On the Logic of Demonstratives