It seems like your view means that, for example, none of the sentences written in Linear A (a lost language) are either true or false. Yet, when they were written, they were (some of them at least) true or false. And were we to find a translation key like the Rosetta stone for Linear A, they would once again become true or false. Suppose one of the sentences we translate comes out to “Crete is rocky and dry”. When it was written, it was true. It is true now. But for three thousand intervening years, this sentence was meaningless?
But for three thousand intervening years, this sentence was meaningless?
It was not meaningless within the model that you describe (see one-place vs two-place functions): Linear A is a language which supports sentences equivalent to English “Crete is rocky and dry”. Whether this model is a good one will be determined when the translation key is found.
It seems like your view means that, for example, none of the sentences written in Linear A (a lost language) are either true or false. Yet, when they were written, they were (some of them at least) true or false. And were we to find a translation key like the Rosetta stone for Linear A, they would once again become true or false. Suppose one of the sentences we translate comes out to “Crete is rocky and dry”. When it was written, it was true. It is true now. But for three thousand intervening years, this sentence was meaningless?
That’s weird.
It was not meaningless within the model that you describe (see one-place vs two-place functions): Linear A is a language which supports sentences equivalent to English “Crete is rocky and dry”. Whether this model is a good one will be determined when the translation key is found.
A translation key is only possible if it is a good model (though I also think all languages are necessarily inter-translatable).