Radical translation is the setting of a thought experiment conceived by W.V. Quine in the late 1950s. In that setting a linguist undertakes to translate into English some hitherto unknown language – one which is neither historically nor culturally linked to any known language. It is further supposed that the linguist has no access to bilinguals versed in the two languages, English and (what Quine called) ‘Jungle’. Thus, the only empirical data the linguist has to go on in constructing a ‘Jungle-to-English’ translation manual are instances of the native speakers’ behaviour in publicly recognizable circumstances. Reflecting upon the fragmentary nature of these data, Quine draws the following conclusions:
- It is very likely that the theoretical sentences of ‘Jungle’ can be translated as wholes into English in incompatible yet equally acceptable ways. In other words, translation of theoretical sentences is indeterminate. On the assumption that a sentence and its translation share the same meaning, the import of indeterminacy of translation is indeterminacy of meaning: the meanings of theoretical sentences of natural languages are not fixed by empirical data. The fact is, the radical translator is bound to impose about as much meaning as they discover. This result (together with the dictum ‘no entity without identity’) undermines the idea that propositions are meanings of sentences.
- Neither the question of which ‘Jungle’ expressions are to count as terms nor the question of what object(s), if any, a ‘Jungle’ term refers to can be answered by appealing merely to the empirical data. In short, the empirical data do not fix reference.
The idea of radical interpretation was developed by Donald Davidson in the 1960s and 1970s as a modification and extension of Quine’s idea of radical translation. Quine is concerned with the extent to which empirical data determine the meanings of sentences of a natural language. In the setting of radical interpretation, Davidson is concerned with a different question, the question of what a person could know that would enable them to interpret another’s language. For example, what could one know that would enable the interpretation of the German sentence ‘Es regnet’ as meaning that it is raining? The knowledge required for interpretation differs from the knowledge required for translation, for one could know that ‘Es regnet’ is translated as ‘Il pleut’ without knowing the meaning (the interpretation) of either sentence. Beginning with the knowledge that the native speaker holds certain sentences true when in certain publicly recognizable circumstances, Davidson’s radical interpreter strives to understand the meanings of those sentences. Davidson argues that this scenario reveals that interpretation centres on one’s having knowledge comparable to an empirically verified, finitely based, recursive specification of the truth-conditions for an infinity of sentences – a Tarski-like truth theory. Thus, Quine’s radical translation and Davidson’s radical interpretation should not be regarded as competitors, for although the methodologies employed in the two contexts are similar, the two contexts are designed to answer different questions. Moreover, interpretation is broader than translation; sentences that cannot be translated can still be interpreted.
The semantics of indicative conditionals: A counterexample to the horseshoe analysis
