Abstract
The argument is re-examined that the program of deriving the rule for the state change caused by a measurement from the Schrödinger equation holding for the object-apparatus composite system falls into a vicious circle or an infinite regress called the von Neumann chain. It is shown that this argument suffers from a physical inconsistency concerning the causality between the process of reading of the outcome in the apparatus and the state change in the measured object caused by the measurement. A consistent argument which accomplishes the above program without falling into the circular argument is presented.