X says: “Wendy Doniger translates all the Sanskrit words into English and thereby ends up distorting their meaning. Dharma becomes religion, Varna becomes color, and apparently she ends up even translating the names of some tribes into English.” Assuming that this is the case (assumption is on my part because I have neither read her translation nor do I intend reading it), what is the problem? Out of the conversation, the following considerations came to the fore: (a) if she had retained some of these Sanskrit words, the Indian scholars could have made more sense of the translation; (b) Dharma and religion are not co-extensive in meaning, any more than Varna is with color; (c) consequently, there is a great distortion in translation. Underlying this discussion was a (fragmentarily explicit) view about translation that if we do not find ‘accurate’ translations of words form one language into another, we should retain the original words, assign them the status of `technical’ terms, and not translate them any further.
Let me begin with the view about translation. I have heard it so often from Indians, I could say that this is their `standard’ view using which they criticize the likes of Wendy more often than not. I have problems with this view: I think it naïve, indefensible and false.
We need to put this issue in its generality first. Are there ever `accurate’ translations of texts from one language to another? That is to say, could we ever say that a translation has rendered words from one language to another `accurately’? The answer to this question revolves around two things: (a) whether we have some way of accurately knowing the meaning of a word (any word) in a foreign language; (b) whether we have some way of quantifying the accuracy of the meaning of words.
Regarding (a): In philosophy of language, there is a thesis called `indeterminacy’ of meaning, a thesis original to the American philosopher called Quine. He showed (in his book “Word and Object”) that, when it comes to translation across languages, words in one language cannot be shown to map onto the meaning of another word in another language. That is, we cannot ever vouch for a `correspondence’ in meaning between two words in two different languages. He was talking about the possibility of translating the word `Gavagai’ (the word used by a foreign tribe whenever they saw a white rabbit) into `white rabbit’, `whiteness’, `rabbit’ and so on. He showed that no matter what kind of empirical tests we use to circumscribe the meaning of `Gavagai’, there would always be indeterminacy with respect to its translation into English. As far as I know, no one has refuted this thesis and almost all knowledgeable people accept this thesis in one version or another. Needless to say, I subscribe to it as well, even if I am not sure I want to accept Quine’s theory of meaning.
Its lesson is rapidly learnt. If the translation of the word `Gavagai’ (which is produced only when a white rabbit is seen) is itself subject to indeterminacy, what about words like `Dharma’ and so on? Obviously, it is only greater: the very fact we are all easily `dissatisfied’ with this translation indicates that the latitude is very great. Each one of us can now sit down and compile our hobby-horse list of the untranslatable Sanskrit words: Deva, puja, laukika, atman, brahma, varna, dharma … Compiling such a list is uninteresting and trivial because it is utterly uninformative.
Regarding (b). If this is the case with respect to (a), there is not much of an answer to give about `quantifying’ the accuracy of translation. At best, one could have a problem with the `quality’. But one can express this dissatisfaction ONLY if one has a better alternative: “`Religion’ should not translate `Dharma’. A better translation would be `…'” (For the simple reason that one is dissatisfied with the `quality’ of translation and this requires comparison.)
Our problem, as Indians, is only beginning. Consider the statement: “The word `religion’ does not accurately translate `dharma'” (or any of its ilks). This statement presupposes that the speaker knows the meaning of the word `religion’ in English and the meaning of the Sanskrit word `Dharma’. Otherwise, he cannot criticize the translation of these two words. One could assume that if English were to be our native language, we have the intuitions of a native speaker with respect to the use of the word `religion’ (or its meaning). For most of us, this is not the case: English is not our mother-tongue. Neither is Sanskrit. We have learnt both languages through the medium of other languages (like Hindi, Marathi, Bengali…) and so on. That is to say, our understanding of the meaning of the words `religion’ and `dharma’ is itself susceptible to the indeterminacy thesis. One cannot anymore argue that the intuitions of a native Bengali speaker with respect to how he uses the word `Dharma’ in Bengali tracks the meaning of the Sanskrit word `Dharma’. Exactly the same argument applies to his knowledge of the word `religion’ in English. This being the case, how can an Indian confidently assert that he knows what `religion’ means in English or `dharma’ means in Sanskrit in an accurate fashion? He simply cannot. That is why many Indians do think that `religion’ translates `Dharma’ accurately; some others disagree. Do we really know what the various meanings of `religion’ in English are? I have yet to come across an Indian who knows what he is talking about when he talks about religion; but I have come across indefinitely many who believe that they `know’ what religion is and put across pompous and half-baked `theses’ about religion.
Added to this, is the nature of a word like `Dharma’ in some Indian language or another (or even with a Sanskrit pundit whose mother tongue is Sanskrit). Everyone routinely assures us that this is a complex word with multiple meanings. (People forget or do not know that this is also the case for a word like `religion’.) If this is true, you can never assign a technical meaning to this word and leave it untranslated. Because, technical terms (as anybody with any modicum of scientific training knows) are univocal: their meaning is rigidly defined in the context of a theory. If `Dharma’ has to be used as a `technical’ term, one has to unambiguously define its meaning and that meaning must not be equivocal but univocal instead. But that is precisely the problem, is it not? These considerations apply to all the pet lists we can make about the `untranslatable’ Sanskrit words.
If one does not translate such words into English but leaves them intact, what exactly do you translate then? Each sentence would be replete with `untranslatables’, and we might as well say that one should not translate texts from one language into another. Think of it the next time you want to read Homer or Tolstoy or any of the classics of the western world.
Not only should we bear the above points in mind when we criticize the western intellectuals for translating Indian texts, but it should also be clear what lessons it holds with respect to this thread: what are these Indic `categories’ that we should use to describe the world and how could the user of these `categories’ know that he knows the `correct’ meaning of Sanskrit words? If he does not, what exactly is being talked about? Who will understand him, when most of us either do not know Sanskrit or only have learnt it through the medium of other languages, including, sometimes, English?
- Why use Indic categories to describe the world?
- Fallacy of Equivocation: Indian Secularism