My interest here is in what legacy modern translators of Sanskrit texts are leaving for future generations, and what we can learn from the legacies left by Chinese translations and by Tibetan translations of Sanskrit texts in history. What Gadjin Nagao said about four different Chinese translations of a Sanskrit text that were made in the past can easily apply to four different English translations of a Sanskrit text that were made in the present. After describing the four Chinese translations of the Mahāyānasaṃgraha, those by Hsüan-tsang, Dharmagupta, Paramārtha, and Buddhaśānta, Gadjin Nagao wrote (An Index to Asaṅga’s Mahāyānasaṃgraha, Part One, p. xiii-xiv):
“As one reads through the Chinese translations of the Mahāyānasaṃgraha many points come to one’s notice. We remark with surprise that various translators deal entirely differently with what we may suppose to have been more or less the same original. This is in virtually diametric opposition to the practice of the Tibetan translators who in most cases followed standardized renderings, for instance those listed in the Mahāvyutpatti. In the Chinese versions it is not rare even for the same term in the same paragraph to be rendered with a variety of, albeit synonymous, terms. This is one reason that the Chinese-Sanskrit portion of the index has grown to such an unexpectedly huge size. Hsüan-tsang’s attempt to standardize his rendering of a given Sanskrit term must be seen, in this light, as an exception rather than as the rule.”
This could just as well say: “As one reads through the English translations of such-and-such a Sanskrit text, we remark with surprise that various translators deal entirely differently with the same original.” There are many instances of this at present. In fact, what led me to study Sanskrit was comparing several translations of Patañjali’s Yogasūtra, which in some cases were so different that you could not tell they were translating the same Sanskrit sūtra.
Gadjin Nagao made a similar observation regarding the Chinese translations (p. xv): “Not a few times as we engaged in the work of compiling this index we were puzzled by how to understand a certain difficult passage or how to interpret the disagreement between the various versions.”
So how can current and future generations understand English translations of Sanskrit texts when these translations disagree with each other? When the translations disagree, they do not provide a stable basis for serious engagement with the ideas found in the Sanskrit texts. I suggest that we can learn from the Tibetan translations of Sanskrit texts, translations that were more literal and used standardized translation terms, and thereby provided a stable basis for serious engagement with their ideas.
The issue of accuracy in translation, of course, is not limited to Indological texts. It has a long history in Bible translations. With no cognizance of the history of translations of Sanskrit texts into Chinese and into Tibetan, Bible translation has on its own moved from less literal to more literal. Today, the New Revised Standard Version has gained wide acceptance, and has been adopted in influential publications including the New Oxford Annotated Bible. Released in 1989, the New Revised Standard Version was produced by a team of thirty translators under the mandate, “As literal as possible, as free as necessary.”