It is remarkable that after almost a half century of development and continuing success, there still is no general agreement on how the quantum theory is to be interpreted, or indeed whether a consistent interpretation, using no external concepts is possible. We are aware that the meaning (or the interpretation) of the theories we invent can change. Consider, for example, the equation Paul Dirac proposed for the electron — surely one of the most beautiful creations of the twentieth century. Conceived originally as a wave equation for the electron consistent with the principle of relativity, it is now interpreted as the field equation for a particle of spin 1/2, while almost every structural relation as originally written down by Dirac is maintained. We are also aware that we can proceed even when it is likely that our theories are inconsistent. This could not be more clearly demonstrated than by the post World War II success of the renormalization programme in quantum electrodynamics — the union in the quantum theory of the fields of Dirac and Maxwell. There the incredible numerical agreement between theory and experiment is balanced by the strong possibility that at their roots the axioms of the theory as usually stated are inconsistent.