Follow by Email

Thursday, December 22, 2011

Are translations harmful?

Filippo, in a recent comment (on this post) proposed a provocative view:

The more you resort to translations, the further you get from becoming even a good sanskritist. Translations retard progress in the mastery of both text and language.

Of course, he agrees that translations are useful for academic purposes ("it is impossible to be an academic indologist without using translations, since very few non-pandits are capable of reading in the original the breath of texts required for writing academic papers"). More important, he also admits that they are useful for one to acquire the broader context one always needs in order to read each text ("Even if your ultimate goal is the deepest possible relation with a text, you'll need translations to get you to the point where that becomes possible, and you can kick the ladder away behind you"). Still, the main point remains: if one uses translations, one is bound to be kept far away from the text.
I suppose that part of the problem lies in one's approach to research. Probably I am in favour of the usage of translations —just like I am in favour of each kind of tools (even electronic dictionaries, at least at the beginning)— since I am a gradualist and believe in slow progress. Filippo seems, by contrast, to be an "instantaneist". To put it metaphorically, I would teach one to swim through long sessions in the swimming pool, Filippo would throw one in the water, expecting that she will learn to swim if she must.
I imagine that he might object that the longer one remains close to a "substitute" of the original text, the more difficult it becomes to get rid of it and dare "jump" into the Sanskrit. This might be true, but it is true also of commentaries. Would he get rid of them, too? And what is then the limit between directness and arrogance?

As for a sub-point, namely the need for translations, I tend to think that they are often useful (just like Sanskrit or Hindi commentaries are) even for Sanskritists while dealing with texts outside their competence. Dharmakīrti seems to be one of these examples, but the same holds true for technical texts on Grammar, alaṅkāras, astronomy, etc.

What do readers think about it? What is your experience with translations? Did they ever bring you closer to the text?

The discussion started on this post. I dedicated far too many posts to translations, a list of the main ones can be found below this post.


shreevatsa said...

Minor point: I don't understand why you say "even electronic dictionaries, at least at the beginning". To read texts in the original Sanskrit, you will almost never completely stop needing a dictionary, and as dictionaries go, electronic dictionaries are much more convenient.

So using electronic dictionaries is in fact a way of ensuring you continue to read in Sanskrit, while always using translations is the opposite. :-)

elisa freschi said...

Hi Shreevatsa, does it mean that you agree with Filippo about translations? As for electronic dictionaries, I tend to warn my students about their usage, since they only select some meanings, offering "meaning" without context, texts where they occur, etc. One becomes lazy using them and some even never learn to use the PW (or the first edition of Āpte's). In sum, electronic dictionaries, so far, are dangerous, because they might convey the idea that you know more than you actually do.

shreevatsa said...

I don't have any strong opinion about Filippo's comment. I think one should definitely read texts in Sanskrit. Of course, it's not possible to read everything quickly, so it is fine to use translations, as general reading or to locate content, or to see how translators interpreted the text. Just like commentaries, as you said. But when one is going to make statements about a text (that it says XYZ, or make arguments based on its contents) then definitely I think one should read it in the original language, at least to confirm.

About electronic dictionaries, I was thinking of electronic versions of good dictionaries, like Monier-Williams or Apte. I agree that unreliable dictionaries are dangerous.

elisa freschi said...

Well of course one needs to read the original before writing about something! Unless one declares one is doing "creative philosophy", inspired by something and not assuming to be reconstructing it.

I have nothing against MW, and the electronic version you mention (unlike mine) has the advantage of mentioning the texts where a certain usage is found. I will certainly recommend it!

Jayarava said...

In my early days of reading translations I was struck by how different they were from each other. This made me suspicious enough to teach myself a little Pāli, and to study Sanskrit. I came to see that no translation is a perfect representation of the text. I now distrust translations - I don't comment on anything without having understood the original.

Very often translators introduce their own bias when interpreting ambiguous words or phrases: translating "dharma" in a Buddhist context is never easy for instance. Translators also do crazy things when the text gets difficult. I have more than once come across passages where the translator has not translated the difficult passage at all, but has tacitly translated the commentarial gloss! The trouble seems that a translator always has to some up with *something*, and this forces them to compromise. Not every translator provides insights into their process.

Finally I would say half the fun of the subject is quibbling with other people's translations ;-)


Anonymous said...

Always has to come up with something: t'was ever thus, no, with the pandits? I remember, from years ago, a comment, in some major commentary, on one completely lucid verse of the Ramayana: गच्छति गच्छते

Anonymous said...

Where sanskrit commentaries on sanskrit texts are concerned, there is no issue of LINGUISTIC alienation from the original, and the chronological gap between original and commentary, being intra-cultural, is certainly less distancing (this is true even of commentaries and translations in other indian languages). The price of the instantaneist method is a years-long and often very discouraging and boring apprenticeship during which one's chosen text is heard only faintly through the murk of the student's ignorance. The translation-dependent student has the clarity of others' translations to keep him company on the way, but my suspicion (which may not hold true for the most linguistically brilliant scholars) is that the harder, slower path of face-to-face engagement (inevitably mediated, yes, alas, by at least the dictionary) leads finally to a clearer sense of being present with the text.

nOe said...
This comment has been removed by the author.
elisa freschi said...

Dear Anonymous,
I see your point re. instantaneism and as I said, I think our different stance has to do with different characters and learning-styles. By contrast, I cannot buy the one about Hindī or Bengalī or late Sanskrit commentaries. A commentator might be really far away from the text and the tradition it deals with. It is surely interesting to read a late commentator, but one cannot claim that whoever writes in an Indian language is, ipso facto, closer to the text. Take, for instance, ancient commentaries in Tamil. Are they not to be read because they are linguistically far away from the original Sanskrit? And waht about Western translations and studies by scholars trained in a traditional way? Why should they be more "alien" than studies or translations done by careless people who happened to write in an Indian language?

However, our disagreement might have to do with different perspectives, I am deeply in favour of reading commentaries, etc., because I like studying the history of the ways a text has been understood. You seem to be more interested in engaging with a "big", rocky author on your own.

A last point: I will mention the risks you name during my next classes, in case I see people like I assume you might be among my students. Thanks for making me aware of the issue.

elisa freschi said...

@ Jayarava,
I am very sorry I'm answering just now! I sense that the problem of over-interpreting translations might be a major one in the case of Buddhism, where conflicting schools might wish to have the Buddha say what their school maintains he said. Things might get less conflictual in other fields, where differences are rather due to non-understanding or to different ways of trying to make sense of a statement.
Best wishes,

Anonymous said...

Elisa, I will think about your good points about the possibly poor claim that any premodern commentary or translation written in an indian language is in some sense closer to the root text than a modern one written in a western language (I am conscious that I am slightly refining my position here). Some background: I have lately been under the influence of a tamil friend of mine who reads english and numerous indian languages, and who told me that he feels that indian commentaries and translations, even tamil ones, are certainly closer culturally and linguistically than western ones. It is to be remembered that, even though we know that tamil and other dravidian languages are theoretically no closer to sanskrit, fundamentally, than european langauges, the fact is that their linguistic and cultural association with sanskit is far deeper and longer than that of western languages, with the result that sanskrit vocabulary is immediately meaningful to the speakers of these languages, on a deep primary level, in a way than it can never be to even the most deeply learned western indologists. This brings these texts immeasurably closer, in a visceral way, than any non-indian translation can ever be. And there is also the separate question of method, indian or indological: a question of culture and period rather than of language. It might be mentioned that, in a recent conversation, this same friend told me that, as an exception to this trend, he found Patrick Olivelle's translation of the Manusmrti superior to any that he had found in an indian language, so this may show that my friend is not blinded by cultural chauvinism. It's true that Olivelle is himself south asian, but his language and method are of course completely western and modern.

Anonymous said...

“Take, for instance, ancient commentaries in Tamil. Are they not to be read because they are linguistically far away from the original Sanskrit?”

I take them to be preferable, in their way, to modern European commentaries because, as I said, both the language and the culture of such commentaries are steeped in the same juices as the original in a way that non-indian texts can never be. Incidentally, it has long struck me that this kind of distinction between sanskritic and Dravidian indian languages is quite unreal, if one’s mind has not been prejudiced by linguistic theory. I read marathi (given enough time and necessity, I even speak it), and have at least tried to learn tamil and kannad, and it really seems to me that no unprejudiced reader would feel that marathi and hindi are really any closer to sanskrit than tamil and kannad: the grammars of marathi and hindi, separated from sanskrit by millennia of stages of profound evolution, are totally unlike that of the older language, despite the abstract historical fact of their derivation from it, and even their basic vocabulary is not apparently sanskritic; whereas both northern and southern languages, in their literary aspect, are massively endowed with sanskrit derivatives. In short, they belong to the same world, and one of the features of that world that links them to their originals even more profoundly than their method, is the presence of faith: these indian commentators, even though they might be dealing with the high philosophical end of the traditions they were writing about, also believed in the whole pre-rational level of those traditions (the pantheon, rebirth, and so on) on the most concrete, literal, un-abstract level, in a way that almost no modern indologist does; and even if an indologist is actually a born and believing hindu, academic method forbids his faith to play an open part in his treatment of the text, which is treated as a historical artifact which he is forbidden to engage with subjectively and creatively -- and in any case, he must be “rational”.

“And what about Western translations and studies by scholars trained in a traditional way? Why should they be more "alien" than studies or translations done by careless people who happened to write in an Indian language?”

Perhaps because they are generally written by literary archaeologists whose ethic compels them to preserve artefacts without leaving the slightest trace on them, whereas the indian commentators were, like their root authors, philosophers, who cared more for the analysis of reality than the mothballing of one individual’s vivakshitam.

elisa freschi said...

Dear Anonimi,

I hope I can fully grasp your point about sharing the author's background and I do think that it makes a major difference. (A few days ago I discussed with a student of linguistics who asked a Japanese native speaker to translate into Japanese "God will come in flesh at the end of the world", in order to study the usage of the future tense in Japanese. She did not notice (!) that this sentence presupposes a whole world-view and cannot be translated "neutrally".) One of my favourite books about Mīmāṃsā is F.X. Clooney's Thinking Ritually, which makes the point that in order to understand Jaimini we must shift our way of thinking into something quite different and unusual to us (but quite familiar to his cultural milieu).
Yet, please consider the following two points:

1. history plays a role and we cannot *postulate* that an Indian commentator necessarily shares the world-view of the author whose text he is commenting upon. Take the case of Sayana's commentary on the Vedas. Indologists might at least be aware of the problem and should at least try not to project their world-view back into the text. Of course, this is only done by the ones who work in a self-critical way.

2. Philosophers (just like mathematicians, physicians, etc.) are not only human beings sharing a certain world-view. They also want to address it in a critical way and in this sense many of them demand critical readers. This is certainly the case of Maṇḍana or Dharmakīrti (at least, this is my feeling about them), who are not just expecting to find emphatic readers, but also readers able to join their critical enterprise. In this sense, to be an "insider" is not enough. This does not mean that being an "outsider" is an advantage. Rather, one needs to think along the text.

aleix said...

I agree with Elisa that it is not about a limitation of language (i.e. the problem is not that the translation is not Sanskrit, or that reading Sanskrit gives you the exact sense) but a limitation of concept (i.e. we disagree with the translator about his/her understanding).
In fact, a perfect translation from Sanskrit to English, or Spanish, etc. is possible: rAmaH Agacchati "Rama comes". Where's the problem? I think the problem comes when I say "gurudevam namAmi", "nirvikalpakapratyakSa" or things like this.
My conclusion: translations are not harmful by themselves, but they can become harmful in the same way as a hammer does.

Anonymous said...

“rAmaH Agacchati "Rama comes". Where's the problem?”

Just possibly, with the polysemy of the verb “to come”? Could be pretty consequential here. Eccetera eccetera…

Anonymous said...

I can’t help remembering how, years ago, while reading Uttarakanda with my doctoral supervisor, I translated one verse (apparently unproblematic enough) as “Lakshmana mounted Sita on the chariot”, and he replied, “You might want to reword that.”

aleix said...


Ok, I'm ready to agree with you. But, the same applies to every language. And even if you read directly in Sanskrit you might not know what Agacchati means. Therefore even reading the "original sources" becomes useless. My point is that translations are not the problem. The problem is understanding the text.

Vidya Jayaraman said...

Thank you for this post and the many interesting viewpoints. As I see it there are two different aspects here:
1.Use of translations as a student while learning
2.Use and creation of translations by scholars in publications

On 1 , I agree with Filippo about not relying on translations
as a student. In fact, the problem with many contemporary
papers and texts is that the translations often display a lack of proficiency/depth of language. Methodology and analysis without this element make for unreliable texts. theory A lot of this has to do with the way academia structures sanskrit syllabus. Sanskrit readers and an introductory book does not teach much. So in that sense training students to read without translations give them a deeper sense of immersion and command over the language.

On 2, For the purpose of publication, texts, papers
at a scholarly level, translations do provide a much-needed additional perspective. Of course these translations would work better if the earlier methodology of accompanying 'anvaya' is used. This is one of the reasons why Sanskrit Commentaries exist and have their value in addition to fostering accelerated intra-disciplinary learning. Expanding and expounding upon a text and having to think through a text in terms of same and /or another language - ie in the process of translation, the very act of
looking for the right word and interpretation is in itself
an enriching learning experience!

elisa freschi said...

Thank you, Vidya, for the nice summary. You are right, translations work in a different way for students and scholars. But what do you think of their usage on the part of scholars of other fields (e.g., a Vaiyākaraṇa scholar who reads a text of mathematics along with its translation)?

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