In this article I argue that Russell's multiple-relation theory of judgment is a continuation of the campaign against Frege and Meinong begun in “On Denoting” with the theory of descriptions. More precisely, I hold that the problem of false belief, to which the multiple-relation theory is presented as a solution, emerges quite naturally out of the problem context of “On Denoting” and threatens to give new life to the theories Russell purports to have laid to rest there, and that Russell's (...) solution to the much neglected third puzzle of “On Denoting” contains the gist of the multiple-relation theory. The failure of that theory to solve the problem of false belief thus signifies the failure of the theory of descriptions. (shrink)