michael Z. 2013-04-17 18:05 z

ʹ for ь[edit]

Symbol ʹ seems a far better choice than or '. A transliterated word can selected with a double-click - try double-clicking on bjulleténʹ (бюллетень). Although ' (single quote) is available on any keyboard but it may cause problems with pairs of single quotes used for bold and italic (e.g. try showing the symbol in boldface on its own) and it's not a supported symbol. Are there any objections in replacing with ʹ? If yes, perhaps need to do the same for Ukrainian (WT:UK TR) and Belarusian (WT:BE TR). --Anatoli (обсудить) 04:54, 29 February 2012 (UTC)

If you do this, please also change the transliteration of the hard sign to ʺ. -- Liliana 05:38, 29 February 2012 (UTC)
Sounds good to me. For the moment, will just add it as an alternative. подъезд would be transliterated as "podʺjézd". --Anatoli (обсудить) 05:45, 29 February 2012 (UTC)
There's also ʼ – modifier letter apostrophe, which works as well as the modifier letter prime. Try selecting bjulleténʼ.
Most, but not all, of the technical transliteration standards specify the primes. But most professional typesetters use the apostrophes, and so readers are probably more familiar with them. Michael Z. 2012-03-05 07:26 z

Tools[edit]

Are there any convenient tools for Russian transliteration? It's tedious to go here and copy-and-paste the letters for ш, ь, ъ, щ etc. There should be some tool, either automatic or with screen-buttons on which you could click with a mouse to print out the transliterated words character-after-character, and then copy and paste it as a whole. --CopperKettle (talk) 02:57, 2 July 2012 (UTC)

When you’re in the editing mode, there is an "edittools" section down at the bottom. You can set it to "Edittools Latin/Roman" and you will find š,č,ž there. If you are using the Firefox browser, you could add the add-on called "Character Palette 0.4.1". You can add the special characters you need to "Character Palette 0.4.1", which will appear at the top of your screen. —Stephen (Talk) 03:13, 2 July 2012 (UTC)
Thanks, Stephen! --CopperKettle (talk) 03:19, 2 July 2012 (UTC)
Some more tricks I sometimes use - use Google translate and translate Russian into Korean or Macedonian but the source should be English. It doesn't work for letter "щ", though but others work OK, need to replace я, ю with ja, ju. Also, the Czech keyboard has all the letters required, including the stressed vowels but you need to know the layout. I will check the character palette plug-in, never used it before. --Anatoli (обсудить) 02:21, 3 July 2012 (UTC)
In Windows XP, you could use a screen keyboard: it shows the Czech (or what ever) keyboard in a separate window. Clicking a key behaves like pressing that key.
In Linux (X Windowing System), you could modify your existing keyboard layout. For example: you could put č on RightAlt+c.
In Firefox, you could try the add-on "abcTajpu" (a collection of alphabets) or the add-on "Clippings" (good for inserting predefined text fragments). In both cases, one can insert the characters via the context menu of any edit field.
In Windows, you could try Autohotkey to insert (for example) č when you press a special key.
Normally, I use the modified keyboard layout. If this is an option for you, I could explain this more detailled. --MaEr (talk) 16:56, 3 July 2012 (UTC)

Transliteration rules[edit]

Hi!

I am a native Russian speaker and I has been recently reverted when I was trying to correct the Russian transliteration. I changed the word zemlja to z'eml'a because there is no 'j' sound in the word, but I was pointed to this page which due to unknown to me reason tells that 'я' should be always transliterated 'ja', even where it represents only 'а' sound, while 'e' is transliterated depending on its position. This approach seems very confusing for me. For example, words белья and the writer's surname Беляев would be transliterated with 'ja', while in fact the vowel 'я' in the second word represents only one sound and not two. This makes correct reading of such words impossible and also disallows to reconstruct the original spelling. I suggest to change the rule and indicate the softness with apostrophe character as it is adopted in Russian national transcription system.--Anixx1 (talk) 07:58, 12 October 2012 (UTC)

In my understanding, transliteration means substituting characters by other characters. In this case, Cyrillic ones by Latin ones. So it has nothing to do with the pronunciation. Example: I would transliterate "сегодня" ("today") as "segodnja" because "Г" is "G" (not "D"). I would not write "sivodn'a" or whatever pronunciation one assumes.
A correct transliteration does allow the reconstruction of the Cyrillic word in nearly all cases. From "segodnja" you will get to "сегодня" (which is correct). But "sevodnja", "sivodn'a" would give "севодня", "сиводня" or similar (which is wrong).
The system on the project page, however, is a mixture of Latin and Greek characters that indicate both Cyrillic characters and Russian pronunciation. —This unsigned comment was added by MaEr (talkcontribs).
As I already pointed out the current system does not allow to reconstruct the original spelling like in белья (of underwear) vs. беляево (a metro station named after the writer) with both transliterated with the same 'ja' according the current rules. Also the system is inconsistent. Why it transliterates е depending on the position and pronunciation as 'je' or 'e', bot in the conpletely similar case of the letter 'я' it always requires to transliterate as 'ja' independently of whether it represents one or two sounds ('a' or 'ja') in Russian?--Anixx1 (talk) 22:49, 12 October 2012 (UTC)
It's just the way it is. There is no perfect transliteration. Please follow our rules. --Anatoli (обсудить/вклад) 23:56, 12 October 2012 (UTC)
@Anixx1Anixx1: I will just say a word about this, but I don’t want to get into an argument about it. First, reconstruction of original spelling is a concept that used to be important, because prior to the computer age, few people or companies outside of the Cyrillic-using areas had the capability of typing or typesetting Cyrillic and so it was important to have a system whereby the Roman alphabet could serve as a container for the Cyrillic orthography, so the transliterations could evenually be reconstructed. It is no longer of any value. If you have a word such as человечество, I can transliterate it any way at all here on Wiktionary (for example, chelovechestvo, tschilawetschistva, čelav'ečistva, or any other way, standard or otherwise) and it is easy for anyone to retrieve the original spelling in Cyrillic: simply highlight the word человечество, press Ctrl-c, move the cursor to the position where you want the original spelling to appear, and press Ctrl-v: человечество. Magic! We always provide the original spelling.
Next, a translit. system such as "b'el'ja/b'el'ajevo" is natural and simple for native Russians, but it is not natural for Americans or British. The English Wiktionary is not here for the use of the native Russians, they have the Russian Wiktionary. English Wiktionary is designed for native English-speakers. Besides the bad fit of a system that writes "b'el'ajevo" (which an American would analyze as бэлаево, since we would ignore the apostrophes), the whole concept of hard and soft consonants is extremely problematic for native English-speakers, and few of us are able to learn to discern the difference between "bje" and "b'e"...to American ears, they sound identical (let me say that I know the difference, since I have spoken Russian for most of my life, but most English-speakers never master the Russian pronunciation).
If Native speakers will wrongly read the word as "бэлаево" it would not be worse than reading it as "бельяево" - the both readings are wrong, so what is your objection exactly? I personally would prefer "бэлаево" though.--Anixx1 (talk) 01:40, 16 October 2012 (UTC)
And next, the way we transliterate "e" and "je" is very consistent for us, even though it might seem confusing for a native Russian. We treat the Russian letter е differently from the other soft vowels such as я because я is relatively rare, while е is extremely common. In transliterating Russian, we don’t have to write "ja" very often, but if we always transliterated е as "je", the page would be filled with "je" everywhere and the text would be obscured by all of the j’s: abjevjega, pjevjets, pjelʹmjenʹ, pjensionjer, pjerjebjezhchik, pjerjednjeje bronjestjeklo, pjerjepjelinaja boljeznʹ, pjerjesmjeshnik, pjerjeshjejek, pjeriodichjeskaja sistjema eljemjentov. This is unacceptable. Instead, we transliterate я as ja, but е has an implied softening "j" so that we don’t write the "j" except in a few places to differentiate it from э: etot, jesli.
And finally, we realize that this system is designed for native English-speakers and that it is confusing and unnatural for native Russians. Some of our native Russian contributors do in fact transliterate like "b'el'ja/b'el'ajevo", and we tolerate this because we understand that it is confusing for you. Gradually, over the years, we will find and fix all the cases of "b'el'ja/b'el'ajevo" by changing them to "bel’ja/beljajevo". So we understand that you, as a native Russian-speaker, have difficulty with our transliteration and we will tolerate the way you need to do it. The one thing we ask is this: please do not change any extant transliterations from "bel’ja/beljajevo" to "b'el'ja/b'el'ajevo".
This is all I’m going to say on this matter because I am currently engaged in languages other than Russian, and I am not doing any Russian editing these days. I hope this helps, but I expect it will just generate more questions which I don’t have time to answer. I hope someone else can field any further questions on the matter. —Stephen (Talk) 00:09, 13 October 2012 (UTC)
Well. Since your transliteration does not follow any established standard, can I suggest to change something? What about transliterating it as 'beleajevo', 'Feodor' etc? This looks to me better than 'beljajevo', 'Fjodor' because the letter suggests iotification which is not present in these words. Also if we take the historic development we will find that in fact Russian name Фёдор was borrowed from Greek Theodor, so that the ё is just Russian way to write 'eo'.--Anixx1 (talk) 12:55, 13 October 2012 (UTC)
Note also that since Russian has no sequences "ea" and "eo", these sequence can be easily back-transliterated to "я" and "ё" while writing "ja" independently of whether there is iotation or not creates confusion whether it was "ья" or simply "я" in the original. So I suggest writing "ja" and "jo" where there is iotation in Russian (if я or ё are after vowels, soft and hard signs and in the word beginning) and write "ea" and "eo" in other cases.--Anixx1 (talk) 13:04, 13 October 2012 (UTC)

It seems, there are very different ideas about transliteration. According to me, it should represent the Cyrillic word in Latin characters. According to others, it should represent the pronunciation. A third idea is, that it should be as it is now. According to Anixx1, it should represent both pronunciation and etymology.
I would prefer to use IPA (or something similar) to represent the pronunciation, and keep the pronunciation and etymology out of the transliteration. --MaEr (talk) 13:11, 13 October 2012 (UTC)]

As I already noted, if you substitute 'ja' for 'я' everywhere, you cannot given the "ja" guess what letters were in the original, "ья", "я" or "йа". So it does not represent Cyrillic word in latin characters.--Anixx1 (talk) 22:53, 13 October 2012 (UTC)

Transliteration rules change[edit]

So if there are no substantiated objections, I will go on and change the transliteration rules. I am contributing and will contribute in the area of Russian language so I want the transliteration to be closer to the original's pronunciation and also better allowing the reverse transliteration, which the current rules do not allow at all. The proposal is to use 'ea' for 'я' and 'eo' for 'ё' when not iotized (beginning of the words, after vowels and after soft and hard signs), and 'ja', 'jo' otherwise.--Anixx1 (talk) 01:24, 16 October 2012 (UTC)

Don’t do that! It is the stupidest idea for a transliteration system that I have ever heard of. If you can’t follow the rules, then don’t contribute. —Stephen (Talk) 01:28, 16 October 2012 (UTC)
What is stupid with it? What you don't like? Actually it is the current rules that are stupid and I already described, why.--Anixx1 (talk) 01:32, 16 October 2012 (UTC)
I have serious objections as well, besides from a person who comes just to challenge or change the rules. Follow the rules or be blocked. I don't see the need to explain details yet to a person who hasn't contributed much. --Anatoli (обсудить/вклад) 01:35, 16 October 2012 (UTC)
This is is a clear violation of WP:OWNING besides it is a personal attack.--Anixx1 (talk) 01:42, 16 October 2012 (UTC)
And again, I did not come here just to change the rules. You can see that from my contributions from my another account: [2]. I changed the name because it seems the previous name is already occupied in Wikipedia. It seems that your insults against me is a violation of Wikipedia:Assume good faith and WP:CIVILITY. I still expect you to comment constructively.--Anixx1 (talk) 01:51, 16 October 2012 (UTC)
No, I have not insulted, I have just warned you. We have many thousand entries and translations, which use the system, changed over the time and tweaked a little, agreed on by contributors who were/are actively working with Russian, comparable with transliteration system of other Slavic, Cyrillic-based languages. Editors who cause more trouble than add value, who make point, not difference get blocked as well, just because we don't have a big team of contributors willing to endlessly discuss the obvious and the lack of efficient methods of resolving disputes. Here's the deal: you may contribute WITHOUT any transliteration if you dislike it (we have some contributors who refuse to add translit.), if you edit without following the rules, your edits will be corrected or reversed. I find your suggestion ridiculous but as I said, no system is perfect. Systems have their advantages and drawbacks. There is no universally accepted translit. system for Russian that fits Wiktionary. The system, which is USED, understood and consistent is the best. It's the method used currently at Wiktionary, please follow or don't edit. Sorry if it sounds harsh and thank you for understanding. --Anatoli (обсудить/вклад) 02:05, 16 October 2012 (UTC)
@Anixx1: This is not Wikipedia. See Wiktionary:What Wiktionary is not and Wiktionary:Wiktionary for Wikipedians. —Stephen (Talk) 02:09, 16 October 2012 (UTC)
@Atitarev I think I already showed that the system you currently use in NOT consistent so I propose an adjustment. What is your objection against this particular adjustment? Note that I do not suggest to change the system entirely, I want to improve the current rules. From this talk page I see that other users also explained dislike regarding your attempts to edit this page unilaterally. I expect a civil response rather than personal attacks.--Anixx1 (talk) 02:54, 16 October 2012 (UTC)

Clarification needed[edit]

The description of this system is not at all clear.

. . . the following exceptions to show the pronunciation are allowed

I presume “allowed” mean that these rules are all optional. Is that correct?

e.g., -ого and -его

If this rule is optional, what factors affect the choice? Can anyone show examples where it is correct to apply it, and where it is correct to not apply it?

The few words where the letter ч is pronounced. . .

Is this a closed set of words? If so, where can we find a list of them? Is this rule also optional? Factors? Pro- and counter-examples?

The few words where the letter ч is pronounced. . .

Same questions as above.

Combinations жю and шю, e.g., in жюри or брошюра, can be transliterated as either žjurí/brošjúra or žurí/brošúra where ю doesn't produce the usual pronunciation.

The sentence isn’t clear. Is this a correct interpretation?:

  1. only when жю is pronounced /ʒu/ and not /ʒju/, it may at an editor’s discretion be transliterated either according to the table žju, or alternatively as žu
  2. Likewise, only when шю is pronounced /ʃu/ and not /ʃju/, it may at an editor’s discretion be transliterated either according to the table as žju, or alternatively as žu

Again: factors? Examples? Michael Z. 2013-03-15 00:40 z

ISO 9 redux[edit]

I was referred to this page from User talk:Nicole Sharp#WT:RU TR and after going through the discussions, see a number of problems. It seems as if the Wiktionary transliteration scheme wants to have a universal Cyrillic transliteration, which can be problematic. Also there seems to be a desire to have transliteration reflect pronunciation, which is even more problematic. Transliteration in another alphabet is not the same as writing the word phonetically. We already have the International Phonetic Alphabet to write pronunciations, that does not also need to be reflected in the transliteration and trying to do so means you will quickly leave an easy one-to-one correspondence scheme that Roman readers can easily recognize as their equivalent Cyrillic letters. This is even worse if you are trying to create a single scheme for multiple languages. Most nations already have official transliteration schemes, I would strongly suggest using the official scheme for each Cyrillic-written language individually (ISO 9:1995 is official in the Russian Federation and the Commonwealth of Independent States), and the same (international and open-source) ISO 9:1995 standard for modern languages which do not have their own official or preferred system. Historical languages may diverge from that, but ISO 9 or scientific transliteration should still apply. Creating a new transliteration scheme just for Wiktionary is confusing and does not reflect how the language is actually represented outside of this particular site. Nicole Sharp (talk) 15:43, 14 April 2013 (UTC)

Yes, and also failing all other known criteria for romanization.
But ISO 9 is not the best choice because it imposes Russian phonology on other languages. It's great for international document retrieval, but the traditional scientific romanization is more suitable for lexicography. That said, we should also consider additional romanizations that are widely used, e.g., for place names, etc. Michael Z. 2013-04-14 16:46 z
I agree we should avoid "Russification" at all costs, had enough of that during the Cold War. This talkpage is only for Russian, but I would suggest using official standards adopted by the nations speaking each language (e.g. Russian standards for Russian, Belarusian standards for Belarusian, etc.). Most nations use ISO 9, but place names and personal names often use different standards for transliteration for historical reasons. I think transliteration according to official local standards would be more beneficial than a universal but more difficult to learn and use standard. Whatever language the author is working in needs to be transliterated according to that language's own schema for that particular context (e.g. standard, geographic, etc.). Also, I don't see any reason why more than one transliteration can be given if a particular word is commonly transliterated different ways, with a notation to what standard each transliteration is in. E.g. Mandarin Chinese words still often use both Pinyin and Wade-Giles, since Wade-Giles is important for historical reference, despite being unofficial in Mandarin. Nicole Sharp (talk) 22:06, 14 April 2013 (UTC)
A reader’s choice of transliteration methods would be good, if it leads to relatively more standardization than improvization. Michael Z. 2013-04-15 23:22 z
@Nicole Sharp and @Mzajac. There's no point on imposing your views if you are not going to work with the Russian language and your views are not going to be followed. The current document is for practical transliteration, which includes phonetic elements for exceptions in the pronunciation. @Nicole Sharp, please leave politics outside of this project and refrain from Russophobic remarks. --Anatoli (обсудить/вклад) 23:42, 15 April 2013 (UTC)
I didn't mean to state anything political or Russophobic. Nearly every language has some sort of problem of a more dominant language interfering with it, e.g. orddelingsfeil (word split error) in Norwegian or German from the popularity of English. National languages should be able to determine their own standards for phonology and transliteration. For many Cyrillic languages, that is the ISO 9 standard by international agreement. Unfortunately linguistic definitions are often politically determined, e.g. Montenegrin and Kosovan now being considered their own languages due to their independence, whereas the Sinitic languages are officially designated as dialects of Standard Mandarin. Nicole Sharp (talk) 23:54, 17 April 2013 (UTC)
Anatoli, you may as well stop telling me that I’m unwelcome. Someone has to coordinate and improve standardization and openness in things like language codes, CSS, text encoding and multilingual romanization in this multilingual dictionary, and I will to do so. I also try to help this dictionary serve the practical needs of our readership, not just the desires of individual editors. There’s very little “practical” for anyone about this very complicated document, but I hope to make what it is as clear as possible anyway. Michael Z. 2013-04-18 01:31 z

ё = Jó, jó/ó[edit]

I see new changes.[3] I assume чё = čó only when we are indicating syllabic stress. So when we are not indicating stress, for example in an extended quotation, does чё = čo, čö, or čjoMichael Z. 2013-04-15 02:31 z

Module:ru-translit has been changed to transliterate ё as jó/ó. It's far more important to cover 99% of cases when indicating stress is important for long words and inflected forms than having to bear with a stress symbol on monosyllabic and multipart words. That's why we override the behaviour of automatic translation. Monosyllabic words can stay as jo/o as before but Module:ru-translit will transliterate "пёс" as "pjós".
Another, a small problem, is compound words like "четырёхугольник" become "četyrjóxugolʹnik" if used without an acute accent and "četyrjóxugólʹnik" with and accent and some extremely rare loanwords where "ё" is not stressed or there is more than one occurrence.
Sample invocation of the module on words with "ё": "пёс": [MODULE CALL REDACTED], "четырёхуго́льник" [MODULE CALL REDACTED]. Monosyllabic words don't cause any problems, besides stresses are put in declension tables even if there is a zero-ending, which will also solve problem with compound words.
Letter "ё" is stressed in 99% of cases, which can be considered a rule. The importance of transliterating "ё" as jó/ó by default, which is not supplied with an acute accent is demonstrated at the conjugation table like ковать (some present tense forms). --Anatoli (обсудить/вклад) 02:56, 15 April 2013 (UTC)
I don’t understand all of that, but it does not answer the question. How does the new version of the system romanize ё when stress is not being indicated? Michael Z. 2013-04-15 06:02 z
Sorry, I missed your last edit to the table. I think it would be clearer to indicate plain jo/o and not mix stress in for only one letter. Stress is mentioned separately, and that simple rule covers every case. May as well mention that ё is almost always stressed. Michael Z. 2013-04-15 06:10 z
Re: May as well mention that ё is almost always stressed. I thought The vowel “ё” is normally stressed in native Russian words made sense?
In texts designed for children and foreign learners, when the word stress is indicated with an acute accent, words with "ё" don't get it. The two dots (две точки) over "е" indicate both the pronunciation and the stress (adult Russian speakers normally replace "ё" with "е" in writing). Our templates don't normally add acute accents to words with "ё" (unless it's a rare exception as I explained above). Does that answer your question?
The issue is now that automatic transliteration module doesn't put stress on words with "ё" because the Cyrillic form doesn't have it and it's not necessary with Russian words. I can change back to "jo/o" but I'd like the module to transliterate it as "jó/ó". --Anatoli (обсудить/вклад) 22:51, 15 April 2013 (UTC)
I think I understand. I’d still suggest that showing the transliteration as “jó/ó” is confusing, because for the 36 of 37 letters stress is dealt with in a different section of the page. If you want to keep the stress marks in the table for one letter only, then you should add another note clarifying that this is stress and not transliteration, but piling on still more notes is potentially less clear. This table explains to readers and editors, and the technical details of the transliteration module don’t really matter here. Michael Z. 2013-04-15 23:20 z
How about the module transliterates ё as follows: it is (j)ó, unless the word already contains an acute accent or another ё somewhere, then it's (j)o. It could also be made to remove accents from monosyllabic words (words which only contain one vowel). —CodeCat 23:25, 15 April 2013 (UTC)
Yes, thank you that sounds very good. Do you think the document describes clearly the usage of "ё" - the two points: 1. stressed in most cases 2. no acute accent in most cases. Do I need to add "3. usually replaced with "е" (Cyrillic) in the running text."? --Anatoli (обсудить/вклад) 23:48, 15 April 2013 (UTC)
No. 3 could be relevant to someone transliterating a Russian text, so I would say it bears mentioning here. Michael Z. 2013-04-16 02:55 z

э[edit]

Why is this letter transliterated as "e"? That makes it impossible to tell the difference between "мэр" and "мер". —CodeCat 14:37, 12 October 2013 (UTC)

Very few native Russian words have э. Mainly it is used in the этот/это words, and it also appears as the first letter of some adopted words, such as эффективный. Being the initial letter, it is transliterated "e", while е would be "je". It is quite rare to find э in the interior of words, such as мэр, and those are all loan words from other languages. If the word had been borrowed with the spelling мер, it would still have the pronunciation of э. And since мер is Ukrainian, е is their э. The Ukrainian soft vowel is є. —Stephen (Talk) 16:36, 12 October 2013 (UTC)
The point is though, that the transliteration currently doesn't distinguish between those two vowels, both get transliterated as "e". That makes it ambiguous which one is meant, and thus which one is supposed to be pronounced. —CodeCat 16:42, 12 October 2013 (UTC)
мер is Ukrainian. Ukrainian е is equivalent to Russian э. мер and мэр are the same, except that one is Russian and the other is Ukrainian. Both vowels are hard e, pronounced the same, and both indicating that the preceding consonant is hard. The Russian sound ме would be written мє in Ukranian. In ме/мє, the preceding consonant is soft, but in the case of Russian, we do not indicate the soft consonant when followed by е or и, because they are too common. The other soft vowels, я, ё, ю, are relatively rare, so it is not too disruptive to indicate мя, мё, мю as mja/mjo/mju. Mja/mjo/mju are somewhat deceiving, because the j is not pronounced at all...it only shows that the consonant is palatal. If we did that with е and и, then almost every Russian syllable would have a j in it. It would be too much of a bad thing. —Stephen (Talk) 17:02, 12 October 2013 (UTC)
э is mostly used as an initial vowel, mainly appearing in the этот/это words. The transliteration is different from that of initial е, because е is "je": ещё (ješčó) versus этот (etot). —Stephen (Talk) 17:06, 12 October 2013 (UTC)
In words where е is not palatal, we use ɛ. Why don't we do that for э also? мэр should be mɛr to indicate that it's different from мер (which would be transliterated as mer in Russian). —CodeCat 14:59, 13 October 2013 (UTC)
ɛ is never palatalizing, so eto/be/ve/ge/de/me, etc., accurately represent it and it does not cause confusion (in practice, э has very rarely ever been used except as the initial letter: это, эффективность; only recently have a small number of words been borrowed, mostly from English, with э as an interior letter as in мэр...it happens very rarely). е is always preceded by j when е is the initial letter of a word, is preceded by ъ or ь, or is preceded by another vowel; е is almost always palatalizing when preceded directly by a palatalizable consonant, so there is no confusion by representing it as be/ve/ge/de/me, etc. The only instance of confusion and frustration is the very rare case when е is preceded immediately by a palatalizable consonant, yet the consonant is not palatalized. These are the only cases that cause trouble, and the trouble they cause is significant...and so these are the cases that need to be clarified with ɛ. The use of ɛ does not mean that it is pronounced ɛ, and it does not mean that the other cases are NOT pronounced ɛ, it only means that the preceding palatalizable consonant is NOT palatalized as expected. The э and its transliterations cause no problems; it is only the е which does not palatalize a palatalizable consonant that is a problem. Traditionally, in most paper dictionaries and more recently in online dictionaries, this problem is solved by transliterating the nonpalatalizing е with the symbol э: animэ́, intэrnэ́t, modэ́l’, bordэ́l’, otэ́l’ (сленг can be either sleng or slэng, as the speaker prefers). But here we just make one small change in this simple model...instead of transliterating a nonpalatalizing е as э, we transliterate it as ɛ. But э would work just as well, and would be easier, since it is on the Russian keyboard and does not have to be copied and pasted. But we have been using ɛ for this to avoid mixing a Cyrillic letter in with Roman letters. —Stephen (Talk) 16:43, 13 October 2013 (UTC)
Why are you talking about Ukrainian мер, instead of the Russian one? Mer ne prinjal mer - what does it stand for? Мэр не принял мер, or Мер не принял мэр? --Panda34 (talk) 16:09, 9 December 2013 (UTC)
@Stephen, you seem to have misunderstood that мер refers to the Russian word (genitive plural of мера) rather than the Ukrainian version of мэр. —This unsigned comment was added by Wikitiki89 (talkcontribs).
No, Stephen understood correctly. They were both intended as Russian words to demonstrate how they, despite being spelled and pronounced differently, are transliterated the same by our current scheme. —CodeCat 16:52, 9 December 2013 (UTC)
Are you sure? His first post in this section, especially the sentence "And since мер is Ukrainian, е is their э." makes me think otherwise. --WikiTiki89 17:03, 9 December 2013 (UTC)
I made the original post mentioning these two words, and I intended them both to be Russian. —CodeCat 17:08, 9 December 2013 (UTC)
Yes, and I think Stephen took the second to be Ukrainian or a non-existent Russian alternate spelling, when in fact there is a Russian word spelled мер (pronounced /mʲer/). --WikiTiki89 17:16, 9 December 2013 (UTC)

Final ъ[edit]

We have a few entries like Іорданъ that use older spellings. In the transliteration, the ъ has apparently been forgotten. How should this be transliterated? Should we use the normal sign ʺ here? —CodeCat 14:56, 13 October 2013 (UTC)

No, it should be Iordán. The normal sign ʺ actually carries with it a distinctive pronunciation, and when it is used after a final consonant such as Іорданъ, there is no effect whatsoever on the pronunciation. That’s why it was dropped from that position. In a word such as въехать, your ear will hear the ъ as a schwa. That is, въехать is not pronounced /ˈvɛxatʲ/, but as /vəˈjɛxatʲ/. Final ъ has no effect on pronunciation and should be ignored in transliteration. —Stephen (Talk) 17:03, 13 October 2013 (UTC)
Stephen, it's /vˈjɛxatʲ/, not /vəˈjɛxatʲ/, there is no schwa --Anatoli (обсудить/вклад) 19:19, 13 October 2013 (UTC)
It might not be a schwa, but I don’t think it is /vˈjɛxatʲ/. Perhaps the /v/ is syllabic...or maybe it’s a sound that is shorter than a schwa. I don’t know the best way to describe it in IPA, but I don’t think /vˈjɛxatʲ/ is accurate. —Stephen (Talk) 19:30, 13 October 2013 (UTC)
Letter ъ doesn't introduce any vowel in any position, even short, only /j/ between the consonant and the following vowel. If you pronounce any vowel, I'd say you have a strange accent. --Anatoli (обсудить/вклад) 20:01, 13 October 2013 (UTC)
No, not a strange accent. The consonant в terminates before the jɛ begins. There is a short gap between the в and the jɛ. I don’t see how the gap could be anything but a vowel. —Stephen (Talk) 20:10, 13 October 2013 (UTC)
Before we digress too much, we could move this discussion to another place. I'll let you judge in the following recordings:
(file)
,
(file)
. The "gap" may be used by teachers to teach students to pronounce the consonant separately but there is no stop and the consonants are pronounced in the normal length. --Anatoli (обсудить/вклад) 21:52, 13 October 2013 (UTC)
But it's a transliteration, not a phonetic transcription. We already have IPA for that. —CodeCat 18:13, 13 October 2013 (UTC)
The IPA is not so easy for a lot of people. A transliteration as Iordán makes perfect sense to me, but I would be confused by Iordánʺ and I see no benefit in it. —Stephen (Talk) 18:54, 13 October 2013 (UTC)
So you think that the transliteration module should be adapted to remove the hard sign from the end of a word? —CodeCat 18:58, 13 October 2013 (UTC)
Yes. --Anatoli (обсудить/вклад) 19:19, 13 October 2013 (UTC)
Yes, remove if it’s Russian. Not for Bulgarian or other languages where it actually has a purpose. You might ask Anatoli what he thinks, he might have a different idea. To me, Iordánʺ looks confusing and I don’t like it, but then I do not bother with these obsolete spellings. In my opinion, I don’t think we should even host the pre-reform spellings here. Every single Russian word that ends in a consonant would have an alternate spelling with a final ъ that has no purpose.
But if we’re going to allow these spellings (if anyone ever feels up to adding any of them), then this Transliteration of Titles in Pre-Reform Orthography page, although it uses a different transliteration system, agrees with me that there is no need to transliterate the final ъ. —Stephen (Talk) 20:10, 13 October 2013 (UTC)
I think we all agreed that there is no sound in the final position (if it's Russian). What Stephen means is that pre-reform spellings are not encouraged. --Anatoli (обсудить/вклад) 20:12, 13 October 2013 (UTC)
You can't "encourage" or "discourage" entries to be included in Wiktionary. Everything that passes WT:CFI, and attested pre-reform spellings do, should be included. --Vahag (talk) 20:48, 13 October 2013 (UTC)
I have no problem with that. --Anatoli (обсудить/вклад) 21:52, 13 October 2013 (UTC)

If you continue turning transliteration into phonetic transcription, then how do you intend to discuss the differences in spelling? The discussion you’re having here about historic spellings will be completely inaccessible to readers and editors who don’t know the Cyrillic alphabet.—This unsigned comment was added by Mzajac (talkcontribs).

I know, I don't agree with it either. —CodeCat 00:48, 19 October 2013 (UTC)
It’s simple. какъ is the pre-reform spelling of как (kak). The letter ъ (hard sign) was dropped, since it has no pronunciation and no effect on anything. Before the reform, words that ended in an unpalatalized consonant received the termination ъ. If a word ended in a palatalized consonant, it ended in the termination ь (soft sign). After the reform in spelling, words ending in palatalized consonants were still marked with final ь, but words that ended in a hard consonant were left unmarked, like как (kak).
On the other hand, if you tried to explain this using kak" versus kak, it would not be simpler...possibly more confusing. —Stephen (Talk) 01:17, 19 October 2013 (UTC)
It's more confusing if both какъ and как are transliterated the same way. It defeats the purpose of transliteration, which is to be a reading aid for people who are not familiar with Cyrillic. To someone who can't read Cyrillic, our entry essentially says "kak" is a pre-reform spelling of "kak". Which doesn't make much sense. —CodeCat 01:48, 19 October 2013 (UTC)
какъ and как are treated as you see in that paragraph. That’s how you discuss the differences in spelling. The purpose of transliteration was not defeated, it was not confusing. As for the entry for какъ, it won’t say that "kak is the pre-reform spelling of kak", it will show какъ (kak) in the headline, and "pre-reform spelling of как" in the definition line. —Stephen (Talk) 02:24, 19 October 2013 (UTC)
Stephen seems to be saying that because we don’t support a usable scheme to romanize Russian spellings, they can only be indicated in Cyrillic script. Of course that’s not true. In the rare instance it is needed, we can simply use the conventional w:Scientific transliteration of Cyrillic documented in Wikipedia. It’s easy to clearly indicate the differences between the spellings как (kak) and какъ (kak″).
But I am concerned that we are not serving our readers well enough because our so-called “transliteration” section offers a phonetic transcription derived by some baroque ruleset, and not a romanized spelling. A complicated, proprietary rendering of a word doesn’t serve Wiktionary’s principles of openness and accessibility. module:ru-translit is useless, because every single use must be manually checked by a native-level Russian reader who values this scheme (a very exclusive little club). The level of discourse and openness in discussions about Russian lexicography suffers, because wiktionarians are stubbornly clinging to this private research project. What is the point of giving the reader some string of latin characters that might never occur in any other publication or website? Michael Z. 2013-10-19 16:21 z

муравьиный[edit]

This word has -ьи- in it, and this is transliterated as -ʹi-. But -ье- would be transliterated as -ʹje-. Shouldn't it therefore also be -ʹji- for consistency? —CodeCat 21:11, 15 October 2013 (UTC)

/j/ is pronounced in -ьи-, so yes, it can be changed so, WT:RU TR should reflect this too. --Anatoli (обсудить/вклад) 21:59, 15 October 2013 (UTC)

Proposal for a revised/simplified transliteration scheme[edit]

The debates we've had in the past, to me, indicate that our transliteration is too complicated because it focuses on pronunciation and not on representing the written form. This makes it unpredictable and requires us to override it in many cases. I don't think that we should go out of our way to make changes to the standard pronunciation just to account for differences in pronunciation which are irregular anyway, even to a Russian speaker. Such differences should be handled in the Pronunciation section, not in transliteration. After all, we don't provide similar "pronunciation respellings" for Latin-based languages, do we? Our Lua module would be perfectly capable of generating transliterations all the time, if only we didn't have these strange rules. Compare WT:UK TR and WT:BE TR for example, which are perfectly straightforward by comparison. There's no reason Russian can't be transliterated the same way.

So I propose adopting the following rules:

  • The iotated vowels я, ё, ю are transliterated with a preceding j, which may indicate either palatalisation or the sound /j/. As in the Russian orthography, the difference between the two is indicated by a preceding hard or soft sign, which is transliterated in all cases it's written.
  • The vowels e, i are assumed to be palatalising by default so they don't get a j. Rather, the palatal/nonpalatal distinction is indicated by a different symbol, using e, i for the palatalising vowels е, и and ɛ, y for the nonpalatalising vowels э, ы. Thus, е > e even in loanwords when it is not palatalising, and -ье- > -ʹe-.
  • All pronunciation-based exceptions are eliminated. г > g regardless of pronunciation, same for ч > č.

There are two points that may still need discussing:

  • It may look "clearer" if e, i are written je, ji all the time, to be consistent with the other three iotating vowels. But this can make the transliteration look messy. An alternative is to write je, ji only after vowels and hard/soft signs, and e, i elsewhere. This is more or less what we do now, but it can be misleading in cases where е is actually to be read as nonpalatalising э. Thus, writing just e in all cases is the most neutral solution that makes the least assumptions.
  • Should the j be omitted after certain always-hard or always-soft consonants? While жа and жя are pronounced the same (as far as I can tell), they're still spelled differently, so we should indicate this difference in transliteration too: ža, žja.

CodeCat 17:22, 19 October 2013 (UTC)

While what you're proposing is better than what we have now, I don't understand why we need to reinvent the bicycle. I am for using w:Scientific transliteration of Cyrillic, without any exceptions whatsoever. --Vahag (talk) 19:55, 19 October 2013 (UTC)
I second that. Michael Z. 2013-10-19 20:03 z
That transliteration scheme doesn't fit nicely with accents. э is transliterated as è, so what happens to э́ then? è́? That looks silly. In all other respects, what I proposed is pretty much the same anyway. The only other difference is ё which becomes jo rather than ë. —CodeCat 20:41, 19 October 2013 (UTC)
I really, really hate ɛ. I see no problem in transliterating э́ as è́. Looks weird but there will be few cases anyway. --Vahag (talk) 20:48, 19 October 2013 (UTC)
We could also use ɘ (a backwards e, like э) or use an underdot ẹ. Both of those work with accents: ɘ́ ẹ́. —CodeCat 20:58, 19 October 2013 (UTC)
That's customization, again. Transliteration schemes are not playthings. We should use a common, well-documented scheme used in actual publications. --Vahag (talk) 21:04, 19 October 2013 (UTC)
Transliterations aren't exact sciences either. Things don't break if you change something, it's meant to be human readable not computer readable. —CodeCat 21:05, 19 October 2013 (UTC)
Romanization schemes are exact standards, and in fact, ours very much ought to be. This “system” has been continuously changed for over half a decade. As long as we use something that we made up to please ourselves, editors will see fit to keep “improving” it. This can’t even be glorified as original research, because there is no external basis for any of it. Romanizations in entries will never match the guideline, and readers will never twice see the same transliteration representing the same original text, nor be able to match what they see here to any other publication or use it a guide to any writing.
Enough transliturbation. We should pick an established system, and a reference describing it, and make all of the romanizations respect it. No made-up system can serve our readers. Michael Z. 2013-10-19 22:20 z
Then let's use the scientific transliteration, but amended to substitute the character è purely for technical reasons (because two accents on the same character presents readability issues; to me it looks almost the same as ě). I suppose using ë for ё, rather than jo, makes some sense too because it lets you distinguish ё from йо. —CodeCat 23:09, 19 October 2013 (UTC)

I'm strongly against any changes to the transliteration method and ignoring the exceptions. If you go ahead, then do it without me. --Anatoli (обсудить/вклад) 08:17, 20 October 2013 (UTC)

I agree with Anatoli. The current system works well and is not complicated. If you change it as you are contemplating, I’m quitting the Russian effort on this Wiktionary. —Stephen (Talk) 08:36, 20 October 2013 (UTC)
Then let's just all use our own transliteration scheme, as it's clear there's no consensus for anything. —CodeCat 13:25, 20 October 2013 (UTC)
it's not true. Own "transliteration schemes" get corrected manually and you know which category contains the exceptions, which can't be handled by the module. I see no difference from Japanese particles は and へ, which are transliterated manually as "wa" and "e", not "ha" and "he" and the way long "ō" is distinguished from "ou". It's possible to transliterate English into Cyrillic letter by letter but what's the use of such transliteration? Russian is more phonetic but deviations shouldn't be simply ignored. No one has to know the exceptions, they are added manually by using "tr=". --Anatoli (обсудить/вклад) 19:34, 20 October 2013 (UTC)
“what's the use of such transliteration?” – the ISO,[4] the United Nations,[5] the Russian government,[6] and libraries and publishers the world over[7] believe that letter-for-letter transliteration of Russian is very useful. Wiktionary is not a platform for you to advance your own theories. Michael Z. 2013-10-20 21:48 z
Can we not discuss the relative merits rationally? With defensible statements and facts?
“The current system works well”
  1. Romanization cannot be automated, because it depends on pronunciation.
  2. Romanization cannot be derived by editors, because it depends on a native-level knowledge of written and spoken Russian. It is designed to put its designers in full, sole control.
  3. Readers cannot use transliterations as a reference because it is different from all standardized transliteration schemes.
  4. It doesn’t represent Cyrillic spelling because it is derived from pronunciation.
  5. It is unstable, changing continually over the last six years. So all the tens of thousands of transliterations in entries and etymologies have to be checked again by Stephen and Anatoli after the next change is introduced.
Stephen, what aspects of it work well?
“...and is not complicated.”
  1. All other romanization systems for Russian and for other Cyrillic-alphabet languages depend on a simple table, mainly consisting of one-to-one correspondences, and perhaps one or two simple rules or exceptions. This one has multiple exceptions for pronunciation and context.
  2. All other romanization systems for Russian can be applied based on the written text only. This one requires native-level knowledge of spoken Russian and of the specific word being romanized.
Stephen, what aspects of it are simple?
Disadvantages of this system: it contravenes just about every criterion for a transliteration system established by an international standards body.
So far the only factual argument supporting the current system is along the lines of “play my way or I’ll take my ball and go home.” It is not persuasive. Michael Z. 2013-10-20 16:31 z [updated]
(after edit conflict) I agree with Michael that transliteration should be just that — transliteration, not ersatz pronunciation. I've said as much before: WT:T:ARU#солнце.
As has been said elsewhere on this page, Wiktionary's Russian content does not exist just for certain contributors, it also exists for Wiktionary's broad audience to read and use... and for that audience, I agree with those who say it would be helpful if we included pronunciation information in the pronunciation section and transliteration in the transliteration field, rather than including pronunciation information in both places and leaving out actual transliteration.
It is unfortunate that certain contributors have a "my way or the highway" attitude towards changes. Such an attitude is ultimately incompatible with the collaborative nature of Wiktionary. - -sche (discuss) 06:13, 22 November 2013 (UTC)
Okay, so now that we have finished with all the arguments for and against, who is in favour of adopting straight scientific transliteration? Michael Z. 2013-11-22 06:09 z
  • supportMichael Z. 2013-11-22 06:09 z
  • Support. - -sche (discuss) 06:13, 22 November 2013 (UTC)
  • Support. Pronunciation exceptions can be shown in another way. For example, many Russian dictionaries show non-palatalized е like this мате́ [тэ]. We can do the same in the headword line or in the Pronunciation section. --Vahag (talk) 06:35, 22 November 2013 (UTC)
  • Strong Oppose.
    If "что", "кого́", "сего́дня" are transliterated as "čto", "kogó" and "segódnja", this transliteration is wrong and useless. Show me a published dictionary, which uses such transliteration. The same way as it is wrong and useless to transliterate "English" as "Энглисх" (rather than "И́нглиш") into Russian, Japanese particle "は" as "ha". Why is it wrong to transliterate "English" as "Энглисх" but OK to distort Russian? Why Japanese, Korean, Arabic, Persian languages are allowed to have exceptions but Russian is not? Is it because Russian has only a few exceptions, so let's ignore them? If you want to learn the Cyrillic letters and the Russian alphabet, use other pages, such as Appendix:Russian alphabet, no need to transliterate letter by letter words, which are pronounced differently from expected.
    Standards for transliterations, such as ISO are used to transliterate geographical names, which seldom have exceptions in pronunciation. As a result, we have "Gorbachev", "Krushchev" type of mistransliterations, which are OK for Wikipedia or legal documents but wrong for dictionaries. Neither dictionaries, nor textbooks use such transliteration or use them as a base only. Wiktionary is a dictionary, it serves practical purposes to teach languages, provide information on the languages, not ISO and other standards.
If the proposed method is implemented, I'd prefer no transliteration at all than misleading transliteration. The mixture of transliteration and transcription is nothing new for practical purposes. It's especially important to transliterate phonetically unpredictable exceptions.
I wish Mzajac would be more zealous to add linguistic contents, such as Ukrainian vocabulary, rather than constantly attacking this project page and impose his views. It is very discouraging. --Anatoli (обсудить/вклад) 07:35, 22 November 2013 (UTC)
Anatoli, it is ingenuous to pretend that only Mzajac is a proponent of a scientific transliteration. On this page alone VZakharov, Nicole Sharp, MaEr, CodeCat, -sche and me have expressed dissatisfaction with the current custom scheme, preferred only by you and Stephen. You are in the minority. The only reason the community has not switched to a more established system, is because we respect you two a lot and we don't want to upset you. --Vahag (talk) 08:29, 22 November 2013 (UTC)
Anatoli, I don’t love harping on this all the time. I very much respect your huge contributions, but I just can’t tolerate this inconsistent non-transliteration system based on nonsensical arguments. I contribute how I can – among other things improving transliteration, which is meant to serve readers like me, and not a pair of editors. It is not fair for you to attack my contribution while ignoring my argument, refusing to acknowledge the difference between transliteration and pronunciation, or the fact that bilingual dictionaries mainly don’t use any transliteration at all.
Горбачёв → Gorbačëv and Хрущёв → Xruščëv would be correct scientific transliterations, not the ones you pulled out of your hayloft. Michael Z. 2013-11-22 18:17 z
  • Strong Oppose. No transliteration at all would be better than this proposal. The only reason anyone supports it at all is because its name contains the magic word "scientific." Their second choice, obviously, would be any system whose name contained the magic word "scholarly." Ooooooh, how wonderful! —Stephen (Talk) 08:08, 22 November 2013 (UTC)
    Reading the exchange between Anatoli and Vahag a few lines up, I'm struck by the degree to which you and Anatoli are debating straw men. I suppose I can't speak for the others, but I supported the principle of transliteration being transliteration (and said as much on WT:T:ARU) long before I ever heard that principle described as "scientific". - -sche (discuss) 09:08, 22 November 2013 (UTC)
  • Support. I don't really understand the arguments for opposing. It should be kogó because that's how it's actually spelled. We don't "transliterate" Latin-script languages to reflect their real pronunciation. I mean, if Russian were written in the Latin script I don't think we'd still insist on transliterating "kovó" so why does that change with the script? —CodeCat 14:09, 22 November 2013 (UTC)
    I agree with every in CodeCat's last post.
    • Support г > g and ч > č. I hurts my eyes to see kovó.
    • Undecided about iotated vowels.
    --WikiTiki89 16:19, 22 November 2013 (UTC)

Atitarev's change to the page[edit]

I really disagree with this change. It's clear from the above discussion that there are more people who oppose pronunciation-based transliterations, than there are that support it. It's clear that there is no agreement, so this should be treated as a contentious issue and changes should be discussed before they're implemented. —CodeCat 21:55, 9 December 2013 (UTC)

The page is meant to reflect our current policy, which is not to transliterate the hard sign at the end of words. You can disagree with the idea, but the change to the page was fully justified. My argument for not transliterating it is that even back then it was ignored in transliterations, and conversely the hard sign was used to transliterate into Russian any foreign word ending in a hard consonant. --WikiTiki89 22:05, 9 December 2013 (UTC)
Final "ъ" is not transliterated in the "scientific" systems as well, like Library of Congress and I wonder, which one does as there is little mention of this. it doesn't need any graphical transliteration it's like Arabic sukun, lack of anything, end of a word. --Anatoli (обсудить/вклад) 22:33, 9 December 2013 (UTC)
Just pointing out, the Library of Congress system is not "scientific". It is meant for use in a library. A scientific one would the kind used by linguists. --WikiTiki89 22:38, 9 December 2013 (UTC)