A HUMBLE RETRACTION OF THE ELIWLAD = EILWLAD PROPOSED ETYMOLOGY

Eagle in an Oak Tree

Having just put out there a radical revision of my book THE ARTHUR OF HISTORY entitled THE BATTLE-LEADER OF RIBCHESTER, I find myself now having to recant the proposed personal name etymology I used as the linchpin of my argument in that title.  This is in one sense embarrassing, but in another quite freeing.  For I had enough doubt about that etymology to continue working on the name behind the scenes.  The goal was to make sure I could accept the hypothetical derivation and that such acceptance was based on a consensus opinion of leading Welsh and Celtic linguists.  As Eilwlad allowed me to identify Arthur's father Uther with a historical figure who could be fixed geographically, a lot was riding on the meaning of this single name.

I faced four problems with the idea that Eliwlad was originally Eilwlad.  The first was not insurmountable, although it did cast some doubt on my theory.  Dr. Simon Rodway, one of the foremost experts in the world on Old and Middle Welsh, preferred a late spelling for the name which terminated in -lod instead of -lad/-lat.  According to Rodway, even though -lod only occurs in late MSS. and was used by a couple of late poets, for the form to exist demanded that Eliwlad must be an instance of the application of the principle of lectio difficilior.  In other words, at some point scribes had opted for Eliwlad because it looked more familiar to them than Eliwlod.  This was, perhaps, because they could make nothing of Eliwlod, and opted for (g)wlad as the second element of the name instead.  -lod itself had to come from something like llawd or blawd.  Alas, no workable etymology could be arrived at for such an exemplar.  Other scholars backed me in supporting the Eliwlad/t spelling.  

Late spellings in -lod also dropped the initial E-, but that is considered to be a "perfectly predictable" (Rodway) development and is not related to the -lad vs. -lod problem.  

The second and more serious problem involves my preferred metathesis of Eil- for Eli-. While I was able to demonstrate this happening in MSS., even good ones, to sustain my argument meant that I had to find an example of Eilwlad extant somewhere.  That I could not do.  For while in the MSS. with eil for eli (or vice-versa) there was never any doubt which form was actually correct.  For an Eilwlad name to have become Eliwlad would imply that the original form was miscopied at some point, the original form completely lost from all sources and the new, incorrect spelling was taken as the exemplar.  Assuming the existence of such a lost form is a dangerous proposition.  Because of all the contextual supportive "evidence" I seemed to possess, I was willing to take the risk.  But only if I was absolutely certain no other, better etymology was available.

Third, I could not account for why an otherwise unknown eliwlad, 'other land' (for 'pilgrim') would have been substituted for the Irish ailithir, 'other land' (aile + tir) when Welsh itself from very early on had tir.  As Professor Peter Schrijver pointed out, Irish aili- would have become Eli- in Welsh, but why would gwlad be used instead of tir?

And four, even if there had been an Irish compound aile + flaith, it would not have been spelled aili-, and the second component would have been altered in such a way as to prohibit it from becoming gwlad in Welsh.  I have this from Professor Jurgen Uhlich:

"Such a compound could of course have existed. Its shape (and thus its reflection in spelling), however, would not feature aili-, as e.g. in ailithir, since while the schwa /ə/ of the second syllable in alilithir is followed by a palatal consonant and needs to be spelled <i> between two palatals, what follows in your example is neutral, hence *aile(f)laith or indeed, if with regular unstressed processing, *ailelaid."

So I went back to drawing board.  All further attempts to etymologize the name failed - until I went back to an old idea.  One that I had considered very early on and then dispensed with.

There is some evidence that gwlad in Welsh once not only meant 'land, kingdom' and the like, but that it also had the same meaning one finds in Irish, i.e. 'prince, lord, ruler.'  Dr. Simon Rodway was kind enough to send me the information on this possibility and I include that here in full:

"‘The Date of Culhwch ac Olwen’, pp. 50–1:

Another possible old word that was not understood is gwlad in the sense of ‘lord, ruler’. In

one of the poems ascribed to Taliesin, his patron, Urien In one of the poems ascribed to

Taliesin, his patron, Urien of Rheged, is praised by comparison:

gwacsa gwlat da wrth Urföen.

In the context, this ought to mean, ‘Useless is a good lord compared with Urien’. Although in

Middle and Modern Welsh gwlad means ‘country’ or ‘kingdom’, its Irish cognate, flaith, has

a triple meaning, ‘lordship; kingdom; lord’. This example makes it likely that in early Welsh,

gwlad could have at least a double meaning, ‘lord’ and ‘country’. A further likely example is

in Culhwch. The phrase mab brenhin gvlat teithiawc in lines 90–1 has a parallel in line 95,

mabyon gwladoed ereill, where ereill shows that the text is referring back to the earlier

phrase. This makes it likely that brenhin here is an embedded gloss, so that the contrast was

between map gvlat teithiawc and mabyon gwladoed ereill. The mabyon gwladoed ereill were

to be housed in the yspyty, whereas the mab (brenhin) gvlat teithiawc would be allowed

through the gate so as to enter the hall: hence the gwladoed ereill would appear to be rulers of

lesser rank than a brenhin or gwlat teithiauc. This in turn makes it likely that gvlat in lines

90–1 and 95 should not be taken in the later sense of ‘major kingdom’, such as Gwynedd or

Powys. The adjective teithiawc was regularly applied to a person or an animal but not to a

country."

Taking this for a -wlad or 'princ, lord, ruler' as the second element of Eliwlad, I began searching for an initial element that would satisfy all the requirements of a close compound name.  El- did not work for reasons I had explored before.  Neither did Eli-. In the end, there were only two words that conceivably could have fronted -(g)wlad.  One of them (eiliw/eilyw, 'appearance, aspect,' etc.) had to be discounted because it broke the language rule in which the first element in a close compound modifies the second, not the other way round. This is a rule of composition and never varies (Rodway).

But, fortunately, we have eiliw/eilyw, 'grief, pain, sadness.'  Eiliw in Old Welsh can be spelled eliw and this ei to e can even happen in Middle Welsh in certain circumstances.  There is no problem with the two /w/s becoming one.  The /g/ of gwlad is, naturally, lost. I have now checked this with several linguists and it has been deemed quite allowable.  The most succinct response came from Professor Peter Schrijver, who said merely: "Yes, that works well linguistically."  From other notable Celticists I have:

"Grief-lord is certainly possible." Erich Poppe

"No problem with Eil- > El-, pretonic reduction, there would have been secondary stress on the 2nd syllable.  So, yes, Eiliw + gwlad is fine for the name." Alan James

"I have no objection to that etymology if we are to restrict ourselves to -lad. It's a good fit." Simon Rodway

I have queries out to others and will add to these comments as responses come in.  

If the meaning of the name had been known at some point, it may have contributed to the motif of Arthur's dead nephew as a ghost-eagle in a oak tree.  Certainly, in the parallel story of Lleu as death-eagle in an oak the prevailing sense is sadness or grief.  Worms and rotten flesh fall from Lleu. In the englyn of "Math Son of Mathonwy" we are told that the flowers of Lleu (feathers) make the sky gorddufrych.  This last is an interesting word.  GPC has gloomy, sombre, very dark, dusky; swarthy.  The first part gorddu is very black, very dark, sombre, pitch-black; somewhat black, darkish; dire, very sad. 

Thus the idea that Eliwlad is 'grief-prince' fits the context of "The Dialogue of Arthur and the Eagle."  And we have similar personal names in early Welsh, like Efrddyl (afrddwl), 'sad, joyless, woeful', and Eiliwed, 'lamentation, sadness, grief.'

And so, if I abandon Eliwlad = Eilwlad, where to now?

Well, the only other apparent identification of Uther with Sawyl Benisel comes from the emendation of kawyl to sawyl in the "Marwnat Vthyr Pen."  But while Rodway thinks kawyl is best altered to sawyl solely through the probability of eye-skip versus a greater number of changes to allow for kanwyll, the context of the poem itself clearly shows a marked preference for kanwyll.  We would have a reference to the star in the poem, something used by Geoffrey of Monmouth as the dragon-head star of Uther (see https://mistshadows.blogspot.com/2019/08/httpsmistshadows.html, https://mistshadows.blogspot.com/2020/02/uthers-star-and-comet-of-442-ad.html). 

And this leaves me back where I started: no good pedigree for Arthur stemming from Uther Pendragon.

Unless, of course, I resort to what I'd long held to be true: that Arthur hailed from the Banna/Birdoswald Roman fort on Hadrian's Wall.  This theory, detailed in my THE ARTHUR OF HISTORY, had shown that the majority of important Arthurian sites were on the Wall or up and down Dere Street.  The problem with an Arthur hailing from the Sarmatian Ribchester of Sawyl Benisel is that I had to somehow account for why Arthur's theater of operations was centered much further north.  I could only guess that he had gone to serve militarily on the Wall, perhaps as a mercenary captain, and had then risen to power and eventually put himself in charge of the entire region.  But this was not an entirely satisfactory explanation.  Especially when we consider that both Banna and Camboglanna/Camlan were within the Irthing Valley of the Artenses, the 'People of the Bear.'  The name Arthur was from early on associated with the Welsh word for bear, 'arth.'

Uther Pendragon as the 'Terrible Pendragon', if belonging to Banna, would be a reference to the Dacian presence there, with their draco standard.  And I may even have been able to prove that Ceidio belonged there.  He was my earlier candidate for Arthur (see https://mistshadows.blogspot.com/2020/01/ceidio-son-of-arthwys-and-powcady.html).  Ceidio is a hypocoristic form of a name which would have meant 'Battle-leader' and this could well represent a designation for Arthur that became the dux bellorum of Nennius.

While I would prefer a real name for Arthur's father and something more than the placement of him at Banna due to that fort's traditional use of the draco standard, I do have to take into account the fact that Arthur's death site (Castlesteads), burial site (Burgh By Sands) and even a possible Grail Castle prototype (Drumburgh) are all to be found on the western end of the Wall. Archaeology also supports the idea that someone powerful was ruling from there at the right time.  

My plan right now is to unpublish THE BATTLE-LEADER OF RIBCHESTER and to reinstate the earlier edition of THE ARTHUR OF HISTORY.  This last will stand as my final say on the historical Arthur question.  

I do apologize to my readers.  It is never my intention to mislead and, certainly, I do not want to be seen as 'wishy-washy.'  The problem with true research is that one is always finding out new things.  In addition, if one is properly applying the scientific method, that means experimentation.  And experimentation requires repetition.  I've never flinched from changing my mind when evidence and/or strong deductive argument demand that I do so.  And I have no plans to depart from such an approach.  Sometimes maintaining a high degree of intellectual honesty is a hard row to hoe.  Yet there is only one garden worth working that will help us harvest the truth.