Einstein's quantum clocks and Poincare's classical clocks in special relativity

Abstract

Einstein's reply to Weyl about the importance in General Relativity of the identity of the sources of spectral lines is well know. We show that, already in Special Relavitity, Einstein's definition of the unit of time from the frequency of atomic emission is deeply incompatible with Poincare's classical conception of time in his special relativity based on the duality "true time-local time". By using Weisskopf's analysis of quantum nature of the identity of atoms, we show that the concept of identity is implicit in Einstein's principle of relativity. On the contrary, in Poincare's relativistic logic, the use of the Lorentz transformations compensates the differences (contraction of lengths and dilation of durations) that are real by principle. The borderline "quantum-classical" passes therefore clearly between the two conceptions of the clocks in special relativity.