The author feels that although the problem of pronominalization and of reference in general is at the very heart of syntactic investigation, transformational grammarians are unable to deal adequately with these problems. He finds their theory of referential indices (Chomsky, "Aspects of the Theory of Syntax") "completely beyond repair." In this paper he presents a few examples of the failure of referential index theory and indicates the direction in which he thinks a solution will ultimately be found: counterpart theory. (DO)