Abstract

I wish first to motivate very briefly two points about the kind of context sensitive semantics needed for attitude reports, namely that reports are about referents and about mental representations; then I will compare two proposals for treating the attitudes, both of which capture the two points in question.