Writing a post on my phone, which is hard to do on blogger. Bare with me 🙏
Quine’s indeterminacy of translation thesis specifically is about the lack of empirical evidence that determines what makes a good translation. In the Dennett interview, Quine clarifies this by stating that the indeterminacy of translation is entailed by the possibility that the mixture of two successful translation manuals is not necessarily a successful translation manual.
Quine had famously had difficulty explaining or giving examples of radical translation failure in this sense. Indeterminacy of translation may seem to be an airy doctrine which pulls its apparent strength from actual translation difficulties without being related to them.
However, if one pursues the Donald Davidson track of trying to understand ordinary language as a way of communicating which truth conditions are cleared1, then the situation becomes more clear. A sentence generally has a complex halo of truth conditions, which generally depend on distant contexts outside the sentence as a marker. As an example, the pseudo-aphorism
‘Du gehst zu Frauen? Vergiß die Peitsche nicht!’
seems to immediately bring some kinky image in mind. If one reads the whole chapter one discovers the one who speaks this - ‘das alte Weiblein’ (who is also Zarathrustra’s ‘Bruder’) is using femaleness as a metaphor for human creativity. The Frauen are Zarathrustra, though he fails to understand, as is typical for him.
A simple example of indeterminacy is the ancient Egyptian sentence fragment ‘snt nn snw.s’, which can be translated ‘Sister without rival’.
Obviously, the english word ‘sister’ and the er en kemet word ‘snt’ (pronounced senet) have different truth conditions. A sister is “a woman with the same parents, broadly construed, as someone else or a nun”. A snt is more like “a same age female relative” - that is, sister or wife. The predicates in double quotes are meant to point to truth conditions rather than be english language translations
Clearly switching between translation manuals with different rules about snt - sometimes ‘sister’ sometimes ‘wife’, each with the understanding that these are to be broadly construed - can create incoherence.
Of course, during translation - especially in a scientific context - we often ask that the translation do more than correctly track the truth conditions of the sentences. Partly this is because sentences in a scientific theory typically have a less complicated halo of meanings than artistic sentences, but also because of the needs of science.
In particular, we ask that a translation map projective predicates onto projective predicates. For instance, Henry Mosley & Niels Bohr translated the ordinal numbers of elemental periodicity into the cardinal numbers of basic positive charges in the atomic nucleus. The first set of predicates is in the language of chemistry and the latter set of predicates in the language of physics.
Not only do both of these predicates have the same truth conditions, but also both of these predicates are projectable. This successful translation is - along with the Boltzmann/Gibbs of heat sentences into stability sentences - deservedly the paradigm of what is known as ‘reductionism’.
The success of these translations is ontic. The number of protons in a nucleus is what the empirical and ordinal atomic number of Mendeleev measured.
Frequently in sciences, such as the social sciences, we attempt causal reductions rather than ontic. A large price change may be caused by a (usually effervescent) change in social psychology, but the price change isn’t the change in social psychology.
Seems bad to me but I also think I figured out how to post to Substack, so who is to say?
I have often summarized Davidson’s philosophy as being Aristotelian about action and radically non-Aristotelian about language. This of course, leads him to being very sympathetic to Freud and Lacan.