So an adequate theory of truth for a language is necessary and sufficient for communicating in the language. This is the same of knowing the consequences of sentences, e.g., the disquotational statements such as 'Snow is white' iff snow is white. But this also demonstrates a problem with Davidson's approach since a statement can have many provable consequences in a language; which of these is selected as the meaning of the sentence. Another objection from Scott Sonds? is that young children understand language without concepts of truth and reference, while we understand "I think I'll take a nap on the sofa"; where is the truth is this?