Sanskrit: visarga pronunciation

marrish

Senior Member
اُردو Urdu
Dear friends,

Recently the question of Vedic pronunciation of some sounds was discussed in extenso but here I have a question about Classical and Modern pronunciation of visarga (of course when it is in pausa or in other non-samdhi positions).

The most popular because it is the Modern Indian pronunciation is with the echoing of the preceding vowel (aha, ihi, uhu).

I have some doubts about it. Perhaps they might be cleared if we can ascertain whether such a pronunciation fits into the metric schemes?

Otherwise, what is the international practice (in universities giving courses of Sanskrit) to pronounce it?

I feel very comfortable with the unvoiced variant of 'h' without the echo but it is perhaps wishful thinking.

Thank you, I know that responding is not as easy as posting a query.
 
  • sanskrut_bhashik

    Member
    Marathi
    For simplification..
    देव: देवह (a god)
    देवा: देवाहा (gods)
    देवयो: देवयोहो (genitive-of both gods/. Locative- in both gods)
    नदी: नदीहि (a river)
    धेनु: धेनुहु (a cow)


    I was taught like this. I learned Sanskrit for 6 years as school and college subject. There might be variation regionally.


    Usually, visarga is pronounced as the 'ह' sound, but adding same vowel to 'ह' as in previous letter.


    Only exception is when the vowel preceding the visarga is 'ऐ'. In this case, visarga is pronounced as following


    तै: तैहि (plural instrumental of तत् - by them)
     
    Last edited:

    sanskrut_bhashik

    Member
    Marathi
    No idea about standard practice. I think you are professional. I am only language enthusiast. No idea of classical and modern use either.
     

    Au101

    Senior Member
    England, English (UK)
    This pronunciation throws out the metre completely, but I think it's been around for a while and certainly is what I was taught.
     

    Dib

    Senior Member
    Bengali (India)
    At least in the traditional Sanskrit pronunciation of Bengal, there is never any echo vowel. Unless it is heavily Bengali accented, it's just a devoiced h, and that's how I learnt it in school too. That's how it is presumably taught everywhere in Bengal. Also, even not-so-educated rural priests use the same pronunciation there.

    ----
    EDIT: Just for clarity, in Bengali accented Sanskrit, the visarga is often assimilated to the following consonant (e.g. duHkha- > dukkhɔ).
     
    Last edited:

    sanskrut_bhashik

    Member
    Marathi
    I maybe biased, but Marathi retains most accurate pronunciations of 'tatsama' Sanskrit words. At least more than that in Hindi. That's why I think they teach most accurate "Sanskrut" there.
     

    Gope

    Senior Member
    Tamil
    Just a couple of observations: देवयोः is pronounced dEvayOho witha short o (there is no representation of this sound in Devanagari script, but Dravidian languages have a short o.)
    similarly, देवाः is pronounced dEvAha with a short a at the end, and not dEvAhA.
    Also हरेः, ablative and genitive sing. of हरि is pronounced HarEhe with a short e at the end. This short e sound is present in the Dravidian languages, but cannot be represented in Devanagari script.

    Now about the visarga in the metric system. The thing to remember is that the sanskritic tradition has always been oral and the pronunciation has been preserved intact across the millennia. I am referring to the pronunciation not tinged by the influence of the mother tongue of the speaker, the pronunciation which you hear from scholars who have rigorously learnt the language. In this context it is interesting to note that the entire Ramayana of Valmiki (about 25000 verses) is set in a metre called śloka metre, patterned after the malediction he uttered spontaneously on a hunter who before his eyes slew a male sarasa crane while in the act of lovemaking:
    मा निषाद प्रतिष्ठां त्वमगमः शाश्वतीः समाः
    यत्क्रौञच मिथुनादेकं अवधीः काममोहितम् ।।
    So the model itself has a visarga at the end, and has not been thought of as not fitting into the metre.:)
     

    sanskrut_bhashik

    Member
    Marathi
    Yes, visarga does not have any representation in script because visarga itself is part of Devanagari script. For simplification I wrote in that way. Visarga usually has short and subtle pronunciation.

    Also, there is no long & short 'O' vowel in Sanskrit and Sanskrit-derived languages. Even इ-ई, उ- ऊ differentiation is not stressed in Sanskrit-derived languages especially Gujarati, Marathi; which is actually a wrong practice.
     

    Dib

    Senior Member
    Bengali (India)
    ...
    मा निषाद प्रतिष्ठां त्वमगमः शाश्वतीः समाः
    यत्क्रौञच मिथुनादेकं अवधीः काममोहितम् ।।

    So the model itself has a visarga at the end, and has not been thought of as not fitting into the metre.:)

    Assuming that this comment is made in reference to Au101's comment: "This pronunciation throws out the metre completely", I'd say, you probably misunderstood him. What he meant is, adding "echo vowels" after the visarga messes up the meter, not the visarga (realized as a sort of -h only) itself. Just count the number of syllables. The meter works because it contains 4 paada's of 8 syllables each:
    mā niāda pratiṣṭhāṃ tvam
    agama śāśvatī samā
    yat krauñcamithunād ekam
    avadhī kāmamohitam

    As you can see, as soon as echo-vowels are added after the visarga's, the syllable counts become 8-11 || 8-9. I am pretty sure, the Ādikavi did not mean it that way.
     

    tarkshya

    Senior Member
    Marwari
    I know that that modern pronunciation of visarga is with an h sound at the end of the word. However, I wonder why the inventors of the Devnagri and related scripts devised a separate symbol ( : ) for it. I mean, देव: can easily be written as देवह. Why the need of a separate symbol? Somehow redundant symbols in any script makes me suspicious. I raised same suspicions over Urdu script too earlier, and discovered that each of its unique letter did represent a unique sound in the beginning.

    Could it be possible, and this is just a wild hypothesis, that visarga was originally used to create aspirated consonants from the consonant preceding it?
     

    tarkshya

    Senior Member
    Marwari
    I maybe biased, but Marathi retains most accurate pronunciations of 'tatsama' Sanskrit words. At least more than that in Hindi. That's why I think they teach most accurate "Sanskrut" there.

    I believe Sanskrut vs Sanskrit question needs a separate thread of it own.
     

    Au101

    Senior Member
    England, English (UK)
    I know that that modern pronunciation of visarga is with an h sound at the end of the word. However, I wonder why the inventors of the Devnagri and related scripts devised a separate symbol ( : ) for it. I mean, देव: can easily be written as देवह. Why the need of a separate symbol? Somehow redundant symbols in any script makes me suspicious. I raised same suspicions over Urdu script too earlier, and discovered that each of its unique letter did represent a unique sound in the beginning.

    Could it be possible, and this is just a wild hypothesis, that visarga was originally used to create aspirated consonants from the consonant preceding it?

    As our good friend Dib (thank you for clarifying my comment, by the way, you were quite right) mentioned over in the other visarga thread:

    Just a very tiny addition: while the normal "h" [ह] was voiced (ghoSha-dhvani) [(i.e [ɦ])], the visarga was supposedly always voiceless [(i.e )] , which can be also guessed at from the environments, it occurs in.


    (Blue additions are mine)

    It is worth noting also that, as mentioned upthread, the echo of the vowel in देवः is not really pronounced exactly like the full अ in देवह, it should be noticeably shorter.

    Thirdly, as we discussed over on your original thread, there were originally three different phonetic realisations: /x/, /ɸ/ and /h/. The /x/ and /ɸ/ could be written with their own, separate characters, or visarga could be used for the lot.

    Finally, I would say visarga, which arises from word-final -s and -r is - phonologically - a long way from ह. From the viewpoint of Sanskrit language, Sanskrit linguistics, Sanskrit metrics, Sanskrit phonology and Sanskrit grammar, अः is really as different from अह as it is from अक, even though - because of natural linguistic development - their pronunciation is these days very similar.
     

    Gope

    Senior Member
    Tamil
    Assuming that this comment is made in reference to Au101's comment: "This pronunciation throws out the metre completely", I'd say, you probably misunderstood him. What he meant is, adding "echo vowels" after the visarga messes up the meter, not the visarga (realized as a sort of -h only) itself. Just count the number of syllables. The meter works because it contains 4 paada's of 8 syllables each:
    mā niāda pratiṣṭhāṃ tvam
    agama śāśvatī samā
    yat krauñcamithunād ekam
    avadhī kāmamohitam

    As you can see, as soon as echo-vowels are added after the visarga's, the syllable counts become 8-11 || 8-9. I am pretty sure, the Ādikavi did not mean it that way.
    Dib dada. I was responding to marrish's thread.
    You will hear the poem recited as
    maa nishaada pratishThaam tvamagamśśaaśvatiissamaah, i.e. 8-8 and not 8-11.
    yatkrauñcha mithunaadekamavadhiihkaamamohitam is again 8-8, not 8-9.
    I think marrish's question was about the pronunciation of visarga in a pause or non-sandhi position, i.e. In this case समाः, and not about agamah, or śaaśvatiih, or avadhiih. One might hear, one usually does hear. समाह but as sanskrut_bhashik observed, here it has a short and subtle pronunciation, very unlike full ha. We might even imagine in a noise free world of a couple of millennia ago this being ponounced as samaah, assuming they all had keener ears then!:)
     
    Last edited:

    Dib

    Senior Member
    Bengali (India)
    Dib dada. I was responding to marrish's thread.

    Oh, sorry for getting the context wrong.

    We might even imagine in a noise free world of a couple of millennia ago this being ponounced as samaah, assuming they all had keener ears then!:)

    Well, as long as वराः and वराह are distinct... :D
     

    marrish

    Senior Member
    اُردو Urdu
    I listened to many recitations. Marathi influenced was visarga 'h' or assimilating but at the end with echo.
    Hindi the same.

    Bengali I would love to hear.
    Southern I haven't had the pleasure.
     

    Gope

    Senior Member
    Tamil
    I listened to many recitations. Marathi influenced was visarga 'h' or assimilating but at the end with echo.
    Hindi the same.

    Bengali I would love to hear.
    Southern I haven't had the pleasure.
    i only noticed today that my post has been deleted, because youtube links are not allowed.
    Here is a link to a southern audio rendering of Kalidasa's lyrical poem Meghaduta, Cloud Messenger, containing readings of the poem as well as the commentary in prose:

    http://www.vedabhoomi.org/MeghaSandesha.html
    In the menu that appears, from the title PM (pūrvamegha) onwards the commentary and text proper begin.
    This recording has been made by scholars who are native speakers of Telugu, spoken in Andhra Pradesh (and Telengana) and is fairly representative of two other states of the South as well, namely Tamil Nadu and Karnataka. :)
     

    marrish

    Senior Member
    اُردو Urdu
    ^ Thank you very much, its hours of listening which I will do some day. Nevertheless I have noticed the way of pronunciation.
     

    marrish

    Senior Member
    اُردو Urdu
    Still, referring to the recordings, it seems quite funny when those syllables with oho ihi etc. are pronounced in haste as if they weren't important for the metre. Another funny thing is that the ending 'aH' at the end of the sentence is pronounced "ahaa".I have read about the visarga sound that it is a voiceless 'h' as opposed to the voiced one ह.

    I can realise that the reconstruction of the original pronunciation might be not an easy task but still, I wish to see the sources which say that it was the unvoiced sound. Secondly, the process of adding the 'echo' vowel is totally obscure to me. There must have been some underlying reasons.
     

    germanictamoon

    Senior Member
    Hindi (West Uttar Pradesh)
    When a language is dead i.e. have no native speaker, it loses its phonology too, just like Proto Indo European language. Same is the case with Sanskrit. Every constructed sound is biased by the constructor's native language, even if the constructor's native language derives from the dead language.
    Here are three sounds in Sanskrit which are completely lost with time,
    1. Vowel Ri
    2. so-called murdhanya Sh
    3. Lri .

    Anyway fortunately visarga is not one of them.

    Visarga was originally a Glottal Stop, having no sound. It means speaker has to stop here with a short and subtle jerk.
    To pronounce it like 'h' sound is utterly wrong.
    It is present in Hindi.

    One can realise it between Do(give) and Do (Two)
    Do(give) is almost always pronounced with glottal stop, while Do (two) is pronounced continuously.

    Same can be demonstrated for word 'du:kh' (sorrow)'s original pronunciation. It has a visarga between du and kh. Now pronounce it and feel it.
     

    Dib

    Senior Member
    Bengali (India)
    Anyway fortunately visarga is not one of them.

    Visarga was originally a Glottal Stop, having no sound. It means speaker has to stop here with a short and subtle jerk.

    That's an unexpected suggestion! Yes please, give us further reference to find more details about this idea.

    One can realise it between Do(give) and Do (Two)
    Do(give) is almost always pronounced with glottal stop, while Do (two) is pronounced continuously.

    Again, this is also an unexpected suggestion. I do not feel this way. Normally, I do hear the glottal stop after long vowels, e.g. in many Arabic letter names, without problem. But I never seem to have heard it in Hindi. Do you have some linguist's analysis, or something similar that you can give us reference to for further details? Thanks.
     

    Qureshpor

    Senior Member
    Panjabi, Urdu پنجابی، اردو
    [...]Visarga was originally a Glottal Stop, having no sound. It means speaker has to stop here with a short and subtle jerk.
    To pronounce it like 'h' sound is utterly wrong. It is present in Hindi.

    One can realise it between Do(give) and Do (Two)
    Do(give) is almost always pronounced with glottal stop, while Do (two) is pronounced continuously.

    Same can be demonstrated for word 'du:kh' (sorrow)'s original pronunciation. It has a visarga between du and kh. Now pronounce it and feel it.
    Interesting. Is what you are saying the same as what I am saying in the thread below, albeit I thought it might be "tone" whereas you are saying it is a "glottal stop"?

    http://forum.wordreference.com/showthread.php?t=2625997&highlight=ghantii
     

    Au101

    Senior Member
    England, English (UK)
    I suppose I have two or three things I want to add to this at the present time. The first is that germanictamoon is not wrong - indeed they are quite right - to point out to us that, when reconstructing sounds, we must always be aware of our limitations. We will never know with the kind of certainty I can have about the existence of the laptop in front of me (which some philosophers might want to argue is not even all that certain :p) what languages we have to reconstruct really sounded like. We will never have any primary evidence, we are always forced to extrapolate and interpolate and, in the end, to make what can only ever be our best guess, based on the available evidence.

    To behave like we have nothing to go on and we can't make pretty compelling conclusions is equally mistaken, however, in my view. To put it rather briefly we can look at transcriptions, we can look at how Sanskrit words were transcribed into Ancient Greek, for example. We can compare sound correspondences, we can compare similar words in different languages which all have similar meanings, e.g.:

    English: father
    Sanskrit: pitā
    Latin: pater

    We take note of the vast separation between these languages in time and location. We note the incredible similarity, despite subtle differences. We remark upon the strange fact that there is nothing intrinsic in these words that force them to mean father. Other languages do not have similar words for father, e.g. in Hebrew it is אָב ʾāḇ, in Finnish it is isä, in Nama it is ǁgûb (ǁg represents an alveolar lateral click). I'm summarising, but based on extensive correspondences, observations about how languages change from written records, as well as from what we can see in our own lifetime, knowledge of the history of these languages and their speakers, we deduce that these languages all derived from a common proto-language and that in that language was a word for father (or a word that at least had a similar meaning, such as parent, ancestor, kinsman) which gave rise to all of these "reflexes".

    So how might we make deductions about what that word would have sounded like. Well we need to start by gathering as much evidence we can and look at many, many languages. Three is not enough. We need to decide, what was the original first sound, was it p, or was it f? Or was it something else? Well it's not very likely to be something else, because it turns out most reflexes start with a /p/ or an /f/ sound, in fact most of them start with /p/.

    In some cases it is enough to simply take note of the majority. The reason this can be enough is that it's difficult to see why lots and lots of languages would all have changed and all in the same way. But we need to look at phonetic plausibility too. In the case of /p/ and /f/ such a change is commonly seen in the development of languages, others are less common. Historical linguistics is not a totally arbitrary, nor a fledgling field, we have a lot of information about changes which are known to have taken place, changes you can see in the written record - and in the spoken one too. We also know a little bit about phonology and we have a lot of knowledge about processes such as assimilation, dissimilation, lenition, fortition and so on and so forth.

    But we don't just make comparisons between languages, we can look at changes within a language itself and, using our knowledge of the language in question, other languages, phonology and so on, we can make deductions about what changes are likely to have occurred.

    We must never get carried away, we must never be too pleased with our theories, because we will never have any raw data, but there are likely possibilities, there are unlikely possibilities and there is sheer fancy.

    The second thing I want to say is that, as I've said before, with Sanskrit we are not completely blind. In fact, we have, rather happily, a meticulous, written account of the production of the sounds of Sanskrit. The prātiśākhya and other śikṣā texts, as well as the schools teaching recitation allow us to talk with a fair amount more authority about Sanskrit pronunciation than, say, PIE. It is every bit as facile to wave this evidence away as it is to think that it allows us to be absolutely sure (given the enigmatic nature of the śikṣā texts and the changes we know pronunciation in the Vedic schools has undergone).

    Speaking specifically about the visarga it is important to remember that it is a very late innovation of Sanskrit, deriving primarily from final -s and, secondarily, from final -r.

    Which really leads me onto my third point, which is more explicitly addresses marrish's original question.

    I quote Michael Meier-Brügger's "Indo-European Linguistics":

    "L 309. 2) In Greek, PIE *s is preserved when adjacent to plosives or in word-final position. Word-initially, it becomes h- (i.e. spiritus asper <῾->)."

    We may note from his book (L 310.) the following (I shall note quote him, for he delves into a detail I dare not summarise for you here):

    Compare the Greek future héksō with Vedic sáhate, Old High German sigu;
    The Greek hístāmi with Latin stāre, Vedic sthā-trá-, Old High German stān, Lithuanian stóti, Old Church Slavonic stati.
     

    Dib

    Senior Member
    Bengali (India)
    Thank you, Au101, for the extensive post. However, would you, please, explain a bit more in detail what you are trying to show in your last point (PIE *s > Gk s/h)?

    I totally agree with your second point.

    I also agree with your first point, BUT I also have some objection towards the way of presentation of your first point. Actually, it is not directed personally to you, you are just following a long tradition of how IE-ists have been presenting their case to lay people all along; but I believe, it is seriously misleading. Presented this way, it sounds no better than the similarities proposed by people like P. N. Oak. It fails to highlight the methodology to solve the problem of chance similarities, which, unintuitively, are actually very common between almost any two languages in the world, as long as their phonetic systems are "sufficiently similar". I personally feel, we should never present evidence of linguistic relation in such a dumbed down manner. A lot of sincere inquisitive people are getting misled by this dumbing down into believing that historical linguistics is nothing more than the flipside of P. N. Oak. It's a different matter, what is the best way of presenting it to the casual reader. I guess, this won't be the right place to discuss that, but I am open to discussing it with anybody else through, for example, private messages or on any other suitable medium.
     

    Au101

    Senior Member
    England, English (UK)
    I brought up the fact that PIE *s > Gk h in word initial position as it is analogous to the case of Sanskrit final /-s/ going to /h/ (assuming visarga is pronounced /h/). I suppose I just brought it up because /s/ to /h/ is a change that we can chart in Greek too and so it is certainly a plausible change to postulate for Sanskrit. We can be very confident from the paradigms and saṃdhi that visarga is from original s (sometimes r). We know that the śikṣā texts classify it as a voiceless spirant, we know that it is today pronounced /h/. Case closed? Well, of course not, but one must establish the plausibility of a change from s to glottal stop (or else challenge the idea that visarga is from original s/r) if one wants to claim that the original pronunciation of visarga was the glottal stop. This is obviously in addition to motivating such a theory in the first place. Mostly, though, I was hoping to add a little flesh to the bones, for I understood marrish - in his question which prompted this - to be seeking a bit more information about what evidence there is for the /h/ pronunciation. Unfortunately I never was in the possession of the resources to really get to the heart of that question with a focus on Sanskrit, and I certainly have even fewer tools at my disposal here at the moment, but - for what it's worth - I thought it worth adding a little evidence for the change of /s/ to /h/ Indo-European languages. I couldn't find any reference specifically to the origins of visarga, but when I spotted this whilst flicking through my books, I thought it worth a brief mention.

    Yes, I accept your criticism, it's been too long since I had this at the tips of my fingers and, anyway, I didn't fancy undertaking an introduction to historical linguistics at 3 am in the morning, so I opted to be cursory. Nevertheless it could plausibly be argued that if you're not prepared to do the job properly, you may do better not to undertake it in the first place :p I wouldn't argue with what you've said :)
     

    marrish

    Senior Member
    اُردو Urdu
    I am extremely grateful for all of those who care to contribute to this thread in a manner that exceeds my boldest expectation. It's a real pleasure to profit from in-depth discussions and different points of view. Indeed, it is true that I am eager to know more about the background of the assertion that it used to be a voiceless spirant (or glottal stop) and whether the current pronunciation in most of the parts of India has any roots in those deliberations. At last I wish we could make it till we can understand the processes leading to the latter pronunciation. For both the initial points I think that comparative evidence is significant.
     
    Last edited:

    Dib

    Senior Member
    Bengali (India)
    I brought up the fact that PIE *s > Gk h in word initial position as it is analogous to the case of Sanskrit final /-s/ going to /h/ (assuming visarga is pronounced /h/). I suppose I just brought it up because /s/ to /h/ is a change that we can chart in Greek too and so it is certainly a plausible change to postulate for Sanskrit. We can be very confident from the paradigms and saṃdhi that visarga is from original s (sometimes r). We know that the śikṣā texts classify it as a voiceless spirant, we know that it is today pronounced /h/. Case closed? Well, of course not, but one must establish the plausibility of a change from s to glottal stop (or else challenge the idea that visarga is from original s/r) if one wants to claim that the original pronunciation of visarga was the glottal stop. This is obviously in addition to motivating such a theory in the first place. Mostly, though, I was hoping to add a little flesh to the bones, for I understood marrish - in his question which prompted this - to be seeking a bit more information about what evidence there is for the /h/ pronunciation. Unfortunately I never was in the possession of the resources to really get to the heart of that question with a focus on Sanskrit, and I certainly have even fewer tools at my disposal here at the moment, but - for what it's worth - I thought it worth adding a little evidence for the change of /s/ to /h/ Indo-European languages. I couldn't find any reference specifically to the origins of visarga, but when I spotted this whilst flicking through my books, I thought it worth a brief mention.

    Ah, Ok. If it is only about plausibility, then there are examples closer home too, e.g. the ubiquitous s>h of Iranian. It also happened in some environments in Middle Indic. That's why we have, e.g. Bengali sat = seven, sottor = seventy, but ba-hattor = seventy-two. Parallelly, I am curious, do we know of any language anywhere in the world where an s>ʔ has happened? Anybody?

    Yes, I accept your criticism, it's been too long since I had this at the tips of my fingers and, anyway, I didn't fancy undertaking an introduction to historical linguistics at 3 am in the morning, so I opted to be cursory. Nevertheless it could plausibly be argued that if you're not prepared to do the job properly, you may do better not to undertake it in the first place :p I wouldn't argue with what you've said :)

    Thank you for agreeing, though I guess, I overreacted - likely goaded, at least partially, by a recent long and tiring thread in the EHL forum. But I truly believe, historical linguists need to find a better way of presenting their case to the uninitiated.
     

    Au101

    Senior Member
    England, English (UK)
    Thank you for agreeing, though I guess, I overreacted - likely goaded, at least partially, by a recent long and tiring thread in the EHL forum. But I truly believe, historical linguists need to find a better way of presenting their case to the uninitiated.

    Nah, don't worry about it, it's fine, I wasn't trying to make a show of humility, your criticism was fair and needed to be acknowledged.

    Thanks for the extra information on *s > h :)
     

    Ali Smith

    Senior Member
    Urdu - Pakistan
    I have read about the visarga sound that it is a voiceless 'h' as opposed to the voiced one ह.

    That's what I've read too. However, in Hindi, ह is always voiceless. So, I doubt Sanskrit speakers who are native speakers of Hindi would pronounce ह as a voiced glottal (non-sibliant) fricative. They would almost certainly pronounce it as a voiceless glottal (non-sibliant) fricative.
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    Secondly, the process of adding the 'echo' vowel is totally obscure to me. There must have been some underlying reasons.

    I wonder if the reason for the "echo" vowels isn't just related to phonology of the modern IA language which forms the background -- something related to the phonotactic constraints governing in what situations word-final h is possible. For example, when I hear people read Hindi texts out loud hyper-meticulously, it sometimes sounds to me like the word वह is pronounced with a light echo of a schwa at the end (something like [ʋəɦə]) even though this is a violation of Hindi's usual word-final schwa deletion. I would guess this is because something like [ʋəɦ] would not be very phonotactically permissible in these speakers' idiolects, so they compensate by adding an echo vowel to get something that is more phonotactically permissible. (Of course, this basically only happens in hyper-meticulous speech, since it's invariably [ʋo:] otherwise.)

    I would guess that similar phonotactic constraints governing word-final h were in effect in parts of South Asia in the pre-modern era, resulting in the echo vowels when pundits from these parts recited Sanskrit. European linguists came in, heard these pundits reciting the visarga with echo vowels, and the phenomenon got recorded in some early Sanskrit primers. The echo vowels presumably didn't exist classically, but maybe that's something people figured out later, after we knew more about historical linguistics and could compare the visarga to homologous elements of other IE languages. But I don't know --- just a shot in the dark! :)
     

    Dib

    Senior Member
    Bengali (India)
    That sounds reasonable to me. In Bengal, echo vowels are not used, probably because Bengali does allow (and may I say, encourage!) word final (devoiced) h in interjections, though, I believe, not in content words.
     

    aevynn

    Senior Member
    USA
    English, Hindustani
    I wish to see the sources which say that it was the unvoiced sound.

    The Āpiśali Śikṣā is supposedly pre-Pāṇini. There's a text of it here, but I don't know how if this text has been critically verified. In any case, here are some bits from this text that seem relevant to pinpointing the phonetic value of visarga. I've attempted (partial) translations, but my Sanskrit is trash. But we have some people here who can translate better! :)

    अकुहविसर्जनीयाः कण्ठ्याः
    हविसर्जनीयावुरस्यावेकेषाम्

    Attempt: a, ku (= the ka-varga?), ha, and the visarga are all kaṇṭhya
    h and the visarga both (...??...)

    य रौ ल वौ चतस्रोऽन्तस्थाश्चःक श ष सःप हाः
    षडूष्माणो विसर्गोऽनुस्वारो ळो नास्यपञ्चकम्

    Attempt: ya, ra, la, and va are the four approximants; and upadhmāniya, śa, ṣa, sa, jihvāmūliya, and ha
    are the six fricatives ...

    [I'm guessing here that ःक and ःप is notation for jihvāmūliya and upadhmāniya, the two allophonic realizations of the visarga...?]

    वर्गाणां प्रथमद्वितीयाः शषसविसर्जनीयजिह्वामूलीयोपध्मानीया यमौ च प्रथमद्वितीयौ विवृतकण्ठाः श्वासानुप्रदाना अघोषाः

    Attempt: The first and the second of each varga, śa, ṣa, sa, visarga, jihvāmūliya, upadhmāniya (...??...) voiceless

    ---

    Anyway, assuming the text on this website is to be trusted, this seems to be a rather early source telling us that visarga is a voiceless fricative whose place of articulation is kaṇṭhya. I think /x/ and /h/ are the only two such consonants, depending on whether kaṇṭhya is interpreted to mean "velar" or "glottal" (and the distinction between jihvāmūlya and upadhmāniya is velar vs glottal...?)
     

    Qureshpor

    Senior Member
    Panjabi, Urdu پنجابی، اردو
    I'll just copy paste a short post of mine from the past. I don't know if it will add anything to what has been said already.

    Whilst at university and in between my studies I was having a go at trying to decipher "Teach Yourself Sanskrit" (!!!) by Professor Coulson of Edinburgh University. I am afraid I did n't get too far with the language. Professor Coulson described the "visarga" (if I remember rightly) as a voiceless h. This I presume is how Urdu-Hindi h is pronounced now, in words like "ham, haT". However, ह was described as a voiced h. This is when I thought..ah, this must be our Punjabi h. and possibly the h in the Punjabi word k_har (ghar) for house.
     

    Dib

    Senior Member
    Bengali (India)
    Thanks, aevynn for the quotations and translations. Here's the completion for the missing parts from the translations:

    अकुहविसर्जनीयाः कण्ठ्याः
    हविसर्जनीयावुरस्यावेकेषाम्

    Attempt: a, ku (= the ka-varga?), ha, and the visarga are all kaṇṭhya
    h and the visarga both (...??...)

    "h and the visarga are urasya (pectoral, from the chest) for/according to some people".

    य रौ ल वौ चतस्रोऽन्तस्थाश्चःक श ष सःप हाः
    षडूष्माणो विसर्गोऽनुस्वारो ळो नास्यपञ्चकम्

    Attempt: ya, ra, la, and va are the four approximants; and upadhmāniya, śa, ṣa, sa, jihvāmūliya, and ha
    are the six fricatives ...

    This finishes the list of consonants, started in the previous verse. After the fricatives, it lists "the visarga, the anusvāra, ळ and the five nāsya's (=pure nasals, also nāsikya's)". See below for more on the nās(ik)yas.

    वर्गाणां प्रथमद्वितीयाः शषसविसर्जनीयजिह्वामूलीयोपध्मानीया यमौ च प्रथमद्वितीयौ विवृतकण्ठाः श्वासानुप्रदाना अघोषाः

    Attempt: The first and the second of each varga, śa, ṣa, sa, visarga, jihvāmūliya, upadhmāniya (...??...) voiceless

    (...??...) = and the first and second yama-s [see below] are open-glottised, "breathy-emitted" and

    ---

    I am, unfortunately, neither super-knowledgeable in Phonetics nor in the Sanskrit grammatical/phonetic traditions. So, I had to dig around a bit to make sense of some of the things, and still, I am not sure that it is all clear. The two sources, I found, quite useful are these:
    Phonetic Texts of Early Indians - K. Meenakshi 2002-3
    Patañjali's Mahābhāṣya with a running Hindi translation (p. 211-2)

    Anyway, I do not know exactly what विवृतकण्ठ (open-glottised) and श्वासानुप्रदान (breathy-emitted) - to take Meenakshi's explanation - correspond to in terms of our modern understanding of phonetics.

    As for the pure nasals, they seem to be some sort of "nasal glides" - four of them are called "yama-s", i.e. (nasal) twins of the previous stop. It seems - by combining the examples (but not the explanation, which sounds wrong) from the Hindi translation of Patañjali and the definition of Monier-Williams - that they occur between a stop and an immediately following (non-homorganic?) nasal. They are supposed to retain the voice and aspiration of the preceding stop, but not its place of articulation, which is purely nasal in anticipation of the following sound. This provides us 4 yama-s. According to Patanjali's statement: "नासिकावचनोनुनासिकः इतीयत्युच्यमाने यमानुस्वराणामेव प्राप्णोति", i.e. He gets only the yama-s and the anusvāra if it is just said that "anunāsikas/nasals are those pronouced by the nose" [i.e. as opposed to both mouth and nose - मुखनासिकावचन as in Pāṇini 1.1.8], the fifth pure nasal is clearly the anusvāra. This is, however, somewhat in conflict with the आपिशलिशिक्षा since it lists the anusvāra separately from the five pure nasals. I am not sure what the fifth nāsikya is supposed to be here.
     
    Last edited:

    aevynn

    Senior Member
    USA
    English, Hindustani
    Thanks for completing the translation, and for the references! Another useful reference on Sanskrit phonetics that we might add to the list is

    Manmohan Ghosh's critical edition of the Pāṇinīya Śikṣā.

    These references seem to indicate that Sanskrit phoneticians weren't in complete agreement about terminology, which makes things quite unclear... Even with this variation in terminology, though, it does seem rather hard to interpret the आपिशलिशिक्षा completely consistently. Then again, it's supposedly a rather early text, so maybe we shouldn't expect perfection of it :)

    कण्ठ्य, उरस्य

    I wonder if the "some"/"others" according to whom "हविसर्जनीयावुसरस्यौ" were trying to use the word उरस्य to distinguish a glottal place of articulation from a velar one. In other words, maybe some phoneticians thought that others were being sloppy by lumping together कु (which is in fact a shorthand for the क-varga, cf. note 22 on p. 63 of Ghosh) and ह all under the label of कण्ठ्य, so they proposed the word उरस्य to describe ह and विसर्ग. If this is correct, these phoneticians correctly noted that the place of articulation is further back than the place of articulation of कु, but they "overshot" in that they went all the way back to the chest instead of the glottis.

    There's a nice table on p. 62 of Ghosh which shows how a number of early Sanskrit phonetic texts classified the places of articulation of Sanskrit's phones, though it doesn't talk about the आपिशलिशिक्शा. Interestingly, a number of these other texts do seem distinguish between the place of articulation of कु and ह, though the place of articulation of the former is the जीह्वमूल while the place of articulation of the latter is the कण्ठ.

    विवृतकण्ठ, श्वासानुप्रदान, अघोष

    I'm also not very knowledgeable about the physiology of human speech, but apparently, voicelessness is characterized by "lack of vibration of the vocal cords; arytenoid cartilages usually apart." It seems plausible that विवृतकण्ठ is referring to the the arytenoid cartilages being apart (ie, the glottis being open), and अघोष to the lack of vibration of the vocal cords.

    Notes 12-13 in Ghosh (pp. 55-57) discuss प्रयत्न and अनुप्रदान (called आभ्यतरप्रयत्न and बाह्यप्रयत्न by later Sanskrit phoneticians). The distinction seems roughly about what the mouth is doing vs what the larynx is doing. Ghosh mentions that श्वास is an example of an अनुप्रदान but doesn't say much about what this means. Meenkashi says at one point that "breath (श्वास) [is the emission feature] of unvoiced aspirated consonants" (p. 198), but this is in the context of a different Sanskrit phonetic text and this is clearly not exactly the correct interpretation in the आपिशलिशिक्षा (since वर्गाणां प्रथमाः श्वासनुप्रदानाः). In the context of yet another text, Meenakshi says the श्वासनुप्रदान is "when the throat is open" (p. 198). This apparently fits the usage in the आपिशलिशिक्षा, but it then seems completely synonymous with what I proposed about विवृतकण्ठ above. I don't know if it's plausible to think that the author would have just repeated themselves but using slightly different words.

    Anyway, all three of these terms seem to be hovering around voicelessness...

    विसर्ग, जीह्वमूलीय, उपध्मानीय

    The आपिशलिशिक्षा seems to present a rather unclear picture of the relationship between these terms. Based on some of the discussion in Ghosh, जीह्वमूलीय is velar [x] and उपध्मानीय is bilabial [ɸ], and both are distinct from विसर्ग, which is glottal [ h]. This fits with the fact that विसर्ग, जीह्वमूलीय, and उपध्मानीय are listed separately in the third verse I quoted in #33 above. But then विसर्ग should be the seventh ऊष्माण in the second verse, and जीह्वमूलीय should be another addition to the list of the कण्ठ्य consonants in the first verse.

    यम

    Thanks for your research and explanation of this! I also have no idea what the fifth nāsya after the four yamas might be, if not the anusvāra. The table on p. 62 of Ghosh seems to think that a couple of other early phonetic texts do classify the 4 yamas together with the anusvāra.

    I was a bit surprised to see that voiceless nasals are in fact a thing, though of course this isn't what the yamas are since they have nasal place of articulation...

    ---

    On an unrelated note:

    Parallelly, I am curious, do we know of any language anywhere in the world where an s>ʔ has happened? Anybody?

    I tried to look for this and came up short. But I did notice on the Wikipedia article on fortition that the Southern Ryukuan language Yonaguni seems to have had a [j] -> [d] transition word-initially. That's a fairly surprising change to me...!
     
    Last edited:
    Top