Talk:Assamese language

From Wikipedia, the free encyclopedia

This article is within the scope of WikiProject India, which aims to improve Wikipedia's coverage of India-related topics. If you would like to participate, please visit the project page.

This article has been rated as B-Class on the quality scale. (add comments)

[edit] Spectral Analysis on unique phonemes of native Assamese speakers -by Dr Rabin Deka

The web link [3] (Moral 1992 - http://www.iitg.ernet.in/rcilts/asamiya.htm) of IIT Guwahati - the author in the link bundled up three distinct phonemes of the modern Assamese language into one phoneme.

For your information IPA, i.e., International Phonetic Alphabet authority never did any spectral analysis on any Assamese phoneme.

Although the X.wav file in link from IIT Guwahati [4] (http://www.iitg.ernet.in/rcilts/Wave/word/assamese/X.wav) is badly recorded, i.e., voice clipped exceeding 16-bit limit, the file is still good enough in order to assess, or conduct, a spectral analysis enebling to predict that this phoneme does not belong to the frequency range of the claimed voiceless velar fricative range.

Any with the knowledge of Digital Signal Processing may quickly conduct an experiment on the said X.wav file in link (use an FFT, Fast Fourier transform, with a standard window function, a minimum 2048 point-FFT is recommended with a Blackman-Harris window; then compare with a standard voice frequency chart).

The contributor of the above four bullets is Dr. Rabin Deka, who is currently a Professor of Digital Signal Processing and Communications. He had visited IIT Guwahati in year 1996 (from where the web links originated) and presented a scientific talk on "Human Voice Conversion Techniques and Algorithms". Currently he lives in San Jose, California.

[edit] Difference between Assamese and kamrupi

I feel the discussion is not appropriate in the main page and should be carried out here. Chaipau 00:48, 21 September 2005 (UTC)

The IIT Guwahati author of the web link [5] (Moral 1992 - http://www.iitg.ernet.in/rcilts/asamiya.htm) violates the definition of dialect of language itself. Because, referring Kamrupi as a dialect of Assamese is as good as telling Latin as a dialect of Italian.

The IIT Guwahati author in the link also carelessly overlooked the original work of Dr. Banikanta Kakati.

Not only that the author in the link bundled up three distinct phonemes of the modern Assamese into one phoneme - read more in Phonetics' section of the link Assamese

A phenomenal difference between modern Assamese and present Kamrupi is the usage of present perfect form with Sanskritize pronunciation which is missing in modern Assamese.

This form in Kamrupi is composed using Sanskritized pronunciations while this grammar is dominant in currently spoken versions of Kamrupi throughout undivided Kamrup, Goalpara and all the way further west to Cooch Behar.

"In addition to Kamrupi wisdom, Kamrupi lokageet, Goalparia lokageet and Vaishnava divine song bargeet are also sung in present perfect form. It is spectacular - over 90% of the sloka found in the yester year Kamrupi scriptures were composed using this grammar found in today's spoken Kamrupi." - quoted directly from the paper entitled "Some Kamrupi Wisdom and Recitation Norms" author Dr. Rabin Deka published by Asom Sahitya Sabha (Assam Literary Socity) North America branch in "Luitor Pora Mississippi", 20th publication, July 2003.

The linguist Dr. Banikanta Kakati also took the above-discussed grammar into account in order for him to define the language spoken in Upper Assam area as a dialect of Kamrupi. The Ahom occupation was one of the main factors behind the formation of this dialect (now called modern Assamese) of Kamrupi.

Volumes of ancient scriptures were written in Kamrupi.

[edit] Writing

The native Assamese speakers wrote only buranji puthi in the native Assamese language which is now the state language of Assam or modern Assamese. But the Kamrupi was the language used to write volumes of yesteryear scriptures.

A lot Kamrupi words have been miss-spelled in Hemkosh. It is also observed in some cases that Hemkosh does not follow Sanskrit pronunciation/style, e.g., manush is written as manuh in Hemkosh, but this is not the case in Kamrupi.

The words listed in Hemkosh are spoken in the old district Sibasagar in upper Assam. As a result, it is difficult for ordinary native Assamese to make the meaning of yesteryear Kamrupi scriptures.

The above discussion portion cut and pasted from main article Chaipau 11:23, 18 October 2005 (UTC)

[edit] Transliteration Thoughts?

Hi all,

I am not a native speaker of Assamese, but I am familiar with it as I have studied phonological systems of eastern Indic languages, and I wanted to work on expanding the linguistic description of the Assamese language (as I have done for the Bengali language article). However, if I am to do this, I want to first get some feedback from the native Assamese speakers here, and from anyone involved with maintaining this article.

Should I use only IPA transcription in describing Assamese phonological phenomena, or should we work on a simple transliteration system? For example, the complex vowel system can be transliterated as:

i [i], e [e], ê [ɛ], a [a], å [ɒ], ô [ɔ], o [o], u [u]

This way, we can avoid using just "o" and/or "a" to describe the four different Assamese vowels people seem to mix up all the time in transcription. The consonants can be transliterated in a style based on the IPA transcription. By adopting such a system, we can show the commonalities among Eastern Indic languages (Assamese, Bangla, Oriya, etc.) while still preserving the unique vowels of Assamese (like å [ɒ]). What do you all think? If people are okay with a system like this, or would like to add/edit something about it, please let me know! I'm very open to suggestions/comments about the idea!

--SameerKhan 04:02, 24 February 2006 (UTC)

I appreciate your editing this page. Please go ahead and use the standard system (the one you have used for the other languages). If we need another system (if others object), we may have it sitting side-by-side where appropriate. Thanks. Chaipau 04:36, 24 February 2006 (UTC)

I have looked at the phonetic and other transcription schemes in Bengali_language, and it might be a good idea to put up a more romanized transcription scheme for Assamese too (along with the IPA based one). Unfortunately, no standard transcription for Assamese exists today. People debate these issues very passionately at times. Is it a good idea to use Wikipedia to come to a kind of agreement on this? Chaipau 15:32, 24 February 2006 (UTC)

Yes I think it's best to have both IPA and some sort of neutral transliteration, so that it's easier to write about the language without relying on complex IPA symbols all the time. For the Bengali system, we didn't go with the previous "standardizations", which basically tried to map the Bengali letters one-to-one with English letters - even silent letters! Since the Assamese writing system also has issues with silent letters, and some symbols being used for more than one possible sound, it'd be better if we could NOT rely on the Assamese spelling system as a basis for a Romanization. Instead we should just have a sound-by-sound correspondence like the one we finally settled on for Bengali. I think that we could apply this same scheme for Assamese, of course with some important changes to reflect the unique sounds of the language.

Here is a tentative scheme. If people find it a good starting point, please send me your feedback and I can make revisions to this scheme before applying it to the article.

**Vowels**
	Front	Central	Back
High	i		u
High-mid	e		o
Low-mid	ê		ô
Low		a	å

**Consonants**
	Labial	Alveolar	Velar	Glottal
Voiceless stops	p ph	t th	k kh
Voiced stops	b bh	d dh	g gh
Voiceless fricatives		s	x	h
Voiced fricatives		z
Nasals	m	n	ng
Approximants	w	l, r

--SameerKhan 08:12, 25 February 2006 (UTC)

So Sameer Khan how will I write the following Assamese words in IPA?

Axom, Dax, Xabha, Xahitya, Xaikia, Xanjib, Xarma, Xatra, Xatriya, Xaurav, Xaxadhar, Xewali, Xujal

Kurmaa 02:45, 26 March 2006 (UTC)

Well since I am not an Assamese speaker, I can't say for sure, since I don't know all of the words based on the spelling given. But from my knowledge of Bengali, I can guess what words you're talking about.

Of course, there is no perfect IPA transcription of any language; the IPA must be adapted somewhat to each language to be truly accurate. For example, with English, certain phonetic aspects of the release of alveolar stops are simply not transcribed, even though they exist in English and not in other languages. With respect to Assamese, the transcription of the consonant often written "x" is also debated. Some people think of it as a fricative, and thus transcribe it as [x] (voiceless velar fricative) or [χ] (voiceless uvular fricative), while others think of it as a voicless velar approximant [ɰ] (I don't know how to attach the devoiced diacritic on Wikipedia). Because [x], , and [ɰ] are all approximations of the same sound in Assamese, most linguists simply choose the most familiar symbol [x]. I am not making any claims in my use of [x], and I know that there are reasons to believe it is not [x], but nevertheless I am using [x] for the purposes of the following transliteration. If you prefer a different symbol, you can replace all instances of [x] with [χ] or [ɰ]. Anyhow, here is what I have for you (I only included words I felt I might know). I wrote everything (1) the way you spelled it, (2) the way I would spell it in Romanization, (3) a broad IPA transcription, and (4) an English equivalent of what I think you meant.

Axom ôxôm [ɔxɔm] 'Assam' Dax dôx [dɔx] 'ten' (or did you mean dax [dax] 'slave'?) Xabha xôbha [xɔbʰa] 'society, group, committee' Xahitya xahityô [xaɦitːjɔ] 'literature' Xatriya xôtriyô [xɔtrijɔ] 'Kshatriya' Xaurav xôurôbh [xɔurɔbʰ] 'fragrance'

I hope this helps. If anything seems odd, or if you have any questions or comments, please let me know. I know a lot about phonology and the IPA, but I am not a speaker of Assamese so a lot of my work is done looking from the outside in. --SameerKhan 04:02, 2 April 2006 (UTC)

Dear Sameer, as you are writing Sam'ee'r in your name (instead of Samir)..must be bcuz you use 'deerghoeekar' in bengali. thrfr, it is understandable tht we must hv distinctions between 'hroswaeekar' n 'deerghoeekar' when we attempt to write an Oxom'ee'ya word in English.

This is not necessarily binding. The horsho- and dirgho-ikaars are arbitrary in Assamese spelling, and we need not follow it in transliteration schemes. Also, distinguishing the different [x] sounds and dontyo- and murdhonyo- would create a big mess. Chaipau 17:05, 6 June 2006 (UTC)

I agree with Chaipau. The spelling of my name has nothing to do with the Romanization of Assamese (or Bangla for that matter). (Either way, my name is Arabic!) Yes, if I were to write my name in Bangla, I would use dirghikar, but that would have no bearing on the pronunciation of course. If I used hroshshikar, it would still be pronounced the same in Bangla. Therefore, the Romanization of either spelling would be samir. I see no benefit in preserving archaic redundancies from Assamese and Bengali when describing the pronunciation of words to those who don't read/write/speak Assamese or Bengali. --SameerKhan 05:57, 7 June 2006 (UTC)

Sameer n Chaipau, Oxomeeya is a very soft language intensively influenced by local components, so finding length of hroswaeekar n deerghoeekar @ of other western languages / sanskrit or hindi is not logical. Oxomeeya hroswaeekar n deerghoeekars definitely have differences...chaipau try micikeeya --- mi,ci (same way)ky/kee (definitely different)...if it would hv been micikiya (thts mi, ci, ki all in the same way)..for me its difficult to pronounce...its becoming a very hard / no more soft/ no more Oxomeeya.../micikeeya just came to mind....these r our wealth... n beauty of softness n sophistication remain thr...

Moreover, we may create hocoeeka n deeghoeeka (as pronounced in many places of Oxom)instead of hroswaeekar or deerghoeekar, but see still u find a very clear 'ee'....infact a mix of hroswaeekar n deerghoeekar in different Oxomeeya words is the real 'jadu' tht we hv in Oxomeeya..

I believe the vowel length distinctions you're referring to are totally stress-related. Assamese has penultimate stress, and thus all second-to-last vowels are longer than other vowels. This has nothing to do with the vowel being written one way or another. Shortening that vowel sounds strange to you because it breaks this consistent rule in Assamese for lengthened vowels in certain positions. However, total contextual licensing of length is evidence of a lack of true vowel length distinction in the language, and thus should not be transcribed. --SameerKhan 07:08, 8 June 2006 (UTC)

I removed the transliterations. Masica (1991) and Goswami & Tamuli (2003) use IPA more or less. Basically, where other IA languages necessitate romanization conventions because of messy visuals such as dental t̪ diacritics and affricate tʃ ligatures, Assamese doesn't, because it has neither anyway. Tuncrypt (talk) 19:35, 19 January 2008 (UTC)

[edit] Creating Assamese Literature

I have created an article Assamese Literature from the History section of this article. This is because this section mostly describes the literature in the language. I believe now we may compress this section here considerably and focus on Assamese literature exclusively in the new article. I shall begin compressing this section in a few days time, after other editors get a chance to see this. Chaipau 13:19, 24 September 2006 (UTC)

[edit] translators needed at Wikipedia:WikiProject India/Translation

Wikipedia:WikiProject India/Translation--D-Boy 19:30, 21 November 2006 (UTC)

[edit] Assam --> Asom

If Assam will soon change its name to Asom, does this mean that the word Assamese will change to Asomese? --Akut 10:35, 21 December 2006 (UTC)

Asomiya? Chaipau 17:08, 21 December 2006 (UTC)

Maybe. In German the name is already Asamiya. --Akut 14:43, 7 January 2007 (UTC)

The name change is not official yet and is fraught with discord among the people about "rebranding" Assam. Consider the recent [National Games] in Assam (not Asom!). There are many other instances of the debate being taken up even on [international fora] supported by extensive research. -- Deepraj 11:00, 14 March 2007 (UTC)

[edit] Anachronistic and potentially racist description of Assamese

I am going to delete the words "giving it a characteristic expressiveness and charm" from the following sentence from "Formation of Assamese:"

Since the time of the Charyapada over the passage of the centuries it has been influenced by the languages belonging to the Sino-Tibetan and Austroasiatic families giving it a characteristic expressiveness and charm.

How does one judge that a language is expressive and charming? Aren't all languages expressive? Isn't that what languages do? And, the word charming strikes me as being a little nineteenth century, as in, "The charming natives use an expressive language." —The preceding unsigned comment was added by Gordonmonaghan (talk • contribs) 20:32, 11 January 2007 (UTC).

[edit] Disambiguation needed for Assamese

I propose a disambiguation article for Assamese which currently redirects to Assamese language. Please check out the Talk:Assamese section for more details. The reason for adding this comment here is to ensure no surprises for the current path to get here. Deepraj 13:25, 14 March 2007 (UTC)

[edit] Writing System

Should we redirect the Writing System to Eastern Nagari script? After that, may be, we can merge Bengali script and Assamese script as these two articles essentially deal with the same set of scripts. -Bikram98 06:48, 3 April 2007 (UTC)

I have tried to merge Bengali script and Assamese script in the past. It does not work. It is best to develop Eastern Nagari script as the mother article that tackles the development of the script and to let the two (or three) daughter articles to specify the language specific features. The Assamese-Bengali script for Manipuri will be replaced by the Meitei script soon. Chaipau 13:06, 3 April 2007 (UTC)

[edit] Recent edits

I noticed there have been several edits to this page today, and most of them are in an attempt to bring the IPA transcription of Assamese to match the transcriptions of Masica and other linguists, which is an admirable goal. However, IPA transcriptions for Assamese are not universal: the back vowels, in particular, are given different labels by different linguists. Even Peter Ladefoged changed his initial transcription of ও from [ɒ] to [ʊ] in later recordings. The transcriptions we were using here (and that I've reverted to) reflect the transcriptions in many of Shakuntala Mahanta's and Peter Ladefoged's systems. Shakuntala Mahanta is an Assamese linguist researching the vowel system of the language, and the late Peter Ladefoged was one of the field's most respected, if not the most respected, phonetician, and I would take both of their transcriptions to be superior than those from Masica, who basically compiled transcriptions from texts that largely predated modern phonetic transcription. Anyhow, that's just my take. --SameerKhan (talk) 20:30, 19 January 2008 (UTC)

Okay, I didn't see this before I reverted. Hold on. Tuncrypt (talk) 20:31, 19 January 2008 (UTC)

You're talking about phonetic values, which is fine. Those disagreements can be worked out. But you reverted my removal of the transliteration, which is another thing. I wrote this in your "Transliteration thoughts" section above:

IMO transliteration's just so pointless for Assamese. I hope you agree. Tuncrypt (talk) 20:36, 19 January 2008 (UTC)

I don't agree. I don't see why Assamese needs a transcription any less than Bengali, Hindi, or other Indic languages. If anything, it would need it more, because Assamese words are currently very often transliterated in a Sanskritic way, thus not sticking to the current Assamese pronunciation at all. The reasons why I think it's important to use Romanization instead of IPA as the primary representation of Assamese words in English are (1) it would be easier to incorporate a Romanization of Assamese words into English sentences rather than IPA (non-linguistically trained people can read "ô", "ê", and "û" without knowing what the diacritic means, as opposed to [ɔ], [ɛ], and [ʊ]... and issues such as capitalization would not be a problem with a Romanization, while they would in IPA), (2) all other Indic language articles use IAST or language-specific Romanizations in addition to or instead of IPA, and (3) since different studies have used very different IPA transcriptions of the vowels, it would be better to have our primary representation of Assamese not be in IPA. No matter what my reasons are, I think we should consult other Wikipedians who have been working on this article before making so many major changes. I'm going to revert back a couple versions and wait for discussion first. --SameerKhan (talk) 20:50, 19 January 2008 (UTC)

But I just told you why it needs a romanization less than Bengali, Hindi, etc. Those languages have sounds/symbols such as t̪ ʈ tʃ iː, which end up neatly and by sourced convention depicted as t ṭ c ī. Assamese has none of these. In any such romanization system then all that'd be left are such measly differences as aspiration Cʰ Cʱ becoming Ch, and the three vowels ɛ ɔ ɒ/ʊ for which you are making up your own scheme of ê, ô, û. And really, what's the point of the latter? How is it better or any more helpful than just using ɛ ɔ ɒ? Because that is what you're doing, concocting something, a layer of meaningless with ê ô û, when sourced examples are what we are to follow, which do use the phonetic symbols ɛ ɔ ɒ.

I'll get back to you. --SameerKhan (talk) 22:49, 19 January 2008 (UTC)

So I guess I don't understand what the difference is between choosing a Romanization for Hindi instead of IPA and choosing a Romanization for Assamese instead of IPA. But although the specific choices of Romanization characters were chosen partially by me, the decision to have a Romanization in the first place was not entirely my decision. Other Wikipedians had asked for a Romanization like that used for Bengali, because the IAST and other current Romanization schemes would be misleading for anyone unfamiliar with Eastern Indic sound changes.

And regarding ɒ vs ʊ, where are your Ladefoged and Mahanta references? You rebutted Masica (1991) as a compilation, but then there are other recent and specific sources such as Goswami, G. C. & Tamuli, Jyotiprakash (2003), "Asamiya", in Cardona, George & Jain, Dhanesh, The Indo-Aryan Languages and Baruah and Masica (2001), "Assamese", in Garry, Jane & Carl Rubino, An encyclopedia of the world's major languages which list ɒ. Tuncrypt (talk) 21:37, 19 January 2008 (UTC)

I know, sorry, I didn't give the actual references! I'll get back to you with that as well. --SameerKhan (talk) 22:49, 19 January 2008 (UTC)

Though I've made my response, invoking source (which seals the deal as far as I know), I still would like to add more anyway. You made two appeals to reason which I'd like to quash. For example, I really don't understand what this means, "non-linguistically trained people can read "ô", "ê", and "û" without knowing what the diacritic means, as opposed to [ɔ], [ɛ], and [ʊ]". Non-linguistically trained people are not going to know anything either way, and with ô, ê, and û, the linguistically trained too won't. With the the proper symbols ɔ ɛ ɒ, those who know will know, and those who don't won't, or will have to learn.

I disagree. I have taught the IPA in phonetics classes to non-linguists and linguists alike, and in my experience, there's a big difference between using a diacritic on a familiar character and using a completely different character. You and I know what ɔ ɛ ɒ mean, but people who have never seen the IPA will be totally lost. Sure, to me, they look like "o", "e", and "a", because I know how the characters of the IPA were put together, but most people will not have a clue (although I guess they might recognize ɛ as a type of "e" if they are familiar with Greek). Unfamiliar diacritics, however, can be easily ignored by people who are unfamiliar with the Romanization scheme, and ô ê û will just look like o e u to such people. Sure, the pronunciation won't be perfect, but it's close enough. This is just like the use of retroflex dots or length diacritics in IAST, instead of the IPA-style hook and colon. When I taught the IPA, many students couldn't even recognize some of the retroflex letters as t d n l r with hooks, and had to be told "Look, these are just t d n l r with hooks at the bottom", and the colon for length isn't exactly intuitive for someone who only reads English orthography. That's why adding diacritics in the IAST system works for many Indic languages - it encodes the information for people who know about the system, while not alienating people who are unfamiliar with the system. Of course, IAST doesn't work for Assamese or other Eastern Indic languages, unless one simply wants to reproduce the spelling of the Assamese word, ignoring the pronunciation. That's why a couple of us put together the system in place now. Anyhow, I'm not super tied to the scheme, and if other editors agree to just stick to IPA, I'm fine with that. But those are a couple of my arguments for the Romanization. --SameerKhan (talk) 22:00, 22 January 2008 (UTC)

Your second point, "since different studies have used very different IPA transcriptions of the vowels, it would be better to have our primary representation of Assamese not be in IPA", is an interesting one. Basically you seem to be arguing that the uncertainty in having to choose ɒ/ʊ can be staved off by creating a neutral placeholder construct of û, intended to mean "the vowel variously described as ɒ or ʊ or whatever." But even that fails in its purpose and takes sides in the argument. û, being what it is (u with a topi), is certainly inclined to hint more at ʊ than ɒ, and in any case you explicitly chose û based on ʊ when you edited them in. Tuncrypt (talk) 23:47, 19 January 2008 (UTC)

I suppose I wasn't clear before. The issue is not just "Is ও [ʊ] or [ɒ]?" There is a much bigger issue. If you look at all sorts of phonetics articles on Assamese, you'll see that everyone agrees there are four back vowels, but the symbols for them are more numerous: [u] [ʊ] [o] [ɵ] [ɤ] [ɔ] [ɒ]. In the Assamese Phonetic Reader (Baruah 1947), the transcription is as follows: উ [u] ও [o] অ' [ɔ] অ [ɒ]. This matches Masica's system, which I believe he adopted from Baruah. Peter Ladefoged's transcriptions of Assamese change across publications, and the recordings used in "A Course in Phonetics" (Ladefoged 1975) and "Vowels and Consonants" (Ladefoged 2001), which can be found here [6], are transcribed using the system উ [u] ও [ɒ] অ' [o] অ [ɔ]. Note that the IPA symbols are the same as in Baruah, but rearranged, so that [ɒ] represents the vowel ও and not অ, thus diverging from Baruah. Further transcriptions from Ladefoged, which can be found here [7], uses a different system: উ [u] ও [ʊ] অ' [o] অ [ɔ]. Shakuntala Mahanta, a linguist of Assamese background and who works primarily on aspects of the Assamese vowel inventory has also revised her system over time. In "Some Aspects of Prominence in Assamese and Assamese English" (Mahanta 2002), the transcription is as follows: উ [ʊ] ও [ɤ] অ' [o] অ [ɔ], while in later work by Mahanta, including "Edge Asymmetries in Assamese Vowel Harmony" and "On Local and non-local blocking in vowel harmony" (Mahanta 2006), the transcription is as follows: উ [u] ও [ʊ] অ' [o] অ [ɔ]. Mahanta revises her system again in "On the Convergence of Positional Markedness and Morphemic Presence" and "On Consonantal Intervention in Vowel Harmony" (Mahanta 2007): উ [u] ও [ɵ] অ' [o] অ [ɔ]. In summary, IPA transcriptions for Assamese are extremely varied. The vowel উ can be [u ~ ʊ], the vowel ও can be [ʊ ~ ɵ ~ ɤ ~ o ~ ɒ], the vowel অ' can be [o ~ ɔ], and the vowel অ can be [ɔ ~ ɒ]. Conversely, the IPA symbol [ʊ] can refer to ও or উ, the symbol [o] can refer to ও or অ', the symbol [ɔ] can refer to অ' or অ, and the symbol [ɒ] can refer to ও or অ. As you can see, the IPA isn't necessarily a more reliable way to transcribe Assamese, due to the various transcriptions given using the same symbols. So the fact that so many sources use [ɒ] loses its meaning when you see that people have used the symbol to refer to different vowels of the Assamese language anyway. That's why, in my opinion, it would be much easier to use a system that doesn't have to assume some exact phonetic value to be understood. উ can be "u", ও can be "û", অ' can be "o", and অ can be "ô". This keeps the particular phonetic values somewhat more vague (even though the current version of the article has IPA equivalents drawn from some of Ladefoged's and Mahanta's work), while illustrating the phonological correlations across related languages (e.g. Bengali অ is also Romanized "ô", and the equivalent of Bengali অ' - written simply অ is the Bengali system - is "o"). Anyhow, I've already taken up far too much space, but I think you get the idea. Like I said before, though, I am not personally attached to the system, and if people want to change it or do away with it, I'm not going to keep reverting the changes. I just want to make sure that other Wikipedians agree with any changes, and that people know the merits of an IPA system and a non-IPA system. --SameerKhan (talk) 22:00, 22 January 2008 (UTC)

I would say that your second point (this block above), demonstrating the uncertainty of the vowel system, is substantially strong and convincing in the case for a romanization. So you've achieved a level of agreement from me, but there's still the question of "Where do we go from here?".

For example, romanization or not, wouldn't the uncertainty of the vowels would still have to be addressed, tackled, and adequately resolved, in IPA, with any such phonology section or article solely and inevitably written in it anyway? Of course other things such as a grammar article would fall in the domain of romanization. Tuncrypt (talk) 14:59, 23 January 2008 (UTC)

As far as I can see, there is no uncertainty in the vowels, only in the system it is depicted in. If there exists an uncertainty in the vowels themselves, then that is in the realm of real linguistic research. I doubt Wikipedia can handle it. Chaipau (talk) 17:33, 23 January 2008 (UTC)

Wait, so it isn't a disagreement of values... but of just symbols? If THAT's the case, then there's no problem... we use the most updated and currently applicable symbols... Tuncrypt (talk) 22:03, 25 January 2008 (UTC)

I interpreted Chaipau's comment to mean that there is no controversy as to which word has which vowel in it, but that the controversy is in how to transcribe that vowel in IPA or any non-native script. When talking, Assamese speakers don't need to know whether the vowel in কোলা is high-mid back rounded, lower-high back rounded, low back overrounded, high-mid back unrounded, or lower-high central rounded. They just need to know that it's not the same vowel as in কুলা, in ক'লা, or in কলা, etc. Thus, in a sense, it's just an issue of symbols, but "just an issue of symbols" is not that minor. If the vowel in কোল can be transcribed in IPA as one of five symbols (o ɒ ʊ ɵ ɤ) depending on the author, especially when some of those symbols are used to describe completely different vowels in other authors' systems... I feel like choosing one IPA scheme over another will just be confusing and controversial. I would prefer using a system that doesn't make strong claims as to the exact pronunciation of the vowels, and I feel like choosing something that looks non-IPA-style and that focuses on the phonological status of different vowels (e.g. the use of the circumflex diacritic to show the phonological relation of the -ATR vowels ও û এ ê অ ô to their corresponding +ATR vowels উ u এ' e অ' o, with which they are often in phonological alternation... and the cross-linguistic connection between Assamese অ ô and Bengali অ ô) would focus on the important information without leading the reader towards IPA transcriptions that can be confusing if people mix systems. --SameerKhan (talk) 01:10, 26 January 2008 (UTC)

Thanks for this discussion, which I found very informative. As a non-linguist, I see now the problem with transcription for Assamese. SameerKhan, it seems to me, has established that there does not exist a standard, and that individual linguists, like Mahanta, have used different evolving schemes, though consistently every time they used them.

I now think there is no need to change the transcription scheme as it stands now. There is an intrinsic appeal in using one that is useful for linguists as well as non-linguists. And I am largely happy that this transcription scheme brings out the phonetic characteristics of Assamese against the other Indo-Aryan languages not only for the linguists, but for the non-linguists as well.

After this discussion concludes, we hope to adopt this standard in the Wikipedia:WikiProject Assam as well as in as:.

Chaipau (talk) 14:50, 23 January 2008 (UTC)

And then there's another thing. While "original", but still "verifiable", we could help ourselves out in choosing between the different sound schemes. We could quite simply have Assamese Wikipedians or whomever record alleged minimal pairs such that we could judge. It wouldn't change the fact of scholarly disagreement and the need to note it, but what this would do is frame the matter. It would inform as to what to choose between "... ɒ, but so and so thinks it's ʊ..." and "... ʊ, but so and so thinks it's ɒ...". Know what I mean?

A phonemic inventory exists in Asamiya. I don't think one should attempt this again for WP. One could, otoh, provide sound samples for words, names, etc. whenever appropriate. Chaipau (talk) 16:59, 23 January 2008 (UTC)

For I do believe that a case for romanization hinges on the uncertainty of vowel symbols, and that if the uncertainty weren't there (or at least were toned down to the level of establishable standard versus noted objection as I was getting at), then the matter might reduce itself to my earlier arguments of there not being much of a point to setting up a romanization scheme as Assamese IPA otherwise conveniently lacks the major trappings (bridges, hooks, colons, ligatures) that'd necessitate it. Tuncrypt (talk) 15:22, 23 January 2008 (UTC)

I would support keeping the transliteration we currently have, but add in the information on the varying IPA transcriptions of the language. Unfortunately, I'm extremely busy right now writing my dissertation, so I can't do this right away. If someone else wants to work on it (the article or my dissertation), feel free! --SameerKhan (talk) 19:20, 25 January 2008 (UTC)

Yeah, it'll stay as it is for now. I will read all that stuff you posted. Tuncrypt (talk) 21:31, 25 January 2008 (UTC)

I made a reply to one of Chaipau's posts. Just pointing that out in case it gets missed. Tuncrypt (talk) 22:05, 25 January 2008 (UTC)