IMSLP Forums

Hello,

Would it be possible to make a global change of transliterated spelling within IMSLP for the Russian/Soviet publisher name "Музыка"? It often appears in IMSLP as "Muzika," which is not an accurate transliteration. The second vowel should not be transcribed as "i" (at least not without a diacritical marking), because the Russian vowels spelled "и" and "ы" are different from each other.

The simplest way to transcribe it (i.e., without diacritics) would be "Muzyka."

"Muzyka" is how the publishers themselves spell their name in English, so Lyle's suggestion is a good one.

There are a few different systems of transliteration from the Russian, which (for understandable reasons) aren't always applied consistently on IMSLP. Maybe we should follow the policy of our sister site Wikipedia, as outlined here:

http://en.wikipedia.org/wiki/Wikipedia: ... of_Russian

Speaking of consistency, we also have some Russian composers are listed by their full names (eg. Pyotr Ilyich Tchaikovsky, Modest Petrovich Mussorgsky) while others aren't (eg. Sergei Prokofiev, Mily Balakirev). Shouldn't they all be one way or the other?

And don't get me started on Rachmaninoff vs. Rakhmaninov!

I agree. As Davydov mentioned, they used the spelling Muzyka themselves, which really settles the issue as I see it. Since we have a substantial collection of their scores at IMSLP, a global change would be ideal. However, do keep in mind that Feldmahler is dealing with a number of more pressing issues right now.

OK. I've been through and manually changed all pages that linked to "Muzika" to point to "Muzyka". It took well under two hours, but at least it won't add to Feldmahler's busy workload This process wouldn't find other lurking references to "Muzika" that weren't enclosed within links.

At the moment "Muzyka" redirects to "Muzgiz", and both publishers are shown on the one page. However, I've retained this distinction in name in case the two are ever disaggregated in future (which might be desirable as they had distinct sequences of plate numbers, for example).

If anyone has any other thoughts on standardisation of Russian names, etc., they'd be very welcome.

I, for one, would prefer that Cyrillic alphabet be used whenever transliteration is not absolutely necessary, and only in modern orthography (except for illustration or explanation).

A transcription system that avoids diacritics as much as possible would be best, IMHO, because it is easier to type and search with. Both Library of Congress and the British system (as in Grove) use diacritics.

As with those systems, however, there would have to be an accommodation for common spellings of names, especially when searching by names as headings. (E.g., Tchaikovsky, Arensky, Rimsky-Korsakov, Cui, etc.).

When pure transcription is required (as with song titles, quotations, etc.), I would tend to support the system represented at the Lied and Art Songs Texts Page. It avoids diacritics and is almost completely unambiguous with regard to ascertaining the original Cyrillic spelling.

There's a problem with what you suggest, Lyle. Not every browser or OS is capable of correctly displaying non-Latin characters. It also raises the question of how everything should be titled, and in what language.

Generic items like Sonatas, Symphonies and the like are fairly easy. The three most common Western European laguages (French, Italian, and German) cover many of the remaining items, but leave the question of what to do about the Russian repertoire and other composers like Dvorak, Bartok and Sibelius who entitled works in the less-common languages. Do we use the original-language title for Tachaikovsky's The Nutcracker, Op.71 (Shchelkunchik), or do we simply use the far more widely encountered English translation? We would probably get a lot of complaints if I were to move the workpage title to Shchelkuchik - even more if it were put into Cyrillic characters. Not to mention the annoyance of a new contributor who'll set up a whole new page under either the English, French or German title.

The over-riding rule is to make things easy to find for the end user. There's always a certain amount of tension between staying faithful to the original title and the use of translations and transliterations. Most musicians looking for Rachmaninoff's Op.29 would look for titles Isle of the Dead or perhaps Toteninsel rather than Ostrov Mertvikh. I'd say we should therefore limit the use of Cyrillic to those items which aren't going to effect either an end user's search or the automated search feature hooked in to Amazon.

Carolus,

I wasn't advocating changing the main entries of Russian works to Cyrillic or to transliteration (Rimsky's "Dubinushka" begin an obvious exception, because that's how it's known in English).

If a non-generically titled Russian work is better known in the West under another name (priority to English, I guess, unless the work is titled in French or German or other as default), then that would be what the main entry should be.

In my previous post I meant to provide access to additional titles, info, etc. through searching.

Sorry for any confusion.

I'd agree that there is scope in the "Alternative Title" boxes and some work lists of Russian composers for titles to appear in Cyrillic, along with transliterations for those people whose browsers don't show Cyrillic characters.

As far as transliterations themselves go, IMSLP currently uses a variety of methods, particularly in the names of Russian composers. Lyle's point about diacritics is well taken, which is why I suggested the English Wikipedia system (where they aren't used).

For textual transliteration (i.e., outside of commonly spelled names, etc.), the Wikipedia Romanization system has some serious drawbacks:

• the letter "y" in this system transliterates not only for ы but also for й and the initial yot of the letters е, ё, ю, я. This creates extreme ambiguity.
  
  In addition, inserting the "y" before "e" is not necessary; all that's needed to distinguish э (which is much less frequently found) from е.
  
  Also, the omission of transliteration of the soft and hard signs (ь, ъ) would make reverse transliteration additionally difficult.

It's much better to use separate transliterations for ы and й, e.g., "y" and "j", respectively. The system that I mentioned before is much cleaner.

I take your point, but the use of the letter 'j' to stand in for a vowel will cause confusion among native English speakers

Seriously though. the Wikipedia system is itself a slight modification of another romanization method in widespread use:

http://en.wikipedia.org/wiki/BGN/PCGN_r ... of_Russian

This may be preferable, in that it avoids some of the problems you mentioned. And whatever system was used for text, they could always be accompanied by the orignial Cyrillc version, making reverse transliterations unnecessary.

With regard to composers, what's your view on how we should deal with names such as Liadov/Lyadov, Scriabin/Skriabin/Skryabin, Rachmaninoff/Rakhmaninov, Ziloti/Siloti, Cui/Kiui, Prokofiev/Prokof'ev, Sergei/Sergey, Nikolai/Nikolay, etc.? Should we aim to be consistent, or take each name on a case-by-case basis?

This situation with "J" is something that English speakers, in my humble opinion, should get used to. Speakers of Spanish and French, for instance, would have a comparable adjustment to make.

"J" in transliteration of Cyrillic is not difficult to grasp at all -- especially if one has to look at Polish, Czech, or Croatian names. Besides, a transliteration is not meant to provide phonetics; correct pronunciation requires knowledge beyond spelling, anyway. (The ambiguous and ubiquitous "Y" in the Wikipedia transliteration system simply looks ugly, is inefficient, and does not do justice to the original. If IMSLP were to host PDFs of vocal texts in transliteration, I'd hate to see this kind of transliteration used as the standard.)

As to names, the consistency should be globally retained within IMSLP for each name, I would guess (in information sections, etc.). As to which (English) version to use, whatever is the most commonly used version probably should prevail (although I can't stand the ones ending in "-off" or "-eff").

For hardly known composers, I'd think that it would have to be decided whether to construct a transliteration from the IMSLP system used for texts, modify that to harmonize with common endings of better known composers (e.g., "-sky" and "-skaya") or take a Roman spelling from their publications, which could be in French or German, most likely.

It seems to me that all the conventional systems have their flaws, but if we take some common Russian names then maybe we reach a consensus on the preferred form? For example:

* Nikolai / Nikolaï / Nikolaj / Nikolay
* Sergei / Sergeï / Sergej / Sergey
* Anatolii / Anatoliï / Anatolij / Anatoliy / Anatoly
* Dmitrii / Dmitriï / Dmitrij / Dmitriy / Dmitry
* Evgenii / Evgeniï / Evgenij / Evgeniy / Yevgeniy / Yevgeny / Evgeny
* Petr / Pëtr / Piotr / Pjotr / Pyotr
* Mariia / Marija / Mariya
* Aleksei / Alekseï / Aleksej / Alexei / Alexey / Aleksey

I think there are strong arguments to be made for the first and last entries on each line. The last is closest to the versions already in most frequent use on the site (and compatible with Wikipedia), although my personal preference would be for the versions listed at the start of each line.

Davydov writes:

With regard to composers, what's your view on how we should deal with names such as Liadov/Lyadov, Scriabin/Skriabin/Skryabin, Rachmaninoff/Rakhmaninov, Ziloti/Siloti, Cui/Kiui, Prokofiev/Prokof'ev, Sergei/Sergey, Nikolai/Nikolay, etc.? Should we aim to be consistent, or take each name on a case-by-case basis?

We really need to stick with the most widely used transliteration for composer names, which I expect is the one employed by Library of Congress. Rachmaninoff, Siloti, Ciu, Prokofiev, Sergei, Nikolai would be the ones in the list you posted.

I think the system mentioned by Lyle at the Lied and Art Songs Texts page is a very fine one overall. Davydov's point about the complaints arising from the use of "j" is a valid one nevertheless. The transliteration "ju", "ja" for ю, я might be easier on the majority of English-speaking users if it were "iu" and "ia", or even simply "u" and "a". The common ending ый, for example, would be transliterated "yj" under the LAST system, which strikes me as very cumbersome.

We should use the LAST chart as a starting basis, with a some modifications of our own, IMO. BTW, thanks to both of you for raising this issue. IMSLP has no consistent policy on the transliteration of Cyrillic, and I for one appreciate that you're taking the time to think about it.

Speaking of Cyrillic, we should remember that the above discussion has concentrated on transliteration of Russian. There are differences in transliterations of Cyrillic (even in Library of Congress) among the different Cyrillic-spelled Slavic languages (don't ask me about Cyrillic that is used for non-Slavic languages. Oy!).

But, since Russian composers seem to predominate, it makes sense to begin there.

Were it much easier to use diacritics on the Web (or with my computer, take your pick), I'd prefer to transliterate Russian using haceks, etc. (and "J"!), as with Croatian and Czech. But that's just me.

The major problem with the Library of Congress system is that one can omit the diacritics (especially the ligatures for "iu" and "ia") and leave a highly ambiguous result. For instance, the ending -ий --> -ij (my preference) or -ii (LC without diacritics); whereas the ending -ии --> -ii (no diacritics needed in either system).

For practical reasons, my dissertation on Cui's operas adopted the LC system without diacritics for those passages requiring transliteration, but I wish I could have used something better. (Fortunately Cyrillic was available, too.)

Carolus wrote:

We really need to stick with the most widely used transliteration for composer names, which I expect is the one employed by Library of Congress

Unfortunately the LoC versions of Russian names can be wildly inconsistent with their own transliteration system. For example, they use "Tchaikovsky, Peter Ilich" but they have the composer's brother as "Chaĭkovskiĭ, Modest Ilʹich" The New Grove also has major differences from the LoC standards, and Baker's dictionary.

As a relative newcomer to IMSLP I don't know if it's by accident or design that most Russian composers' names happen to correspond to the versions used on Wikipedia (which often differ from the LoC's preferred standard, e.g. Pyotr Ilyich Tchaikovsky). The folks at Wikipedia have had a little longer to thrash out all these issues, and in most cases seem to have achieved a good degree of consensus. Some of their manuals of style make interesting reading for us when considering what we should do (especially the "Talk" sections):

http://en.wikipedia.org/wiki/Wikipedia: ... _of_music)
http://en.wikipedia.org/wiki/Wikipedia: ... s_(operas)
http://en.wikipedia.org/wiki/Wikipedia: ... s_(people)
http://en.wikipedia.org/wiki/Romanization_of_Ukrainian
http://en.wikipedia.org/wiki/Wikipedia: ... of_Russian