Wiktionary talk:Votes/pl-2014-07/Allowing well-attested romanizations of Sanskrit

Attested or mentioned?

Attested in usage or mentioned? If it's used then it's a normal English word. We already have a number of such already, particularly in IAST. --Ivan Štambuk (talk) 21:22, 26 July 2014 (UTC)

My thinking here is directed towards words like mahā, for which citations like this exist - which, in the RfD, were deemed by the community to be insufficient to have an entry in any language under our current rules. I think the use/mention distinction becomes a bit fuzzy where the word might occur as part of a foreign phrase or in a string of foreign words as part of an otherwise English text. bd2412 T 00:54, 27 July 2014 (UTC)

mahā there is used as an ordinary English word. There are gazillions of such Sanskritisms used in every text dealing with Indian religions, history and languages (e.g. grammars). They are allowed per CFI as any other English words. --Ivan Štambuk (talk) 09:51, 27 July 2014 (UTC)

The thrust of the opposition at Wiktionary:Requests_for_deletion#mahā was that, for example, "mahā is not a word in any language". bd2412 T 14:55, 27 July 2014 (UTC)

The opposition seems to be confused. You argued in the RfD for the restoration of mahā as a transliterated Sanskrit term used for Sanskrit in Latin script, when in fact we're dealing with a word used in English. This is completely different from the case made for Gothic.

The same person who claimed that mahā is not a word also claimed that महा (a form which only occurs in compounds and never on its own in isolation) is, so I wouldn't bother too much with what was said there. There are citations, it passes CFI, end of story. --Ivan Štambuk (talk) 15:45, 27 July 2014 (UTC)

Reading this section, and recalling all the previous discussions, I can't escape the observation that every user seems to have a different view of what constitutes a "word", what constitutes a "use" and/or what constitutes "English". I'm sceptical that enough users would find the citations at Citations:mahā to be sufficient for it to pass RFV as an English word. (Re महा: I was operating under the assumption that महा is a word; if it's not, that's a separate problem. Merely being limited to compounds does not necessarily make it not a word, however; compare prefixes like ent-.) - -sche (discuss) 21:31, 14 August 2014 (UTC)

Perhaps the first thing we need is an RfC on the meaning of "word". bd2412 T 12:33, 15 August 2014 (UTC)

Unlike lesser evolved languages Sanskrit has formalized rules for compounding and any nominal (noun, adjective, participle) can be used in compounds. We don't have separate entries for bound morphemes except for "well known affixes" like the one you linked for German, so I don't see why should the same be valid for Sanskrit. mahā is not a word on its own - outside its context it doesn't exist. Admittedly, helping users desandhify Sanskrit sausages could be a goal by itself should we decide to do it, but that could increase the number of entries for any word by 5-10 times. --Ivan Štambuk (talk) 14:07, 16 August 2014 (UTC)

I'm not sure I understand what you mean. What, exactly, is it that would be increasing the number of entries by 5-10 times? bd2412 T 15:13, 16 August 2014 (UTC)

When forming Sanskrit compounds members (word stems) undergo sandhi, i.e. the final sound (or sounds) of the first member changes depending on the initial sound of the following word. This also occurs in ordinary language. Normally Sanskrit nominal lemmas are in bare stem form and we don't have to worry about that (except in case of consonant-stems but they're an insignificant minority), as we have no inflected forms at all for the same reason, but if all Sanskrit compound members were to be added as separate words in all forms they appear or could appear, that would significantly increase their number. Though in case of mahā the concern is legitimate because that seems to be some kind of defective stem worthy of a separate dictionary lemma. --Ivan Štambuk (talk) 15:54, 16 August 2014 (UTC)

Under this proposal, attestation would be required first. We would not have theoretical forms, only those for which use could be demonstrated. bd2412 T 17:24, 16 August 2014 (UTC)

Going forward with this vote.

Absent any further comments, I intend to go forward with this vote. Cheers! bd2412 T 19:04, 14 August 2014 (UTC)

Frankly, I don't think this is a very good vote. A better vote would be to allow IAST romanizations of attested non-romanized forms; on the previous vote, some people opposed because any romanization, not just IAST, would be included. But you can give it a try. I don't think I am going to oppose, but if I come up with a reason, I will. --Dan Polansky (talk) 19:18, 14 August 2014 (UTC)

Allowing well-attested romanizations does not foreclose allowing an expanded selection in the future. It is a narrowly defined issue. bd2412 T 19:35, 14 August 2014 (UTC)

By the proposal, you allow English-sentence-attested non-IAST Sanskrit romanizations as well, as "romanization of" entries. I am not sure people will like it. They may say that these are in fact English words anyway. Remember the RFD now archived at Talk:tovarish? --Dan Polansky (talk) 19:56, 14 August 2014 (UTC)

I am aware of that. It is odd that no one is arguing that zhāng and shima are merely English words, but then again, no one is arguing that internal consistency is our strong suit. bd2412 T 20:53, 14 August 2014 (UTC)

An IAST romanization may be an English word and a romanization. Via English-sentence-attestation, it is an English word; via the corresponding non-romanized form attestation, it is a romanization. You want to use English-sentence-attestation as a basis for inclusion-as-romanization rather than inclusion-as-English, which I find odd. --Dan Polansky (talk) 20:58, 14 August 2014 (UTC)

Hmm. If one accepts the idea that citations can't "double count", i.e. that a particular instance of a string as an English word that belongs under an ==English== L2 cannot simultaneously be an instance of that string as romanized Sanskrit that belongs under a ==Sanskrit== L2, then I think the wording of the vote is possible to interpret in a coherent way. (I admit it's liable to be interpreted differently by different people and so perhaps suboptimal in that way, but it's not immediately clear to me how it could be reworded.) It concerns itself with "citations [...] showing that a romanization of a Sanskrit word [exists]", so citations showing that an English word exists continue to do just that. And "a romanization of a Sanskrit word" can be "well-attested in English text" without being English: it can for example be mentioned and then glossed, like mahā in the 2013 at Citations:mahā. And the vote (correctly, IMO, given its aim) only requires romanizations to "appear", not to be "used". - -sche (discuss) 22:14, 14 August 2014 (UTC)

Now that you highlight the word "appear", I see things about the vote I don't like. It uses the term "well-attested"; what is that? We have the word "attested" defined, not "well-attested". And then, a bracket after "well-attested in English text" states something that looks like our good old WT:ATTEST until one notices the word "appear". I don't like this. --Dan Polansky (talk) 22:26, 14 August 2014 (UTC)

Well-attested is defined in the vote - appearing in three independent texts across three years. I don't think that is any different from our current CFI. This presents a higher standard than that, not a lower one (since allowing merely CFI-attested transliterations has already been rejected). However, if it is a sticking point, I'll be glad to change it to "used to convey meaning in", since I consider that to be synonymous. bd2412 T 02:11, 15 August 2014 (UTC)

CFI uses the term "attested" in the same meaning, from what I can see, so you should not be using any different term, not even slightly different. --Dan Polansky (talk) 03:43, 15 August 2014 (UTC)

@-sche: And "a romanization of a Sanskrit word" can be "well-attested in English text" without being English: it can for example be mentioned and then glossed, like mahā in the 2013 at Citations:mahā - In the 2013 cite mahā is used in English, just as it is in all the other citations. Unless this vote argues that we reformat IAST-spelled English Sanskritisms as ==Sanskrit==, I don't see its purpose. --Ivan Štambuk (talk) 14:13, 16 August 2014 (UTC)

You think it's English; I'm sceptical. But it's simple to come up with other, clearer examples of romanizations being found in English texts not as English words: "'Subham janmadinam,' he said when I entered." Several of the citations at User:-sche/svobodnyx/Citations are romanizations of Russian words found in the middle of English texts. The aim of this vote seems to be to let the existence of citations like those (except for Sanskrit, not Russian, obviously) support the existence of soft redirects to the Devanagari spellings. - -sche (discuss) 17:12, 16 August 2014 (UTC)

But these are not usages in English, but Russian. This are complete Russian sentences, not individual Russian words used in English as is the case with Citations:mahā. --Ivan Štambuk (talk) 18:12, 16 August 2014 (UTC)

Do you think zhāng is as English as tovarish? As the RFD showed, tovarish is used in the middle of otherwise English sentences, and is inflected (tovarishes), etc. And other dictionaries, e.g. Merriam-Webster, treat it as English. I have doubts that the same is true of any sense of (lowercase) zhāng. (I find it plausible that Zhāng might be attested similarly to Zhang and so might merit a surname entry. But only citations would determine that for sure.) - -sche (discuss) 22:14, 14 August 2014 (UTC)

That's on Google. They need to step up their game when it comes to searching for words distinguished by capitalization. bd2412 T 02:16, 15 August 2014 (UTC)

Software alternative?

It seems to me that what those not proficient in Sanskrit would need would be a reliable means to proceed from any (or "any common", or "any standard") transliteration, especially those with no diacritics, entered in a search box to the actual Sanskrit entry, in Devanagari. This is work that ought be be done in software because it can be done more completely than any insertion of Roman-character transliteration entries. I don't doubt that it would be difficult and that it would be error-prone, but it seems a worthwhile goal and likely to become achievable in a reasonable period of time (ie, less than 10 years). If we want, in the meanwhile, to add a few hundred or a few thousand transliteration entries, it could be permitted. When the technical solution I imagine and hope gains support from those capable of implementing it, the enthusiasm to add such entries would rapidly dwindle. DCDuring TALK 18:46, 7 April 2015 (UTC)

Do you mean something like a redirect? bd2412 T 19:07, 7 April 2015 (UTC)

This vote

is just one epic failure. From the dubious naming (Sanskrit is not a script that can be romanized), dubious justification (romanizations weren't technically forbidden before, just frowned upon), criteria (for extinct languages a single attestation is enough). Listed rationales include English Sanskritisms not Sanskrit usexes, and all but one of 12 supporters haven't edited Sanskrit entries at all. It ignores countless more pertinent issues such as which of the dozen transcription schemes to standardize on, whether to provide soft-redirects or full-blown entries (which is a NPOV issue, since Sanskrit has no "native" script), and why start this trend with Sanskrit when there are countless readily-available languages written in Cyrillic and Arabic with exactly the same set of issues relating to easiness of lookup and scholarly acceptance of romanizations when citing.

For all practical purposes this vote can be ignored. --Ivan Štambuk (talk) 11:42, 3 August 2015 (UTC)

It is worth noting that after this vote was initiated, a separate vote was initiated and closed on the question of "excluding romanizations by default"; that vote failed (with a majority in opposition). The vote at issue here arose due to efforts to remove an actual entry that reflects an actual transliteration of a Sanskrit word, and which was needed to resolve links made to that word on Wikipedia. Absent a definitive set of rules with respect to this practice, editors can continue to make such entries, and the efforts to remove them through RfD will generally continue to fail. bd2412 T 13:22, 3 August 2015 (UTC)

Sanskrit entries were added in Latin transcription before, but it was decided that the preferable to have them in Devanagari. There are no rules general or specific policies to forbid them, they are just "strongly discouraged".

What word/link are you talking about? Wikipedia has plenty of links to Devanagari spellings for Sanskrit here. For Sanskrit loanwords in English, it should link to the corresponding English entries.

This policy effectively does nothing. We have zero Sanskrit entries written in Latin script, and I'll delete every one that is created. --Ivan Štambuk (talk) 13:52, 3 August 2015 (UTC)

The discussion was about mahā. --WikiTiki89 13:55, 3 August 2015 (UTC)

I have also created tat tvam asi, which is noted in Wikipedia (not by me) to be "a Sanskrit sentence". The citations provided show the first use of this phrase in transliterated Sanskrit, and later uses in English, German, and Spanish text. I have also initiated a discussion at w:Talk:Tat Tvam Asi#Is this a "Sanskrit" sentence? on this topic. bd2412 T 14:07, 3 August 2015 (UTC)

@Wikitiki89: Well the entry is still there AFAICS. There is no ==Sanskrit== header there, since it's not a Sanskrit word (neither in transcription). --Ivan Štambuk (talk) 14:39, 3 August 2015 (UTC)

The RFD discussions archived at Talk:mahā discuss deleting the ==English== entry and creating a ==Sanskrit== one. I'm not advocating either way, only answering your question, "What word/link are you talking about?". --WikiTiki89 14:51, 3 August 2015 (UTC)

It's not a sentence, and neither is that Sanskrit proper. You should stick to editing languages that you're familiar with. --Ivan Štambuk (talk) 14:39, 3 August 2015 (UTC)

@Ivan Štambuk, I hope you mean you'll propose them for deletion through RfD, where it is up to the community to determine whether they are deleted. Anything deleted out of process can just as well be undeleted. bd2412 T 13:57, 3 August 2015 (UTC)

It is a general consensus among Sanskrit editors that entries in Latin transcription should not be created. This takes precedence over this illegitimate and logically unsound vote. --Ivan Štambuk (talk) 14:39, 3 August 2015 (UTC)

Bullshit. You don't get to decide what constitutes a word. We go by the sources, we don't censor them. bd2412 T 14:46, 3 August 2015 (UTC)

It's not my opinion, that's how things worked until you decided to push this nonsensical vote several times, and use Polansky to extend it until it passed. And in the meantime creating nonsensical entries that are not even valid Sanskrit, demonstrating the lack of basic domain knowledge. There is no "censorship" going on. It's this vote that is retarded and which you use to create "Sanskrit" entries such as tat tvam asi with quotations from several different languages where the phrase isn't even used. It's a pure massacre of rational thought. --Ivan Štambuk (talk) 15:10, 3 August 2015 (UTC)

You need to calm down and face reality. How can I be using "quotations from several different languages where the phrase isn't even used" when the phrase is used in the quotation? Are you engaging in some kind of doublethink exercise where your eyes see the words "tat tvam asi" in a sentence, but your brain refuses to process their existence there because it doesn't fit into your notions of what should be there? bd2412 T 15:18, 3 August 2015 (UTC)

What my eyes are seeing is that all of the citations you added to tat tvam asi are of the phrase used in the running text of other languages. This does not count as an attestation for Sanskrit. --WikiTiki89 15:27, 3 August 2015 (UTC)

You are mistaken. The 1835 citation is entirely running transliterated Sanskrit. bd2412 T 15:31, 3 August 2015 (UTC)

I wonder what you mean by "Sanskrit editors." If you mean the scholarly community, then we may as well ignore them, but if you mean Wiktionarians who actually edit Sanskrit entries, then I do think their opinions might be weighted above those of us schmucks who just see this thread and develop an opinion. Finding things like these votes may be a secondary concern for editors focused just on editing entries, so if you could invite some of the members of this "consensus" to share in this thread that would be helpful. Aperiarcam (talk) 15:21, 3 August 2015 (UTC)

Even "Wiktionarians who actually edit Sanskrit entries" collectively can't decide that a well-attested transliterated phrase doesn't exist just because the raw fact of its existence defies their preconceptions of what should exist. We might as well excise ain't from the dictionary because English prescriptivists would say that it isn't a word. bd2412 T 15:26, 3 August 2015 (UTC)

The issue of tat tvam asi is entirely distinct from the systematic creation of Romanized Sanskrit entries which redirect to their Devanagari equivalents. Ivan Štambuk is right in his scrutiny of tat tvam asi; it is somewhat absurd that it should contain a host of quotations in various European languages. Also, the issue of ain't is completely distinct from this. This is a formatting issue; the fact that Sanskrit has been occasionally attested in Latin script does not necessitate our carrying of Latin entries any more than the existence of w:Gwoyeu Romatzyh compels us to start making entries for every word attested in such publications. Aperiarcam (talk) 15:46, 3 August 2015 (UTC)

I am not interested in making (or justifying) Sanskrit transliterations for every word that can possibly be attested. I created mahā because it was regularly being linked to from Wikipedia articles, so there needs to be something there for it to point to. Tat tvam asi is a phrase important enough to have its own encyclopedia entry. Other possible entries are irrelevant at the moment. These exist widely enough that we should define them for readers who may come across them. bd2412 T 15:52, 3 August 2015 (UTC)

I agree that tat tvam asi ought to be defined, certainly at least in Devanagari (a search in Latin script would turn up the Devanagari entry anyways). Anyways, at least until this vote has concluded, there should not be a Sanskrit entry for tat tvam asi in Latin script, especially one that dubiously cites usage examples in English, German, and Spanish. This is, in any event, precisely what we are voting on here, and we should wait until a decision is made, as the standard before the vote was not to create such entries. There is an argument to made that "tat tvam asi" deserves entries in English, German, and Spanish rather than Sanskrit. Aperiarcam (talk) 16:01, 3 August 2015 (UTC)

If we're going that route, why not call it Translingual? Is there any other transliteration of it in any Romance language (other than the hyphenated version of the same)? bd2412 T 16:15, 3 August 2015 (UTC)

I already explained it: due to the peculiar nature of such Sanskritisms originating in religious/mystical usage, they are most often cited in the scholarly transcription rather than being phonetically adapted, since something like that would amount to sacrilege of ancient sounds. Often diacritics are dropped, but that's only common with well-known terms that have penetrated the general vocabulary. What you are conflating in your examples is the 1) complete and published transcriptions of Sanskrit works 2) such isolated 2a) usages and 2b) mentions of Sanskrit terms in various languages. For the first case we don't want entries in transcriptions. For the second case 2a) should be formatted as ==English== or whatever (lika mahā), 2b) could be formatted as Translingual, unless there is evidence of phonetic adaptation (such as alternation in transcription) in which case it should be formatted as the recipient language. --Ivan Štambuk (talk) 22:30, 3 August 2015 (UTC)

The existence of works using w:Gwoyeu Romatzyh does compel us to accept words in that orthography. It does not compel us to start making them; it will get done if we have a volunteer that wants to do them.

Wiktionary is a huge, sprawling populist dictionary. We should continue that tradition of providing definitions for words people can find and might want to look up, instead of elitistly ignoring them on the grounds they're in the wrong script.--Prosfilaes (talk) 21:49, 3 August 2015 (UTC)

It does not compel us to do anything. Dictionary is by definition an elitist project. That's why every one has lengthy usage notes. Usage concerns of statistically irrelevant userbase is also irrelevant. --Ivan Štambuk (talk) 22:19, 3 August 2015 (UTC)

As an elitist project, I propose we ban anyone who writes as ungrammatically as that. Isn't that what elitists do?

Besides being ungrammatical, "Dictionary is by definition an elitist project" is silly. There's nothing at dictionary that defines it as being elitist, nor are dictionaries in human society rare or exclusive. Works like Merriam-Webster's Elementary Dictionary illustrate that they are considered valuable for even the youngest readers, and works like the Urban Dictionary and Wiktionary illustrate the deep populist desire to create a dictionary that reflects the needs of popular communities. Lengthy usage notes are there to clarify the often complex connotations of words for the service of our audience.

Talking about our "statistically irrelevant userbase" is typical of pseudoelitism, where large words that sound important are used without thinking about their meaning. It is not even wrong; I can't think of a way in which it makes sense to say a user base is "statistically irrelevant".

I'm comfortable that people are using Wiktionary, based on seeing people cite it on places like Duolingo. If there's nobody using Wiktionary, then working on it is a waste of time, and Wikimedia shouldn't be wasting the money to support it.--Prosfilaes (talk) 23:35, 3 August 2015 (UTC)

How a person writes or speaks is a matter of personal freedom. The notion of "ungrammatical" writing is made up by authorities imagining themselves to have the right to deduce the general rules of language usage either by generalizing statistically frequent patterns, or (more frequently) by making such arguments contrary to the general practice on the basis of some spurious historical arguments. Personally I tend to often write ungrammatical Croatian by inventing words or combining them in unusual constructs that can or cannot be justified - it serves as an excellent way to gauge emotional involvement of the interlocutor to such issues at worst, or at best to derail them into a grammar Nazi rage fit, especially the ones who do it out of various ideological motivations. It's pretty fun, like this little convo.

The thing is, Wiktionary is not an elementary dictionary. It's an all-inclusive dictionary. We don't just do words for children, collegiate vocabulary or advanced ESL learners. We do everything from the history of time till today. Our goal is OED and beyond. Have you seen the entries for basic words at OED? They have dozens of meanings, most of them dead and obscure. Many Sanskrit words are like that - huge number of meanings, and you need to tag them by the author and period if you want to make any use of them while looking up. And that's the state of affairs for pretty much every language with a few million speakers and above, with a developed literature spanning more than a few centuries. Complete entries would be include countless dead meanings, obscure derivations that 99% of people have never heard of, dozens of regional pronunciations, obscure spellings and impenetrable etymological explanations. Each such language easily has half a million+ words attested, meaning that ordinary person hasn't seen or heard 95% of them. There can be no dumbing down if we are operating on the level (and we are).

The userbase of people randomly looking up Sanskrit words and not knowing anything about Sanskrit is statistically irrelevant. Those who read religious or philosophical texts that contain such random phrases, are at least sufficiently knowledgeable of its phonology to make a proper query. When we're dealing with running Sanskrit text in English works such as those mantra collections it's almost always decomposed word by word with a commentary. No one is gonna encounter tat tvam as in a foreign language text without it being translated on the same page. That doesn't happen.

People come to Wiktionary because it has information that others don't. It's usually for inflections, etymologies and obscure words. In other regards, competition is pretty far off, and that includes English coverage. The purpose of this project is not enhance learning or educate the general population. It's to collect and organize lexical knowledge, even if it's useless. The thing with languages is that most (in lexical terms) fits that uselessness category. Once the general usability threshold is passed, it's done for the amusement of editors and language nerds. People don't care, and will readily eat up whatever is served to them if this is the right place to find what they seek. All this talk of audience and users is irrelevant - the point is in the process and the result, not the consumers. This is not a corporation or a popularity contest. There is no usability experts or QA vetting the comprehensiveness and understandability of entries, which does make it an elitist project by default. If you tell people that you write dictionaries in your free time, I'm pretty sure they'll think of you as a snob (not that there's anything wrong with it, writing a dictionary not being a snob that is). --Ivan Štambuk (talk) 01:09, 4 August 2015 (UTC)

Ah, so you're a troll. Thank you for admitting that. You do realize that exercising your "personal freedom" that way encourages others to exercise their personal freedom to dismiss everything you say?--Prosfilaes (talk) 07:07, 4 August 2015 (UTC)

I'm not a troll, but I'm neither Mother Teresa. Now that we reduced each others editing motivations to tangible explanations (abstract idealism of knowledge for knowledge's sake, egotistical fun, educate the planet) we can focus at the issues at hand. The big problem with discussing anything with others is that it has been proven (with measurements) that for many people discussion itself gives a feel-good hormonal boost, rather than realization that they have been possibly wrong, which ultimately manifests itself as a strengthening of their original position. This is often the problem when discussion various "touchy" issues which don't have any hard metrics other than personal estimates, such as "what number of users is gonna look up randomly encountered Sanskritism in a transliteration out-of-context". --Ivan Štambuk (talk) 09:05, 4 August 2015 (UTC)

Thank you, Prosfilaes, and bd2412 for providing these intelligent responses, and showing the posts you are responding to for what they really are, for those readers who do not see it by themselves. --Dan Polansky (talk) 23:52, 3 August 2015 (UTC)

"It is a general consensus among Sanskrit editors that entries in Latin transcription should not be created." If that refers to Sanskrit editors at Wiktionary, it's untrue. Some of us who work on Sanskrit here would very much like there to be entries in Latin transcription, even if they say nothing more than "Romanization of महा" and "Romanization of तत्त्वमसि". We do allow romanization for Gothic, Chinese, and Japanese, and we ought to allow it for Sanskrit, Avestan, Hittite, and any other ancient language that scholars more often cite in romanization than in the native script. (Not that Sanskrit has a native script, but we're allowing Devanagari to fulfill that role.) And we should allow Sanskrit entries—again merely pointing the reader to the main Devanagari entry—for all other scripts in which Sanskrit is attested and which are available in Unicode, which includes almost all South Asian scripts. —Aɴɢʀ (talk) 10:21, 4 August 2015 (UTC)

Well you only started editing Sanskrit after this vote. As for your argument - Ancient Greek, Hebrew, Arabic, Armenian, Cyrillic-script languages are all also much more cited in romanized form in scholarly sources written in Latin script (such as English), so why not apply the same language to them as well? Can I create ethike with {{romanization of|ἠθική|lang=grc}}? I be Joe Sixpack, dunno how to type doz funny letters. There are like gazillion of attestations for transliterated Greek phrases used and mentioned in Latin-script written languages, so why not add them a well. That argument can be applied to every language not written in Latin script, but for some strange reason Sanskrit gets singled out.

Chinese and Japanese romanization are native scripts (people actually use them to write their speech) and that case doesn't apply. Sanskrit has no "native" script - all Indic scripts were used concurrently, because the emphasis was on speech not writing (rules of pronunciation were well defined, so it didn't matter how you wrote it as long it was uttered properly). As for Gothic though - I don't really support that decision, but given its small corpus the damage is minimal. --Ivan Štambuk (talk) 11:36, 5 August 2015 (UTC)

I've been editing Sanskrit since at least 2011, though to a smaller extent than other languages. I'm not thrilled about having Romanizations for modern languages like Chinese and Japanese, and I wouldn't encourage it for modern Hebrew, Arabic, Armenian, and Cyrillic-script languages either. And Ancient Greek and Biblical Hebrew are almost never given in romanization in running texts in the source language (as opposed to mentions of individual words). But whole texts of Sanskrit have been published in romanization, as they have in Hittite and Gothic, and it is precisely because Sanskrit has no native script that the use of the Latin alphabet for Sanskrit, which has been going on for 300 years, is as valid as the use of all the Indic scripts. As for Gothic, the damage isn't "minimal", it's nonexistent; there is no damage whatsoever, but only benefit, as there would be only benefit to including entries for romanized Sanskrit. Likewise for all the other ancient languages whose native scripts are much less widely known than Devanagari, such as all the languages written in cuneiform, Mycenaean in Linear B, Oscan-Umbrian in Old Italic, and so on. It's very cool that we can have entries in that native script for these languages, but it's counterproductive to exclude romanizations when even the scholars who specialize in those languages use romanization for all purposes. —Aɴɢʀ (talk) 15:01, 5 August 2015 (UTC)

@Aɴɢʀ: I agree with your view entirely. — I.S.M.E.T.A. 10:25, 4 August 2015 (UTC)

• As to whether this vote is a "failure", "epic" or otherwise: it is the opposite. This vote succeeded in what votes are so good at: show objective verifiable evidence of the scope of consensus for a particular proposal. And the extensions helped the vote do this job better: at the original end of the voting period, the vote had 10 participants and lasted one month; now, it has 25 participants and lasted 6 months so far. Therefore, right now, the vote is much better evidence of scope of consensus than it was, and than we had when the vote was not there. And it is not mere voting; people post comments with their votes. We even have an oppose comment, that of Angr, that sees the proposed criteria as too stringent. --Dan Polansky (talk) 10:50, 8 August 2015 (UTC)