For describing position of a particle we need vector coordinates r=(x,y,z). But when you go to higher physics like lagrangian formalism also basis vectors are introduced. The position vector then becomes r=xi+yj+zk, where i,j,k are basis vectors. Why do we have do do that? Why is not enough just to use the r=x+y+z?