Pavel Tichý, in a recent article, 1 argues that Saul Kripke's purported examples of necessary a posteriori truths2 are unsuccessful. I am sympathetic to some of Tichý's assumptions and to some of his conclusions. But his arguments seem misguided to me, and I will try to explain why.

In Section II, I discuss Tichý criticism of Kripke's treatment of the Hesperus-Phosphorus example. That criticism is seen to rest on the assumption that sentences which express the same proposition are interchangeable in epistemic contexts, an assumption Kripke would not accept. Nevertheless a revised version of Tichý's criticism may be successful.

In the rest of the paper I adopt Tichý's assumption and explore what someone who, unlike Kripke, accepts it might say about the necessary a posteriori. In Section III I quickly develop the idea that we have attitudes to some propositions indirectly in virtue of out attitude to other propositions. In Sections IV, V, and VI, I emply this idea to criticize some of Tichý arguments. Section VII provides a brief conclusion and a tentative suggestion about how one might reformulate the positivist doctrine that the necessary and the a priori coincide.