Wikipedia talk:Naming conventions (Mongolian)/Archive 01

Choices for ж

But why 'j' for 'ж' and not 'ch' or 'č' ? It is after all /ʧ/ and not /ʤ/ (at least according to Mongolian language). - Francis Tyers · 12:49, 14 February 2007 (UTC)

Ok, so I just looked again, and 'ch' is already used for 'ч', /ʧʰ/. Omniglot at least has 'ч' as /ʧ/ and 'ж' as /ʤ/ and /ʧ/ [1] I guess in this case it is ok to use 'j' (as Tajiks do). But it's still wierd why our Mongolian language page doesn't mention /ʤ/. - Francis Tyers · 12:55, 14 February 2007 (UTC)

The actual choice for ж is between J, č, and Zh. č is is only comprehensible for linguists and Czech language speakers, so I'm ruling it out. And as far as I know the pronounciation in Mongolian is with more voice than in Russian, which would justify the J instead of Zh. As you may have noticed, I'm not a linguists, so this may or may not make sense... --Latebird 13:40, 14 February 2007 (UTC)

Actually iirc 'č' is quite widely used as /ʧ/ in transliteration into Latin for other languages, e.g. Serbo-Croatian, Macedonian, Bulgarian, Russian etc. I agree with you that we should use 'j'/'J' for the letter, but what I was wondering is why the page on Mongolian language does not mention the /ʤ/. - Francis Tyers · 13:45, 14 February 2007 (UTC)

Zh as in "Zhuulchin" was used in socialism, I think. You don't really want to write Zhargalsaikhan or Terelzh, do you? Yaan 13:49, 14 February 2007 (UTC)

I don't know, do we? How would they be written in Mongolian? - Francis Tyers · 13:54, 14 February 2007 (UTC)

I'm afraid I can't speak for the authors of Mongolian language. As a non-scientific addendum: J is simply the most commonly used replacement for the Mongolian ж, inside and outside of en:WP. That may serve as yet another reason for us to agree on it. ;-) --Latebird 13:59, 14 February 2007 (UTC)

/ʧ/ is usually written as "ch" or "č" only when it is in opposition to /ʤ/. When /ʧ/ is in oppostion to /ʧʰ/, it is often treated as equivalent to /ʤ/ and there written "j" (or "dj", possibly "zh").—Nat Krause^{(Talk!·What have I done?)} 19:23, 18 February 2007 (UTC)

IPA

Also, it would be good to have the IPA in the romanisation table. - Francis Tyers · 12:56, 14 February 2007 (UTC)

I don't think that listing IPA would be of much use in a WP-internal romanization guideline. That rather belongs into article space. --Latebird 13:40, 14 February 2007 (UTC)

I think it would be useful to describe on what we are basing this. - Francis Tyers · 13:53, 14 February 2007 (UTC)

As the proposal says, it's based on a compromise between BGN/PCGN and Wikipedia:Romanization_of_Russian. I don't think the guideline needs to specify the linguistic rationale for each indivual character. We can do that here in the discussion for the non-trivial cases. -Latebird 14:03, 14 February 2007 (UTC)

BGN/PCGN 1964

Romanization system BGN/PCGN 1964 (recommended in WP http://en.wikipedia.org/wiki/BGN/PCGN_romanization) exists. Why you, Latebird, are ignoring it? This system is in use on American maps, on Mongolian maps too. Why you don't put recommended system http://www.eki.ee/wgrs/v1_3/rom2_mn.pdf on this project page? The main task for this project is to help WP user orientation not only in WP texts, but in information (romanized) coming from official, private, mapping sources, from Google Map/Earth, Expedia, MSN maps etc. So we can make some possible steps:

1) to analyse present situation: to find what romanization systems are really in use, and where (offical documents in Mongolia, other countries, in business and/or private use, on maps etc.)

2) to analyse possible recommendations:

a. To make a one choice among existing systems

b. Select two or more systems, which will be used in different areas (one for toponims, other one for other words etc)

c. To criticize existing systems, but does our critics will change present situation (does WP can have an influence on BGN/PCGN or UNGEGN or Mongolian and foreign institutions?)

d. To create own new system, but don't you afraid - we are not linguists at all? Will one more system introduce order but not increase chaos?

To my opinion we have two widely used systems: BGN/PCGN 1964 (on maps) and its modification (other areas):

Mongolian spelling	English BGN/PCGN 1964 romanization	English BGN/PCGN 1964 modification (if present)
А (а)	A (a)
Б (б)	B (b)
В (б)	V (v)
Г (г)	G (g)
Д (д)	D (d)
Е (е)	Ye (ye)
Ё (ё)	Yo (yo)
Ж (ж)	J (j)
З (з)	Z (z)
И (и)	I (i)
Й (й)	Y (y )	I (i)
К (к)	K (k)
Л (л)	L (l)
М (м)	M (m)
Н (н)	N (n)
О (о)	O (o)
Ө (ө)	Ö (Ö)
П (п)	P (p)
Р (р)	R (r)
С (с)	S (s)
Т (т)	T (t)
У (у)	U ( u)
Ү (ү)	Ü (Ü)
Ф (ф)	F (f)
Х (х)	H (h)	Kh (kh)
Ц (ц)	Ts (ts)
Ч (ч)	Ch (ch)
Ш (ш)	Sh (sh)
Щ (щ)	Shch (shch)
ъ	'	(omitted)
Ы (ы)	Y (y)
ь	(ǐ)	(i)
Э (э)	E (e)
Ю (ю)	Yu (yu)
Я (я)	Ya (ya)

BGN/PCGN 1964 is in mapping use (American topomaps, Mongolian maps, in Web: Google Map/Earth, Expedia, MSN maps)

BGN/PCGN 1964 modification is most widely used in offical/commercial/private use Bogomolov.PL 15:57, 15 February 2007 (UTC)

My proposal explecitly references BGN/BCGN, so why do you think I'm "ignoring" it?

There's a large number of different romanization systems available. Even BGN/BCGN seems to exist in several variations. But I'm not sure which system your table shows, because that's clearly *not* BGN/BCGN. While there may be institutions that recomment certain systems, what people use in practise is often entirely different. We've already found that even US military maps transliterate Mongolian names very inconsistently. And using different systems for geographical and other names seems silly at best.

So I'm not quite sure what you're suggesting here. The steps that you recommend have all been discussed over more than a year, and my current proposal is the result of that discussion.

If you have specific arguments to change some details of my proposal, please say so. But I don't think that going back to the basic presumptions is still a useful option at this stage of the process. --Latebird 01:03, 16 February 2007 (UTC)

I am sure, but maybe I'm in a mistake, we will make a good job if first will find and note in comparative sheet really used romanizations. It will be useful not only for romanization system finalizing (it is possible, I hope) but give WP users a tool for transliteration texts coming from different sources: official government/parliament sources (togrog, aimag, soum, bagh, khoroo, hot ail), maps etc Bogomolov.PL 17:14, 16 February 2007 (UTC)

Such a table (Mongolian.pdf) has been given very early in the first discussion on the topic. In the same discussion, I have also given the rationale for my choice to use a modified PGN/PCGN here, and for each of the modified characters. For convenience, I'll summarize the updated list again:

* Е -> ye (most common pronounciation)

* З -> z (simplify)

* Х -> kh (equivalent choice between h and kh)

* Ъ -> (omitted) (no phoneme of its own, only in Russian words)

* Ы -> y (simplify)

* Ь -> (omitted) (no phoneme of its own)

* Ю -> yu (most common pronounciation, simplify)

The most important goal was to make it simple and easy to use. The second goal was to stay close to WP:RUS where reasonable, eg. when there were two nearly equivalent choices.

Other sources (governement use, maps, etc.) are extremely inconsistent (sometimes within the same text or map), so that we can't really use them as a reference. Most governement agencies don't really use the systems recommended by their own governement in real world documents. --Latebird 12:41, 18 February 2007 (UTC)

A few questions

1. How effectively does the modern Cyrillic script transcribe standard Mongolian? I tend to prefer a transcription over a transliteration, so if there are a lot of cases where Cyrillic implies the wrong pronunciation, we may want to take that into account.
2. Does Mongolian really have palatalised consonants? That's what Mongolian language says. If so, how are they represented in this proposal?
3. Is Ы pronounced the same as И in Mongolian? If so, perhaps they should be transliterated with the same letter.—Nat Krause^{(Talk!·What have I done?)} 19:29, 18 February 2007 (UTC)

Tricky stuff you're asking there...

1. My proposal is based on BGN/PCGN, because for the most part I believe that to best represent the actual prounciation. I'm not sure if Cyrillic implies anything different for Mongolian (it's certainly a lot closer to the contemporary pronounciation than the classical Mongolian script).
2. Not being a linguist, I have troubles answering this authoritatively. As far as I understand, omitting the soft sign means to ignore (some cases of) palatalization. I don't think this is really a problem, but I may of course be wrong.
3. И, Й, and Ы all sound very similar, but again a linguist would be required to describe the exact difference (latin vowels in both English and German represent the Mongolian vowels only in a rough approximation). I have chosen to transcribe Ы as Y because that seems to be the most common choice, and to simplify from the diacritics in BGN/BCGN. --Latebird 20:19, 18 February 2007 (UTC)

If palatalisation occurs, but Mongolians don't construe it as phonemic, then it is perhaps unnecessary to include it in a romanisation. For instance, if Mongolians treat [kar] and [kʲar] as "the same" word (or would it be [kʲar] and [kjar]?). It's unclear how examples such as "Rinchinnyamyn" are supposed to be pronounced. In Russian (so far as I understand it—I'm not an expert) "н" is pronounced [n] and "я" is pronounced [ja], but "ня" is not [nja], it's [nʲa], which they regard as clearly a different sound. If Mongolians make a similar distinction, it would be preferable to capture this in a romanisation.

There's nothing in Mongolian language which indicates that Mongolian has an additional vowel Ы. This might just be an influence of Russian transliteration, since Ы is definitely a distinct vowel in Russian, and this is conventionally written "y".

Also, what are we doing about Mongolian's long vowels? It looks like the proposal uses a double vowel (aa), but I wonder if a macron (ā) wouldn't be more pleasing to the eye. On the other hand, that would mean two diacritics in some cases (ǖ and ȫ), and the accepted form seems to be the double vowel, so it's probably better just to leave well enough alone.—Nat Krause^{(Talk!·What have I done?)} 21:34, 18 February 2007 (UTC)

I do not know what [nʲa] is but "Rinchinnyamyn" is pronounced exactly as it is written. Can you write how Russians would write [nja] in cyrillic if it is different than [nʲa]? Perhaps then I will have some idea about the difference.

I am not an expert but to me, Ы is really a different vowel. I think either it is different than И or it makes the consonant before it sound different. For example, МАЛЫН and МАЛИЙН would sound something like МАЛЪИН and МАЛЬИН. In Mongolian language vowels are grouped into two main groups (vowel harmony) and any word can include vowels from only one of these two groups. So the words are divided into "male" and "female" words. The third (neutral) group consists of И, Ы and when you want to use long "i", ИЙ is used for "female" words and Ы is used for "male" words. Temur 00:11, 19 February 2007 (UTC)

My understanding is that [nja] does not occur in native Russian words, but if it appears in a loan word, it is written нйа.

Given that Mongolian does include palatalised consonants, it makes sense that Cyrillic would use Ы to write [i] when it follows a hard consonant. Ы was originally two letters "ЪІ" (and І was the same as И), before it developed into a distinct vowel sound, so it makes sense to use Ы in place of ЪИ.—Nat Krause^{(Talk!·What have I done?)} 07:43, 19 February 2007 (UTC)

Do we actually have Mongolian words that only differ by a soft-sign? It's theoretically possible, but I'm not aware of any.

Or are you suggesting we explicitly specify palatalization even where the cyrillic ortography doesn't? Treating specific character combinations differently than the individual characters alone also looks like making things more complicated than useful. I share the desire to get as close as possible to the actual pronounciation, but a straightforward romanization system has its inherent limits, as it is necessarily based on the written form and not the spoken one. In those cases where the exact pronounciation really matters (or simply, is accurately known), we can still include it with IPA.

Long vowels are written as double characters in cyrillic, so I don't see any reason to do differently in latin. --Latebird 22:20, 18 February 2007 (UTC)

But Cyrillic orthography does specify palatalization. That's why they have ten vowel letters and Ъ and Ь. As I mentioned, I favour transcription over transliteration, so it would be nice to include all of the phonemes of standard Mongolian. You're right, of course, though, that no system is perfect, so I won't press this point if it proves too inconvenient to show palatalisation.—Nat Krause^{(Talk!·What have I done?)} 07:43, 19 February 2007 (UTC)

As I figured out by now, Ъ is not used for Mongolian words.

The latin script doesn't have any mechanisms to specify palatalization, so that any replacement for Ь may not have the desired effect directly. On the other hand, the transcription used in the de:WP is based on the transcription given by the standard dictionary of the former "brother state" GDR. It replaces the Ь with an J, which is pronounced like I or Y in English. I don't know if an I after a consonant turns on palatalization in the latter in English. If that should be the case (or a sufficient approximation), then we might have another option here. I don't like the apostrophe or the ĭ suggested by BGN/PCGN, because those would only confuse normal readers. --Latebird 16:16, 19 February 2007 (UTC)

I thought that a few examples might be helpful. This is what I found in WP. Unfortunately, we don't have the mongolian spelling for most articles about people, so that only geographic names are present here:

-	i	Mongolian	Context
Darganga 14	Dariganga 667	Дарьганга	mountain, sum in Sükhbaatar, ethnic group
Chandman 1590	Chandmani 827	Чандмань	sum in Khovd & Gobi-Altay
Möngönmort 22	Möngönmorit 13	Мөнгөнморьт	sum in Töv
Sagil 2350	Sagili 90	Сагиль	sum in Uvs

A quick and unscientific Google check (excluding WP clones) revealed that the versions with i are less often used than the ones without. The one exception is Dar(i)ganga, where the ethnologists seem to strongly prefer the version with an I. Among the versions without the i, most omit the soft-sign completely, and others replace it with an apostrophe or a dash. (Dar'ganga, Dar-ganga).

Because of the small sample size, it is difficult to jump to definitive conclusions from this. But for me, it confirms that omitting the soft-sign is at least a reasonable option. Of course, it will be a good idea to include an "also spelled as..." with the most common alternative in the introduction of the affected articles. --Latebird 17:41, 19 February 2007 (UTC)

Well, as I said, I'm not going to press the issue too much, but if palatalisation is an important, contrastive feature of Mongolian pronunciation, I would prefer to include it in Wikipedia's transliteration. Writing the soft sign as a vowel has the negative effect of implying additional syllables where none exist. A better option, I suppose, would be either an apostrophe or a diacritic. Examples:

apostrophe	diacritic	Mongolian	Context
Dar'ganga	Daŕganga	Дарьганга	mountain, sum in Sükhbaatar, ethnic group
Chandman'	Chandmań	Чандмань	sum in Khovd & Gobi-Altay
Möngönmor't	Möngönmoŕt	Мөнгөнморьт	sum in Töv
Sagil'	Sagiĺ	sum in Uvs

Nat Krause^{(Talk!·What have I done?)} 18:30, 19 February 2007 (UTC)

About [nja] in Russian. Really it looks be present in loan words only, but possible in Russian is not only нйа but нъя with the same vocal result (I found нъя in transliterations from Vietnamese).

I don't like ' or ĭ (last one especially). Palatalization presence in Mongolian is transscription question, but not transliteration. For us is clear, that Ъ comes with Russian words and so is transliterated respective rules (for Russian) using. Ь is not so widely used in Mongolian, but is present in native Mongolian words after consonants in the end of the word and between two consonants. Ь is a very short vowel, really close to [i]. Mongolians are reducing vowels strongly. May be it is why in old traditional forms (Gobi) coming from ancient Mongolian we find и instead of ь. Only this makes for me more preferable transliteration form ь -> i. Я and Е transliteration is not any question: is it Б palatelization in Бямба; but how latinize this letters: ya/ye or ia/ie or a/e Bogomolov.PL 08:56, 21 February 2007 (UTC)

...after consonants in the end of the word and between two consonants. ь is a very short vowel, really close to [i]. Very useful information, thanks! With punctuation marks and other special characters getting less and less popular, this strengthens the option with I(i) again. --Latebird 09:57, 21 February 2007 (UTC)

Checking In

Hi folks--coming late to the discussion on Nat Krause's invite. Not that I speak Mongolian--but I AM a linguist.

I can tell you that there are subtleties in the traditional Mongolian script that the Cyrillic obscures.

I have almost daily contact w/ Mongolians (my students), but they tell me it's still all very ad hoc. A romanisation standard has yet to emerge. I'd be happy to check with my students for any specific questions. Just post on my talk page.

Meanwhile, we might want to try checking ubpost.mongolnews.mn , the English-language newspaper there to see how they deal with the problem on a daily basis. I expect there's some young ex-pat copy editor there who can speak more authoritatively on these matters.

I don't know "the proposal," but here are the answers I got from my student to Nat's questions:

A few questions How effectively does the modern Cyrillic script transcribe standard Mongolian?

Well--pretty well, but not perfectly--but (this is nhrenton) doesn't the Roman alphabet do an equally horrid (if not worse!) job with English? We--like the Mongolians--have gotten used to our alphabet. The problems that would be created by a reversion to Mongolian script would be ten times worse.

I would say that laymen may think that some perfect linguistic solution to Romanization exists--but it really doesn't. Linguists are less confident. IPA is more a piece of art then science, I'm afraid.

The trend in romanizations, however, is to adopt the ones suggested by the non-Roman alphabet language-speakers. They're the ones who have to use it more often. Romanizations float on a scale between ease for English-speakers and ease for foreigners.

Does Mongolian really have palatalised consonants?

Yes. I'll need more specific examples. My student's name is Lkhagvasuren, with the Lkh representing the palatized /l/.

Is Ы pronounced the same as И in Mongolian?

No--but there is only a slight difference. The И is a little longer; both vowels are usually romanized as i. The И might be doubled i (i.e., ii) occasionally. I was told thay y is definitely wrong.

I just got my student to sign in. Maybe you can ask her--don't be shy--it's her homework!Nhrenton 18:14, 20 February 2007 (UTC)

Thanks for your input! So essentially what you're saying is that we're free to chose whatever is most practical for WP... Except for one point: I was told thay y [for ] is definitely wrong. Told by whom and why?

To be honest, I don't really expect the UB Post to use a consistent transcription system, but then I've been surprised before. I happen do disagree that the romanization system needs to be adapted to the native speaker of the source language, at least within WP. Our target audience are english language natives. --Latebird 19:28, 20 February 2007 (UTC)

I am not an expert, but am a native speaker. I disagree with "y is definitely wrong". Y is a good candidate for Ы. Temur 20:59, 20 February 2007 (UTC)

Can we get some kind of consensus on whether Ы is actually a different vowel itself, or whether it affects the preceding consonant differently? Wikipedia's article on Mongolian language implies the latter.—Nat Krause^{(Talk!·What have I done?)} 21:09, 20 February 2007 (UTC)

According to my Mongolian textbook, Ы is characterized as a "syllable character" (together with Е, Ё, Ю, Я, and Й), which means that it doesn't need a consonant to form a complete syllable. This makes me think that it won't modify a preceding consonant either. In addition, such a familiarity between Й and Ы could serve as an argument to transcribe both with Y. --Latebird 21:43, 20 February 2007 (UTC)

But Е, Ё, Ю, and Я do modify the previous consonant ... at least, in Russian they do. If palatalisation in Mongolian is at all the same, I'd imagine the consonant-modifying sounds would be written with those characters. As for Й, what sound does it make when it is alone in a word? Normally, I thought it was [j], but [j] isn't usually a syllabic consonant.—Nat Krause^{(Talk!·What have I done?)} 21:50, 20 February 2007 (UTC)

The Й never appears alone, but only in diphtongs (including ИЙ). I can't find anything about palatalization in my book right now outside the context of ъ and ь. At the same time, it *does* indeed seem that certain character combination influence each other's pronounciation. But the cyrillic spelling doesn't offer any extra hints about that, which means our romanization probably shouldn't either. --Latebird 23:00, 20 February 2007 (UTC)

Right; I misunderstood you—I thought you meant that these letters can form words with no other letters. Still, I must continue to quibble with the statement that "cyrillic spelling doesn't offer any extra hints about that". Cyrillic indicates palatalisation clearly ... at least, it does for Russian, and I see no reason to think it would treat Mongolian differently.—Nat Krause^{(Talk!·What have I done?)} 06:41, 21 February 2007 (UTC)

So there are palatalizing consonant/vowel combinations (besides Ъ and Ь) that are spelled differently than they would if no such phonetic codependency happened to exist? Or are we just quibbeling over the wording here? --Latebird 08:35, 21 February 2007 (UTC)

I have an example: АМЬ (life essence) has a palatalized consonant as it is indicated by Ь. If you want to say "life essence's" then it will be АМИЙН. АМ (mouth) has no palatalization, and "mouth's" will be АМЫН. Temur 18:28, 21 February 2007 (UTC)

This is somewhat inconvenient, but still good that you found it. The probability may be low, but we should still expect to eventually get two such words in WP then that we need to distinguish from each other. That's a problem when we omit the soft-sign. Based on previous discussion, and particularly Bogomolovs most recent comments, I'll change my proposal to I(i) for now. --Latebird 21:04, 21 February 2007 (UTC)

So, АМЬ woud be spelt ami? In that case, it would written the same as АМИ. Also, readers will assume that it is two syllables.—Nat Krause^{(Talk!·What have I done?)} 21:14, 21 February 2007 (UTC)

To be honest, I have no idea how to correctly pronounce a palatalized M in any language... Bogomolov explained that the soft sign is typically pronounced as a very short [i]. If I understand this correctly, then the word will indeed sound like it had two syllables. I'll ask a few Mongolian friends to give me sample speeches, which will hopefully give me a better idea of the difference. --Latebird 21:26, 21 February 2007 (UTC)

It seems to be the case that Ь in Mongolian is a holdover of a sound that was once a full vowel; the same, I believe, is the case in Russian. In Russian, I don't think this is considered to be an additional syllable; but perhaps it is different in Mongolian. In any event, it seems clear that АМЬ is pronounced differently from АМИ—Nat Krause^{(Talk!·What have I done?)} 21:48, 21 February 2007 (UTC)

It was great example with АМЬ - АМИЙН АМ - АМЫН. Like with ГОВЬ - ГОВИЙН the "Ь" looks like "hidden" и and palatelization disclose its presence. The same was with Ъ ("hidden" Ы), which would be after every nonpalatelized consonant at the end of the word (this rule was in Russian until 1918 reform, Bulgarians use Ъ instead of Ы, palatalization in this language is absent), but it is superfluous to note nonpapatalization if much more rare palatalization is noted by soft sign... It is physiology of palatalization process - muted short vowel "i" is physically present (if you really CAN palatalize consonant - not every language has palatalization). Yes, AMI is pronounced differently from АМЬ (it is not transscription), but it is not confused with АМ Bogomolov.PL 06:52, 22 February 2007 (UTC)

Ok, I've asked a mongolian friend (not a linguist, but with some Mongolian-English interpreter training): Apparently, the historical pronounciation of АМ and АМЬ was "ama" and "ami" (roughly), which is reflected by the spelling in the classical Mongolian script. The Pronounciation today can be approximated as "am" and "em", where in the first case the "m" is clearly defined, and in the second case it sounds kind of open-ended, hinting at a muted "i". With other consonants, the "i" seems to be less muted, but it has always left some traces in the pronounciation.

It seems that all instances of the soft sign in the Mongolian language can be traced back to an "i" in the classical script. In other words, there will never be a conflict between words that only differ by those two characters. In our example, АМИ is not a Mongolian word, because the word that was historically pronounced that way has since morphed into our АМЬ. This makes it safe to substitute the Ь with I(i). --Latebird 22:49, 22 February 2007 (UTC)

Well, if it's the case that there are never minimal pairs for Ь and И at the end of words, then it is a reasonable solution to use i for both of them. This does, however, seem like more of a native-speaker focus and less of an English-speaker focus, because very few readers will realise that the i is not actually pronounced. I also wonder whether this solution will be feasible in cases where the Ь is not at the end of the word, such as Dariganga and Möngönmorit, mentioned above.—Nat Krause^{(Talk!·What have I done?)} 04:40, 23 February 2007 (UTC)

In Dariganga and Möngönmorit the i's sould not be pronounced. Only the consonants before them changed. I think in any word, the empahsis is on the first vowel only, and all the other short vowels are almost not pronounced. "Dariganga" has two words in it, "Dari-ganga", so "i" and the last "a" are not pronounced. If you remember this rule, the ambiguity problem of Ь and И can be solved: if you want to make an И to be pronounced clearly, say, in the end of a word, the word should be a non-mongolian word, or one of Ы (y) and ИЙ (iy or ii) should be written where you want И to be. ("I am not a linguist" is implied) Temur 07:00, 23 February 2007 (UTC)

---

Personally (and not intending to deprecate any of our excellent Mongolian contributors here), I think we need a bit more input from a Mongolian speaker who is fairly familiar with linguistics or from an outsider who has put in some considerable study of Mongolian before we can comfortably make a decision on this. I personally, don't feel that I have a clear enough idea of how Ь is pronounced (both in formal and informal speech, as well as historically), to have an educated opinion about how we should spell it—the same goes for other unstressed vowels, too.—Nat Krause^{(Talk!·What have I done?)} 23:07, 26 February 2007 (UTC)

Ending vowels and pronounciation hints

(subheading for editing convenience)

In other words, the Ь gets treated no different than any other vowel at the end of a word. This peculiarity of the Mongolian pronounciation rules can't be solved by a romanization system. We just have to live with the fact that most of our readers will pronounce the ending vowels in Erdene, Selenge, Chandmani, Soyombo, etc., where a Mongolian would not.

We could, however, give pronounciation hints (where known) in the introduction to individual articles.

• The Soyombo script (Mongolian: Соёмбо бичиг, soyombo bichig; pronounced: Soyomb)...
• Dariganga (Mongolian: Дарьганга; pronounced: Dar-gang)...

This should only be necessary with ending vowels or some special cases of composed words like Dariganga/Darganga. Of course, adding IPA would be even better, for those who are fluent enough in it. Is this a reasonable approach? --Latebird 09:44, 23 February 2007 (UTC)

Generally speaking, any romanisation which requires you to routinely respell it for pronunciation is questionable. Also, I wasn't aware that Mongolian never pronounces vowels and the ends of words. That does put things in a different perspective.—Nat Krause^{(Talk!·What have I done?)} 00:08, 24 February 2007 (UTC)

Well, the ending vowels are not completely ignored, but they become very short. In some cases, if you do not know an exact pronounciation of a word, it is better (understood if you) clearly pronounce the ending vowels than completely ignore them. Temur 07:20, 24 February 2007 (UTC)

The important point here is that transliterating Ь as i means not to treat it differently than other ending vowels. Giving extra pronounciation hints may or may not be questionable. Those hints don't necessarily have to be bolded, btw., if that makes them look too much like a "respelling" (just changed that in one of the examples above to compare the effect). And if we get a consensus that they're not necessary at all (or not appropriate), then we can also leave them away. The question to decide this is: Are they more helpful for english language readers, or more confusing? --Latebird 11:05, 24 February 2007 (UTC)

Do many articles for other places with foreign (like french) names have spelling hints? I think anyone who knows how to pronounce "sain baina uu?" (btw. I strongly prefer writing 'I' for 'Й') or "bayarlalaa" has a clue that you don't really pronounce all vowels, and the other probably don't really need to. Plus they would still need to know where the stress is to pronounce the word entirely correct. Yaan 14:11, 26 February 2007 (UTC)

I couldn't find any french names with spelling hints in a quick scan, which kind of surprised me. Am I worrying too much about this? It would certainly make things simpler to leave them away... In the ideal case, all articles about foreign language subjects would have an IPA specification, but that is obviously not realistic in the near future. --Latebird 22:46, 26 February 2007 (UTC)

Renton to Latebird

Thanks for your input! So essentially what you're saying is that we're free to chose whatever is most practical for WP... Except for one point: I was told thay y [for ] is definitely wrong. Told by whom and why?

You're welcome. Re I was told by my student--I can't really judge as to her authority--she is college-educated, though, and speaks quite good English. She DID seem sure about it. I got the impression that when Mongolians text-message they go for the easiest form.

To be honest, I don't really expect the UB Post to use a consistent transcription system, but then I've been surprised before.

You're right there; my student told me that Mongolians often found the Romanizations inaccurate.

I happen do disagree that the romanization system needs to be adapted to the native speaker of the source language, at least within WP. Our target audience are english language natives.

Latebird, Yeah, I see your point. But these things change over time: I know Mongolians are pretty insistent that we use Ulaanbaatar and not Ulan Bator (did we change that recently? My 2005 Oxford Atlas (12th Ed.) only gives Ulaanbaatar. Taegu or Daegu? (I know Korean better). Canton or Guangzhou? Frankfort or Frankfurt? Livorno or Leghorn? For most entries I'd suggest following the native speakers' lead. This has been the case in Korean, I know. BTW, the Oxford Atlas sort of did something funky with the Korean--the editors use Sǒul. Huh? Not Seoul?

Fortunately, we don't have space limitations on WP. We need merely pick a canonical Romanized spelling (based on editorial consensus) with variant forms in bold. Redirect the variants to the canonical form. Write also "commonly also" next to highly competitive variants. Always give the Cyrilic for the native speakers (and pedants!) out there. (I see someone added to my paltry listing of UB districts in the UB article and included--among lots of lovely data--variant romanizations--anyone here? Anyway, thanks!)

Mind you, I think pronouncing them correctly under the given variants there is harder--I suspect they worry too much about trying to represent the Cyrilic. I mean h instead of kh?

If the English spelling/name is a passionate issue--it should be included as a discussion at the beginning of the article. The Sea of Japan article deals with something similar.

My quick tuppenceNhrenton 13:43, 22 February 2007 (UTC)

Renton, while I value the opinion of your student, she seems to be in a minority with it. The most commonly used romanization for Ы in English is Y, with only a few sources using I. In text messages, people in all languages will use the form that can be typed with the least effort on a numerical keypad. I don't think we can follow that path here.

---

Another reason could be that Mongolians often need the 'y' for sentences like 'sain bna yy?'Yaan 14:10, 26 February 2007 (UTC)

How is "sain bna yy" pronounced, or how is it written in Cyrillic?—Nat Krause^{(Talk!·What have I done?)} 21:45, 26 February 2007 (UTC)

It is written САЙН БАЙНА УУ. Some people use y for У (x for Х, v for Ү, etc) because of the similar appearance. IMO, this practice should be ruled out. Temur 21:57, 26 February 2007 (UTC)

---

As explained in the proposal, Ulan Bator is not a romanization but an english word (translation). The criteria for chosing romanisation vs. translation are not part of the proposed guideline, though. They are given in WP:NAME :

Generally, article naming should prefer to what the majority of English speakers would most easily recognize, with a reasonable minimum of ambiguity, while at the same time making linking to those articles easy and second nature.

I'm trying to follow those requirements as best as I can in all aspects of this proposal. The future will reveal my rate of success or failure... --Latebird 23:00, 22 February 2007 (UTC)

Well, I certainly prefer transliteration, as few people will manage to correctly pronounce Mongolian words without ever having heard them. The word-final short vowels of contemporary Mongolian writing are not pronounced, and linguistically they don’t exist (in some instances they hint at other linguistic factors, e.g. behind n and behind g in a word with back vowels) or are loanwords written with one vowel yet lengthened in speech, e.g. lesbi. ы does only appear in the accusative and genitive cases where it contrasts with ий. Phonemically, both are instances of one archiphoneme further determined by morphophonemics: words with “back vowels” take ы, words with front vowels take ий. If the consonant preceding the suffix is palatalized, it is always ий and the soft sign would be dropped. Phonetically, they are pretty close to each other.

ы and й as “y” is unfortunate. Both only appear as vowels: й is the way to write и when it is preceded by any other vowel. Phonetically and phonologically, these combinations are to be regarded as diphthongs in Ulaanbaatar Khalkha. So one would best transcribe both и and й as “i”. ы is problematic. It does not correspond to the consonant “y” as in “yu” ю etc. However, as mentioned above, there are some few contexts where treating ы and ий alike would blur linguistic information. On the other hand, writing them differently would blur their far greater similarities, and to access any linguistic information, readers would have to have considerable knowledge anyway. So I’d propose to transcribe ы as “ii”. By the way, most Mongolians do so (e.g. Google: Baataryn: 122 vs. Baatariin: 461, the stem is Baatar, not Baatar’).

Now how to deal with ь? On the one hand it alternates with и, e.g. банди or, somewhat less frequent, бандь. Yet, even if we take its derivational form бандидах that is a non-compound and where using a soft sign would be orthographically wrong, there is still no [i] and the word has only two syllables. On the other hand, people unacquainted with Mongolian will pronounce an “i” as an [i], and they couldn’t get it more wrong. Therefore, I’d prefer to use a diacritic like an apostrophe, so that people who are aware of its function can pronounce it correctly and people who aren’t can simply ignore it. G Purevdorj 00:28, 28 February 2007 (UTC)

Thanks for your comments, GP. It begins to seem that и in Mongolian is the same as ь, but that they are both the same as Russian ь (palatalisation), rather than Russian и ([i]). Perhaps a diacritic would be best. It also seems that the distance between Mongolian Cyrillic writing and speech is quite significant. Do I have this right? This does indeed create a difficulty in balancing transliteration and transcription.

I think the logic of writing ы and й both as "y" is that ы is the vowel "y" and й is the consonant "y"—completely different sounds in English. If ы is written as "ii" and й is "i", then ий will be the same as ы.—Nat Krause^{(Talk!·What have I done?)} 19:50, 28 February 2007 (UTC)

That's very interesting and helpful information indeed. On the other hand, let's remember that it's not really our goal to reinvent too many wheels. We might end up with a system that is "too perfect", when it's very different from most other established systems. I tried to make only small simplifications from BGN/PCGN for exactly that reason. Is it not possible to reach a reasonable solution by staying close to that? --Latebird 23:14, 28 February 2007 (UTC)

Й (й) -> I (i)

(subheading inserted for convenience)

"It also seems that the distance between Mongolian Cyrillic writing and speech is quite significant." The difference is slight, but the Mongolian script assumes phonological representations that ceased to exist in the course of the last century. Actually, the difference between long and short vowels only exists in the first syllable now; in other syllables, vowels written as long vowels are short and vowels written as short are only epenthetic vowels totally underspecified and determined by their consonantal environment and vowel and rounding harmony. Contemporary Khalkha goes to great lengths to avoid open syllables with an epenthetic nucleus. For this and other phonotactical reasons and for morphophonological reasons, epenthetic vowels may be inserted in other places than where phonemic short vowels once used to be, or not at all. On the other hand, Mongolian orthography is being changed by its prudent users. Bagshiin ih surguuliin erhem gishüüd nar would propose that it is тус-ла-ж and not тус<а>-л-ж, but they're fighting a losing battle: the first one gets 3190 googles, the second one about 13500! So what can we do to let English readers correctly pronounce Mongolian words? Present them with IPA, have them read, understand and learn by heart the first ninety pages of Svantesson et al. 2005, or teach them Mongolian classes.

What can we do short of this? ы as "y" might not be the issue; I somehow didn't consider the English vowel "y" when I wrote my last comments. It's not phonemic and it's not pretty but proper transliteration uses this symbol as well. The soft sign may be i or ', that's not that important, though I favour the latter. But be so good and abandon "y" for й. й isn't some kind of [j], but an /i/ appearing in the second position of a diphthong, and ий is symply a long (as in niitiin tetver /niitin titvr/ (i and e in the first syllable have merged, thus titver)) or a phonemic i in a non-initial syllable (Other vowel graphemes are repeated to indicate length , but in cursive handwrting, ии is easily confused with both ш and т, thus ий). So й ought to be i. G Purevdorj 01:12, 1 March 2007 (UTC)

Yeah, I support i for й. There is another issue less serious: For example, я is ya so яа should be yaa, which is rather odd. I have never seen such a spelling. People always use ya for both long and short. Temur 01:37, 1 March 2007 (UTC)

I can see й being essentially the same as и, and the cursive handwriting is an interesting explanation why it even exists. This looks like an easily justifyable simplification, so I've changed it to i now.

I also understand that many people are too lazy to write Yaa for яа, but I don't think we should create a special case just because of that. --Latebird 02:15, 1 March 2007 (UTC)

If the Cyrillic spelling clearly transcribes valid pronunciation as of as recently as a 100 years ago, then that's good enough for me; a more-or-less direct transliteration of the Cyrillic is probably adviseable, especially if a) we lack a reliable source for more recently pronunciations (particularly if they are not standardised) and/or b) more exact transcription would be awkward or otherwise difficult for some reason.

As for "ий", I think "ii" is great. However, we might consider using "iy" instead, just because it's the more common spelling (if it is, in fact, the more common spelling).—Nat Krause^{(Talk!·What have I done?)} 06:34, 1 March 2007 (UTC)

I think "ii" is the more common spelling, although the aestetically more pleasing might be "iy", look at this example: Hentiy instead of Hentii. Temur 08:26, 1 March 2007 (UTC)

Google doesn't reveal a significant difference in popularity between ii and iy, one or the other seems more popular depending on the examples chosen. That may not be scientific proof (and will include results from other languages than english), but it's close enough to justify using the simpler variation. As for esthetics, that is highly subjective and I wouldn't want to impose mine on anyone. It's unfortunate that on WP the iy has been used quite consistently so far, but we'll have to touch most names anyway. Let's declare this specific quiestion closed then. --Latebird 10:48, 1 March 2007 (UTC)

It is a lot of vulgar (sorry) latinization forms used in Cyrillic zone: SMS wrighting or Internet forums conversation very often is out official latanization rules - for native speaker it is very easy to understand in a contex. (Cyrillic у -> y ; Cyrillic х -> x etc.) Bogomolov.PL 16:06, 9 March 2007 (UTC)

Translated words in composites

In the section Use conventional names there is still an open question about composite words, where we have a conventional english name for one of the parts. Currently, we have two prominent examples that are treated differently. For names that contain the word Khan, the english translation is used quite consistently (with the Chinggis Khaan International Airport as a noteable and justified exception). On the other hand, we have the Gobi desert, but many place names where the transliteration "Govi" (or even "Govĭ") is used instead.

Personally, I'd prefer to always use the translated english form in such cases, unless there are specific reasons against it. This would result in titles like Dornogobi province (Mongolian: Дорноговь аймаг, translit.: Dornogovi aymag, pronounced: Dornogov), Dundgobi, Gobi-Altay, Gobisümber etc. For what it's worth, the german language WP also handles it that way (even if I'm not aware of an explicit guideline about composites there). Any opinions? --Latebird 11:31, 23 February 2007 (UTC)

IMO not a good compromise, it makes -gobi look like the correct mongolian form, when it should be -govi. It's like using Nord Rhine-Westphalia instead of either Nordrhein-W. or North Rhine-W. Yaan 14:10, 26 February 2007 (UTC)

I agree with Yaan on this. It's better not to mix Mongolian and English forms together like that.—Nat Krause^{(Talk!·What have I done?)} 21:46, 26 February 2007 (UTC)

I agree with Latebird on using Gobi, but translating the word inside a composite word seems to be going too far. Should Dornogovi be written as Eastgobi? Temur 22:03, 26 February 2007 (UTC)

Ok, I see the point of consistency, and changed the relevant section accordingly. --Latebird 22:32, 26 February 2007 (UTC)

Soft sign and pronounciation hints, second round

It seems that within the transliteration table, the soft sign is the last open point. I don't see any strong opposition against i anymore, but there are some arguments for an apostrophe. Personally, I think that an apostrophe stands in the way of making linking to those articles easy and second nature, so I'd prefer the i for the article name. But it probably makes sense to give the apostrophe version as an alternative:

Ь (ь)

I (i) / '

Дарьганга = Dariganga (also Dar'ganga)

Example:

Dariganga (Mongolian: Дарьганга; also Dar'ganga).

This should make it clear that there's something funky going on in the cyrillic original, and that the soft sign is not just a regular vowel. At the same time, I've come to the conclusion that the "pronounciation hints" were probably a silly idea. The pronounciation rules for ending vowels are not part of ortography, neither in cyrillic nor in latin, so they shouldn't be part of this guideline. Of course, if someone can figure out (or look up) the accurate IPA form, then that would be very welcome anyway. --Latebird 12:00, 1 March 2007 (UTC)

I agree. Could we find an article where "pronounciation hints" are useful and move them there? Temur 16:49, 1 March 2007 (UTC)

What I would like to see is a system of romanisation which will show all of the linguistically relevant information contained in the Cyrllic spelling. It's still not clear to me whether there is a relevant distinction to made between и and ь in Mongolian.—Nat Krause^{(Talk!·What have I done?)} 20:32, 1 March 2007 (UTC)

Somewhat simplified: If last consonant palatalized: Two word-final consonants -> i, One word-final consonant -> soft sign. However, new syllabification rules: arhi syllabifies as a-r'<V>h. Thus, the difference is linguistically irrelevant. G Purevdorj 20:46, 1 March 2007 (UTC)

Thanks for the linguistic details. I have made the respective changes to the proposal. --Latebird 14:29, 3 March 2007 (UTC)

Hard sign

It’s not the case that the hard sign is not used at all in Mongolian. It may be used after <ya>, eg харъяалах ‘to have sovereignty over’. However, most Mongolians do not recognize the use of this sign (I am actually not aware if it has any, and I don’t know of any contemporary research that deals with so marginal a question). So they often replace it – with the soft sign! (Damdinsüren would turn over in his grave!) Google gives 430 харъяалах and 1060 харьяалах, and the dictionary of Bawden (a decidedly descriptive dictionary) has only харьяалах. And of course, there is the voluntative in –ъя, the non-emphatic, most common pronunciation of which is -[i]. Its most common romanization is –ii, alternatively -ya, eg Google yavii 1110 vs yavya 727, hairlii 229 vs. hairlya and hairalya combined 27. The occurrences of har(i)ya(a)lah are insignificant, but we have Google gavyat 376 vs. gaviyat 455. As we really don’t know if it has any function and no function is perceived by many Mongolians, it is most convenient to drop the hatuugiin temdeg in transcription, but we cannot bluntly state that it isn’t used at all in the Cyrillic orthography of native Mongolian words. G Purevdorj 12:13, 3 March 2007 (UTC)

The information that the hard sign is only used in Russian words comes from Mongolian.pdf. My dictionary says to omit it in non-scientific transcription. Do we actually have any articles yet where it appears (or should appear)? If we give an example, then I'd prefer to take one that actually has a chance of turning on WP. --Latebird 14:41, 3 March 2007 (UTC)

Does the hard sign have any effect on pronunciation where it does appear? Also, as a more general question to G Purevdorj, do we know how standardised modern pronunciation is or isn't? Is there some body regulating the Mongolian language, or publishing dictionaries with recommended pronunciations?—Nat Krause^{(Talk!·What have I done?)} 18:17, 3 March 2007 (UTC)

To Latebird: I can’t think of something name-like right now, but there are other examples you are very likely to encounter. There are 9000 googles for гавъяат (and further 19000 googles for гавьяат), eg Монгол Улсын Гавъяат багш „Distinguished teacher of the Mongolian State“. The transcription would be Mongol ulsyn gavyaat bagsh, ignoring the hard sign as your fine dictionary does. I’m not aware if a title like that does appear at Wikipedia right now, but I did a quick search for the influential politician Bat-Erdenii Batbayar (who refused to accept the title Distinguished Writer) and the famous singer Ariuna(a) (who should be something like gavyaat urlagchin or the like) and couldn’t find them. So while I shan’t make sure that a hard sign has already been used somewhere, I bet it ought to!

To Nat Krause: “Does the hard sign have any effect on pronunciation where it does appear?” As I said, I don’t know for sure. I’d suppose it will “block” palatalization, and I know that at least for some speakers there is no palatalization of the preceding consonant, so this “prediction” is carried out. By not marking the hard sign, we wouldn’t indicate any palatalization, so that is fine. “…do we know how standardized modern pronunciation is or isn't? Is there some body regulating the Mongolian language, or publishing dictionaries with recommended pronunciations?” There should be a normative book on phonology, but I never used it and I can’t access it (nor any other book) right now, and there are some publications later on. They constitute a pronunciation standard that is not entirely descriptive and probably not even entirely congruent unto itself, but I’ve never dealt with this question either. No serious dictionary I know provides pronunciations, including Tsevel and Choimaa. The academy of sciences had a commission for terminology, but its publications provide only written forms. (I’m not aware if this commission still exists, but I shall learn this in time.)

G Purevdorj 19:49, 3 March 2007 (UTC)

Thanks for your info, G-Purevdorj. I was thinking that a classic example of a romanisation system that reflects pronunciation is Hanyu pinyin, which is able to reflect pronunciation specifically because the government specificies a "correct" pronunciation for each word and actively promotes this as standard. If Mongolian is not standardised, then there will never be a "correct" pronunciation for any given word. Even, so I would be amenable to developing a Wikipedia system of transcription that would match the pronunciations given in one or more of the major textbooks for students of Mongolian. However, this is not a task that I would be able to do myself, and certainly do not insist on it if everyone else wants to go with a Cyrillic-transliteration-based approach (as it appears they do).

I do, however, prefer that we incorporate the hard sign into our system of transliteration, since it appears likely that it does affect pronuncation.—Nat Krause^{(Talk!·What have I done?)} 23:17, 4 March 2007 (UTC)

The question might be: should <yV> be interpreted as palatalizing? If V is a front vowel, palatalization is considered non-distinctive. (A questionable opinion for one reason I’m not gonna discuss here, but the state of actually available research.) <yu> and <ya> remain. <yu> doesn’t seem to appear as a non-initial letter of any first or non-initial long syllable in any word of Mongolian origin (can you prove me wrong, Tömörchin?), thus there’s just <ya>. And behold, it does indeed palatalize its preceding consonant when in initial syllable. <nyalkh> has but four phonemes, but <gav’’ya> has five. However, this is so for simple diachronic reasons (*/nilqa/ [n’ilqa] > /n’alqa/). If we presuppose that <y> does only then indicate palatalization, if NO COMPLEX ONSET makes any other interpretation impossible, then the hard sign behind <ya> is indeed devoid of any useful information. Any indication of the it will only incite confusion. But I have an alternative proposal: <ya>, if it is the second letter of any root morpheme, should be transcribed as <ia> instead. That would mend the synchronic inconsistency in Mongolian orthography that writes <nyalkh>, but at the same time <xiam>. Moreover, it would be helpful always to use the hard sign (that is, zero) and not the soft sign that now often replaces it.

If our goal was indeed to represent pronunciation first, we would have to represent the Mongolian vowel shift. To give one example: уруул would be <orool>, and it only gets more difficult if we add API diacritics. Moreover, no Mongolian would succeed in recognizing any words any longer. For me, that is not really the way to go. G Purevdorj 00:52, 5 March 2007 (UTC)

Pronounciation is a concern, but not the first one. I've already cited the primary requirements from WP:NAME further above:

Generally, article naming should prefer to what the majority of English speakers would most easily recognize, with a reasonable minimum of ambiguity, while at the same time making linking to those articles easy and second nature.

I think that we have already reached a reasonably balanced optimum towards those requirements. Any further linguistic research on ephmeral questions like this one is likely to only confuse matters more than necessary. For what it's worth, WP:RUS also omits the hard sign. --Latebird 03:33, 5 March 2007 (UTC)

Historic names

I've added a short section to mention that historic names may take different forms than what the modern cyrillic Mongolian version would suggest. --Latebird 12:07, 5 March 2007 (UTC)

It is good idea to create historic names list. About Ulaanbaatar: it looks like Ulan Bator cames from Russian Улан-Батор, but in Russian this name (diffrent fron Mongolian Улаанбаатар) is the same pronounciation (as it possible in Russian) like in Mongolian (unaccented "о" in Russian is spelled "a"). Bogomolov.PL 12:25, 9 March 2007 (UTC)

There is a potentially very large number of historic names (more people than places) to take into account, so it may be difficult to compile a comprehensive list. We'll have to look at each example individually anyway. It is correct that Ulan Bator is of Russian origin, but it has become the common name in English as well. --Latebird 12:38, 9 March 2007 (UTC)

Yes, but I mean existing WP articles texts: is it possible to give OUR latinization in brackets? Or we will use only traditional form? What will be with latinizations realy in use in other sources? Gobi - gov' - govi or other forms? Ulaanbaatar - Ulan Bator - Ulan-Bator? I think is reasonable to use traditional form with other, our also, forms in brackets. Bogomolov.PL 16:16, 9 March 2007 (UTC)

Your examples are modern names, not of the historic type I had in mind. A better example of a place name would be the ancient Karakorum, in contrast to the modern Kharkhorin. People with historic names include Qabul Khan (turkic), Köke Temür (classical mongolian), or Demchugdongrub (chinese, despite the cyrillic transcription in the article), etc. In most cases, it is quite obvious that they can't be directly transliterated from modern Mongolian, and that section in the proposal serves merely as a reminder of that. If a modern Mongolian form exists (and is known), that should of course also be given in the introduction. --Latebird 17:40, 9 March 2007 (UTC)

Now an official guideline

I've been bold, remaned the page to Wikipedia:Naming conventions (Mongolian), and declared it an official guideline. Let's hope it will be as useful as it is intended to be! --Latebird 09:15, 16 March 2007 (UTC)