This paper offers a response to Ramsey's famous challenge to draw a logical, as opposed to merely syntactical, distinction between names and predicates. Three attempts to meet this challenge are considered, one from Dummett, one from Geach, and one from Brandom. It is shown that none of these adequately addresses Ramsey's challenge. The paper concludes with a response to Ramsey that avoids the pitfalls of the other three replies. The advantage of the view proposed is that there is nothing preventing (...) the set of sentences in consideration from consisting only of atomic sentences. The upshot of my solution is that the distinction between names and properties rests on the presence of relational predicates in a language and the validity of the structural rules of identity, contraction, and cut. (shrink)