APPENDIX IV: IUSAY SON OF CEREDIG

Many years ago I floated the idea that Iusay, son of Ceredig son of Cunedda, may be a form of the family/tribal designation Gewissae or Gewissei. While a proposed relationship between these names was not well-received (or, rather, for the most part ignored!), I would like to briefly revisit the possibility here.

The forms Gewissei and Gewissae are attested (see Richard Coates "On some controversy surrounding Gewissae / Gewissei, Cerdic and Ceawlin").

The later Welsh forms Iwys or Iwis for the Gewissae would appear to derive from the Anglo-Saxon form of this word.  Simon Rodway has confirmed for me that "Iwys is the Welsh form of Gewissae (Armes Prydein, ed. Ifor Williams, English version by Rachel Bromwich (Cardiff, 1972), pp. 49-50)."

Alfred is king of the "giuoys", i.e. Gewissae, in Welsh Annal entry AD 900.  Asser says in his LIFE OF ALFRED: "Cerdic, who was the son of Elesa, who was the son of Geuuis, from whom the Britons name all that nation Geguuis [Gewissae]."

Iusay (variant Usai) has not been successfully etymologized by the Celtic linguists.  Recently, I sent queries to several, all of whom were forced to admit that they could not come up with an acceptable derivation.  I myself have tried everything I could think of, including Classical and Biblical names. This attempt ended in failure.  Although there are some forms of Biblical names as recorded in Irish texts (like Usai), the initial /I-/ of Iusay prohibits us from identifying such with the Welsh name.  A Ius- might suggest a Roman name like Justus, but then we cannot account for the ending of Iusay/Usai.

Of course, it is possible Iusay and Usai are corrupt or that they represent some Welsh mangling of an Irish name. Neither I nor the language experts have been able to find such an Irish analog.  This is not to say it does not exist, merely that we have been unable to find it.

All of which brings me back to this:

I have shown in previous research that the Anglo-Saxon Chronicle's Cerdic is Ceredig son of Cunedda, that the same source's Cynric is Cunorix son of Cunedda (as Maquicoline) and that Ceawlin, supposed son or successor of Cynric is, in fact, Cunedda (Maquicoline).  Sisam and Dumville have aptly proven that Elesa (= the metathesis Esla) is a borrowing from the Bernician pedigree.  Omitting Elesa, then, permits us to see Gewis, eponym of the Gewissei/Gewissae, as the immediate ancestor of Cerdic/Ceredig.  As the genealogy in the ASC in the main runs backwards, it may be that Gewis/Gewissae/Gewissei is properly the son of Ceredig.

If so, we might be able to account for Iusay after all.  It is well known that the /G-/ of Gewis or Gewissei/Gewissae came to be pronounced as a /Y-/.  This is what accounts for the Welsh forms beginning in /I-/.  /W/ and /U/ regularly substitute for each other, especially when going from Welsh to Latin (cf. gwyn and guin).  If the terminal diphthong in Iusay/Usai represents the /-ei/-ae/ of Gewissae/Gewissei, then we need only allow for a lost medial small vowel /-i-/.  Iusay would then be a Welsh form of not Gewis, but of the group designation Gewissae/Gewissei.

I feel this is a rather elegant solution to the problem posed by the name Iusay.  Additional support of this idea has come from the following top Celticists:

From Professor Oliver Padel Honorary Research Fellow in the Department of Anglo-Saxon, Norse and Celtic, University of Cambridge - 

"In fact I think your suggestion is not only ingenious but also quite convincing. The only difficult bit, I suppose, is how a tribal name came to be thought of as an individual personal name.

The I- for OE Ge- is fine, of course; as for its loss (Iu- becoming U-),  one might think of the wider Welsh loss of I- in words beginning Iu-, such that original iudd (`lord') became udd (I'm using Modern Welsh spellings for clarity), and personal names containing that word as an element did likewise. (You will find details in Jackson's Language & History in Early Britain -- sorry I haven't got it to hand)."

From Dr Ben Guy, Research Associate, Latin Lives of the Welsh Saints Project, Department of Anglo-Saxon, Norse and Celtic, University of Cambridge -

"Your email was forwarded to me by Professor Russell, because I specialise in early Welsh genealogies (I completed a PhD on the subject last year). I'm happy to help if I'm able.

I think you're right that no etymology has been proposed for 'Iusay/Usai' before. What you propose is certainly an intriguing suggestion, but I think that you may encounter a couple of difficulties with it. Firstly, as you point out below, there appears to be one too few minims in Iusay for it to equate to Gewisse/Iwys. Welsh forms of Gewisse, of which the best known is in Armes Prydein Vawr, always appear as Iwis or Iwys (compare the examples listed in GPC online). There are also earlier forms that point to the same thing: 'Giuoys' in Annales Cambriae A, s.a. 899, and Asser's 'Geguuis'. But as you suggest, this is not an insurmountable problem - though the loss would be more readily explained on a palaeographical rather than phonological level. The greater problem is the '-ay/-ai' ending. Comparable endings appear in the English forms because they survive in Latinate contexts - chiefly Bede's nominative plural form 'Geuissae' and a genitive plural 'Gewisorum' (implying a Latin nom. pl. 'Gewisi') in some Anglo-Saxon charters (as mentioned in the Keynes-Lapidge Asser book, p. 229). I don't think that that kind of ending would be expected in an OE context, and it certainly wouldn't in Welsh - GPC takes Iwys as a plural or collective noun whose ending has been influenced by the plural noun ending -wys (< Lat. -enses) found in words like 'Gwennwys'. So in other words, for your proposed derivation to work, Iusay would have to be a version of a Latinate form such as Bede's 'Geuissae'. The question of how that got into the Ceredigion genealogy in the form 'Iusay' would then be all the more complex, and wouldn't be solely a matter of linguistics! That's not to say that you're necessarily incorrect, of course, but it would require a more elaborate, and therefore more speculative, theory of derivation.

There is one further thing you might consider though, if you wanted to pursue this further: the genealogy of St Cadog. This survives in two versions, one appended to the Life of St Cadog, the other in the Jesus College 20 genealogies. The former calls Cadog's great-grandfather 'Solor', the latter 'Filur'. Both of these names were probably copied ultimately from 'Silur' or the like. Given where St Cadog's cult centre is (Llancarfan), this can't be anything other than a representation of the pre-Roman tribe 'Silures', who were resident in that area. But the form 'Silur' is not the result of regular linguistic development from the 1st century AD; it is a form taken at a later stage from a Latin text, with the '-res' ending lopped off. This might help you envisage the kind of process that might have led to a Latinate form such as Bede's 'Geuissae' being included in the Ceredigion pedigree, but one has to make rather more leaps to get there!"

From Professor Doctor Stefan Zimmer, Department of Celtic, University of Bonn -

"Spontaneaously, your idea of interpreting "Iusay" as a W form of OE Gewisse seems quite attractive. One must, of course, check meticulously the palaeographic details. As I am, alas, not a palaeograher myself, I cannot say more. I see no "LINGUISTIC" problems."

From Professor Patrick Sims-Williams, Department of Welsh and Celtic Studies, The University of Wales, Aberystwyth -

"I suppose Ius- is the older form and became Us- like Iustic in Culhwch which becomes Usic. Forms of Gewissae are noted by Williams/Bromwich Armes Prydein pp. xv-xvi. One Welsh form is Iwys, which rhymes as I-wys, and as the diphthong wy can become w, you could get I-ws- which could be written Ius- in Old Welsh and then add  -ae from Latin which almost gets you to Iusay."

From Professor Doctor P.C.H. Schrijver, Department of Languages, Literature and Communication, Celtic, Institute for Cultural Inquiry, University of Utrecht -

"Linguistically, the first thing that comes to mind regarding the initial alternation Usai /Iusay is the pair OW iud, MW udd 'lord' < *iüdd. So OW word-initial j- disappears in front of ü (= MW u). As to your assumption that Iusay may be connected to Gewissae if there is a rule that states that medial -i- is lost, I can tell you that there is indeed such a rule: *wi > ü in non-final syllables (as in *wikanti: > MW ugeint, see my Studies in British Celtic Historical Phonology 159-60). This generates the ü that we need in order to later get rid of the initial j. The only remaining problem is connecting OE Ge- /je/ with OW j-. Barring that, I would say, yes, what you suggest is possible. That still leaves the origin and etymology of the name in the dark (the reconstruction leads to something like *iwissai- or *g/jewissai-), but first things first."

From Professor Doctor Stefan Zimmer, Department of Celtic, University of Bonn -

"Spontaneaously, your idea of interpreting "Iusay" as a W form of OE Gewisse seems quite attractive. One must, of course, check meticulously the palaeographic details. As I am, alas, not a palaeograher myself, I cannot say more. I see no "LINGUISTIC" problems."

From Professor Patrick Sims-Williams, Department of Welsh and Celtic Studies, The University of Wales, Aberystwyth -

"I suppose Ius- is the older form and became Us- like Iustic in Culhwch which becomes Usic. Forms of Gewissae are noted by Williams/Bromwich Armes Prydein pp. xv-xvi. One Welsh form is Iwys, which rhymes as I-wys, and as the diphthong wy can become w, you could get I-ws- which could be written Ius- in Old Welsh and then add  -ae from Latin which almost gets you to Iusay."