ABSTRACT
The attempt to construct a semantic metatheory - a metatheory defining the notion 'possible semantic theory' or translation manual for a natural language - has proceeded for the most part in a way which has been exactly and deliberately analogous to the attempt to construct syntactic and phonological metatheories. (See, for example, Katz 1972: 324; Jackendoff 1976:89-91.) The aim has been to defme a set of substantive and formal universals prescribing, in large part, the semantic theory for individual languages. As we have seen, the success of this enterprise would mean that Quine's indeterminacy claim could be denied by simply pointing to the semantic metatheory as a motive for choosing between alternative systems of analytic hypotheses, or alternative semantic theories, since this metatheory would define 'accessible' or 'natural' systems as would be required.
