It's arbitrary to say that it wasn't. To say that it can't be and any proof to the contrary invalid on account that it has already been decided that it can't be sounds like circular reasoning. This is Kent's position, true, but he does not go into any satisfactory detail
nor make any strong case that I can see why they should be considered Visigothic, and specifically from Narbonne no less.
It took me less than five minutes to come up these examples.
Had we been talking about barbs from a century before there'd be no argument. In those cases it's for the most
part a cut and dried distinction thanks to the consistently careful
work of the official mints. Having a reliable yardstick is
the reason we can point a finger at a coin with a funky
legend and call it barbarous even when all other features seem appropriately in line with official mintwork. But this regularity is missing in 5th century coinage. Sure, there's the outliers from both ends of the spectrum that prove existence of skilled and rough camps, respectively, and the reasonable conclusion being that the
fine ones must be government handiwork while the crude stuff must have come from freelancers. But unless I'm missing some key research that has proven otherwise I don't see how you can look at these coins and say with any degree of confidence "yeah, you see, this
mintmark here absolutely could not have come from xyz" when pretty much the ENTIRETY of western
Roman coinage of the 5th century exhibits at least some barbarous traits vis a vis its coinage from previous centuries. If we are to say that regularity of execution is the threshold by which we differentiate barbarous from
Roman then the fact that we have at least these three, with otherwise period-correct
style and
fabric and whose only distinguishing mark are the reversed
mintmark positions, then it's a real strain to support Kent's opinion without some additional arguments. Either that or we have to go back to the drawing board on how we make the distinctions.
Ras