Follow by Email

Saturday, April 6, 2013

varṇa: 'letter' or 'phoneme'? (Or: How to translate Sanskrit technical terms?)

How to translate Sanskrit technical terms? Do we have to keep their technical aspects or make them more readable through non technical translations?
In my opinion, the answer varies according to the target readership and the text one is translating. I remember having translated thali in a Hindi novel with 'plate' because its specificites were irrelevant. By contrast, I would have not acted this way in case I were translating a cooking manual meant for cooks.

Take, for instance the case of akṣara in Indian linguistics. I cannot understand why, after Wezler's illuminating essay Credo quia Occidentale one can still translate it with letter/lettre or any other European synonym. If one translates it with 'letter', one

a) subscribes to the Western prejudice in favour of the written form of a language (although the truth is that each language is pronounced before being written),
b) projects into a Grammatical text something it did not mean to say (several Śabdapramāṇa theorists explicitly maintain that only spoken language communicates knowledge and the few who address the problem of written language say that it can inferentially communicate something because one infers phonemes out of their written form): if one makes an author speak of "letters" a reader aware of the debate might think that he was a sustainer of the possibility of written language to communicate knowledge. Is this really the case?
c) A naïver reader might think that 'kh' is made of two akṣaras/letters.

Long story short: the meaning (artha) is —in Indian linguistics— not conveyed by 'letters'. It is conveyed by phonemes. It is a distinct achievement of Indian linguists to have understood it instead of falling in the trap their Western colleagues felt in (i.e., the one of the superiority of the written system, aptly denounced by Saussure in his Cours). I really do not want to level their achievements down to the average Western level.
Westerners should start to think more about the topic if they find sentences such as "The meaning of a linguistic expression is conveyed by the phonemes composing it'' wired. And is not this (=leading to further thinking) our purpose?

How do you translate technical terms?

The whole discussion has been triggered by Aleix Ruiz Falqués, who dedicated a thought-provoking post to it. Be sure to read also the interesting comments by Jayarava and Aleix' answers. 
UPDATE: Further interesting comments can be found here.



My answer is still that there should be a margin of freedom, so long the translation is part of the semantic field of the original word, or give a complex definition of the Sanskrit word and then keep it in Sanskrit for the rest of the essay, book, etc.


Hi again Elisa!

One question: according to you, why do Indian grammarians use the same word for "phoneme" and "letter"?

By the way, I never said translating phoneme is wrong, I just pointed out that it is also an incomplete translation. Do you think otherwise?

And one detail: I did not deal with the word varna, only aksara. varna cannot be translated as letter and I would never defend that.

elisa freschi said...

Thanks, Aleix. I am happy to read that you would not go so far as to translate varṇa with 'letter'.

As for why does akṣara mean both, my point is that it does not mean both. It means the oral aspect of the couple and only secondarily its written form. Basically, it is the opposite of what happened in the Western terminology and of the the subservience to the written form condemned by de Saussure.

Licenza Creative Commons
Quest' opera è distribuita con licenza Creative Commons Attribuzione - Non commerciale - Non opere derivate 2.5 Italia.