Telugu-Tamil comparison of some 'core' vocabulary.
C.R.Selvakumar - Electrical Engineering
[ For the SCT readers: I collected telugu equivalents for a selected
set of 100 english words to compare tamil-telugu to carry out
an amateurish glottochrological 'study' thereby one can guess the
approximate date of separation of these two langauges. If any of you
follow up this posting, please take care that SCIT and SCT
folks will be able to follow ]
I'm presenting the telugu equivalents I received for the FIRST list
I posted sometime ago in SCIT. I sincerely thank Prasanna Garu,
and Ramarao Garu for the telugu equivalents.
I've placed a star (*) next to the serial number of the english word when
I think the equivalents in Telugu and Tamil have very different root
forms. ( I've also done this when both telugu and tamil has
nearly same words but in my opinion the word is of non-dravidian origin ,
say skt ). The tamil forms might not give the exact meaning as in telugu
but might denote something related to it which is assumed to be due to
the meaning shifts which occur over a period of time and this is common
even within the same language.
I have provided brief info about tamil words.
Please bear in mind that my attempt to apply glottochronology for
guessing the approximate date of separation of tamil and telugu
is purely amateurish and should not be taken more seriosuly than
it deserves. For indo-eurpean languages it was claimed that
86 % of words were retained per millennium.
I'm quite convinced that the particular set of words given by
Morris Swadesh might not be suitable for all languages nor the
fraction 0.86 retained per millennium might be applicable for
all languages ( this may depend on the nature, frequency of
interactions with mixing populations) and the fraction may not
remain constant.. in any case the predictions for the date of
separation between Spanish and Portuguese ( A.D 1586),
Italian and French ( same date) and Romanian and Italian ( A.D. 1130)
Englsih and Dutch ( A.D. 860) and English and German ( A.D. 590)
are considered to be very close..
So without adjusting anything, let us try to do a similar comparison
for Tamil-Telugu.
I find that about *15 words* in these 50 words that are unrelated.
^^^^^^^^^^
Thus in all 26 ( = 11 + 15) words are of different origin in these
100 words. This means 74 % of words are retained.
This implies that Tamil and Telugu might have got separated
around 2000 years ago, which sounds reasonable.
But as I said this is based on something which is not fully
satisfactory and the method is not established..
But the idea is to see the commonness and I'm amazed at the reasonable
date that get. I think if a professional linguist follows this up
with more rigrous application of this method he/she might get
may be 60% or even 80% of retained words but in this ball park I would
guess.
Even if one word same as in tamil is found in telugu denoting
a related item thats enough and vice versa ( namely if you know some
telugu word which exists similar to tamil word and means something
closely related to the tamil word, then it would be counted as
retained word.. and I'm only basing on the supplied telugu words)
I'll provide a similar input for the first 50 words soon.
English Telugu Tamil
[1] I nenu nAn ( often said as nAnu or nA )
[2] you nevu nI ( often pronounced as nIyu )
[3]* we meeru nAm ( inclusive sense. often said nAmu)
memu/manamu
[4] this ithi ithu
[5] that athi athu
[6] who evaru yAr, evar
[7]* what emeti enna, ethu ( = which)
enti/emiti ( 'en' and 'ennai' also means 'what' but
these are more archaic forms)
[8] not kadu illai ( like telugu lethu; tamil kUdAthu
means 'not'; kUdaa also mean
no-not)
[9] all anni ellAm ( another word is anaiththum)
[10] many chala pala ( chAla means huge, broad in tamil)
enno
[11] one okati onRu ( okka will be used for 'to become
one, to unite)
[12] two rendu iraNdu ( often said as 'reNdu')
[13]* big pedda periya ( also meththa)
[14] long podugu nILa ( In tamil 'pudai' means fatten
podavuga 'follow-up', 'ancillary.
pudavai is a long cloth worn
saree. 'pudai coozha' means
with long entourage accompanied
re kings and other dignitaries)
[15] small chinna siRiya ( also chinna)
[16]* woman aada manishi(ssthri) pen, peTTai
[17] man maga manishi(purushudu) AN, ( magan = son)
[18]* person manavudu AL
vyakti
[19]* fish chepa mIn
[20]* bird pitta paRavai
[21] dog kukka nAy ( also literary word kukkan, kukkal)
[22]* louse purugu pEn ( a small black insect usually on hair)
[23] tree chettu maram ( tamil 'chedi' means plant )
[24] seed ginja vithai ( kAzh is anothe word)
vittanamu
[25]* leaf aaku ilai ( thazhai etc. there are more than 20 words !)
[26] root veru vEr ( often pronounced as vEru)
[27] bark beradu paTTai
peDu
[28] skin tholu thOl ( pronounced as thOlu)
charmamu
[29]* flesh mamsamu Un, iRaichchi ( mAmisam is also used)
[30] blood raktamu raththam ( comes from tamil araththam)
But there are many other words
like kuruthi etc.
[31] bone emuka elumbu, embu
[32]* grease jiddu neY
[33]* egg guddu muTTai, aNdam
gruddu
[34] horn kommu kombu
[35] tail thoka vAl ( thOgai is used only for peacock)
[36] feather eeka iRagu ( Ikku is also used- but lexical)
[37] hair juttu mudi, mayir ( cUttu is for hair on head)
ventruka
[38] head thala thalai ( often pronounced as thala)
[39] ear chevi kAthu, chevi ( chevi is well known)
[40] eye kannu kaN ( often pronounced as kaNNu)
[41] nose mukku mUkku
[42]* mouth nooru vAy ( noRu-kku is an imperative to eat esp.
snaks like murukku)
[43] tooth pannu (pal ( often pronounced as pallu)
[44] tongue naluka (nAkku , nA ( nAl is also tounge)
[45] foot ari kalu (padam); ( kalu is leg. tamil kAl = leg
padamu pAdam, kAl adi are used
[46] knee mo kalu muzham kAl
[47] hand cheyi kai (chey is usually what is done by hand)
[48]* belly boddu vayiRu, thoppai, thondhi
potta
[49] neck meda kazhuththu, kaNdam ( midaRu means throat)
[50]* breast channu nenju, mArbu
I find that some of the words might not be representative of
'core' vocabulary.. but it is another matter.
Thanks again for the telugu friends.
Regards
-Selva
Hari Krishna Tadepalli
>
>
> [ For the SCT readers: I collected telugu equivalents for a selected
> set of 100 english words to compare tamil-telugu to carry out
> an amateurish glottochrological 'study' thereby one can guess the
> approximate date of separation of these two langauges. If any of you
> follow up this posting, please take care that SCIT and SCT
> folks will be able to follow ]
>
>
> I'm presenting the telugu equivalents I received for the FIRST list
> I posted sometime ago in SCIT. I sincerely thank Prasanna Garu,
> and Ramarao Garu for the telugu equivalents.
>
>
> 'core' vocabulary.. but it is another matter.
>
> Thanks again for the telugu friends.
>
> Regards
> -Selva
>
>
I have a lot to comment on the research methodology being
followed here. But let me say something constructive: I believe
the best source of 'core' vocabularies would be primary readers
, viz. those meant for (L|U)KGs, Ist and IInd standard books,
and an excellent traditional primary reader in Telugu known
as 'Pedda BAla Siksha'. IIWY (If I were you), I would procure
as many such primary readers in the both the languages as
possible.
- T. Hari Krishna
--
C.R.Selvakumar - Electrical Engineering
Hari Krishna Tadepalli <tade...@ori.cis.udel.edu> wrote:
>In article <CE39E...@watserv1.uwaterloo.ca> selv...@watserv1.uwaterloo.ca (C.R.Selvakumar - Electrical Engineering) writes:
>>
>>
>> [ For the SCT readers: I collected telugu equivalents for a selected
>> set of 100 english words to compare tamil-telugu to carry out
>> an amateurish glottochrological 'study' thereby one can guess the
>> approximate date of separation of these two langauges. If any of you
>> follow up this posting, please take care that SCIT and SCT
>> folks will be able to follow ]
>>
>>
>> I'm presenting the telugu equivalents I received for the FIRST list
>> I posted sometime ago in SCIT. I sincerely thank Prasanna Garu,
>> and Ramarao Garu for the telugu equivalents.
>>
>>
>> I find that some of the words might not be representative of
>> 'core' vocabulary.. but it is another matter.
>>
>> Thanks again for the telugu friends.
>>
>> Regards
>> -Selva
>>
>>
>
>
> I have a lot to comment on the research methodology being
> followed here. But let me say something constructive: I believe
> the best source of 'core' vocabularies would be primary readers
> , viz. those meant for (L|U)KGs, Ist and IInd standard books,
I agree. As I mentioned in my post, the word-set
may not be suitable for indian or dravidian languages, but
I sought telugu equivalents for that particular set
in order to not change anything which Morris Swadesh
apparently presented. If you lloked in to those words
there were many basic words missing like 'he, she, it,
they, here, there' etc. It appears originally the set of
words used were much larger ( 1000 ?) but which was reduced
to a smaller number ( 200 ?) and finally to the 100 words
I posted. Since the results will be *very* different if
one chooses a different set and since I merely wanted to
know what kind of results we might get witht the original
set, I sought words for that particular set. I entirely
agree with your concerns and please don't this anything
more than a guestimate of when tamil and telugu might have
looked very close or one language almost.
I appreciate your interest.
> and an excellent traditional primary reader in Telugu known
> as 'Pedda BAla Siksha'. IIWY (If I were you), I would procure
> as many such primary readers in the both the languages as
> possible.
>
>
> - T. Hari Krishna
>--
Regards
-Selvaa
Chandrasekhar Kambhampati
I know that you were supplied with the telugu equivalents. BUt there
seems to be a small problem here. (may be I am nit-picking).
>
> English Telugu Tamil
>
> [22]* louse purugu pEn ( a small black insect usually on hair)
Louse (lice) isnot purugu in telugu - purugu is any insect. Louse (lice)
is penu in telugu.
This would change your ratios I guess!
> [24] seed ginja vithai ( kAzh is anothe word)
Seed is also vittanamu - ooops sorry I see it in the next line under
tamil equivalent _ my mistake!
> [28] skin tholu thOl ( pronounced as thOlu)
also charmamu
> [48]* belly boddu vayiRu, thoppai, thondhi
boddu - is navel (i.e BELLY BUTTON). Belly should be potta, bojja etc.
and not boddu.
-regards
chandrasekhar
Ramana Juvvadi
First of all, let me thank you for bringing a little linguistics
to this group.
> I find that about *15 words* in these 50 words that are unrelated.
> ^^^^^^^^^^
> Thus in all 26 ( = 11 + 15) words are of different origin in these
^^
Why that 11? I think it deserves an explanation.
> 100 words. This means 74 % of words are retained.
>
> [16]* woman aada manishi(ssthri) pen, peTTai
There is a Telugu word called 'boeni' which means a woman. It might
have the same origin as Tamil 'pen'. Also 'koedi petta' in Telugu
refers to a hen whereas 'koedi punju' refers to a cock.
> [22]* louse purugu pEn ( a small black insect usually on hair)
'paenu' is there in Telugu also.
u v babu srinivas
%
% is purely amateurish and should not be taken more seriously than
% it deserves.
but as it was done anyway lets try to be more clear about to what we
want from it, and what we can hope to get. we will at least have an idea
of whether it was worth doing it!
I know you have already warned many times that we shouldn't take
this seriously at all. but still, not intending to be rude, I just want
to point out how different the results can be with little changes and
that the figure 2000 years which we arrived at might not mean anything
at all [other than probably making us feel good about having experimentally
arrived at a result we already "know"].
this might even prompt you to post more about the test and the assumptions!
assumptions:
1. telugu and tamil are not of independent origin
2. they are either siblings [originating from a common language which is
no more] or one of them originated more recently [broke off from the
other]
3. that the figure 0.86 holds for all languages
4. that the set of words used here have the same characteristics or
would hold when tested with the same criteria used for selecting the
actual set of words used by Morris Swadesh (?)
5. And all the assumptions that were made by Morris Swadesh (?)
discussion on assumptions:
(a) is #1 ruled out? if telugu is the older sibling [very old, say] of the
two, would the procedure change? if telugu is the proto-whatever language
of the two which went through lot of changes because of varied influences,
would this procedure be the same?
(b) the figure 0.86 -- why do you think it wouldn't matter too much if we
used it for tamil and telugu? what if the figure, considering all
factors like influence of sanskrit, the position of the lands where
these languages are spoken and many other possible factors, is not .86
but 0.75?
in this case --
75 * 100 = 75 1000 years [.75 instead of .86]
75 * 75 = 56 2000 years
35/50 = .70 [retention]
so the separation date was only about 1000 years ago.
now what if it is .89, only little bit different from the previous
figure of .86. this can be case too [more probable]. probably because of
relative isolation of both the lands [telugu and tamil] and because of
lesser number of influences, lets say the retention is much more giving
the figure 89%. in this case --
89 * 100 = 89 1000 years
89 * 89 = 79 2000 years
89 * 79 = 71 3000 years
89 * 71 = 63 4000 years
now 35/50 = .70
that means, the separation was definitely more than 3000 years ago!
(c) now about the word set -- this is the tricky part. I am still not
sure what the criteria was for the word set used by the person who got
the .86 figure. only if we consider the criteria it would be possible to
get a parallel set for telugu and tamil.
in the 100 words given by you, I have noticed that there are words in
telugu which are closer to the tamil equivalents you have given, but
which you have counted as not of same origin. so lets say the actual
total retained words is not 74 but 79. in this case it would mean that
the result is not 2000 years, but only 1500 years or less.
now if we considered all the factors and chose a word set of 100 which
was different from the one you used [the one used by the person you
quote]. now lets say that the words that didn't match are 36 instead of
26 we arrived at. in that case, retention is 64%. now, with .86 we have
the result would mean that the separation was 3000 years ago and not 2000
years.
of course, 1000 years is a lot to ignore!
here is something which will give an idea of the variation:
<0.82> <0.83> <0.84> <0.85> <0.86> <0.87> <0.88> <0.89> <0.90>
<0.60> 3 3 3 3 4 4 4 5 5
<0.62> 2 3 3 3 3 4 4 4 5
<0.64> 2 2 3 3 3 3 3 4 4
<0.66> 2 2 2 3 3 3 3 4 4
<0.68> 2 2 2 2 3 3 3 3 4
<0.70> 2 2 2 2 [2] 2 3 3 3
<0.72> 2 2 2 2 2 2 3 3 3
<0.74> 1 2 2 2 2 2 2 3 3
<0.76> 1 1 2 2 2 2 2 2 3
<0.78> 1 1 1 1 2 2 2 2 2
<0.80> 1 1 1 1 1 2 2 2 2
the figure on the vertical axis is the percentage of word matches. the
horizontal axis is the variation on the .86 figure [retention factor].
the numbers in the table indicate the number of 1000s of years ago the
separation might have taken place. this is too rough as we don't know
the curve for the retention factor between milleniums.
-x-
u v babu srinivas
root] it could also mean that both came from a different source
altogether [like sanskrit]. how would your list deal with that?
Are the words selected in such a manner that that could never happen?
How?
Kathiravan Krishnamurthi
Mother tongues, natural languages do not arise
from sythetic languages that are used/were used only
by few people only for specific purposes.
Please sing a lullaby in skt.
anban
kathir
C.R.Selvakumar - Electrical Engineering
u v babu srinivas <udu...@vanilla.wpi.edu> wrote:
>If two words that appear both in Telugu and Tamil [and have a common
>root] it could also mean that both came from a different source
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>altogether [like sanskrit]. how would your list deal with that?
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Yes, and such words are, as far as I can assess, not counted.
Please see my comments and the list- I've specifically mentioned
this..there are, in fact, such cases like mamisam = flesh which I have
not taken as a common word. But the word raktha = blood, although
some might think is of skt origin, I've
taken as dravidian because it is sanskrit which has borrowed from
dravidian ( look at the tamil words like arakku ( = red wax, red,
honey, wine), arakkal ( to rub ), arakkam ( = blood), araththam ( =
blood, red wax, red cotton, red flower hibiscus, red) , araLi ( a pinkish
red flower) , aran ( Lord Siva - both 'sivappu and aran' mean red),
araNi ( = sun, flame, fire-producing-stick, igniting-stone ) etc.
and there are more 30 words and all having the root sense of 'red'
and the word 'raktha' = blood does not have indo-european cognates.
[ although I am intrigued about the word 'red' itself which could
be related to 'rath' > raktha ? but the words for bood don't
match, but in the case of tamil it is 'ara, ari' that gives
the sense of 'red' .. I've not looked at the origin of the word
'red'..]
>altogether [like sanskrit]. how would your list deal with that?
>Are the words selected in such a manner that that could never happen?
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
As I said the words are not selected by me. It is my assessment
that the tamil-telugu equivalents I've indicated as common are
dravidian words and not skt or other non-dravidian words.
A few might be munda group although I don't think any in the
list are of that type.
>How?
^^^^
It is a well known fact that indo-european ( to which skt belongs)
and dravidian are two distinct language families. Tamil and Telugu
without a shadow of doubt belong to the Dravidian family, although
I know many telugu folks might think that telugu is a sister/daughter
language of skt. Literary telugu might have borrowed much from skt
both in terms of terminologies and literary conventions and practices,
but the _language_ is undoubtedly dravidian.
Therefore words common to tamil and telugu which are of skt origin or
other known languages like english ( :-) ) are of dravidian origin.
It is my sincere hope that by identifying common words and trying to
popularize these ( both in normal usage and for coining new words,
may be scientific, economic etc. ), we will develop greater appreciation,
awareness and understanding.
>
Thanks for the interest,
Regards
-Selvaa
P.S. I'll be off campus until coming Wednesday ( Oct 6th). Therefore
folks, please excuse me for not responding. If you mail to my
e-mail address : selv...@watserv1.uwaterloo.ca - I might try to
give a response - thanks.
u v babu srinivas
% Please sing a lullaby in skt.
lets not get other issues into this, but what you are basically hinting
at [right or wrong], to be sure, is clear to me. forgetting about sanskrit
for sometime, what I am trying to understand is this:
why couldn't the words belonging to the particular set of 100 chosen belong
to the ones that are prone to influence? isn't that an assumption that is
made when using the word set?
instead of appealing to intuition, is it possible to state in concrete
terms what made it possible to choose such a list of words?
C.R.Selvakumar - Electrical Engineering
Ramana Juvvadi <juv...@ulysses.att.com> wrote:
>In article <CE39E...@watserv1.uwaterloo.ca> selv...@watserv1.uwaterloo.ca (C.R.Selvakumar - Electrical Engineering) writes:
>
>
>First of all, let me thank you for bringing a little linguistics
>to this group.
>
>> I find that about *15 words* in these 50 words that are unrelated.
>> ^^^^^^^^^^
>> Thus in all 26 ( = 11 + 15) words are of different origin in these
> ^^
>Why that 11? I think it deserves an explanation.
I found 11 out of the 50 words with serial numbers 51-100
were of different origin ( which I had posted earlier ).
Therefore out of hundred words about 26 are siginificantly different.
But the taking into the input I had recently received especially
from K.V.Baparao Garu, I think even fewer words are of
different origin. As I had emphasized, this is only an amateurish
exercise and the very method is also only empirical with no
solid foundantion to assert the accuracy of the conclusions
except that it seems to be interestingly accurate for at least
indo-eurpoean langauges. I think to surmise that 65-85 % of words
are common between telugu and tamil considering some common
'core' vocabulary ( not necessarily the one given my Morris
Swadesh ) is not unreasonable and it is my guess that tamil and
telugu might have been more like two dialects ( like thirunelveli
tamil and Madras/chengalpattu tamil or corresponding dialects
in telugu-speaking lands [ i enquired about the different dialects
in telugu regions in one of my posts in SCIT and I didn't get
any response except that Rama Rao Garu sent me e-mail with
interesting details]) some 1300-2500 years ago
roughly speaking. There is much that is common and I thought
it might be interesting to know for ourselves.
>
>> 100 words. This means 74 % of words are retained.
>>
>> [16]* woman aada manishi(ssthri) pen, peTTai
>
> There is a Telugu word called 'boeni' which means a woman. It might
> have the same origin as Tamil 'pen'. Also 'koedi petta' in Telugu
> refers to a hen whereas 'koedi punju' refers to a cock.
Thanks for the info.
>
>
>> [22]* louse purugu pEn ( a small black insect usually on hair)
>
> 'paenu' is there in Telugu also.
>
Thanks.
>
>> I find that some of the words might not be representative of
>> 'core' vocabulary.. but it is another matter.
>>
>> Thanks again for the telugu friends.
>>
>> Regards
>> -Selva
>
Many more words seem to be common between tamil and telugu
in this particular set than I estimated based on the original
response. These new correspondences come from people who know
telugu well and who have pointed out telugu words closer to
tamil equivalents. The results are quite interesting, indeed.
I'll be off campus until next Wednesday ( Oct 6th), therefore
please excuse me if I don't respond - thanks
I also warmly thank all those who were kind enough to offer
further insight and details ( esp Baparao Garu, Ramarao Garu,
Prasanna Garu and many others )
Regards
-Selva
Hari Krishna Tadepalli
Ingestion of Sanskrit into Telugu is a relatively new occurrence as I
had written elsewhere in this net. It is also not difficult to understand
that Telugu and Tamil are independent south indian languages. As for the
primacy of one over ther other and inferences regarding the same based
on syntactic and phonological similarities, we seem to have little
proof other than our own respective intuitions and subjective
inclinations. I am not picking with anyone. (I remember a specific
newspaper article I read about 4 years ago which summarized a thesis
submitted in Tamil linguistic studies which hypothesized that the
Mexican language was derived from Tamil. I am not joking.). At the
moment I am inclined to believe that Telugu had some relation with
Tamil in some unseen past & Tamil is one of the several ingredients
that Telugu has grown from. Proof?: I have few sources on Telugu with
me here. At some later point of time, I will post a complete article
on the independent roots of Telugu words, most of which are very basic
in contemporary Telugu and are hardly seen in its sister languages -
Tamil and Kannada.
T. Hari Krishna
--
u v babu srinivas
%
% >If two words that appear both in Telugu and Tamil [and have a common
% >root] it could also mean that both came from a different source
% >altogether [like sanskrit]. how would your list deal with that?
C.R.Selvakumar (selv...@watserv1.uwaterloo.ca) wrote:
%
% this..there are, in fact, such cases like mamisam = flesh which I
% have not taken as a common word. But the word raktha = blood, although
% some might think is of skt origin, I've taken as dravidian because
% it is sanskrit which has borrowed from dravidian (look at the tamil
% words like arakku ( = red wax, red....
I wasn't talking about this particular experiment you have conducted,
but about the ORIGINAL WORD-LIST itself and your USE of it.
If there is a common word in telugu and tamil, it might mean any of
these --
(a) it came from a single proto language of which tamil and telugu are
derivatives
(b) that the word came from an external source to both tamil and telugu
(c) it came to tamil from telugu
(d) it came to telugu from tamil
(e) both came up with it accidentally
so there should be a criteria by which given a common word we should
decide which of the above is true. we are interested only in (a) and the
rest should be discarded in such experiments.
how can one determine, given a root, that it belongs to one family of
languages and not the other?
but obviously, the set is chosen so that only the most persistent ones
get included in the list. the criteria for persistence should be
IRRESPECTIVE OF THE LANGUAGES that are going to be tested. but once you
come up with the set, and once we apply it to our languages [tamil and
telugu here], if we find many words common, we should also be sure about
their being persistent. If we know they are not persistent but are only
diffused, it means the theory behind the word-set is wrong. if it is
otherwise, it means the languages are closely related or of common
origin or whatever.
so the proof is both ways --
(a) that the theory behind the set is accpetable
(b) that the results of applying the theory are such that theory
is shown to be still true
the actual desired result is then a secondary issue.
-