Hindi, Urdu: Pronunciation and transcription of [n] nasal or dental

Discussion in 'Indo-Iranian Languages' started by marrish, May 14, 2014.

  1. marrish

    marrish Senior Member

    اُردو Urdu
    Dear friends of the Forum,

    We've had a short discussion with Dib about the transliteration and transcription of nasals in Hindi and Urdu (cf. sticky thread "Transliteration of Urdu and Hindi as used in this Forum") and Dib posed a pertinent question which I am unsure whether I am competent to answer.

    The question is about the pronunciation of [n] or [N] in Hindi or Urdu in the example word دانت दांत and दान्त. The question arises, is or can be pronounced da:nt/daant through which we have a problem between transliteration and pronunciation?

    Other examples and thoughts are most welcome!
  2. Dib Senior Member

    Bengali (India)
    Well, I shouldn't probably have chosen a real word - "daaNt" (tooth) in this case, to illustrate my point, because our knowledge of the language would disambiguate the situation. Let's say something like "shoNdaa" (it is actually a valid Bengali word but, I am assuming, unknown in Hindi-Urdu). I think, the transliteration scheme used on this forum would interprete it as Hindi Devanagari "शोंदा" and Urdu "شوندا", both of which are ambiguous, in my mind, as far as the sequence "oN" is concerned, whether it is a pure nasal vowel, or whether it is a vowel+nasal consonant sequence. It may be argued that, at least, in the case of Urdu, the use of N in the transliteration actually removes this inherent ambiguity of the native orthography and unambiguously marks it as a nasal vowel. But I am afraid, that is not the case for Hindi Devanagari scheme.

    What I am trying to figure out is whether we can add a new symbol to the scheme (I suggest: ~ put after a vowel, e.g. o~) to unambiguously mark a vowel as nasal. For the most part, I am fine with continuing with the presently popular transliteration scheme with partly ambiguous N for Devanagari, and phonetically better-informed transliteration than the native script in case of Urdu - as far as "transliteration" is concerned. My suggestion is only for phonetic "transcription", without having to resort to full fledged IPA, and reusing the already excellent transliteration scheme used here.

    Anyway, it is a not a pressing issue, but ideas are welcome. I'd probably use the ~ on an adhoc basis as and when necessary, unless there is any other suggestion.
  3. Qureshpor Senior Member

    Punjabi, Urdu پنجابی، اردو
    You have opened an interesting thread, marrish Jii. Sometimes I am not quite sure when we have a fully fledged nasal consonant and when it is a nasal vowel.

    chaNg (claw) or chang (also raNg/rang)

    raNj/ranj (dukh)



    You get the drift?
  4. Dib Senior Member

    Bengali (India)
    The words that I am reasonably sure about from your list:

    raNg/rang - (velar) nasal consonant + g

    daaNt/daant - pure nasal vowel, but a hint of consonantal n is also possible (meaning slightly later closure of the nasal airstream than the oral airstream at the occlusion of t); This is when the word has its usual meaning of "tooth". However, there exists a Sanskrit word "daanta-" (tamed, subdued). I don't know whether it has ever been used in Hindi. But if it is used, I am pretty sure it would be pronounced "daant" (like "shaant" < Skt. "shaanta") with a clear consonantal n.

    guNR/guण - much trickier. My "perception" is that the vowel is indeed somewhat nasalized, especially towards the end of its duration (this may be true also for rang - it just means that the nasal airstream associated with the nasal consonant starts slightly earlier than the oral occlusion associated with it). As for the nasal consonant itself, the English Wiki article on "Hindustani phonology" states, and I agree: "[/ɳ/] has a nasalized flap [ɽ̃] as a common allophone," where the flap in question is transliterated in our scheme here as "R". So, I think both "NR" and some single-letter symbol would be reasonably close as phonectic transcription, though there may be more accurate ways.
    Last edited: May 14, 2014
  5. tonyspeed Senior Member

    English & Creole - Jamaica
    This issue is discussed by Ohala. An early article can be read in full here: http://linguistics.berkeley.edu/~ohala/papers/nasal_epenthesis.pdf

    According to my understanding homorganic n is possible after nasalised vowels mainly when the following consonant is voiced. (t is not voiced) ChaaNd is Ohala's model word for this phenomenon since d is voiced. Therefore, though written as chaaNd, what is actually heard is chaaNnd. In conclusion they argue it results from the "physical restraints of the vocal tract." A similar paper is in M. Ohala's book, which I have not read in full.
    Last edited: May 14, 2014
  6. tonyspeed Senior Member

    English & Creole - Jamaica
    The writing scheme is ambiguous in Devanagri whenever the dot is used. As a non-native learner, I had to learn the difference by dictionary. Native speakers learn to differentiate by ear.

    The accurate spelling of daaNt in Hindi is दाँत (using the anuswara , since there is nothing above the top line to take up the space it needs), but due to laziness in the 1800s, the dot (anusik) is used by many people for both consonants and nasalised vowels.
    Last edited: May 14, 2014
  7. Dib Senior Member

    Bengali (India)
    Thanks a lot for sharing this. I just gave it a quick look, and the results presented seem to conform with my own personal perceptions - for both Hindi and French (Ohala uses weird IPA for some French vowels though). Certainly, the n in "chaaN(n)d" is very pronounced and worth being analyzed as a separate segment, while in "daaN(n*)t" it is not. However, a brief nasal may still exist phonetically. Refer to Ohala's paper page 5/211, last two paragraphs for Hindi and French respectively. Correct me, if I am wrong. I'd, however, argue that the "universality" of epenthesis of a homorganic nasal between a nasal vowel and a following voiced consonant is likely a bit exaggerated. Exhibit: Standard Kolkata Bengali (as opposed to Eastern Bengali, which has very clear epenthetic nasals - it, in fact, works as a shibboleth). I'll be surprised if my Bengali "cha~d" has anything more than a slightly prenasalized d (in line of the t of Hindi daaNt)**. But then, I am not a phonetician. I should probably stop making too bold statements. :p

    ** Removed this because, it is not my intention to assert this, though the paper seems to give some evidence towards this.
    Last edited: May 15, 2014
  8. lcfatima Senior Member

    In a teapot
    English USA
    What about this one: I noticed some people say baiNgan and some say baingan. It sounds like some people even say baigan with no nasal consonant or nasalization on the 'ai' at all, though that may just be my imperception of a light nasalization. [h=2]बैंगन بينگن[/h]
  9. Dib Senior Member

    Bengali (India)
    Yes, baiN(n)gan also fits the bill. It is like "chaaNd", and epenthetic nasal velar is common in this context. A version without any nasal - "baigan" - won't surprise me (though I have to admit, I am not familiar with it), as the corresponding Bengali ("begun") and Oriya ("baigɔNR(ɔ)") words have non-nasal vowels.


    To the topic now: I think, we all realize the problem Marrish and I were trying to highlight. The mission was to decide whether it is worth trying to remedy this ambiguity in transcription, and if so, how.
  10. tonyspeed Senior Member

    English & Creole - Jamaica
    Oxford lists baigan as a pronunciation variant of baiNgan.
  11. littlepond Senior Member

    Note that "daaNt" (going by the "old" transcription scheme, i.e., nasal), for tooth, but, surprisingly, "dant" (not nasal) as in "dant-manjan" (toothcleaning powder/paste).
  12. littlepond Senior Member

    While I have heard both "baiNgan" (the standard) and "baigan", I have never heard "baingan"! Could you shed some light on the speaker's background, if possible, lcfatima jii?
  13. Wolverine9 Senior Member

    American English
    Since this ambiguity is also present in speech for a lot of words, I don't know if there is a remedy. I guess people can just adjust the transliteration to N/~ or n, depending on how they pronounce a particular word, or they can continue following the orthographic conventions of the language.
  14. mundiya Senior Member

    Hindi, English, Punjabi
    Continued from the thread http://forum.wordreference.com/showthread.php?t=2947452&page=2

    That's where I believe you are mistaken, littlepond jii. The words are indeed baaNdh and DhuuNDhnaa. There are no such words as baandh and DhuunDhnaa. If you notice, the correct spelling of both is with chandrabindu (nasal N). As I said before, the pronunciation of nasal N depends on the structure of the word. It can sometimes have a consonantal touch, which is why you are disputing their status as nasalized vowels.

    बाँध bā:dh (nm) a dam; bund; weir; dike; barrage; an embankment; —टूटना a bund to give way; an embankment to be swept away; o, धैर्य का one's patience to be exhausted, patience to come to an end; —बाँधना to erect a dam, to raise a bund, to construct an embankment.

    ढूँढना ḍḥū:ḍḥnā: (v) to seek, to search; to trace.

Share This Page