It might be objected that someone who knows a sentence to be true must also know the proposition expressed by that sentence, on the ground that if he knows enough about the meaning of the word 'true' to be credited with the knowledge that the sentence is true, he must know the connection between knowing something and knowing it to be true ... For instance, he must know that 'Horses are called "horses" ' is true if and only if horses are called 'horses': hence, since by assumption he knows that 'Horses are called "horses" ' is true, he will, if he is capable of performing a simple inference, also be capable of knowing that horses are called 'horses'. But this objection derives its plausibility from ignoring in its premiss the distinction which it purported to demonstrate to be without substance, that, namely, between knowing that a sentence is true and knowing the proposition it expresses. We may justifiably credit someone who does not know what 'horse' means, but who knows that it is a meaningful general term, with the knowledge that the T-sentence, ' "Horses are called 'horses' " is true if and only if horses are called "horses" ', is true; but to assume, as the argument requires, that he knows that 'Horses are called "horses" ' is true if and only if horses are called 'horses' is to beg the question.