Are frame-dependent quantities physically real?

Simon W. Saunders


In the early days of special relativity physicists often spoke of frame-dependent quantities as measurable quantities. Philosophers have followed them; metaphysical questions concerning tense and temporal becoming have often been answered in these terms. I argue that "frame-dependence" is a misnomer; measurable quantities are invariably _invariant_ quantities, typically relations between events. Frame-dependence has been defended in a different context (most notably by Bell), namely as a basis for the explanation of characteristic effects in special relativity such as length contraction and time dilation. Might it similarly be used to explain the phenomenology of time? To this I argue in the negative: the real virtues of the Lorentz-Bell pedagogy (as Brown and Pooley have called it) is (i) to recover Newtonian concepts of force, space and time in a special relativistic setting - and to explain their limitations, and (ii) to emphasize that rods and clocks are dynamical systems that only behave as rods and clocks because the dynamics underlying their behaviour is Lorentz covariant. Of these (i) may be taken over to explain the phenomenology of time, but to a different conclusion than the one usually sought.