Tamil Discussion archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

To Anu, reply to his questions.

Anu: I'll try, quickly, to give a few answers.  I'm posting them on the
Tamil font list in the hopes that they may give some needed perspective.
>        Q.1.  How many distinct sounds are there in the world and
>        how many of them are produced by human beings?
I doubt anyone has the remotest idea.  There are languages in Southern
Africa that have 20 distinct "clicking" sounds.  Things you have never
>        (This is a hypothetical question and may not have an exact
>        answer.  To intiate the thought process from the beginning
>        I have asked this question.  You don't have to answer.)
>        2. How many distinct sounds (phonems) would make a language
>        complete and why?
The answer to this is easy: the number of phonemes a language needs to be
complete is the number it has.
>        3.  How many complete languages and incomplete languages
>        are there in the world?
There are no incomplete languages.
>        4.  What are the sounds tamil lacks to define it an
>        incomoplete language and upon addition would become a complete
>        language?
No language needs extra sounds to be complete.
>        5.  If additional phonems would help tamil, how many additional
>        phonems would make tamil a complete language and why?
>        6.  How did any body choose the magic number four (grantha phonems)?
>        How about five, six, seven, thousand or million?
>        Are these bad numbers fot tamil?
Languages grow historically.  As they grow and change, their phonetics
change.  Very few Tamils, for example, can pronounce the difference between
the 2 r's (though Malayalis and Sri Lankan Tamils can).  This is a historic
change -- 1000 years ago, the phonemes were probably distinct in almost all
dialects of Tamil.  As you may know, the difference goes back to
proto-Dravidian, and (as I remember) is also in old Kannada.
>        7.  Does tamil lack only these four phonems or many thousands?
To discover the phonemes Tamil has, one must look at its speakers without
prejudice.  If most speakers use the sounds "j", "s" and "h," these are
present in the language.  The fact that it may have lacked these phonemes
2000 years ago doesn't mean it lacks them today -- the fact is, Tamils of
all backgrounds and castes habitually use the phoneme "j."  Same for "h"
and, for that matter, "b" (e.g. bayam, from Skt. bhayam).  Recently, Prof.
Emeneau and my wife Kausalya have written a paper showing that initial
voiced stops (g, b) occur in proto-Dravidian in onomatopoetic words.  They
are in modern Tamil, and are no doubt extremely ancient (e.g. gala gala
>        8.  Does the addition of these four grantha phonems would make
>        transliteration from English to tamil or from any other
>        language completely feasible or does tamil need many more
>        phonems?
Writing a language is a matter of convention.  It has become conventional
to use the 4 grantha letters, and it's hard to see how the language would
manage to represent what people actually say without them.
>        9.  As such I do not see any equivalent tamil scripts to
>        the following English scripts, B, C, D, F, G, H, J, Q, S,
>        W, X, and Z.  Also the remaining scripts although may look
>        to resemble some tamil phonems, the way they are pronounced
>        in Rome, London and New york is quite different from tamil
>        phonems.  Do the SW vendors have plans of introducing
>        these additional phonems at a later time point as they
>        do not have time at present?  Or are there any other
>        reasons?
Again, writing is a matter of convention.  We write paasTaan for Boston.
You might argue that having a "b" would not be a bad idea, but I have not
seen anyone suggest this.  Convention dictates against it.  BUT everyone
says baasTan, not paasTan.
>        10.  Do you or other programmers know that time taken by search
>        engines in parsing and pattern identification will increase as
>        number of scripts increases?
I don't think the increase of time is significant.
>        11.  Can you please suggest the situation wherein you prefer
>        transliteration to translation and also substantiate your
>        claim with past evidences?
The fact is, you cannot translate place names.  No one knows what Boston
means, so how could you translate it?  Same for Moscow.  You have to
transliterate them -- no choice.  With regard to other kinds of words, this
again is a matter of convention.  We say uruLaik kizangku, not paTeeToo --
this is a successful translation.  But no one has come up with a
translation of "coffee" that will be accepted.  This is not up to us --
it's a process that languages go through and that is not predictable.  Some
words are accepted, some are not.  The Nazis tried Fernseher for
Television, but virtually all Germans use "Television."
>        12.  If you recommend transliteration of words from one another
>language which language you would recommend and why?
As I said, this isn't up to me or anyone.  It's a process that happens
naturally -- languages evolve in a natural way.  They are always in contact
with other languages, and always influenced by them.  There is not one
language on the face of the earth of which this is not true.
>        13.  If something is to be transliterated from A to B,
>        according to set theory, either A has to be a super set and
>        B a subset.  Is it english a superset and tamil a subset?
Of course not.  They're just different languages.  The flow of borrowing is
usually dictated by political and economic considerations.  Chomsky (I
believe) once said "A language is a dialect with an army and a navy."
>        13.  If both of languages are subsets of some hypothetical superset
>        then transliteration should take place in both directions.
>        To successfully do this, new scripts have to created in both
>        languages to fill in the lacking phonems.  Did such script
>creation ever happen in tamil or English in the past?
>        Or in any other languages in the past?  If not, why?
Most languages have non-native letters for transliteration.  Examples are
z, ph, and x in English, and f in Russian.  The fact is, word borrowing and
sound borrowing are universal.
>        14.  Why  did English or any other language never try to create
>        tamil's unique phonems like "za", "La", "Na", "Ra" in their
>        language?
Languages don't try to create phonemes.  Sounds change naturally.  If
English, before it became standardized through print and the media, had
been in close contact with Tamil, it probably would have created those
sounds.  Singhalese, an Indo-Aryan language, has borrowed several Tamil
sounds, including za.  As you may know, Malayalis pronounce "Tamil" sounds
better than most Tamils, who cannot say "za" properly.  Almost no Tamil in
India can properly articulate "Ra."  The sound has been lost in their
language, though they still write it.  A similar example is the "gh" in
English "night."  It was once pronounced, but the entire phoneme has been
lost in modern English.  All Tamil dialects I know of have lost the
difference between the two n's (as in naay, ini), but Malayalam maintains
it (though, curiously, Malayalam does not write them differently).
>        15.  Can the Govt. make evil drugs legally available to everybody
>        as people would always find ways to get them even if the Govt.
>        does not provide them?
>        16.  If preservation of older objects is an intention, then
>        does every body know that there is a variety of BitMapFile
>        format from Device Independent to scalable (BMP, DIB, RLE, GIF,
>TIF Jpeg and etc),that one can video graph, photograph, scan,
>        draw and store?
Can't answer this.
>        For the time being I stop with these questions.
>        I believe they are all rational and if you find some irrational
>        or emotional please leave those questions.
>        I appreciate your sincere reply.
>        Thank you and with great regards,
>        Anu
>	**********************************************
>	*  Ƒ  æƑ!       *
>	*    ئƑ!  *
>	**********************************************
>		Dr. Anukanth Anumanthan

Home | Main Index | Thread Index