Donald Davidson on translation and alternative conceptual schemes
In his paper On the Very Idea of a Conceptual Scheme, Donald Davidson addresses those who think that others (other human beings) can have radically different concepts to ours or radically different beliefs. Such others have an alternative conceptual scheme. Davidson denies that we can make sense of this possibility. Below I present Davidson’s argument and one objection to it. Note that some commentators think that there is a stronger argument that Davidson's philosophical system entails (see Morris 1993: 167). You can find an explanation of another argument here.
The translation criterion. Davidson thinks it is acceptable for us to assume that others can only have an alternative conceptual scheme if they possess language. He thinks that a good criterion for having an alternative conceptual scheme is a translation criterion: others have an alternative conceptual scheme if and only if they speak a language that cannot be translated into our language. He criticizes various reports of alternative schemes for at once claiming that others speak such a language and then telling us about their thinking, thereby revealing translatability. This criticism suggests that he is working with a strict view of non-translatability: if another language is untranslatable into ours, anything asserted in that language cannot be conveyed at all to speakers of only our language.
The argument. Davidson holds the following view: in order for something to be another language, there must be a significant range of sentences from it which can be translated into our language. This view together with the translation criterion specified above leads him to the conclusion that we cannot make sense of the possibility of others having an alternative conceptual scheme. He tries to justify the view by analysing accounts of what a language is, which he extracts from writings on alternative schemes. He argues that they do not actually allow for others to have a language that cannot be translated into our language. (You can pick any natural language as 'our language' and this result is supposed to still obtain.)
The organizing account. According to the organizing account, something is only a language if it consists of a vocabulary (a set of concepts) for organizing, the thing which is organized being either sensory experience or reality. This is quite an obscure and puzzling account, but Davidson passes over most questions about it. He asserts that the concept of organization only applies to pluralities, using the example of organizing the closet to illustrate his point. If you organize the closet, you organize the items in it. The instruction to organize the closet itself is obscure. On the basis of his point that the concept of organization applies only to pluralities, Davidson focuses on our language organizing a plurality and another language organizing it differently. He says that the two languages must be translatable, but the reason why he thinks this is not altogether clear. He thinks that both languages must have ways of referring to the things that get organized differently, enabling translation. Hence the organization account does not allow for others to have a language that cannot be translated into our own.
The fitting account. According to the fitting account, something is only a language if it consists of a set of sentences which fit something, the thing which they fit being once again either sensory experience or reality. This too is quite an obscure and puzzling account. (Presumably, the sentences will express beliefs taken by speakers to be obvious, e.g. ‘There are objects,’ for English speakers.) Davidson regards the description of sentences fitting as a metaphorical way of saying that the sentences are true. He appeals to Tarski to propose that the concept of truth entails that a true sentence in another language is translatable into our language. Hence if another language fits sensory experience, or reality, it will be translatable into our language. P.M.S. Hacker objects to this analysis of the fitting account (1996: 297). If we regard sentences as things that can be true and we regard one set of sentences, associated with our language, and another set, associated with another language, as each fitting sensory experience, this is not another way of saying that both sets are true. Both fit but, like two theories which both fit the evidence, this leaves room for one or both being false.
Davidson, D. 1984. On the Very Idea of a Conceptual Scheme. Reprinted in Inquiries into Truth and Interpretation. Oxford: Clarendon Press.
Hacker, P.M.S. 1996. On Davidson's Idea of a Conceptual Scheme. The Philosophical Quarterly 46: 289-307.
Morris, M. 1993. The Place of Language. Proceedings of the Aristotelian Society, Supplementary Volumes 67: 153-172.