Talk:Devanagari/Archive 2

Archive 1

Archive 2

Why are there phonetic values in the consonant chart?

If they're retained, it must be specified which language(s) they represent (and it should be all of those that use Devanagari). I think the IAST should be enough, for the chart at least. The phonetic values should belong on specific language articles. ʙʌsʌwʌʟʌ ^тʌʟк 17:40, 26 April 2017 (UTC)

Why remove? Why not just add a refn note in the sentence, just above the table, to clarify? Ms Sarah Welch (talk) 19:35, 26 April 2017 (UTC)

Clarify what? Clarify that these are the phonetic values in Hindi? This article is not only about Devanagari used for Hindi. Ok, I see that it cites a source Sanskrit, probably using Hindi-based pronunciation. If that source doesn't have IPA in it, we should remove it anyways as per WP:OR. ʙʌsʌwʌʟʌ ^тʌʟк 19:47, 26 April 2017 (UTC)

I don't see IPA in the source. Removing it. ʙʌsʌwʌʟʌ ^тʌʟк 19:51, 26 April 2017 (UTC)

@Ms Sarah Welch: Having phonetic values in the chart is biased, if they're just for Hindi. This script is used by many other languages, and the phonetics values can go already on language articles. No need for them here, unless we pick Sanskrit because it's what Devanagari is associated with, but even that's contentious since there are many ways to pronounce Sanskrit. ʙʌsʌwʌʟʌ ^тʌʟк 21:16, 26 April 2017 (UTC)

@Basawala: Phonetic values of the script in one or more language is relevant and useful encyclopedic information. It was there, added by someone long ago, you removed it because it was unsourced, I added it back along with sources. What is your concern now? Per BRD, you need to let sourced content remain in the article. If you know of RS that gives phonetic values for the same script in another language, we can certainly add those too, perhaps as refn or in the table, to address your bias concerns. Do you know of such RS? Ms Sarah Welch (talk) 21:24, 26 April 2017 (UTC)

Actually Ms Sarah Welch, the IPA values you re-added are not from the book you cited, and thus, still unsourced. So no, BRD doesn't apply here. ʙʌsʌwʌʟʌ ^тʌʟк 21:31, 26 April 2017 (UTC)

@Basawala: What do you mean? You don't see it in the two sources!? Or some are different? Give me an example of what you see. Ms Sarah Welch (talk) 21:32, 26 April 2017 (UTC)

@Ms Sarah Welch: /ɸ/, for example. And one of your sources didn't even provide consonant IPA values; please check when adding sources to make sure it matches with the content it backs up. But whatever, that's not the point. I agree that IPA is relevant, but this violates POV right now. Also, there are no IPA for vowels, so that's not consistent. Either IPA for both (in an NPOV way) or neither. ʙʌsʌwʌʟʌ ^тʌʟк 21:40, 26 April 2017 (UTC)

@Basawala: Strange. Both sources provide phonetic values, at length. Please don't delete it. If there is a variation, tag it or update it. If you have other WP:RS or phonetic values with sources, please add it. Please don't be disruptive. Ms Sarah Welch (talk) 22:41, 26 April 2017 (UTC)

@Ms Sarah Welch: No need for us to quibble about the specifics in the sources. My position is still that there should be no phonetic IPA values at all on this page. Once you allow Hindi, you create POV issues if there's only Hindi, and if we allow all Devanagari languages' phonetic values, then it would be too cluttered, and the chart would have to be reworked so that you can tell which values are for which language. There's nothing wrong with omitting them entirely. So I think the only viable options are:

• (1) only Hindi and add IPA to the vowel chart as well (I disprefer, since I think this has POV issues),
• (2a) allow any Devanagari language and rework the chart entirely so that it looks like a table, perhaps similar to those here (I highly disprefer, too much clutter, plain table is too boring)
• (2b) have the chart as is but include IPA for any language, with superscripts to disambiguate when there are multiple IPA values (I disprefer, too confusing),
• (3) no IPA values (my preference).

Anyways, time for me to rest from this page, and I'd like to give time to hear from others too. ʙʌsʌwʌʟʌ ^тʌʟк 23:07, 26 April 2017 (UTC)

(Also, this discussion back in 2011 is relevant, but none of the participants appear active.) ʙʌsʌwʌʟʌ ^тʌʟк 23:17, 26 April 2017 (UTC)

To discuss this, we need to know how much the phonetic values would differ between Hindi and other languages. If Hindi, Sanskrit, Marathi, Nepali (and I'm not even sure we have sources for all of these) all use mostly the same phonetic values for most vowels and consonants, we would only have to mark the ones where they differ, and it wouldn't be too cluttered. Shreevatsa (talk) 02:58, 27 April 2017 (UTC)

Indeed Shreevatsa, that is what I was thinking. For now, Basawala has just raised vague suspicions/claims that the phonetic values for Devanagari script in various languages are different, without backing up with any RS. If we do find RS that identify differences, we can certainly clarify the difference in a number of ways. Such clarifying information with cited sources would be of value to readers. I highly doubt that "क" has a phonetic value of "/s/" or "/t̪/" in some parts of their region, and "/k/" in others. Basawala or my doubts don't matter for this article, reliable sources do. Ms Sarah Welch (talk) 14:32, 27 April 2017 (UTC)

@Shreevatsa: Oh, finding sources for those would be easy. I saw somewhere that Omniglot is not considered a reliable source (I'm not sure how much I agree), but it's already cited on this article. Its content is exhaustive and highly accurate, even if possibly not reliable according to WP:RS. I actually have a new idea:

• (4) use the (constructed?) IPA pronunciation of Sanskrit, instead of Hindi. This is on the basis of what's done at Latin alphabet, but this would mean that we would want to create a separate articled called Hindi alphabet, for example, so that the language specific values get to go somewhere.

This actually seemed to be what was being done, before Ms Sarah Welch added the edit about Hindi. Anyways, as for option (2), the biggest variation is in the vowels. From the Omniglot page (which, even if it might not be RS, is highly accurate, and will reflect what we find in scholarly publications), I clicked on some of the languages that use Devanagari, listed there. अ has the value [ɵ] for Konkani, [ɔ] for Bodo, [ə] for Marathi, [ʌ] for Hindi, [a] for Kashmiri, [ɛ] for Rajasthani. (Even within a language like Hindi, difference sources will distinguish इ vs ई as [ɪ] vs [i] (Aspects of Hindi Phonology, Ohala 1983), [ɪ] vs [i:] at Hindustani phonology, and [i] vs [i:] at Omniglot.) Perhaps option (2) might be more viable if we omit IPA for letters which have too many IPA correlates, I don't know. But I prefer leaving them out from this page (3) or only including Sanskrit (4) (I understand it that even though Devanagari was created post-Brahmi (i.e. post Pali), it was meant for Sanskrit.) And then the IPA values for Hindi would go in Hindi alphabet, Marathi in Marathi alphabet, etc. (There is indeed basis for this both conventionally and for the fact that different sets of nukta/diacritic variants are used in different languages. Most languages using the Cyrillic alphabet have their own pages, for example.) ʙʌsʌwʌʟʌ ^тʌʟк 04:05, 29 April 2017 (UTC)

@Basawala: Omniglot is a wiki-like source, and therefore unreliable. You write, "[Omniglot]] even if it might not be RS, is highly accurate, and will reflect what we find in scholarly publications" Please find those scholarly publications, first. We can go from there. As you probably know, from wikipedia's perspective your or any individual editor's personal opinions do not matter. Only verifiable RS do. Ms Sarah Welch (talk) 04:39, 29 April 2017 (UTC)

Ms Sarah Welch: you have chosen to ignore my perhaps not-bureaucratically-waterproof but perfectly well-intentioned exploratory findings. What I've done here is perfectly practice for informally scoping out options; not fine for adding to WP articles. Let me remind you that right now we're not proposing to add those as sources to Wikipedia, we're just trying to evaluate options (1-4; in particular 2), since we all agree that the current version of this page (with IPA on consonants and not vowels; with IPA for Hindi but not other languages) is not ideal. You, for example, have not demonstrated with RS that the phonetic IPA values for all/most of the graphemes don't vary much across languages; that's a "personal opinion" (in your sense) too. I don't have the burden of proof; you don't either. Unlike me, you haven't shown (whether with RS or provisionary data) an argument in support of 2. Right now, I haven't heard from you which version you support. Obviously if we agree on option (2), RS's will be found. Let's focus on the options. For example, do you object to (4)? That's the I'm in favour of the most, and if chosen, will obviously be supplied with RS's. ʙʌsʌwʌʟʌ ^тʌʟк 05:09, 29 April 2017 (UTC)

Basawala, please quit WP:FORUM-y discussion. See WP:TALK. I don't need to prove anything to you on this talk page! Nor you to me! Yes, I object to (4) because that is not encyclopedic. Find reliable source(s) for other languages, listing different phonetic values. We can then work from there. Don't get me wrong. Your effort is appreciated. But, the right place to start is to find and study the reliable sources on phonetic values for multiple languages, thereafter explore options on how to best present the verifiable information you found. If you or someone does so, that would be most welcome. Ms Sarah Welch (talk) 05:31, 29 April 2017 (UTC)

Didn't I just say that neither of us has the burden of proof? I do not see how my discussion is WP:FORUM-y nor any basis to your accusation, so I'll ask you to be civil. You claim that (4) is unencyclopedic; that's a personal opinion which I will ignore. You don't even provide an argument for why you think so, let alone RS's. (Note that I said that if we chose option 4, all addition will be supported with RS's. In fact, option (4) is no less encyclopedic than what we have right now, which given its inconsistency and Hindi-bias, is not very.) I'll ask a question, which if answered, would be quite constructive: which version of (2) do you prefer, (2a) or (2b)? ʙʌsʌwʌʟʌ ^тʌʟк 05:36, 29 April 2017 (UTC)

Comment: As an uninvolved person, I would like to answer the question. Please correct any *factual* errors in what I say. If the article about Devanagari, a script, it is going to be necessary if the article is to be understood by people like me who can't already read it, to give some indication of how the letters are read. One way to do this is with a conventional transcription, if such exists, for the script. But the problem with conventional transcriptions is that sometimes the uninitiated reader has no idea what they mean. (For example, the distinction between 'ca' and 'cha' could be practically anything.) It helps readers to give IPA, and if the script is one used by multiple languages, this means choosing one language, using the criterion of "reasonable", not to select one which is an outlier either in population terms or in phonetic terms. If you choose Hindi, this may be an excellent choice; though it might be (I've no idea) actually better to choose a different language, which is significant in terms of population, but is closer intuitively to some presumed ancestral pronunciation. Of course, labelling this as "pronunciation of Devanagari" is an error, but (if you have chosen Hindi), labelling it as "pronunciation in Hindi" this is not an error, not is it "Hindi bias". Imaginatorium (talk) 08:34, 29 April 2017 (UTC)

Yes I agree with the same. It is helpful to the reader to indicate some pronunciation in the article, and it doesn't matter too much which language is used, as long as we specify it clearly. We could use both Hindi and Sanskrit for example (distinguishing the two only for letters where they differ). Shreevatsa (talk) 17:42, 29 April 2017 (UTC)

Indeed, Ms Sarah Welch (talk) 18:01, 29 April 2017 (UTC)

I certainly agree that IPA is useful here. However, currently due to the nature of the fact that scripts can be used for multiple languages, the choice of which one(s) to use can be subject to debate. It looks like everyone here, from an aesthetic point of view, is okay with Hindi only, ignoring the bias question. So then it's time to add the vowel pronunciations. It would be useful (for me or others) to create separate articles (in addition to this one) on the Hindi alphabet and Marathi alphabet, for example, and to perhaps migrate some of the language-specific sections and commentary, such as "Allophony of 'v' and 'w' in Hindi", which is much more about phonology than about writing system. ʙʌsʌwʌʟʌ ^тʌʟк 05:02, 3 May 2017 (UTC)

Preference for symmetry

@Imaginatorium: The symmetry can be 1 axis, 2 axes, etc. Devanagari, and nagari letters in general, do have a strong preference towards symmetry. Some modern fonts distort the symmetry more than others. Old manuscripts make this quite obvious. Cardona and Jain book is RS. If you find another RS that disagrees with Cardona and Jain, we can summarize that too. Ms Sarah Welch (talk) 14:45, 29 April 2017 (UTC)

What do you mean by "symmetry", then? (And what does these authors mean by it?) Can you show me one particular character that you think is "symmetrical", and describe in detail where this symmetry is? For that matter, can you point to a "square outline"? If we think of the Roman alphabet even, it seems to me to include far more symmetry than Devanagari, so I am simply puzzled by the claim. Symmetry might be a useful reference. Imaginatorium (talk) 15:53, 29 April 2017 (UTC)

क has near-two fold symmetry. The entire letter is essentially set inside a square box. Ms Sarah Welch (talk) 15:56, 29 April 2017 (UTC)

OK, ignoring the topline, yes, this has rotational symmetry. ("Two-fold" is not clearly defined) Looking at the basic consonant table I can see that ठ is (roughly) symmetrical about a vertical axis, फ rotational symmetry, व symmetry about a horizontal axis. That's 3 out of 33; I count 8 out of 26 capital Roman letters which are not symmetrical. (I also cannot really understand this "essentially set inside a square box" claim... or is this comparing with say "ascenders and descenders"...? Let's try to do symmetry first.) Imaginatorium (talk) 16:29, 29 April 2017 (UTC)

अ has near-one fold symmetry. Many others, if you move the axis around. Ms Sarah Welch (talk) 16:35, 29 April 2017 (UTC)

OK, let's try to sort this out. Please stop using undefined terms, like "near-one fold symmetry". Are you saying that अ, is a bit like "3-1", which is a bit like symmetrical about a horizontal axis? And without the topline, which if I understand correctly in this character has a particular asymmetry, because it is on the right only? Well, what do you understand "a preference for symmetry" to mean? I would expect it to mean that at least a majority of the characters were (near-enough) symmetrical. And I cannot see, nor have you even attempted to demonstrate, that this is true. Note that being an "RS" is something to do with printing books and stuff like that, so does not of itself actually imply that it is pure truth, but in this case I do not doubt the author's expertise in Devanagari, I am pointing out their stunning ineptitude with the English language. They are using a word which has a precise meaning, but in a "warm feeling" sense, probably trying to say something like "compact and balanced". So their statement is, on the face of it, plainly false. I imagine you know much more about Devanagari than I do: can't you find someone who describes it without blundering? Imaginatorium (talk) 09:58, 18 May 2017 (UTC)

Brahmi from Semite?

Are we supposed to believe that the oldest civilization had to derive it's language from the language of a nomadic tribe? HOMERDESI (talk) 07:43, 7 February 2018 (UTC)

Language order

Libhye, can you please document the reasoning behind the language order through in-line HTML comments? If it is ordered by number of speakers, provide sources for the numbers too. Without a documented source, there is no way to verify changes in ordering. —Gazoth (talk) 16:35, 5 August 2018 (UTC)

You're mistaken if you think we need a source in order to not organise lists by alphabetical order. Ordering a list alphabetically is the least informative way of ordering it; literally anything else is preferable. This list was originally added by me and lists the languages by importance, which is of course a vague concept but still infinitely better than alphabetical order, which is completely useless. Libhye (talk) 16:47, 5 August 2018 (UTC)

An objective ordering is always better than a subjective one, but an objective ordering doesn't have to be alphabetical. If you didn't have any objective reasoning for your ordering, don't mind me changing it to one ordered by number of speakers in the future. —Gazoth (talk) 03:15, 7 August 2018 (UTC)

4300 ligatures etc and some serious WP:OR

All this zillion compounding and ligatures based on someone's OR and www.sanskritweb.net as source is WP:Undue, shall I say highly undue. That is not how Devanagari has been or is in mainstream use. I do not understand the rationale for including any of this fluff in this encyclopedic article per wikipedia community agreed guidelines. If someone has reasons, or a source, let us discuss them please. Ms Sarah Welch (talk) 17:58, 28 October 2018 (UTC)

There should be at least an explanation of the term

The article gives some explanations to the second part of the name, telling that nagari means something like civic script. What deva stands for, i.e. divine, is not said. — Preceding unsigned comment added by 2001:16B8:467C:EA00:2DD6:DCCC:CD7B:E12B (talk) 18:58, 25 November 2018 (UTC)

Symmetry?

The introductory paragraph says this script has a preference for rounded symmetrical shapes, yet none of the characters shown is in anyways symmetrical. Unless there is a different idea about what symmetry is than I have. Wschart (talk) 17:42, 30 November 2018 (UTC)

‎Biconsonantal conjuncts?

Why were the biconsonantal conjunt tables removed? This was the only place on the web I had found a detailed list of the main conjuncts in proper unicode. Can the deletion of these be undone? — Preceding unsigned comment added by 2605:E000:4DC2:1500:8D25:893F:4B3E:434B (talk) 21:00, 8 January 2019 (UTC)

Move discussion in progress

There is a move discussion in progress on Talk:Nagari (disambiguation) which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 00:46, 26 March 2019 (UTC)

Semitic Origin of Script

The descent of Brahmi from a semitic alphabet has no scientific or scholarly consensus. It is intellectually dishonest to write that it is descended from the Aramaic or Sinaitic alphabet in the infobox (same for any other indic script). The Controversy is clearly detailed in the article itself, so why is the disputed origin written in the language descent? Any unsuspecting reader would assume this to be consensus. (whence it is not). A simple note linking to the controversy is sufficient. Abh9850 (talk) 14:18, 21 July 2019 (UTC)

Lets discuss this at one central location: Talk:Brahmi script#Semitic descent on the infobox. Abecedare (talk) 14:27, 21 July 2019 (UTC)

Error in the classification of r as moordhanya ===

While Y l and v were probably correctly classified in their respective rows in the table, r is not moordhanya. The equivalent is actually the old L as in marathi as in veLa (time). This used to be in old sanskrit and vedas, and came into marathi, kannada, telugu, etc. but was lost in hindi. this L has a devanagari character that looks a bit like l with circles in the bottom as seen marathi which also morphed into Kannada and telugu. The r is to be placed in its own place elsewhere in the table because it is neither moordhanya nor talavya. In fact r and l are abheda hence it may fall in the same row as l.

South Asian

The article would look more intelligent if the words 'Indian Sub-Continent' were changed into South-Asian. There was no such India before a few centuries. And the very concept of an 'Indian' came only after the arrival of the English East India Company rule, which led to the creation of British-India. This British-India also did not contain the full of current-day India, Pakistan and Bangladesh. It was only around half of the SouthAsia or even less. — Preceding unsigned comment added by 2409:4073:2099:553B:D155:2B2D:B5D2:5B3F (talk) 03:48, 17 October 2019 (UTC)

I don't think South Asian is going to convey the correct meaning. Specifying as Indian Sub-Continent seems better. SmitaVarma (talk) 12:21, 28 December 2019 (UTC)

Better Infobox image needed

The image currently in the Infobox isn't bad, except for one thing: the prominence of the "Chandas" font name, which can be misleading for those who know little about Devanagari and might see that very large title as some sort of descriptor for the writing system, where it clearly is not. I may check with the Graphics Lab, to see if they can remove the font headline from the image, because other than that, the image seems fine. If there is an even better image, though, then I'm fine with changing it. Mathglot (talk) 21:30, 9 February 2021 (UTC)

Script

Why we called brahmi script .we have many evidence that this is dhamma script not brahmi script. Rajiv15071989 (talk) 22:03, 5 June 2021 (UTC)

Dhamma (Dharma, Lanna, Tai Tham script) is what Northern Thai is written in. They're on branches of Brahmic script that diverged directly from Brahmi—Tamil–Brahmi and Gupta, respectively. Largoplazo (talk) 23:42, 5 June 2021 (UTC)