Cynglas: the 'Tawny Butcher'?

clearing up a minor point of dark ages British history

Chris Stewart

One of the many candidates for an historical ‘Arthur’ is a fifth/sixth  century  prince of Rhos in Gwynedd who is named in Welsh  king lists as Cynglas or Cuneglasus. The British monk Gildas speaks about this character in less than favourable terms  in his De Excidio Britannia. He indicates that Cynglas used the nickname “Bear” and also, rather obscurely, calls him “driver of the chariot in the lair of the bear” (commonly taken to mean Din Arth in North Wales) . Together with other compelling biographical convergences,  Mark Devere Davis www.angelfire.com/md/devere/urse.html  cites this as  evidence for identifying Cynglas with Arthur, a name which probably means ‘the bear-man’ in Celtic, and can be similarly derived from the  Latin star name Arcturus – the charioteer or drover of  the Great Bear constellation. 

However there is a notable problem with Gildas’ text at this point. The sentences about Cynglas are part of  a longer passage which denounces a series of contemporary native rulers as immoral usurpers or 'tyrants'. In each case Gildas gives a  Latin translation of their given names. For Cynglas, Gildas uses the Latinate name-form  Cuneglasus, putting it in the vocative case -  Cuneglase - because he is addressing this person directly with his diatribe; (the presumption being that Cuneglasus is a living contemporary of Gildas);  he then writes: “in lingua Romana lanio fulvus” – “in the Roman tongue: tawny butcher”. The problem here is that Cynglas, if it is read as a  Welsh name, does not mean “tawny butcher” but appears to mean “blue dog”!

So has Gildas simply got it wrong? That would be the obvious conclusion. But this has consequences. How reliable does this make Gildas in other matters? It is generally recognized that De Excidio Britannia cannot be regarded as historically precise with regard to dates and events, nor indeed can any dark age text, which is largely why we call this period ‘dark’. But more specifically, if Gildas cannot even interpret his name right, can we rely on the rest of the personal information Gildas gives about Cynglas? It has even been argued on the basis of this perceived error that the De Excidio is either a forgery, or that it is the theologically  polemical work of an exclusively Latin speaking  Roman cleric -  either someone living on the continent of Europe or a foreign resident of Britain  who belonged to a ‘Romanizing’ religious party as opposed to a supposedly native ‘celtic Church’.

The whole question of the so called ‘Celtic church’ can be dismissed as irrelevant, first because there never was  any  ‘celtic’ racial consciousness at this period, such a thing is a very modern phenomenon. In fact no resident of these islands from the 3^rd century BC to the 16^th century AD would have used, or even understood, the term ‘celt’ about themselves. There does appear to have been periodic British insular political organization. In later centuries it was mostly aimed at taking over the Roman Empire rather than rejecting it. Under Vortigern, it is true, there seems to have been an attempt to block re-annexation by Rome, but there is no serious evidence that this went hand in hand with any ecclesiastical self identification as ‘celtic’ or of British Christians seeing themselves  as distinct from Christendom in general. If anything Vortigern might have been a Pelagian in theological outlook, but that is quite a different matter. 

Secondly the separation between insular Christians and Rome has been grossly exaggerated, usually by those who see history through the eyes of the 16^th century quarrels of the Reformation. There was a gradual, accidental loosening of contact between Britain and the continent after the pagan invasions, which made for differing disciplinary and liturgical development, but it was never total. Patrick of Ireland made a point of seeking validation for his mission from Gaul and Rome, and Columba of  Iona exchanged gifts with the Pope. There never really was  a distinct ecclesiastical body which could be identified as a ‘Celtic church’ except perhaps during the few decades of confrontation over these minor issues between the mission of Augustine of Canterbury and the surviving native Christians in the seventh century. To attribute such an agenda to Gildas is deeply anachronistic.

Neither is it credible to suggest that there was a sharp cultural split between ‘native speakers’ and ‘Latin speakers’ by the fifth and sixth century. It is true that Gildas refers to Latin as ‘nostra lingua’ – ‘our language’ - but he does this in contradistinction to the Saxon dialect, when he translates the Germanic phrase ‘long ships’ into ‘nave lunge’. He is not making an opposition between Brittonic and Latin.  In fact it would have been quite natural for a sixth century Briton to regard Latin as ‘our’ language. After all the very term used by British residents to  identify themselves was cumbrogii – ‘fellow citizen’ – ie. citizens of the Roman Empire.

The native dialects were still in use, of course, throughout the Roman period, but they probably became quite intertwined with Latin in daily conversation and  they were never the literary language of the thoroughly Romanized culture of Britain. In the centuries after the withdrawal of the legions, Latin gradually became less widespread and early Welsh did become a literary language by the seventh century. However no native Briton, especially an educated one, of the fifth/sixth century (as St.Patrick bears witness) would have regarded Latin as a foreign language. Bi-lingualism would have been the norm not the  exception for anyone living in these islands at this time, rather like English in post-British India.

It  nonetheless remains true that Gildas appears  to get it wrong with Cynglas’ name when translating between its Brittonic and Latin interpretation, and not just mildly wrong.  It is true that Gildas makes liberal use of word play (a common 'celtic' literary  habit) and loves a good pun, but in this case it is difficult to see what possible basis in either Latin or Brittonic there is for any workable pun between Cuneglase and lanio fulvus. Apart from the bare fact that in both cases we have a noun and a colour, the difference between ‘blue dog’ and ‘tawny butcher’ is pretty decisive! Or is it? It is one of those annoying details that teases the mind, or at least my obsessive mind! I cannot be content just to say: ‘he got it wrong’, and move on.  Even if Gildas was not a native speaker -  despite claiming to be a relation of Cynglas and Maelgwyn and the whole royal house of Gwynedd, and of having grown up with these characters -  there were surely plenty of people around at the time (including Cynglas himself) who would instantly have spotted the error and exposed the fraud. So what is going on here?

In order to shed any light on the matter we must first remember that Gildas and Cynglas  did not speak what we call Welsh in its modern form, but the earlier language from which modern Welsh is derived, which scholars generally call Brittonic. The situation may be further complicated in that Gildas lived at a time when the ancient British language must  have been shifting and changing into a form which had become the earliest version of ancient Welsh by the beginning of the following century. What little direct evidence of the sounds and meanings of Brittonic we have come  from  personal names, place names and a few surviving  inscriptions, most of which are filtered through Latinate literary forms. Despite this, much can actually be gleaned from these scattered sources, and with a fair degree of accuracy, about the way in which words have mutated through early into modern Welsh. Then on this basis we can attempt to reconstruct a reasonable amount of the original Brittonic language by a process of back derivation from modern Welsh, Cornish and  Breton. This process remains speculative and provisional, but it is not without foundation, and anyway it is  the best we can do. 

In order to answer our problem with Gildas and the name Cynglas we will need to take a closer look at both his Latin phrase lanio fulvus and the putative Brittonic words which form the name cynglas/Cuneglasus  to see if any reconciliation or at least a rapprochement is possible between them.

Lanio: There is no doubt that lanio means butcher, although it can also mean executioner by association with the sense of maceration and bloody destruction.

Fulvus: This is commonly translated as tawny, meaning a reddish or brownish yellow.  However  fulvus was actually used to describe the colour of sand, lions and even gold. This is more in keeping with the root of the word (ful-) from the verb fulgere which means to shine. So rather than the muddy sounding tawny  it might be more accurate to translate fulvus as sandy or even, at a pinch, golden; perhaps the simplest modern translation would just be to say orange. Certainly it ought to mean a light and lustrous shade of yellow-red. This might appear to be nit-picking, but as we shall see shortly it is actually an important linguistic point, not because of  any tenuous concern over the precise shade of colour we might be talking about, but because the concept of fulvus as a colour comes from the concept of reflected light. The relevance of this will become clear below.

Cyn: is a common element of early Welsh names and would appear to come from a Brittonic word which has arrived in  modern Welsh as cy; the Celtic word being cognate with the Latin cane, and even more closely with the Greek kynos – all of which mean dog.  

However there is another and perhaps better candidate for the meaning of the early Welsh name prefix Cyn(e)- ; that is the Welsh word cun , (readily interchangeable with cyn) now considered archaic and obsolete, which means lord or chieftan.  This appear the most likely meaning and much more credible than accepting the cyn = dog etymology. Although the use of  animal names, especially with sacred overtones,  is not unlikely in Celtic culture,  the epithet 'chieftan' or possibly ‘warrior’  makes far more sense as  a name form - especially one that crops up so frequently in King lists. We should note here too that the word cun is also found in the (probably even more archaic form) cuniaid, which would imply a Brittonic form such as *cuniato. This longer form of the word could then explain the hanging vowel Cune- in older forms the name  Cynglas/Cuneglasus, which would come from an abbreviated prefix from *cuni- (ato).

Glas: is the Welsh for blue. It is similarly cognate with the Latin: glaucus  and Greek: glaukos which strictly speaking mean sea-blue, or grey-blue, the colour of blue eyes. The linguistic interchange is even easier to see when we understand that in British Latin the letter ‘c’ was normally pronounced as an ‘s’; so glaucus would become ‘*glausus’ in ordinary speech, the equivalent Brittonic probably being *glasos.

Therefore the name written out in Latin form as Cuneglasus could  indeed  mean Bluedog, or even perhaps Greyhound, although it would more probably appear to mean  Blue Chieftan. In any event where then did Gildas get Orange Butcher from? It seems to bear no relation to either of these possibilities. But there must surely be some basis even for a successful pun or even for an error?

The mystery seemed intractable until, while looking up something else in a standard Latin dictionary, I came across the word glaesum which means amber. Here was a highly possible basis for confusion or word play, at least with regard to the colours. Is this what Gildas was hearing (or choosing to hear) in the name ending – glasus? The word glaesum is not the name of a colour as such, it refers to the semi-precious gem itself, but certainly the reddish yellow luster of amber would make it a perfect candidate for using the Latin adjective fulvus to describe it.

It seems highly unlikely that there was no native word for this highly prized jewel and it seems very probable  that it would be similar to the Latin. Something that sounded rather like  *glesu would fit the bill very credibly.   The modern Welsh for amber is gwefr,  which is conceivably a distant mutation or corruption of something like *glesu. We may also note in passing that the Latin/British name for the town of Gloucester was Glevum signifying "shining" city. This probably refers originally to a vitrified hill fort, so it can be also interpreted as "glass castle" (an image  which has passed into medieval Arthurian legend ). The word glevum is clearly of the same root we are discussing and although the town name later evolved into Glouiu caestr hence Gloucester it could also easily be cognate with gwefr too. The point being that this gl- word stem seems to be readily adaptable to mean anything shining or glassy in a variety of colours or none.

The similarity of the Latin words glaucus and glaesum would be made even greater if the British forms of  these words had in effect dropped or swallowed the (s) and (m) endings to give something like *glauso’ and *glaesu’. Since we can expect the corresponding Brittonic forms of these words to have been very close to the Latin they would have been easily interchangeable too. We must remember that Brittonic was never a written language, so spoken forms would be the only source for linguistic analysis, both for Gildas and ourselves. Unlike Gildas, we, of course,  do not know anything about the pronunciation of Brittonic, which must have been subject to numerous dialectic variations, just as spoken Welsh still is. So the sounds of the words for amber and sea-blue may have been very close indeed. We can at least suggest with confidence that they were substantially similar.

All of this is highly suggestive although, it must be admitted, inconclusive, but I then stumbled on a remarkable and decisive piece of evidence in Tacitus’ first century account of the tribes of Germania, where he writes:

“So turning to the right hand shore of the Suebian sea, we find it washing the country of the Aestii, who have the same customs and fashions as the Suebi, but a language more like the British. They worship the Mother of the Gods, and wear, as an emblem of this cult, the device of a wild boar, which stands them in stead of armor or human protection and gives the worshiper a sense of security even among his enemies. … They also ransack the sea. They are the only people who collect amber - glaesum is their own word for it - in the shallows or even on the beach.” (Tacitus De Origine et Situ Germanorum, Chapter 45 emphasis added)

Apart from the fascinating glimpse of yet another continental  tribe of Celtic cousins living on the Baltic shores, whose language is specifically compared to the British rather than say the Gauls other continental Celts,  Tacitus is telling us clearly that  the Latin word glaesum is in fact a loan word from a p-Celtic original! Not only can we be pretty sure therefore that there was a corresponding Brittonic word for amber which sounded something very like glaesum, but this is originally an actual  Celtic word which has been adopted into Latin, presumably receiving the neuter ending  – um in the process, as is common with most loan words.

Furthermore several etymological dictionaries confirm that  glaesum belongs to a  group of Indo-European words which generally have the meaning of something that shines or reflects; it is this word group which gives rise to the English nouns ‘glass’ and ‘glaze’for example. We can also see the root meaning of the term being carried faithfully across linguistic mutations. The noun forms which retain the  ‘gls’ root in various languages are nearly always associated with verb forms of a ‘gl(a)’ type which similarly mean to shine or be bright.  In fact our English word  ‘glow’ is a direct cognate of this same root. This  would make Gildas’ choice of  fulvus (with its root in the Latin word to glow or shine - fulgere) extremely apt as a translation of the Celtic word glaesu(m) – something yellow-red that shines.

Finally we come full circle when we learn that this same word group includes the Greek and Celtic forms which originally  indicated the shining surface of the sea, (a different colour but the same idea) - glaukos/glasos – which we have already examined.  So the Welsh glas is from the exact same root as glaesum and both are well attested Celtic/Brittonic words. It should be no surprise that then that the word for shining amber and the word for grey-blue could end up looking the same when incorporated as a name suffix, and doubtless they sounded very similar too.

I believe that we have now shown convincingly  that there is not only a similarity  but a direct linguistic relationship between the words for sea-blue and amber in Celtic. So which should we choose in interpreting the name Cynglas? It looks increasingly likely that Gildas could actually be right, at least about the colour. He was, after all, in a better position to know which ‘gls’ root word was being used in this name which is contemporary to himself but not to us.  At the very least he is obviously not talking nonsense.

If  indeed the '-glas' element in this name does mean amber, rather than blue, then why did Gildas not simply translate it into Latin as glaesum? Surely this would have been the simplest thing to have done,  and it might have avoided much  later confusion too! One could argue that he avoids a virtual  transliteration from -glasus to glaesum because glaesum is not actually the standard Latin for amber, the more usual word being electrum and sometimes even sucinus; glaesum as we have seen being of Celtic origin in the first place.  In this case then  why does he not translate it as electrum or sucinus? It would make sense if he is not trying to indicate the resinous  product of amber as an object, but the colour of amber which also carries a strong sense of the quality of shining, glowing or radiating with coloured light. To capture this nuance, his choice of fulvus would not only be apt,  but shows a man with  perfect command of both Celtic and Latin linguistics. 

If on the other hand we hold to  the idea that he is indulging in punning and insulting word play, the avoidance of glaesum, electrum or sicinus and the use of fulvus could be a deliberate ploy to play down the bejewelled imagery of  polished amber and switch to  the more humble and prosaic idea of something just being orange, sandy or tawny coloured. Alternatively, if we  wish to keep the original meaning of -glas as blue rather than amber  (remembering that either appears credible on linguistic grounds),  then a punning move to orange/tawny still remains understandable in terms of a similarity of sound  between two Brittonic words.  For the verbal trick to work in this case Gildas would be expecting his readers first to follow the aural pun between glas/blue and glaes/amber in Brittonic, and then further to understand his prosaic substitution of  fulvus/orange for amber when putting the pun into Latin. Such complexity is possible of course, but it must be said that this option now looks  less attractive in terms of conceptual and verbal wit. In any event, contrary to what has been the received perception of him, Gildas begins to appear as an excellent translator from Brittonic to Latin, and his choice of colour adjective here is quite justifiable whether as a pun, a carefully nuanced translation, or maybe a little of both.

But what about the substitution of butcher for dog or more likely  chieftan ? Here too we need to know if there is any linguistic or at least aural convergence or possibility of  confusion in ancient British. Once again we must be cautious and realistic about the limits and pitfalls of linguistic reconstruction, but we may legitimately look for a basis on which to proceed.

We have already looked at the word dog,  which would be something like *cyno’ in late Brittonic, and also the word cun, cune or cuniaid which may have evolved from *cuni-/*cuniato meaning chieftan or lord.  The modern Welsh for butcher is cigydd ;  projection into Brittonic would seem to give *cigud or *cigut. We also find that the Cornish equivalent word for butcher is kyger.  This in turn could  indicate a Brittonic form such as *ciget, or possibly *cinget. The  ‘cig/kyg’ sound can easily be a contraction of a nasal ‘cing/kyng’ sound. Also the final consonant (t/d) having been softened in Welsh, and in Cornish to the point of virtual disappearance, a good case can be made for its softening and elision already in late Brittonic, which could yield a word such as *cingeth or even *cinge’ .

All of this is speculative, of course, although not wildly so, but whichever way we work out the Brittonic forms once again the two outcomes – dog and butcher  – do not appear to be  so very far from each other in sonic terms. However  if this was all the we had to go on, it must be admitted that our case would remain somewhat strained and tenuous. But once again, after having worked out the speculative possibilities above, I found another remarkable piece of solid evidence buried in well attested linguistic history. McBain’s Etymological Dictionary of Celtic Languages interprets the famous Gaulish and British chieftan name Vercingetorix as follows:

ver   -   cingeto  -  rix:

over  -  warrior  -  king       ie. Supreme Warrior King.

The Gaulish word cingeto (this is not really an 'invented' word, it clearly is an element of  a well known name) McBain cites as meaning warrior. This word would  actually be another  acceptable alternative derivation of the common  prefix cyn- or cyne- in British and Welsh names. It would also be a possible fit for an early  Brittonic form of the Welsh cigydd and the Cornish kiger. The fact that  these later forms have come to mean ‘butcher’ rather than ‘warrior’ need be no great difficulty.  The same word would happily suffice for both concepts in their respective contexts. The idea of a warrior as a ‘butcher’, and a butcher as a somewhat tamed warrior, is perfectly credible on a cultural and linguistic level.

When we remember that the Latin lanio can also means executioner by extension,  once again Gildas’ choice of words appears highly apt, especially as we know that  he is not exactly trying to be complementary about this man. In fact he makes it very clear that he regards this warrior prince who calls himself ‘Bear’ as a blood stained destroyer of life and limb. It seems quite credible that the one word-form may  have borne a double sense - warrior/butcher - in Gildas’ time, and may even have had a common bloody association way back in its Gaulish roots -  the idea of a warrior as one who dismembers bodies with his blade With the later demise of the warrior classes the word simply retained its ‘butchery’ significance and mutated into the modern Welsh and Cornish forms noted above. 

Lest this profusion of possibilities is getting confusing it is worth reminding ourselves that *cyn (?dog), *cune (?chief), and *cinge (?warrior and/or ?butcher)  are the three late Brittonic/early Welsh forms we are juggling with as candidates for the first element of the name Cynglas.  But we could simplify things even further if we suggest that the word cun (?chief)  in its older form cuniaid <*cuniato might itself have come from the ancient root word cingeto/warrior, (by the contraction of the ng to n and the loss of the inflected ending). If this is so then the verbal relationship to butcher remains the same as before - chieftan, warrior and butcher  all being cognates of the same root  in Celtic. However, if this is thought to be stretching linguistics a bit far, then  a close pun certainly remains possible between *cuni(ato)/chieftan and *cinge(to)/butcher/warrior, just as it does between *cyno/dog and *cinge/butcher.

It is more obvious with this second element of the name that  some sort of word play is in fact going on when Gildas translates Cyne as lanio - butcher, since this is clearly a calculated insult and highly unlikely to be the real meaning of the man's  name. If  we  take the Cyn(g)e- element to mean  warrior originally,  then it seems that Gildas may only have had  to put the less noble sounding spin on the same root  to interpret it as butcher and achieve his desired effect. If on the other hand we take the Cyn(e)- element to mean chieftan then a Brittonic pun with the word for butcher is likewise readily available, and may also be from the same root,  so the same verbal process may apply. If however we still insist on Cyne -  meaning dog, then, as with the colours,  the pun is still possible on the basis of the sonic convergence of  two Brittonic words - *cyno and *cyn(g)e, - but, as before, this option looks  somewhat  arbitrary and a more strained  in conceptual terms. 

By way of  summary,  if we  finally try to back derive lanio fulvus into a putative Brittonic phrase, we arrive at what may have sounded something like:

*cin(g)e’ glaesu (amber butcher)  

which is very close indeed to the name which, if it is read as meaning Amber Chieftan or Warrior would have sounded like:

*Cyneglaesu.

Or even if we still wish to interpret it as Blue Dog it ought to have sounded more like:

*Cynoglaso.

The fact that the actual name is written in Gildas text in Latin format as Cuneglase  and not Cunoglase is more suggestive of the former being correct  than the latter. The final -e is simply the Latin vocative of course, but the middle -e- is not accounted for this way and is probably not an inflection but an integral part of the original word element. All of which lends weight to our preference for Amber Warrior/butcher. The form Cynglas which is found in the regnal lists of later centuries doubtless reflects a contraction of the name, which had remained in use in this early Welsh style, having lost the inflected ending of Brittonic. At the end of the day, whichever sound and meaning we project as the original form (and we will never know for sure), we can see now that either interpretation is completely plausible, and therefore Gildas' Latin translation/pun is not nearly as surprising or unjustified as it first appears.

What does all this imply?   Gildas' penchant for making insulting puns out of the names of his personal targets makes it very likely that he is punning here too. The question then is: what is the linguistic basis of the pun? For our hero to be  known to contemporaries as ‘Amber Chieftan’ (perhaps even bearing the associative sense of ‘Shining Warrior’) is not at all impossible, and I would contend that this is a perfectly justifiable interpretation of the name Cynglas on the linguistic grounds outlined above. The pun Gildas makes of it,  in that case,  is to modify the name  'Amber Warrior' into  'Orange Butcher'. This not only works well as a short step on linguistic grounds, but we can see even in English how, just by slightly shifting the emphasis of  words, this works as a biting and belittling insult .   We can readily understand how using the term 'butcher' as an alternative term for a warrior is insulting; it would also fit with Cynglas' otherwise attested reputation as a fearsome and bloodthirsty soldier (he is sometimes listed as bearing the additional descriptive name goch – red) who, according to Gildas had turned his ruthless martial tactics to the pursuit of civil war among his fellow Christians ; and the term 'orange', 'sandy' or 'tawny' may  refer disparagingly to the colour of his hair, rather than evoking the royal gleam of amber gems.

Of course it may still be true that the name Cynglas originally meant Blue Dog as it appears to us at first sight, but even if this is so, on the evidence we have explored above, this would still not mean  that Gildas was confused or mistaken about the etymology of the name Cuneglasus. When dealing with the other ‘tyrants’ of Britain, Gildas makes his  puns  by twisting the Latin sounds of their names. With Cynglas it looks as though he  may be doing this first in Brittonic, then translating the pun into Latin.  Whilst saying ‘orange butcher’ instead of  ‘blue dog’ obviously does not work as a pun either in English or  Latin, it would appear quite possible from the sheer sonic similarity that it does make an effective Brittonic one. 

I would still favour the idea that he is punning from 'amber chieftan' or possibly 'warrior' to 'orange butcher'. The move is neater both as a verbal pun and as an insult: emptying out any sense of nobility and heroism from the name and reputation of a proud Celtic warlord by a subtle shift of usage. In fact this particular shift of emphasis is probably only possible when using the interplay of the two languages by making a deliberately clumsy and 'inappropriate' (although not strictly inaccurate) translation from Brittonic into Latin. But either way, if there is any value to these ideas, then Gildas is vindicated as genuine in his knowledge of Brittonic as well as Latin. In fact,  not surprisingly,  he seems to have a better grasp of the two than we do, plus an inside knowledge of his contemporaries and his times which we cannot hope to share and which no doubt prevents us from fully  appreciating his rather convoluted sense of humour.

cynglas@hotmail.com