THROWING A BEAR-WRENCH IN THE WORKS: A NEW PROBLEM WITH THE NAME ARTHUR

I've been having discussions with Professor Roger Tomlin and Dr. Benet Salway about the conventional view that the British name Arthur must derive from Roman/Latin Artorius.  Something has come out of this exercise that I was not expecting, and it is more than a little disturbing.

To begin with Tomlin on the subject:

Artorius is a nomen, not a cognomen, and 'single names' are generally cognomina. They are used within (Roman) families to distinguish a particular member, and if a non-Roman becomes a Roman citizen, he makes his existing name into a cognomen combined with his new nomen, which may be taken from the patron who gained him citizenship, his former owner if a manumitted slave, but very often the Emperor himself – auxiliary veterans, for example. Thus a Celt whose name is Artos ('bear') might take the Roman name of Marcus Ulpius Artos, or even Marcus Ulpius Artus, by adopting the Roman termination –us for the non-Roman –os. 

I could find a hundred instances of Artorius as a Roman nomen, but none of it as a cognomen. Mócsy's Nomenclator claims there are a couple of instances, but he is not altogether reliable, and I don't have the references. Certainly it is of nomen form in –ius, and true, nomina such as Iulius were sometimes used as a cognomen by Roman citizens.

This is the practice until the third century. Benet will tell you about the system breaking down and changing after AD 212, with people using multiple cognomina instead like 'Ammianus Marcellinus', so that 'Quintus Aurelius Symmachus' in the fourth century sounds old-fashioned (if aristocratic). But someone who gained citizenship by favour of Artorius Castus would take Artorius as a nomen, not a cognomen. If he wanted to imply a relationship, he might adapt the nomen as Artorianus. True, there are 'Decknamen', which are Roman cognomina which 'conceal' Celtic names. A good example is Verecundus, which an ancestor of the prefect Iulius Verecundus probably adopted to make himself sound like a good Roman. 

The reverse process, of a Celt taking a single Roman nomen would be hard to parallel. This is what the 'Arthur' people assume, but I don't know that it happened.

Benet may be helpful, and I attach a copy.

The 'copy' Tomlin alludes to is Benet's "What's in a Name? A Survey of Roman Onomastic Practice from c. 700 B.C. to A.D. 700", The Journal of Roman Studies , 1994, Vol. 84 (1994), pp. 124-145,

https://www.jstor.org/stable/300873.  I read the paper and then contacted Dr. Benet with my questions.

His response:

Obviously it is *possible* for any Roman nomen to turn up as cognomen in late antiquity, whether in Britain or elsewhere, but generally the nomina that make this leap are the most common imperial nomina, such as Iulius, Claudius, Flavius, and Aurelius. Given that Artorius is not in this category, this makes it highly *improbable*.

The next bit is going to contain a lot of hypotheticals.

*If* there ever existed a single individual in post Roman Britain who went by a name later recorded as Arthur or similar and *if* this reflects some kind of Latin original, it would seem much more likely to be Arcturus, the name from Greek cosmology/mythology designating the brightest star in the northern celestial hemisphere (the ‘bear-guardian’). To my knowledge this is not found as an anthroponym (i.e. as a personal given name for mortals) but is just the sort of thing that might be applied as a nickname or adopted as a ‘nom de guerre’. And in a British context, the fact that the etymology of the first element would be transparent to Celtic speakers, might be an added advantage. But this only really works *if* there was an awareness of the star name. Of this, I think we can be sure because of the popularity of Aratus’ poem Phaenomena, which was translated at least 4 times into Latin, two of which survive and one of which, the late antique one by Avienus, had a very wide diffusion to judge from the manuscript tradition.

Also theoretically *possible* but unattested, is Arcturius (i.e. ‘having the quality of Arcturus) as a given name.

It is possible that ‘Artorius’ might be used as a Deckname, but not probable since it is never attested used as a cognomen.

Now, it is important for me to say before going further that Arcturus has been proposed before for the name Arthur.  The most extensive discussion of this possibility that I've had access to is found on pp. 187-194 of Thomas Green's CONCEPTS OF ARTHUR, 2007.  The author actually prefers the Arcturus etymology to Artorius (although he was not privy to my newly proposed reading for the L. Artorius Castus stone, which makes for a potential Arthur with a greatly increased measure of fame in Britain).

However, Green does not mention the difficulty of using a uncommon nomen like Artorius as a cognomen - something that weighs more heavily in favor of Arcturus as the origin of the name Arthur.

If Arthur is L. Artorius Castus, for example, why was he not called by his cognomen Castus?

I should also add that it doesn't make sense for us to entertain the notion that a British (or Irish) name meaning something like bear-guard was replaced by Arcturius, as the latter is not a real Roman name.  We have plenty of examples of bear-names in the Irish and Welsh sources and no evidence at all that any were replaced with Roman bear names or Roman names that resembled Celtic words for a bear.  

If we are stuck with a nickname Arcturus, who was it applied to?

Well, I've been working for many years on various Arthurian candidates.  The only one who has notable ursine associations is Ceredig son of Cunedda, AKA Cerdic of Wessex.  Ceredig's kingdom has in its midst an Afon Arth of Bear River, and there was a headland fort at the mouth of the Arth.  Three of his immediate descendents have *Arto- names.  Arthur, through his father Uther (who, if = Pen Kawell of the MARWNAT VTHYR PEN can be identified with Ceawlin-Cunedda, father of Ceredig), is linked to Caer Dathal/Dinas Emrys of the mythological Math, a personal name almost certainly borrowed from the Irish word for bear.  

Cerdic of Wessex and Arthur are exact contemporaries, according to the HB and the ASC, and I once had an entire book published arguing for Cerdic/Ceredig as Arthur.  

I will give this name problem considerable thought and follow up on the subject at a later date should such work be warranted.