Abstract

To single one out of the infinitely many, empirically indistinguishable gauge potentials of classical electrodynamics, and to deem it `more real' than the rest is not trivial. Only two routes are open to one who might attempt to do so. The first leads to a slippery slope: if one singles out a potential solely by requiring it to admit well behaved propagations, and on the strength of this behavior one subscribes to its reality, one inevitably subscribes to the reality of infinitely many. As for the second, it seems to be barred from the beginning. But if, for reasons of metaphysical economy, one insisted on taking it, it would lead to a `truncated theory' that is physically and empirically inferior to the complete.