Definitive guide to Tamil sounds ca, sa, Sa, sha?

Languages used in Carnatic Music & Literature
Post Reply
sridhar_ranga
Posts: 809
Joined: 03 Feb 2010, 11:36

Post by sridhar_ranga »

Is there a definitive guide to the use in Tamil of the sounds ca, sa, Sa, sha? Which specific words use which sound?

My reference words for these sounds:

Ca (also used as cha at times) - as in chandanam / chAlees(40)
sa - as in samam eDuppu!
Sa - as in Sankar, Sakti
sha - as in visham

To simplify the query / focus on the cha/sa/Sa issue alone, let me add the following up front:

1. I do not want to go into other Tamil pronunciation issues related to ka/ga, ta/da, Ta/Da, etc"¦just want to look at the ca/ sa / sha family of sounds.

2. To simplify it further let us look at only when the sound/ letter is at the beginning of the word.

I see a definite trend of moving towards ca and sha at the expense of sa all around me and hence the question.

I have definitely heard the announcer at All India Radio go "sennai vAnoli nilayam"

cmlover
Posts: 11498
Joined: 02 Feb 2010, 22:36

Post by cmlover »

The sa-fication of cha was a chennai/NorthArcot trait. Primarily the cha is properly pronounced/retained in the south of TN (kanyakumari/Tinnevely/..). The northerners will always say 'sonnAr' for 'chonnAr'. Some wag stated that when you are drunk you cannot properly say cha and you replace it with sa :) Occassionally it becomes 'ShonnAr' too; especially if you are chewing paan :) The outrageous statement is 'yonnAr' which I have heard among some tribal communities in TN. The ya-fication of cha is not uncommon. But the extreme (unpardonable) is the zha-fication of saying 'zhonnAr' which occurs among those who sttammer :)

All these sounds being palatal the variations are regional and adhere to native practices!

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

CML is right, one good example is the difference between chendoor (i.e tiru-chendoor) and its murugan deity called chendhil (usually pronounced sendhil in north-eastern Tamil Nadu). The words chendoor and chendil are from the same common source, but are pronounced differently by error.

Wherever a Tamil word has a cha/sa/sha/śa, it should be cha which is correct, since tamil does not have the sibilants sa, śa and sha.

For loanwords (i.e non tamil words used in Tamil) such as śiva/śakti/śruti etc, the original pronounciation is best (instead of tamilized pronounciation - suruthi etc).

arasi
Posts: 16774
Joined: 22 Jun 2006, 09:30

Post by arasi »

sridhar_rang,
Thank you for bringing this up--the puzzling, sometimes hazy matter of 'how to pronounce' our own native tongue. Those from out of tamizh nADu are excused!
The one other variation you did not bring up was Sa. I have a feeling sa is more current (media influence?). In our days, sa was not used as much.
While I do not know how rules of grammar dictate it, I interchange ca (cha) and Sa (the way it occurs in the flow of speech and in the place of the word in the line. Added complication: those who say c(ch)onnAr while they speak, sing SonnAr or sonnAr! What is pleasing to the ears also counts. I don't have the answers. I have a question, though.
How is it that I should sing this line? : c(ch)endUrAn (SendUrAn) unai
Sindaiyil nRrutti? I sing it as cendUrAn unai Sindaiyil! No logic.
I feel that words have something to do with it too. They have their own personality ! No logic again! I need Arun to tell me off :)

By the way, I have heard Sanjay sing summA summA and SummA Summa!

mohan
Posts: 2806
Joined: 03 Feb 2010, 16:52

Post by mohan »

I've always wondered why when writing Indian language words in Roman script a lot of people write "ca" instead of "cha" (eg caranam instead of charanam). I would think the latter is a better representation.

rshankar
Posts: 13754
Joined: 02 Feb 2010, 22:26

Post by rshankar »

Mohan,

I am one that uses 'c' for words like caraNam, and 'ch' for words like chAyA (shadow/image). So charaNam for me will be inaccurate...

cmlover
Posts: 11498
Joined: 02 Feb 2010, 22:36

Post by cmlover »

The problem is complicated by comparing with sanskrit where there are 'ca' the unaspirated and 'cha' the aspirated. Even here there is no agreement some systems consider 'ch' as unaspirated and 'chh' as aspirated...
Fortunately (or unfortunately) in 'pure' Tamil there is only one 'ca' (or is it 'cha') for ca, cha , ja and jha of the sanskrit palatals. In addition we have the sibilants 'sha' (some represent as 'S') 'sa' and 'Sha' all of which are equated to the 'ca' in 'pure' Tamil...
(I am discounting the 'grantha' akSharams that have crept into Tamil)

mohan
Posts: 2806
Joined: 03 Feb 2010, 16:52

Post by mohan »

Thanks ravi and cml for the responses but as a language ignoramus its all Greek to me :)

arasi
Posts: 16774
Joined: 22 Jun 2006, 09:30

Post by arasi »

Mohan,
If it is greek to you, it is latin for me and a tower of Babel for us all.

A confession: I used to wrie it as charaNam all these years, but have switched to caraNam. When we (mainly Ravi) go to the trouble (not really) of giving the meaning to songs, it is better that we and those who request them stick to a somewhat standardized way of writing and reading of the words.
Here is the way I go about it: ca or Sa for caNDai or Salippu, as in calanATa or SyAmal. sha strictly for anything which sounds like usha and shut in english.

gmohan
Posts: 125
Joined: 06 Apr 2007, 01:58

Post by gmohan »

Shankar when shortened always becomes changu!?

rshankar
Posts: 13754
Joined: 02 Feb 2010, 22:26

Post by rshankar »

gmohan wrote:Shankar when shortened always becomes changu!?
Please don't tell me!! And I do not want to relive the conversation/argument I had with a guy from Madras University on how I wanted my name written on my degree!

cmlover
Posts: 11498
Joined: 02 Feb 2010, 22:36

Post by cmlover »

arachi chankar cholppaDi cheykiRAr :)

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

Arasi wrote:How is it that I should sing this line? : c(ch)endUrAn (SendUrAn) unai Sindaiyil nRrutti? I sing it as cendUrAn unai Sindaiyil! No logic.
cendUrAn is right. Sindai (short form of "sintanai" or more properly "cintanai" is borrowed from a Sanskrit word cintanA) and hence also should be pronounced chintai (since it has a "ca" even in sanskrit).
Arasi wrote:By the way, I have heard Sanjay sing summA summA and SummA Summa!
As per tamil phonology, there can never be a 'sa' (or any of the other sibilants), in any word. It may be admissible in words that are not native to Tamil but borrowed into the language. Not for native tamil words.
Mohan wrote:I've always wondered why when writing Indian language words in Roman script a lot of people write "ca" instead of "cha" (eg caranam instead of charanam). I would think the latter is a better representation.
Mohan, the standard orthography for latin transliteration of Indic scripts is explained in http://en.wikipedia.org/wiki/IAST

It is used widely for Sanskrit and most other Indian languages, but to a lesser extent for tamil (since dravidian sounds like ழ, ள, ற, ன, ஒ, எ, and ஃ do not neatly fit into IAST). Some people use a modified form of IAST to include some additional characters to denote these sounds that IAST does not represent, but it is not standardized yet.

Here in IAST ca (nonstressed) is sought to be differentiated from cha (stressed/aspirated) because these are distinct in sanskrit (similar to ka vs kha, ja vs jha, ta vs tha, Ta vs Tha, pa vs pha etc). Tamil has only the ca (and ka, ta, Ta, pa respectively). Since Indology seeks to harmonize the latin representation for both dravidian and Indic language families, they seek to use the same orthography for the common sounds).
cmlover wrote:arachi chankar cholppaDi cheykiRAr
Actually the word 'arasi' is derived from arasan (which itself is derived from rAsan and ultimately from Skt rAjan). I would not pronounce arasi as arachi since pronouncing it so would be to err both on the side of convention as well as on the side of accuracy. So let it atleast adhere to the convention/tradition and remain "arasi".

chankar is again not accurate, it should be Śankar (the common practise in tamil to write Śa is to use the grantha letter, similar to the grantha letters ஷ, ஸ, ஜ which are used to write sha, sa and ja in tamil).

What we call sangu (conch) in tamil is "Śaṃkha" in sanskrit (normally pronounced as "Śankha") whereupon it became sangu in tamil. ṃ is anuswara, Ś is taalavya ushman (palatal fricative)

ragam-talam
Posts: 1896
Joined: 28 Sep 2006, 02:15

Post by ragam-talam »

Malayalam sounds may provide a clue to the 'original' Tamil sounds, considering the two languages have shared roots.
In Malayalam it is: chummA, chArAyam, chandran, etc.

The 'S' sound - is this restricted mostly to brahmins? You do hear this a lot from Palghat Iyers, who 'sit on the fence, as it were.

But then, if English has several versions today (American, Canadian, Australian, etc. in addition to Brit English), perhaps we should allow similar latitude in Tamil pronunciation too?

It's Arasi's rAsi indeed! ;)

arasi
Posts: 16774
Joined: 22 Jun 2006, 09:30

Post by arasi »

arasi, rasi, is more like it :) By the way, I have never been inside rAsi, the saree shop!

CML,
But for the timely rescue from srkris, your quip might have inspired someone to 'en peyarai aracca tuvaiyalAkkum paNi'! In the old sangeetham forum, someone did call me Arisi once. annam is far better than a side dish!
Last edited by arasi on 28 May 2009, 07:40, edited 1 time in total.

keerthi
Posts: 1309
Joined: 12 Oct 2008, 14:10

Post by keerthi »

srkris wrote:
What we call sangu (conch) in tamil is "Śaṃkha" in sanskrit (normally pronounced as "Śankha") whereupon it became sangu in tamil. ṃ is anuswara, Ś is taalavya ushman (palatal fricative)
Shankha is one of those words of doubtful parentage, in sanskrit, as it's etymology isn't well established.. It is as likely as not, that the word came from a dravida source, and was absorbed by sanskrit..

More classic words used for conch are kambu and dara...

S.Govindaswamy
Posts: 47
Joined: 23 Oct 2006, 06:48

Post by S.Govindaswamy »

Mr.sridhar_rang started this discussion asking for guidelines about "the use in Tamil of the sounds ca, sa, Sa, sha"

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

Govindaswamy,

I can normally pick a phonetic error from almost every line of every song from the rendition of most of the top level carnatic singers of today. This includes Tamil songs as well (although not to the extent as in other languages)

Goes to show phonetics is one of the most neglected areas by carnatic musicians.

erode14
Posts: 726
Joined: 21 Jan 2007, 21:43

Post by erode14 »

sa is there in thamizh.

it is sendhoor only. chendhoor comes when you connect thiru and sendhoor, thiruchchendhoor.

enna solli azhaiththAl & varachcholladi...

samaiyal, semmozhi, sari (cha-ri-ga-ma is horrible, isitn't?)

sometimes, it is said chO.... chweet... like that it happens.
Last edited by erode14 on 29 May 2009, 16:40, edited 1 time in total.

ragam-talam
Posts: 1896
Joined: 28 Sep 2006, 02:15

Post by ragam-talam »

Jesudas talks about sa-Sa-cha etc. (starts at 5:37)
http://www.youtube.com/watch?v=0iznqMZy5qY

cmlover
Posts: 11498
Joined: 02 Feb 2010, 22:36

Post by cmlover »

Thanks R-T for that reference.

ragam-talam
Posts: 1896
Joined: 28 Sep 2006, 02:15

Post by ragam-talam »

cml, the complete interview is quite interesting. He also talks about his days with Chembai et al.

cmlover
Posts: 11498
Joined: 02 Feb 2010, 22:36

Post by cmlover »

Yes! I noticed. Quite revealing.. What a man!

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

erode14 wrote:sa is there in thamizh.

it is sendhoor only. chendhoor comes when you connect thiru and sendhoor, thiruchchendhoor.

enna solli azhaiththAl & varachcholladi...
You are correct, it is prevalent in current-day spoken tamil, it is present in colloquial use. We were speaking of sa not being present in "correct" tamil (old-tamil).
erode14 wrote:samaiyal, semmozhi, sari (cha-ri-ga-ma is horrible, isitn't?)
semmozhi = chemmai + mozhi

Sa ri ga ma is not tamizh. So it need not be pronounced as cha ri ga ma. By the way, ga is also not there in "correct" tamizh. We can understand what is correct tamizh by looking at how it is written. Writing does not change as fast as the spoken language, so what is written (at least for tamil) is more authentic than spoken language.

In some cases, ga is intentionally written and pronounced as ka in tamil (one example is the word kaNitam = mathematics). This word is actually gaNitam in sanskrit (and also in vedic), derived from the root gaN (which means to count/enumerate). Other derivative words from this root are gaNa (group), gaNapati/gaNesa (lord/god of the group) etc.

Some languages (Vedic for example), cannot be written accurately, because it is too complex for writing (where meaning of the words can differ by the pitch/accent at which they are pronounced).

sridhar_ranga
Posts: 809
Joined: 03 Feb 2010, 11:36

Post by sridhar_ranga »

Spoken Tamil existed for several centuries probably before written Tamil came about - and whoever adapted the 'Aryan' script first into Tamil (Brahmi? early Jain/ Buddhist monks?) did a shoddy job of it :)

IMHO the script cannot be a basis for conclusions about the non-existence of sounds like sa in spoken Tamil.

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

Spoken tamil is not a constant, it has added sounds over the millenia (since most languages borrow sounds and words from other languages all the time). In fact the tamil script (originally tamil was written in brahmi) provides us a gazillion clues about old tamil's phonology.

There are express instructions in the grammar of how to reduce complex sounds in borrowed words into simpler sounds to make them suit tamil phonology. Thus agastyam should be pronounced as akattiyam.

Those who wrote tamil down using initially brahmi, and later other scripts, therefore did us all a lot of service. I dont think it was a shoddy job at all. The grammars of tamil (nannool, tolkappiyam) totally support the thesis that tamil has added a lot of sounds, and has made it possible to differenciate old tamil from later tamil just on the basis of phonology and grammar. If there were no written tamil, the sounds that are unique to tamil would not have been preserved.

One cannot therefore form prima-facie conclusions about old-tamil merely based on the way people talk it now. As I said, written tamil (rather than spoken tamil) has remained more unchanging and representative of old tamil.

sridhar_ranga
Posts: 809
Joined: 03 Feb 2010, 11:36

Post by sridhar_ranga »

srkris,

Absolutely agree with the point that written tamil has remained unchanged and more representative of old tamil while spoken tamil has undergone a lot of changes over the centuries.

However it is difficult for me to believe that sounds like ga, Da, da, ba didn't exist in old tamil and it was ka, ca, Ta, ta, pa all the way in tamil of those days.

If we say that both ka and ga were represented by a single letter, (and similarly Ta and Da, pa and ba etc), don't you think it is likely that more than one sound was represented by the letter ca as well?

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

ka and ga (and so on) were not represented by a single letter in old-tamil, because there was no ga in the first place. Using the same letter for two different sounds is:

1. An innovation in later spoken tamil (was not present in old-tamil)
2. to be used mostly (or only) for borrowed words, not native tamil words.

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

srkris,

Can you quote some references w.r.t the information that softer form of these consonants weren't part of tamizh (even in the middle of a word - are you implying this above?) in old tamizh?

If it was hard consonants throughout, what role did mei-ezhuttu form of a consonant preceding the consonant itself play in old tamizh (particularly when conjoining words). In fact

Thanks
Arun
Last edited by arunk on 02 Jun 2009, 23:18, edited 1 time in total.

S.Govindaswamy
Posts: 47
Joined: 23 Oct 2006, 06:48

Post by S.Govindaswamy »

I have not read the rules for pronuncition which might be available in any of the old grammar books. But my understanding is as below. This could be wrong.
Most of the confusion arises in respect of the 6 hard consonants (வலà¯ÂÂ

rshankar
Posts: 13754
Joined: 02 Feb 2010, 22:26

Post by rshankar »

[quote="S.Govindaswamy"]
2 When two of these occur they take the second sound kha, cha, Tha. Tha and pha. These sounds are produced from the middle of the tongue. However I think in a word like அகà¯ÂÂ

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Fascinating stuff. What Govindaswamy wrote sounds right to me since all that checks out when I try them, with the exception that the second sound of Ka seems to be not as hard as the Kh sounds of sanskrit. Iit does have the 'plosive' quality to it, though.

We say all these words without thinking about it much. I have a feeling we are paying too much attention to the script while looking at today's Tamil, though that is an excellent way to specualte about how old tamil was spoken.

Tamil is not as phonetic as other indian languages, so be it. It is always known that it has a lot of context sensitive sounds than the alphabet allows. Was old tamil a lot less context sensitive than current tamil? If so, how do actually know that?

The 'sa/Sa/Sha' which is the main topic of the thread is a sensitive topic in the carnatic circles for some reason. We have talked about it a lot before so I really do not want to stir that pot again but since we are talking acadamic precision here, let me venture here reluctantly: In 'vaNDADum sOlai', the 's' is unmistakably an 's' and not 'S' or 'Sh' etc. It sounds very close to the English 'so'. Any thing else sounds so so. MS, in singing that great song, pronounces it pretty well ( according to me ) and a few others say it with a sanskritish 'S' color and a selected few take it all the way to the Sanskrit 'S'. Here I am assuming 'sOlai' is a native tamil word and not borrowed from somewhere else.

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

vasanthakokilam wrote:In 'vaNDADum sOlai', the 's' is unmistakably an 's' and not 'S' or 'Sh' etc. It sounds very close to the English 'so'. Any thing else sounds so so.
:D

It should be pronounced cOlai (சோலை), not as sOlai.

Let me give other examples - caRukku (சறà¯ÂÂ

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

srkris,

Please provide some references on why ga/ba/da etc. weren't at all part of old tamizh. Your posts indicate that you are quite sure about this and so I would like to know the references that support this.

Your claim that if they were indeed part of the old tamizh, separate letters would have been in the script, is certainly a possibility. However, it also seems like a theory which to me does not sound super-strong (although it is intriguing enough that I cannot dismiss it)

In Tamizh script, for non-imported words, it turns out there is really no ambiguity on when sa/ga/da/ba/Da can come vs. when ca/ka/ta/pa/Da can come (although sa/ca is definitely a can of worms in terms of applicability). So the original script designers could have also worked at efficiency to eliminate redundancy. In other words, they could have found a way to represent both sounds with same letter since the occurences weren't ambiguous. Of course, the continual influx of imported words made things very very ambiguous.

Also can you please post an audio as to how to properly pronounce nERRu and inRu as you have deemed the common one "misunderstandings/misconceptions and lay misinterpretations". (although I will agree that nERRu cannot be fully equivalent to nETRu as in that case the spelling would have simply reflected the latter. Perhaps you are referring to this - in which case pl. disregard this)

Arun
Last edited by arunk on 04 Jun 2009, 21:13, edited 1 time in total.

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

I think govindaswamy is referring to the somewhat unavoidable mixing of the "ha" sound for the "ka" part of akka due to the preceding stop (i.e. mei k). Try comparing akka with aka (of course cannot be part of tamizh) or kal. I agree that this is not the same as kha in other languages, and I dont think it is considered a separate phoneme from ka and ga in tamizh.

Arun
Last edited by arunk on 04 Jun 2009, 21:17, edited 1 time in total.

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Arun, now I am quite confused what the transliteration 'ca and sa' mean. I will re-read this thread and straighten myself out.
How do we know that writing is much more conservative than speech? Its because if there were additional sounds in spoken tamil that existed before a particular script came into use - when the script was invented, it would have included separate symbols for those additional sounds.
srkris, That is based on a rather bold and possibly unsubstantiated assumption that the inventors of script strived for context-free highly phonetic script. I do not think that should necessarily be the case. It is not out of the realm of speculation that they took a fairly abstract approach with the objective of limiting the size of the alphabet and put in clear rules for how the letters needed to be sounded in various contexts.

I am not saying that is what actually happened, but just pointing out that you are assuming that the context freedom was the objective of the script inventors and it will be good to state that assumption explicitly. If there is any evidence that the script was at one time very phonetic and context-free, then I grant that your thesis is much stronger.

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

vk,

Regarding ca/sa - not sure where your confusion is and whether it got worse after my post :). Anyway, what I mean is that for all other cases, for native tamizh words
1. the "harder form" (ka/pa/Ta/ta) is the one that can start words.
2. The harder form can come in the middle only when preceded by an "appropriate" mei. This "appropriate" mei of course includes the mei of the consonant (i.e. Ekkam, veppam, nattai, kaTTam). But it can be others too (e.g. nuTpam , vETkai,)
3. The softer form (ga/ba/Da/da) occurs in the middle of the word in all other cases. Based on #1, the soft form never occurs at the start of a native tamizh word.

Now if we consider the case of ca (as in paccai) and sa (as in isai), then it must be the "harder form" ca that should be used at start of all native tamizh words. But I believe it is rarely so in today's tamizh. In fact ca I think is mainly used in case #2 above i.e. as in paccai etc. - although some people do use it at the beginning of the words (and some use it perhaps erratically (?)). That is why I called its usage a mess (besides the interchangeability of sa and Sa by some sections of the populace)

Arun
Last edited by arunk on 05 Jun 2009, 00:20, edited 1 time in total.

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Arun, Got it! ( your post did not cause the confusion, srkris's 'stickler to old tamil' post did but since you also seemed to agree with it, I thought I wil direct my question to you ;) )

My meta-confusion was between these two confusions you have pointed out. ( sa and ch, and sa and Sa )

>"But I believe it is rarely so in today's tamizh. In fact ca I think is mainly used in case #2 above i.e. as in paccai etc."
>"besides the interchangeability of sa and Sa by some sections of the populace)"

My meta-confusion is resolved now ;)

Having said that, how do you two distinguish between 'today's thamizh' vs 'the other kind' , how far back you are going to make the distinction?
In other words, where are these rules mentioned? In tamil grammar books of the past 60-100 years or going back to tolkAppiyam etc.
Would a tamil pulavar of great repute of the past 60-100 years have uttered these things correctly according to these rules?

Because, for example, I have not heard anyone say the word 'sol' as 'col'. In that case, 'today's tamil' atleast means 'tamil of the last 45+ years'!!

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

I dont know about these rules in particular but tolkAppiyam does lay out the rules (and so I would expect these to be in there in some form). I think from a strict, cold logical perspective, "ca" would make sense at start of words - but then we are talking about languages :) - imo languages break rules as much as they conform. So I would say "I am not sure if it was ca for all word beginnings then" - because if ca became sa - then why did the change apply only to that combination? Because of imports? But Sanskrit has as enough words starting with ca as well as sa.


But I havent studied old tamil etc. - and so I could be off. I just see a nice logical structure for the others (based on my exposure to tamizh which certainly is not exhaustive - only up to 10th std!!!), and I see the ca/sa combo alone deviating from that.


Arun
Last edited by arunk on 05 Jun 2009, 01:19, edited 1 time in total.

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

For sol vs col. See that sol when following tAi becomes col as in tAiccol. So it becomes a harder sound when combined.

Now consider pAl and tAi (although pAl here may not be tamizh native word). They combine exactly the same way => tAippAl. But the sound is hard initially (pAl) and remains so on combination. But sol/col as used today goes from soft to hard. It seems to continually deviate from the "rule" followed by others. So couldnt/shouldnt it be tAi + col => tAiccol just like tAi + pAl => tAippAl?

Arun
Last edited by arunk on 05 Jun 2009, 01:17, edited 1 time in total.

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Arun, I see what you mean

The one thing that I am not able to get over is, 'sa' sound is not a difficult sound, it is an easy and natural one. Why would not pre-tamil script people have used that sound for a word? I agree this is all speculative thinking.

Here is another way to get some more intelligence about this. Let us look at some old tamil poems ( thirukuraL, nAlaDiyAr, the nAyanMar's various works, etc ) where rhyming requirements are strictly obeyed. Let us see if the 'Cha' rule for beginning of the word is obeyed. There are also poem types that are strictly based on rhyming rules ( like veNBA ?? etc ). Unfortunately, I have forgotten most of such stuff I learnt in high school.

May be CML and Arasi can also pitch in here for usages from those old poems that support or violate this rule.

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

I cant think of any rules that would provide clue - but that is because I am clueless in that regard :)

BTW, on some googling I can see that many researchers have opined that the old tamizh contained only harder forms (called surds, the soft ones are called sonants). But then - some think not. Anyway check this book http://books.google.com/books?id=XsC444 ... frontcover on google books, and look at the Table Contents, and look for "Notes on Old Tamil and Jaffna Tamil". This author I believe thinks that sonants (soft form) were there but in discussing so does talk about research(ers) that theorized that sonants were there. Interesting read. There is also an article on the ayuda ezuttu.

Arun

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Arun, I will check that out. Thanks

Beyond that, my concern is really not about old tamil. What is the 'correct' way to speak formal tamil now and where is that documented?

Srkris seems to take a strict line that written tamil has not changed from Old tamil and so a tamil speaker speaking formally ( say a lecture in a tamil university in an academic setting ) should use those old tamil rules for native tamil words. So in such speech, someone should actually say 'inku Cindanai Cei'. That will sound so awkward, but I am really curious to know if that is actually how they speak.
srkris, please let me know if I am interpreting what you are saying wrongly.

I am heading towards this:

If no one follows these old rules and the 'current' rules have changed, then we have to accep the language has changed and we have to stop branding those who speak the currently prevalent formal tamil as 'Wrong way of speaking'!

VK RAMAN
Posts: 5009
Joined: 03 Feb 2010, 00:29

Post by VK RAMAN »

How about those pravasi tamilians settled and brought up in other states of India and other countries including Sri Lanka, Malaysia, Singapore? What will we call their tamil?

arunk
Posts: 3424
Joined: 07 Feb 2010, 21:41

Post by arunk »

I think the old way has gone lo...ng ago. The debate is reallly whether Old Tamil (as in tolkAppiyam) had the softer sounds. There is no debate that it changed over period. There seems some debate as to when it changed but evidence of transliteration of tamizh verses in other languages (see below) with a more flexible script indicates that the change has happened long ago - and maybe from the beginning. At the least we are talking many centuries.

Expecting now to adhere to Old Tamil or insisting that is only correct seems as meaningful as asking everyone now to speak and write in Old English, and insisting that is really the proper english.

Today, as far as I know, in formal tamizh, it is ka/Ta/ta/pa at start of words, and when preceded by the approp mei, and ga/Da/da/ba etc. in the middle. As far as I can tell, it is sa at the beginning (which morphs to ca when combined as explained above), as well as in the middle. Using "Sa" for sa while acceptable (in fact preferred - somewhat oddly) in cm circles, would probably not be so in formal literary tamil circles where I would think that it would be perceived as an accent.

BTW, from what I can gather - tolkappiyam does *not* address ka vs. ga. This is another reason why people think it didnt exist. But there seem greek inscriptions of tamizh names which indicate morphing (da in the middle, as well as nga etc.) and those inscriptions are middle (?) of first millenium. Later at least in 12-13 century, in a telugu work you have tamizh phrases which clearly is more consistent with what we pronounce today (but it has Ican and NOT Isan or even ISan - remember this is in telugu script)


Arun
Last edited by arunk on 05 Jun 2009, 05:47, edited 1 time in total.

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

Thanks Arun. That makes me feel a bit better about my own Tamil. Let us wait for what srkris has to say!

rshankar
Posts: 13754
Joined: 02 Feb 2010, 22:26

Post by rshankar »

vasanthakokilam wrote:Thanks Arun. That makes me feel a bit better about my own Tamil. Let us wait for what srkris has to say!
And a lot better about mine - glad I can still call it 'tamizh' - While we argue over how to pronounce sounds for which the tamizh alphabet has no letters, what kills me are those self-styled 'perfect' tamizh speakers who mess around with sounds for which unique letters do exist in the tamizh alphabet - nothing nauseates me more than the 'L' sound being substituted for the 'zh' sound, and the 'l' and 'L' sounds used interchangeably!

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

inku Cindanai Cei
cintana is sanskrit.

Even if the script deliberately did not accommodate separate characters for the voiced consonants ga ja Da da ba, grammatical rules should have existed about how to employ voicing in the language. Does the tolkappiyam say that in certain situations, voiced (soft) consonants should be used instead of unvoiced ones?

Does any grammatical work distinguish between ca and sa and say that in certain words ca should be pronounced as sa?

These are all moot points if one considers the grammarians did not have voicing as a possibility at all. So they did not have to make rules for things that did not exist during their time. The same thing applies in the case of employing sibilants (sa, sha, śa).

A third reason is that old-tamil converted the soft/voiced consonants of sanskrit into hard/unvoiced consonants in borrowed words. A pertinent example is gaNita (arithmetic/mathematics) which was converted to kaNitam in tamil. The tolkappiyam has express instructions on this subject (how to simplify the pronunciation of borrowed words to fit tamil phonology).

So unless all these three reasons are exceptional, we dont really have a case for believing old-tamil had the voiced consonants.
in a telugu work you have tamizh phrases which clearly is more consistent with what we pronounce today (but it has Ican and NOT Isan or even ISan - remember this is in telugu script)
Not only that, Ican as an epithet of IÅ›wara/Ã…Å¡iva is a borrowed word (cf. IÅ›ÄÂÂ

vasanthakokilam
Posts: 10956
Joined: 03 Feb 2010, 00:01

Post by vasanthakokilam »

srkris, my point there was not about cintana but the other two words. Are you saying the right way to say it is 'inku <whatever> Cei'. Come on! ;) Who says it like that as far as you know?

your ganita example is fine but does not help here since ga is in the beginning of the word. That rule makes sense. We have agreed that 'ga' can not be in the beginning. The issue is about the prevalent 's' in the beginning and 'ga' in the middle.

Would you go to the extent of calling the use of 's' in the beginning of the word and 'g' in the middle of a word as wrong formal tamil with respect to tamil as it exists today/in the recent past/in the past few centures ( and not how it was 2000 years back ) .

srkris
Site Admin
Posts: 3497
Joined: 02 Feb 2010, 03:34

Post by srkris »

How most people speak now is not the point at all. I am not even giving a prescription that people should speak in a different way to agree with Old-Tamil phonology.

The point is about what the grammar describes (as the tamil of the grammar's time). The point is about not confusing today's speech with Old-Tamil. So however intuitively you put forth the argument that no one today pronounces ingu as inku, or that inku is unnatural, that's not the point.

It is a rule of thumb that in any language, phonetics is the first to get hit by change, while writing remains the same relatively over time. Hence the literary language always stays more conservative compared to speech. People are always subconsciously simplifying their speech, thats human. The differences if any, are only perceptible over time, many times people don't even realize the change.

For example, many sanskrit speakers dont pronounce the अः (ḥ), अं (ṃ), श (ś), ऋ (ṛ) correctly.

kRSNa: is pronounced as krishNaha (horror horror)

Post Reply