We start from our "rigid body notion", where we can define concepts like translation, rotation and other deriving from them, like linear position variables or angles of rotation and velocities. We solve most of the relevant problems with excellent accuracy, although we cannot in general solve all of them. But, what is the legitimation when we go deeper and deeper for writting an object like a Lagrangian density, or for speaking about SU(3)xSU(2)xU(1) symmetries or even talking about Parity (which is strictly connected to the Cartesian view of cosmos)? Can we do such a great jump? What is your opinion?