I am a translation theorist and translator. These two aspects of my work grew organically. While I wrote a dissertation on translation, I decided to translate a philosophical text from Sanskrit because I could not stand the translations on the market. Getting involved with the scholarly challenge of translation had a marked influence on my philosophical views and translation theory. It further undermined, for me, a common view in the literature that the object of translation is the linguistic meaning of what is said and that the content of thought is this meaning. This is the assumed view of philosophers in the Western tradition in both the Analytic and Continental traditions, from what I could tell: it underwrote the kinds of skeptical arguments about translation we find in philosophy (from philosophers as dissimilar as Quine and Derrida) , but also the optimistic views about translation we find entertained (as seen in the moral semantics literature). My research and practice of translation also directed me to the conclusion that a common and influential view in the Translation Studies literature (championed by the likes of Eugene Nida) is mistaken too. According to this account, accurate or good translation reproduces the pragmatic effects of the original texts, but in the target audience. (It’s a modern version of a theory that was long ago entertained and rejected by Russell and Wittgenstein: a functional theory of meaning.) So if a text made people laugh in the original, then a good translation has to do the same according to this account. While this might seem appropriate in the case of fiction or drama, it’s unreasonable to expect that this has any bearing on philosophy, science, law, mathematics–disciplines where truth matters. For the pragmatics of statements is a largely subjective and inter-subjective matter, and the topic of philosophy, law, science or mathematics is not subjective or inter-subjective.
One commonality in the literature, I observed, whether inside or outside of philosophy, was a tendency to describe translation as a one-size fits all exercise, but one that had primarily to do with linguistics.
My response was to observe hat there is no such thing as translation as such: one always translates works of a specific kind. This is why translators must specialize in specific disciplines. This is because:
- Translation is always reconstructing a text with differing semiotic resources, and the criterion of success is text-type relative.
- What is appropriate in the translation of one kind of text (poetry, or fiction) would not be appropriate in another (law or science)
- One has to appreciate the kind of text in question in order to accurately translate.
- Equality of use of semiotic resources according to the relevant kind of text is translation success.
This approach corresponds to our best translation practices employed in the support of research in the academy and in matters such as law. It also entails the following:
- Accessibility or the ability of a target audience to understand a translation is not a universal criterion of translation success, for the simple reason that the target audience may not be equipped with the requisite disciplinary expertise to appreciate the original text.
This implies that there are many processes of semiotic conversion aimed at user friendliness that are not translation, but localization. It also entails other interesting findings for the philosophy of thought:
- Thought is not the meaning of what we say, but the disciplinary purpose of what we say: that is what can be translated.
- Thinking about translation in terms of localization confuses the pragmatics of language use (what is context bound) with the semantics of thought and inquiry (what transcends contexts in translation).
- Linguistic accounts of thought as the meaning of declarative sentences confuse pragmatics of culture bound language use (literal meaning) with semantics (what can transcend contexts of language use).
The only reason for denying these last observation is what seems like a stubborn commitment to the idea that the meaning of what we say in our language is propositional content. The problem with this position is that it entails something funny:
- The more we need a translation because of the linguistic divergence between languages, the less likely we can produce one.
The real problem though for the linguistic approach to thought is that it has to settle on an account of meaning that is bound to linguistic contexts–-specific languages. But the principle means of distinguishing between the pragmatic and the semantic is that the semantic transcends contexts, while the pragmatic does not. So linguistic accounts of thought end up settling with a pragmatic account of meaning, which is ironic. The way we avoid this irony is by identifying semantic content and propositional content as what transcends languages in translation, but this is only possible if we allow that accuracy in translation is judged along disciplinary lines for this disciplinarity allows us to identify the meaning of a text as something that transcends the contingencies of the culture that defines and gives rise to resources that compose it in the first place.
- So in the end, meaning and thought is really discipline bound, and not a property of languages.
We can and do identify the meaning of utterances in languages, and we can and do identify the literal meaning of such utterances in terms of their systematic or basic role in a language. This is something important for disciplines such as lexicography and linguistics, but not for others.
One of the most important outcomes of this approach to thought, meaning and translation is that:
- Cultural variation and radical difference is no obstacle to accurate translation.
- The apparent incommensurability of languages is no challenge to translation: it is only a problem if one takes matching linguistic meaning as translation.
