Skt transcription into ASCII

[Guest guest]

Dear Sri Sundararajan,

 

[in re yr query about a transcription system for devanAgari]

 

If it pleases, here is a system for DevanAgari transcription into ASCII

that i use (in my postings to certain linguistic mailing lists). The main

design goal has been to:

 

(1) to use only ASCII characters

(2) be smooth & pleasing to the eye

(3) the sounds of the transcription-scheme must be close to their English

sound value

 

The system is:

 

---

(1) vowels/semi-vowels:

 

a A/aa i I/ii u U/uu e o R RR L LL

 

M (halantam) H/: (visarga)

---

 

(2) diphthongs

 

ai/y au/w

----

 

(3) Stops and nasals: (V = voiced/voiceless; A = aspirated/unaspirated)

 

V-,A- V-,A+ V+,A- V+,A+ Nasal

--------

velar: k kh g gh n

 

palatal: c ch j jh n

 

retroflex: T Th D Dh N

 

dental: t th d th n

 

bilabial: p ph b bh m

--------

 

Please nota: the velar (ng), palatal (nj) and the dental (nt) nasals

have been collapsed to the same letter {n}. This is because the former

two occur only in specific & exclusive environments.

 

The velar nasal ("ng") occurs only in the immediate vicinity of a velar

stop (k,kh,g,gh: also the affricate "ksh", which basically is fronted by

a velar sound) & the only "n" to occur in the vicinity of a velar stop is

the velar nasal (excepting halantam). Thus, there is no ambiguity in

mapping it to the letter {n}.

 

Likewise, the palatal nasal ("nj") occurs only in the vicinity of a

palatal stop (eg., -jn-; -nc-); and the only nasal occuring in the

vicinity of a palatal stop is the palatal nasal. Thus, there is no

ambiguity in mapping it to the letter {n}.

 

On all other occasions, ie., on occasions when the sound transcribed as

{n} does not occur in the vicinity of a palatal or velar consonant, it is

taken to be a dental nasal.

 

In principle, the velar & palatal nasal cd be mapped to {n^} & {n~},

respectively, but i feel that adding diacritical marks may be a bit

jarring to the eye!

 

------

(4) Liquids, continuants and affricates

 

y r l v C S s

 

h

 

x/ksh

 

J/jn

 

------

 

Sample words:

--------------

 

CrI/Crii (ie., Sri/Shree)

 

rAma:/rAmaH

 

laxmI/laxmii (ie., Lakshmi)

 

hari:/hariH

 

kRSNa

 

vykuNTa/vaikuNTa

 

viSNu

 

nArAyaNa/naaraayaNa

 

Rgveda

 

kwstubha/kaustubha

 

tvaSTR,tvaSTA

 

savitR,savitA

 

vedAntadeCika

 

Cankara (ie., Sankar, Shankar)

 

mahAbhArata

 

CixA/Cikshaa

 

pAThaCAlA/paaThaCaalaa

 

xamA/xamaa/kshamaa/kshamA

 

&c.

 

Please let me know if u wd like me to clarify some point further!

 

Hari Om

srikanth

[Guest guest]

Re: Sri Sundararajan and Sri Srikanth's posts on Skt transliteration

 

For a font with diacritical marks, quite good results are obtained with

the Times_CSX font developed by John Smith at Cambridge and the

Normyn/MyTymes fonts developed by a Prof. Norman .... They are all

available for different platforms (Unix/PC/Mac), for free. Connect via

anonymous ftp to bombay.oriental.co.uk, and go to the pub/fonts

subdirectory. The fonts look the same in hardcopy, across platforms, when

you use word processing software. However, I've found that html documents

written using these fonts do not output the same on different kinds of

machines when you use an internet browser, like Netscape 3 or IE, because

the character assignments seem to be problematic sometimes. I'm not sure

if Netscape Communicator 4 handles them better.

 

Vidyasankar