Wiktionary talk:Transliteration

Currently just has chunks of relevant discussion from Beer Parlour, RFD etc.

Also see Wiktionary talk:Policy - Transliteration/Collection of past discussions Richardb 15:03, 7 Feb 2005 (UTC)

transliteration

The act or product of transliterating, or of expressing words of a language by means of the characters of another alphabet.

Why was a substantial part of this policy deleted?

==Quotations== Quotations should be in a form readable, understandable by the modern English reader, whilst preserving the quotation of the word in its orignal script.

This can be acheived in one of two ways:-

Quote the whole thing in original language and script, with the word in question in bold, followed by a translation, with the word transliterated and in bold

If the quotation is in something semi-intelligible to the modern English reader (perhaps Middle English, simple French etc), then the quote can be in it's original language, with accents, with the word in question in bold, and a transliteration provided in brackets immediately after the word.

==Translations== Translations into languages using other scripts should also provide a transliteration eg: Fuck, in Bulgarian, is: еба (eba).

Might also be nice to provide a pronunciation too, for use as a "phrase book". See Where is the toilet? and see which phrases are usable as a guide to saying the phrase ?

How old should quotations be?

Hello, wiktionarians. I read Wiktionary:Quotations just now, and have a question for you about it. It implies that one of Wiktionary's goals is to document the span of time over which words have been used. Is there a cut-off date for this? Should quotations from Old and Middle English be included? (This question might have wider implications. Which language does this dictionary describe? English in its full history, or only Modern English? Or Middle and Modern English, since Old English has its own Wiktionary?) --znusgy 11:53, 7 Feb 2005 (UTC)

This is the English Wiktionary. It is intended to be readable for current English users. You can put Middle or Old English words in here, just as you can put French, Khmer or Chinese words, under the correct language heading. But, the definitions, quotations etc must be readable by current English readers. If you want to write defintions, explanations, names of parts of speech, quotations in another language (including Old or Middle English), then go to the Wiktionary of that language, if it exists. So, please translate the quotations into English that is readily intelligible to current readers, as much as is necessary, but preserving the purpose of the quotation to illustrate the usage of the word. ((Just my view))--Richardb 14:37, 7 Feb 2005 (UTC)

portokali

Do we need a link that phonetically spells a Greek word with English letters? I fixed πορτοκάλι (not completely) to which this page links, but I don't know that this page will serve any purpose. ???--Alia H 03:38, 4 Feb 2005 (UTC)

• Although I argued to some extent in favour of the idea some time ago, I've done nothing more about it. There are arguments for this in terms of making other scripts more accessible to more readers. Since I have no plans in the forseeable future to develop this idea, I'm not going to complain much if the community wants to dump this. Eclecticology 08:54, 4 Feb 2005 (UTC)

Transliterations might be worth pursuing, but this is a bit of a minefield, in my view. Just sticking with (modern) Greek, would we allow for, say, υ (upsilon) to be transcribed as u, y and i? What about accented characters? "Πορτοκάλι" could become "portokali" or "portokáli", but some words could have numerous transliterations (take "ευχαριστώ" - υ could be u or f, χ could be kh or ch, ώ could be o, ó or even w...). What about Chinese and Japanese, with their various transliteration schemes? I think it all risks getting a bit exponential, with each word requiring several redirects, unless, of course, we come up with standard transliteration schemes and require users to stick to them.

Anyone who wants to find a translation for a word they find on the web can cut and paste it into the search box. Anyone who wants to add a foreign word can use the "Characters:" box at the bottom of the edit page (although this provides only for European languages at the moment).

These facilities don't cover all the bases by any means, but I think the amount of work involved would prevent automatic handling of transliteration from getting off the ground.

— Paul G 10:00, 4 Feb 2005 (UTC)

Help me out, please. Was a consensus ever reached? I see after randomly hitting paidika, and doing some research that this had already been debated back in August. I only just recently stumbled upon Wiktionary, so I missed it, and I cannot find it in the current Beer Parlour discussions. Meanwhile, what do we do with paidika, which doesn't even have the Greek word anywhere on the page? (Sorry I know this discussion probably doesn't belong here any more, but this is where it is, so this is where I'm asking.)--Alia H 05:50, 6 Feb 2005 (UTC)

I strongly disagree with PaulG. This is the English Language Wiktionary. The entries should be readable to English readers. So Transliterations provide a useful tool. A Greek phrase book for English users phrasebooks would not have the Greek Phrase in Greek alphabet, 'cos it would be useless to the English reader. Same goes here. don't get rid of the transliterations, they are vitally useful.--Richardb 14:46, 7 Feb 2005 (UTC)

I'm not trying to be contrary, but I'm not quite understanding the point of transliterations. (This is the first time I've run into them, except for "borrowed" words; which of course are transliterated and then considered English.) Are they for people who speak but don't read a foreign language? I guess I don't see how they are "vitally useful." If you are a native Greek speaker, learning English, you would look up a word in Greek to find its translation. And if you are a native English speaker learning Greek, you would look up the word in English to find its translation. Either way, you should be familiar with both alphabets.

I guess the only practical way to go about this would be to assign letter for letter transcriptions so that υ would always be u even when it sounds like f. Because you would need f to stand for φ.--Alia H 03:20, 8 Feb 2005 (UTC)

• <Jun-Dai 00:41, 15 Jun 2005 (UTC)>

Transliterations are an important feature for any non-phonetic language. This allows people who are not fluent in the language to look up terms. I intend to make these changes to the policy page -- comment if you think I shouldn't: Wiktionary:Policy - Transliteration/Proposal20050614

</Jun-Dai>

anurag

Claimed to be a Sanskrit, but Sanskrit does not use the Latin alphabet. This looks like a transliteration. It needs to be replaced with a page in Sanskrit script. — Paul G 10:11, 7 Feb 2005 (UTC)

Hang on! This is the English Wiktionary ? We allow foreign words, as long as they have explanations in English. This fits that description. It may then have a translation which uses the Sanskrit script, and a link to the Sanskrit Wiktionary (??) written in Sanskrit. Or have I missed something ?? It just needs Wikifying, with a ==Sanskrit== language heading.--Richardb 14:26, 7 Feb 2005 (UTC)

Yes, this is the English wiktionary. Yes, Sanskrit words are to be explained in English. But the Sanskrit word is spelled अनुराग (transliteration: anurāga) not "anurag". —Muke Tever 19:57, 8 Feb 2005 (UTC)

Transliteration heading/tag?

Is there a transliteration tag in foreign language entries? If so, then the entire concept mimics that of English plurals and other senses: yes, the other entries should exist (usually as redirects.) Therefore, the headword that is redirected to in this case would be the un-transliterated version of the word. In the headword entry, multiple tranliterations could then be used. As many transliteration redirect entries as desired could also be added. --Connel MacKenzie 18:14, 7 Feb 2005 (UTC)

Romanizations is the tag/heading I was thinking of. Aren't these the same things? --Connel MacKenzie 03:33, 8 Feb 2005 (UTC)

I agree with Connel, mostly. In places wher ethe word is really only a foreign script word.--Richardb 12:46, 8 Feb 2005 (UTC)

But what about a word like Rouble. (Or is ruble) It's undoubtedly a Russian word, so should the main entry be in Cyrillic ? But for years, maybe hundreds of years, we have used it as though it were an English word, or ast least written it in Latin/Roman/Western characters.--Richardb 12:46, 8 Feb 2005 (UTC)

I think category:English borrowed words is a better place to address rouble, as it seems to be part of the English language now. Very tough to draw the line, I agree. In such a commonly used word (ha!) as rouble I'd like to see both the Romanizied and Cyrillic words each with their own entry, linked by ===See also===. --Connel MacKenzie 14:24, 8 Feb 2005 (UTC)

Example of the problem of those who refuse to acknowledge the need for transliteration for the non-cognoscenti amongst us

See Играть- This is the English Wiktionary, yet this page is largely unintelligible and useless to an English reader or speaker not versed in reading Cyrillic.

There is no pronunciation or even transliteration for an English speaker or reader to use.

It is largely only useful to a Russian reader.

It surely has no place in an English Wiktionary.--Richardb 11:08, 22 May 2005 (UTC)

Although I have some issues about details in the article, I think that the general presentation is very good. Pronunciation and/or transliteration is provided for Играть itself; that should be sufficient. Of course, if you want to add the others in the conjugation table by all means go ahead. My preference would be to turn all these forms into links so that the transliterations could be shown on each of the pages.

I don't think that intelligibility to the English reader is the point here. A person without at least a rudimentary understanding of the Cyrillic alphabet is unlikely to need the conjugation table.

Removal claim

I don't see the point of Richardb's claim on my user talk page. I did revise the page in March, and I'm surprised that you did not notice it until now. I was led to the page by the discussion over portokali on the RfD page. What may have been lost didn't seem necessary. Eclecticology 21:20, 22 May 2005 (UTC)

Proposed revision

<Jun-Dai 00:44, 15 Jun 2005 (UTC)>

Provide comments, etc. Tell me not to put it up, or I'll do so in a few days (of course, we can still talk about it at that point). Also, feel free to make changes. It's only a proposal, after all:

Wiktionary:Policy - Transliteration/Proposal20050614

</Jun-Dai>

<Jun-Dai 01:55, 16 Jun 2005 (UTC)>

After a day of hammering out the Japanese romanization rules, I got impatient and implemented my proposed changes. I've kept the original proposal, which would probably be worth deleting in a week or two, in case anyone objects and reverts the changes.

</Jun-Dai>

pinyin (Shijian)

From the proposed changes page: (Jun-Dai 01:47, 16 Jun 2005 (UTC))

I have no serious complaints since most of this is similar to what I had in mind. There are some minor wording changes that I would now make, but that's often the case with my own work after I have let it lay for a while. The "language consideration" format for the article title should probably be the same for Greek as what you now have for Chinese and Japanese. This will allow us to develop some consistency. Eclecticology 08:09, 2005 Jun 15 (UTC)

Could you also look at Shijian as a possible way of dealing with the pinyin issue. Eclecticology 17:33, 2005 Jun 15 (UTC)

<Jun-Dai 01:53, 16 Jun 2005 (UTC)> Unfortunately I never studied Chinese for more than two weeks, so I'm really not qualified to determine whether the system implemented on Shijian is good or practical. Given that it's a flat romanization (without tonal diacritics), the question remains as to whether the pages with the diacritics should redirect or not. I'd say no, but I don't really know what's best there, and there certainly a large number of "translations" to pinyin romanizations with diacritics (such as giblets) that would have to be fixed. </Jun-Dai>

Surely someone else will need to do the work on the Chinese articles, and once we fing that person he/she may have his or her own ideas. Flat romanization seems to be the best. Diacritics with four tones on each syllable and two ways of representing the tones would mean an additional 4*4*2=32 pages for each disyllabic. A three syllable word could have 128 pages! Eclecticology 07:46, 2005 Jun 16 (UTC)

Since Mandarin is the dominant literary language, and every Mandarin dictionary I know uses Pinyin with tone marks, I would follow their lead. The only other Chinese language/dialect which has a good number of dictionaries that I've seen is Cantonese but I've not been able to discern if there is a norm amongst those.

Another issue of variance in Chinese transliteration, even for just the case of Pinyin with tone marks, is syllable division. There is (almost famously) no standard of whether to include spaces in polysyllabic terms. Some sources use spaces between each and every syllable, some seem to write without spaces all syllables which would form a dictionary headword, some seem to be arbitrary between these two extremes. One more point of variance is capitalisation of Pinyin. The general rule seems to be to retain the capitalisation of the source language - but I don't know how precisely this is followed. Many English words appear capitalised and noncapitalised - eponymous word especially; and if there are Chinese words borrowed from German I doubt that all nouns would be capitalised... — Hippietrail 08:12, 16 Jun 2005 (UTC)

Pinyin needs to be looked at as a romanization rather than a transliteration. Although it largely reflects Mandarin speech patterns it is still not the language itself. The same pinyin romanization would be used in works set in any language using a roman alphabet. The value of pinyin is not that it is more correct than the other romanizations but that it presents a common reference point that allows us to escape from the previous patchwork. Pinyin should not be viewed as a pronunciation guide; that belongs in the body of the article for the characters where a whole range of dialectical differences can be presented.

The value of stripped down titles is to enable people to find things. A dictionary is useless if nobody can find the words they seek. Chinese tones or Russian stress marks need to be explained -- just not in the article titles. A distinction needs to be made between what is really a part of the word, and what is added to make it more understandable.

When you look through a Chinese dictionary you soon see that a majority of Chinese concepts are built on two characters being linked together. The Chinese dictionary that I mostly use (admittedly an old one that still uses Wade-Giles) gives no romanization for the disyllables at all. A native Chinese speaker knows which syllables go together before he learns to read and write; he would also know these even if he remained illiterate for his entire life. It is not obvious to an outsider. In running text an outsider would tend to put the wrong syllables together and end up with resultsa akin to the difference in English between "ice cream" and "I scream". I think we would do better by continuing to represent these terms without spaces, or alternatively with hyphens.

Capitalization is a non-issue. Capitalized Chinese characters don't exist. Thus when we speak of capitalized pinyin we are grafting the standards of some European language to a place where it does not belong. Eclecticology 18:18, 2005 Jun 16 (UTC)

Languages with no standard transliteration schemes

Under the heading "On Wiki-romanization" I see this: "Because most languages have multiple standards for romanization..." but I cannot find anything for cases such as Thai and Khmer and, I think, Burmese Lao, and Tibetan. Languages which I believe have no accepted standard transliteration and which each book dealing with them generally devises its own system. Does anybody have any ideas what to do in these cases? — Hippietrail 08:02, 16 Jun 2005 (UTC)

<Jun-Dai 15:01, 16 Jun 2005 (UTC)>

I don't know enough about the individual languages to speak to them, but as for a general approach, we should devise our own standard based as much as possible on (1) what people would expect and (2) what seems most effective. Our main goal is to be consistent. If we cannot meet all of the varied expectations on what rules our transliterations will utilize, then we must aim for simplicity, effectiveness, and consistency, so that at least the user can come to develop an understanding of how to look these items up in the Wiktionary.

I changed "standards" to "systems".

</Jun-Dai>

I'm inclined to agree with Hippietrail that many of these languages do not have accepted standards for transliterations. We probably will need to devise our own system that will work for us; the guiding principles that Jun-Dai outllines are spot on. If it works for us others would be free to use it in their own situations, and our way would become the standard, but not because we imposed it. WP is already big enough to start influencing the way that others do things. None of this means working in a vacuum. The Library of Congress already has a series of charts that it uses for transliterating the titles of books in the course of cataloging them. This is available on-line, and would be a reasonable starting point. Eclecticology 18:34, 2005 Jun 16 (UTC)

The Library of Congress is a very good suggestion. I would also recommend quoting, at least in a comment, which book you took it from when using a print dictionary or pedagogic material. — Hippietrail 19:07, 16 Jun 2005 (UTC)

Some stuff moved from elsewhere

--Richardb 14:28, 7 March 2006 (UTC)

Voting for Wiktionary:Policy - Transliteration

I would propose that the Wiktionary:Policy - Transliteration policy draft needs to be voted on to make it policy and not a policy draft. It seems that there is much confusion about how transliteration and Romanization should work and having this policy to point at would be very helpful.

Here's what I'm proposing, the voting should be open for two weeks, all registered users can vote, simple majority will carry. The question is simply should this be policy yes/no. The voting would happen on the talk page.

If someone will second my proposal for a vote then I think we could get the polls open. If you think the policy needs more discussion then vote against it. If you think it's good (or at least workable) then vote for it. Kevin Rector 05:17, 5 May 2005 (UTC)

It's a good idea for us to define this policy, but I don't think we're ready enough for it. It would help us try to be consistent, but anything we came up with would not be satisfactory for long, particularly as we would not be likely to get more than a half-dozen opinions on it, which would only be a little better than zero. Even more important, the policy is little more than a stub, so there's not much to vote on. Better that we focus our efforts on collaborating on the policy--possibly by voting on individual details of it--and save large-scale voting on it for later. Jun-Dai 07:18, 5 May 2005 (UTC)

Kevin, you have my strong support if you want to take a stab at further developing "policy" through the suggested stages of Policy Think-Tank, Draft Policy, Semi-Official Policy, Official Policy. Anything to get the discussion of what clearly needs to be a policy out of the ever growing, far too big Beer PArlour. Far too many people are far too happy chit-chatting for ever in the Beer Parlour, and never actually want to commit to doing the basic ground work in some Committee Room. Discussion should then take place in the "Policy Committee Room" (or relevant Talk page) of a particular policy idea.

Whilst I have a pretty strong view that some ideas on transliteration policy are definitely up there in cloud-cuckoo land, designed to suit the academic pointy heads in this world, I actually feel less strongly about any particular policy than I do about, for god sake people, get off the fence and start defining some issues as policy matters, not matters for never-ending idle chit chat in the Beer Parlour.

As with any Wiki Project, no-one can impose rules anyway. And taking an issue out of the Beer Parlour into a "Policy Think Tank" talk page hardly constitutes imposiing rules on people. If I wasn't such a part-time player these days, I'd be stripping so much never-ending chit-chat out of Beer Parlour and putting them into "Policy Think Tank" talk pages.--Richardb 02:02, 22 May 2005 (UTC)

Move to Wiktionary:Romanization

Transliteration, applied strictly, means changing individual letters, and can only be applied to alphabetic writing systems. The result can also be in any alphabetic writing system, e.g. Latin, Cyrillic, Korean, etc.

Romanization:

1. Can be applied to text in any writing system: alphabetic, syllabic, logographic, etc.
2. Results in a Latin-alphabet rendering of the source text: it is useful to English-language readers.

I suggest we move this page to Wiktionary:Romanization to more accurately reflect its intent. —Mzajac 22:39, 6 March 2008 (UTC)

Purpose

The purpose of romanization in Wiktionary is not clearly defined here, and as a result this policy is being applied inconsistently. This is compounded by the definitions in the policy, which are inaccurate. I'd like to fix the definitions, define the goal of transliteration with them in mind, and adjust the details of the policy to fit the goal.

Definitions:

• Transcription is the written representation of spoken language. Wiktionary uses the International Phonetic Alphabet (IPA) to represent pronunciation. More specifically, we use phonemic (phonological) transcription to represent the general pronunciation of a word, and phonetic IPA transcription to represent the precise pronunciation in a particular dialect. Details at Wiktionary:Pronunciation.
• Transliteration is the conversion of letters or words into another alphabet. Some transliteration systems may convey the pronunciation of words to some degree, to readers of a particular language (not necessarily English). Other transliteration systems are intended strictly to represent the original orthography. The term generally does not apply to non-alphabetic writing systems which use ideograms or logograms.
• Romanization is the conversion of text from another writing system into the Latin alphabet. Like transliteration, a particular system may or may not convey pronunciation to some degree.

Illustrative examples:

• w:ISO 9:1995 is a standardized United Nations transliteration system for Cyrillic text, based on scientific transliteration. It applies a strict one-to-one correspondence between letters (graphemes) in Cyrillic and Latin. It is the same for all languages, allows faithful reverse transliteration, but disregards pronunciation.
• w:Scientific transliteration is the most common transliteration system used in linguistics today to romanize Cyrillic text. It was originally intended for use in German libraries, and is based on Croatian Latin orthography (closely related to Serbian, which is written in both Latin and Cyrillic alphabets). Cyrillic orthographies are strongly tied to phonemic principals, and so scientific transliteration rules are modified for each language, and the result corresponds to a phonemic transcription. Its pronunciation is intuitive for many central and eastern Europeans, but not so much for English speakers.
• Appendix:Russian transliteration is scientific transliteration for Russian.
• Wiktionary:About Russian#Romanization is based on the same system, but with extensive modifications intended to convey pronunciation, but which lose the original orthography. These modifications seem to be based on Wiktionary editors' whims rather than on any published reference.

As mentioned above, IPA is already used to represent pronunciation, optionally including syllabification and stress. (Hyphenation in writing is separate from syllabification in speech.) See Wiktionary:Pronunciation.

Romanization should be used in Wiktionary to convey the original orthography to English-language readers, who may be unfamiliar with the original writing system. We list English headwords in standard English spelling, not in IPA for the same reasons.

These principals should be approved by consensus as soon as possible. —Mzajac 23:29, 6 March 2008 (UTC)

For centuries, transliteration has been necessary to represent languages that use a different script, mainly because in the places where a language is not spoken or written, there are no facilities for writing it. Transliteration schemes were devised so that a language such as Greek or Russian could be written in Roman letters, and then a knowledgeable reader could reliably convert it back into the original script. Just in the last decade or so, this need has fallen away. On Wiktionary, a transliteration should help a learner with pronunciation, and the transliteration no longer needs to be etymological or be able to be converted back into the original script, because the original script is actually provided. Etymologically, a Korean ㄹ is "R", and in some environments it is normally pronounced "R", but in others it is pronounced "L"; in a few words, it is pronounced "N". We can actually transliterate ㄹ as "R", "L", or "N", as the case may be, and not worry about how the "L" or "N" might be reconverted to Korean, because the original Korean word is provided in the original script. The purpose of the transliteration is no longer restrained by the need to revert it, it only needs to help someone who does not know the script to pronounced a word in an approximate fashion. —Stephen 01:00, 7 March 2008 (UTC)

I disagree. This has not changed in centuries.

The opposite is the case. "Facilities for writing it" used to be a pen which everyone had, now it consists of a foreign-language keyboard layout, which most people can't use. But foreign text in a book are not romanized so I could "convert it back into the original script".

Romanization is used to permit comprehension, and this need hasn't fallen away. (Q: Hey English-language reader, how do you spell beetle in Ukrainian? A: I can't read it, but it looks like a zig-zag thing, a letter y, and a letter k. Wrong, you spell it zhe, u, ka.) "Zhuk" or "žuk" has some meaning to an anglophone, but жук has none. He can even read the word if it is spoken with sounds, or clicks, or tones that don't exist in his language, or that he can't even distinguish. A written word is a thing, and it exists thanks to its orthography. When you don't understand the orthography, you need a version of it in a familiar writing system—in our case, an orthographic romanization. If the romanized version helps perceive the pronunciation, that is a bonus, but not a necessity.

A spoken word is not the written word. It exists thanks to its sound, and can be represented in writing by one or more pronunciation guides. These belongs only in the pronunciation section of a Wiktionary page.

A headword is the written word. A pronunciation is a phonetic transcription. Two different things, which belong in two places. —Mzajac 01:41, 7 March 2008 (UTC)

Hm, maybe it's much simpler to address Stephen's argument according to its context rather than the esoteric details:

If you are asserting that Wiktionary needs pronunciation guides and no transliterations, then you should propose the removal of the transliteration guidelines. "Transliteration [which] should help a learner with pronunciation" belongs in the pronunciation section of a page.

Until such a proposal is accepted, Wiktionary:Transliteration remains a separate thing from Wiktionary:Pronunciation. It has a different function, and belongs in a different place. So let's discuss what exactly it is. —Mzajac 00:43, 8 March 2008 (UTC)

They don’t have to have "facilities" for writing it, they can copy the original word or phrase right from the page. It is located prominently next to the transcription. As for transcriptions being used for comprehension, you are not taking the English Wiktionary audience into consideration. Native Russians who want to look up a Russian word look to the Russian Wiktionary. English Wiktionary is for English users, and English-speaking linguists and those who have a working knowledge of Russian use the Cyrillic, not the transcription. Except for checking to see where the accent falls, nobody uses the transcription except casual users who are not linguists, do not know Russian, and who do not wish to waste a half hour learning Cyrillic.

I do not assert that Wiktionary needs pronunciations and no transliterations. IPA pronunciations are useful to linguists and a handful of others. Most people who look up a Russian word here, and who do not already know Cyrillic, are not really interested in a perfectly pronounced word, but just what we have been putting. It is better to use the word transcription than transliteration, because transliteration suggests a one-to-one exchange. Transliterations were needed in the past because the word in the original script was not available, or not printable. That is no longer true, and anyone can copy the original Cyrillic word even if they have no Russian keyboard. The etymological transliteration no longer serves a useful purpose, except perhaps in e-mail clients that are not correctly configured for Cyrillic. When people type Russian in Roman letters in e-mails, they spell it all sorts of ways. —Stephen 15:49, 8 March 2008 (UTC)

You seem to know a lot about why people use English Wiktionary and why they don't. I think you're stating a bunch of unfounded assumptions. If I am "not taking the English Wiktionary audience into consideration," please show me where you found such detailed information about every one of their needs.

"...they can copy the original word or phrase right from the page"—d'ya think some people use Wiktionary for more than copying words? Transliteration has been used alongside Cyrillic by linguists much smarter than us for over a century, and is still being used in both print and electronic documents.

Without transliteration, I need to learn two different Cyrillic alphabets to understand the etymological relationship of Polish jego, Russian его, and Ukrainian його. Perhaps it would be easier to just include transliterations: Polish jego, Russian его (jegó), and Ukrainian його (johó). The addition of a phonetic transcription in the "pronunciation" section would add value, noting that the Russian word is pronounced as jevó (/jɪˈvo/). —Mzajac 00:45, 9 March 2008 (UTC)

For one, English Wiktionary is intend for use by native English speakers. Native Russians are welcome, but they have to understand that the Russian pages are tailored to the needs of English-speakers who speak little or no Russian. For two, perhaps it would be better if you found a bunch of native English speakers who also speak Russian, but who nevertheless need roman transcriptions. I do not believe there is a significant number of them...I don’t believe there are any of them.

Certainly people use it for more than copying words. But anyone who needs etymologically true Russian needs the Cyrillic. Transliteration has been used for centuries, but the reasons are the lack of facility to port or print the original Cyrillic. Unicode has removed this ancient need.

Anyone who needs to understand the etymological relationship of Polish jego and Russian его only needs the original script. I cannot imagine the Slavic etymologist who is not comfortable in Cyrillic. Nobody needs the transcriptions except for casual users who do not speak Russian and who do not know Cyrillic and who do not wish to spend half an hour learning. —Stephen 01:03, 9 March 2008 (UTC)

Are you saying we should not romanize words from other writing systems in Wiktionary? —Mzajac 08:21, 9 March 2008 (UTC)

No, I am not saying that. I have no idea how you reached that conclusion. I was about to make a suggestion that had occurred to me this weekend but now I see that the discussion has turned ugly. I am bowing out of it. —Stephen 20:08, 10 March 2008 (UTC)

Restarting the discussion

I realize that I've been getting to wound up about this. I apologize to all for my tone. I'd still like to try to clarify my concerns about transliteration, if anyone would still be willing to listen to me.

I still think this policy needs some clarification, so that transliteration can be used consistently in the various languages. I realized this week that Appendix:Ukrainian transliteration and Appendix:Russian transliteration are inconsistent in their intent and implementation, and I couldn't find any clear guidance here to resolve the conflict. Or maybe the answer is creating a framework to make it easy to add supplementary transliterations in different systems, to accommodate different needs. —Mzajac 21:17, 10 March 2008 (UTC)

Key terms

The "Key terms" section has some inaccuracies and omissions. I hope the following is an improvement. Any specific objections? —Michael Z. 20:18, 18 March 2008 (UTC)

• Writing system (or script): native representation of a language in writing or print. Types of writing systems include alphabets, abugidas, abjads, syllabaries, and pictographic, logographic, and ideographic writing systems.
• Romanization: rendering of written text from a foreign writing system into the Latin (Roman) alphabet, possibly supplemented by diacritical marks or additional characters.
• Romanization system: standardized romanization systems exist for most languages, used in linguistics, library science, geography, publishing, government and legal documentation, and other fields. For a list, see w:romanization.
• Wiki-romanization: a romanization system chosen for Wiktionary. It is usually a common standard of romanization, or based on one and modified for Wiktionary's specific needs.
• Transliteration (literally 'lettering across'): rendering of written text from one alphabet or syllabary into another, letter by letter. In Wiktionary we are mainly concerned with transliteration from a foreign system into the Latin alphabet (a subset of romanization).
• Transcription (literally 'writing across') has several meanings, including transliteration and phonetic transcription, the written representation of spoken language. For the latter, see Wiktionary:Pronunciation.

It's been two weeks, so I will update the project page. —Michael Z. 03:03, 4 April 2008 (UTC)