indeterminacy of translation a pair of theses derived, originally, from a thought experiment regarding radical translation first propounded by Quine in Word and Object (1960) and developed in his Ontological Relativity (1969), Theories and Things (1981), and Pursuit of Truth (1990). Radical translation is an imaginary context in which a field linguist is faced with the challenge of translating a hitherto unknown language. Furthermore, it is stipulated that the linguist has no access to bilinguals and that the language to be translated is historically unrelated to that of the linguist. Presumably, the only data the linguist has to go on are the observable behaviors of native speakers amid the publicly observable objects of their environment. (1) The strong thesis of indeterminacy, indeterminacy of translation of theoretical sentences as wholes, is the claim that in the context of radical translation a linguist (or linguists) could construct a number of manuals for translating the (natives’) source language into the (linguists’) target language such that each manual could be consistent with all possible behavior data and yet the manuals could diverge with one another in countless places in assigning different target-language sentences (holophrastically construed) as translations of the same source-language sentences (holophrastically construed), diverge even to the point where the sentences assigned have conflicting truth-values; and no further data, physical or mental, could single out one such translation manual as being the uniquely correct one. All such manuals, which are consistent with all the possible behavioral data, are correct. (2) The weak thesis of indeterminacy, indeterminacy of reference (or inscrutability of reference), is the claim that given all possible behavior data, divergent target-language interpretations of words within a source-language sentence could offset one another so as to sustain different targetlanguage translations of the same source-language sentence; and no further data, physical or mental, could single out one such interpretation as the uniquely correct one. All such interpretations, which are consistent with all the possible behavioral data, are correct. This weaker sort of indeterminacy takes two forms: an ontic form and a syntactic form. Quine’s famous example where the source-language term ‘gavagai’ could be construed either as ‘rabbit’, ‘undetached rabbit part’, ‘rabbithood’, etc. (see Word and Object), and his proxy function argument where different ontologies could be mapped onto one another (see Ontological Relativity, Theories and Things, and Pursuit of Truth), both exemplify the ontic form of indeterminacy of reference. On the other hand, his example of the Japanese classifier, where a particular three-word construction of Japanese can be translated into English such that the third word of the construction can be construed with equal justification either as a term of divided reference or as a mass term (see Ontological Relativity and Pursuit of Truth), exemplifies the syntactic form of indeterminacy of reference.
See also MEANING, PHILOSOPHY OF LAN- GUAGE , PHILOSOPHY OF MIN. R.F.G.