Who was rhain the Irishman?

Who was Rhain the Irishman?

Department of History, Bil/tent University, Ankara

According to the medieval Welsh chronicles, a pretender called Reyn Scotus or Rein Yscot, that is ‘Reyn. the Irishman’, appeared in south Wales in 1022 claiming to be the son of king Maredudd ab Owain (d. 999), and for a time succeeded in establishing himself in the kingship of Deheubarth, apparently with the approval of the southern Welsh, until he was defeated at Abergwili by Llywelyn ap Seisyll of Gwynedd later in the same year. Scholarly opinion among Welsh historians regarding the claims made by Reyn. have varied somewhat, though to my knowledge no one has accepted them in their entirety. J. E. Lloyd, for example, referred to Reyn simply as an ‘Irish pretender . . . who claimed to be son of Maredudd’,1 whereas others have been less certain, suggesting, for exam-ple, that he was perhaps of ‘a mixed Welsh, Irish and Hiberno-Scandinavian descent’.2 However, recently, and approaching the problem from the Irish side of the equation, Sean Duffy has tentatively posited the ‘possibility’ that Rein. Yscot may be identical with the Irish dynast Roen (or Raen), king of Meath, who died in 1027.3 In this article, I shall argue that a detailed reading of the Welsh sources indicates that there are strong grounds for arguing quite the contrary: that this enigmatic ‘Rhain the Irishman’ was in fact a Welsh dynast and may very well have been the son of Maredudd ab Owain, as he claimed to be.

The starting-point for any consideration ofReyn Scotus must be thefive extant medieval accounts of his deeds — two Cambro-Latin and three Middle Welsh — in the light of an analysis of the textual history of the relevant chronicles. The two surviving versions of the Annales Cambriae (=AC) describe the events of 1022 as follows:4

AC (B): Reyn Scotus mentitus est se esse filium Mareduc qui optinuit Dextrales Britones; quem Seisil rex Venedocie in Hostio Guili expugnauit, et occisus est Reyn. Eilaf uastauit Demetiam. Meneuia fracta est.

1 J. E. Lloyd, A History of Walesfi‘om the Earliest Times

to the Edwardian Conquest, 2 vols., 3rd edn. (London, 1939), i, 347; cf. A. H. Williams, An Introduction to the History of Wales, 2 vols. (Cardiff, 1941—8), i, 167.

2 David Moore, ‘Grulfudd ap Cynan and the

medieval Welsh polity’, in K. L. Maund (ed.), Gruffudd a]; Cynan: A Collaborative Biography (Woodbridge, 1997), 1—59 (p. 29); note also K. L. Maund, Ireland, Wales, and England in the Eleventh. Century (Woodbridge, 1991), 160. Thus, the author (1010 Morganwg?) of the dubious Brut leuan. Breolzfa repre— sented ‘ln' as son of Maredudd ab Owain by an

Irish concubine (‘0 Ysgottes a fu’n oi‘dderch iddaw’):

The h/vyrian Arehaiology of Wales: Collected out of

Ancient Manuscripts, ed. Owen Jones et al., 2nd edn. (Denbigh, 1870), 718.

3 Sean Duffy, ‘Ostmen, Irish and Welsh in the eleventh century’, Peritia: journal of the Medieval Academy of Ireland, 9 (1995), 378—96 (p. 383).

4 London, PRO, E164/1, p. 11 (=B-text); London, British Library, MS Cotton Domitian A.1, f0. 141" (=C-text). For a conflated and imperfect edn. of the relevant texts, see Annales Canzbriae, ed. J. Williams ab Ithel (London, 1860), 23 and n. 2; in addition, the two annals were also printed in Thomas Jones, ‘Historical writing in medieval Welsh’, Scottish Studies, 12 (1968), 15—27 (1). 26).

132 DAVID E. THORNTON

AC (C): Leuuelin filius Seisill rex Venedocie pugnauit contra Reyn, qui se dicebat esse filium Maredut; et deuictus est Reyn in Ostilo Guili. Eilaph uenit in Britanniam et uastauit Dyuet et Meneuiam.

The annal for 1022 falls into two parts: first, the deeds of Reyn and the battle of Ostimn Guili (confluence of the Gwili); and secondly, the raid by Eilaf against Dyfed and (pos— sibly) also St David’s. Neither text gives us any reason to assume that these two (or maybe three) episodes were connected in any way, and 1 am inclined therefore to regard them as separate events during the single year.5 At first glance, the wording of the two accounts seems quite different, but there are a few verbal echoes which maybe indicate a com— mon source.6

There are some important differences between these two short accounts of the deeds of Reyn. The version in the B—text reads rather clumsily: for instance it is not clear whether the grammatical subject of the phrase qui optimtit Dextmles Britones (for which, it should be noted, there is no equivalent in C) is meant to be Reyn. or Mmeclm. More obviously we have the statement in the B-text that it was not Llywelyn ap Seisyll who fought Reyn but a Seisyll: given what we know of the rule of Llywelyn ap Seisyll (d. 1023), this looks to be an accidental scribal omission rather than afirm belief that Reyn’s opponent was called Seisyll. Other differences look less like mere errors during copy— ing. The B—text, for example, stresses that the claim of this Reyn. Scams to be son of Maredudd was false and states that he was killed; whereas the C—text, which omits the epithet Scams and any reference to Deheubarth, states that Reyn said he was the son of Maredudd (without commenting on the truth of this claim) and simply has him being defeated, but not necessarily killed, by Llywelyn. These differences no doubt reflect the wider textual variations between these two later versions of the A'neialesz7 though both derive ultimately from the annals kept at St David’s since the late eighth century, the B—text (probably copied in the early fourteenth century at Neath) was not based on the ‘Annals of St Davids’ directly but rather on a (now lost) version of these annals drawn up at Strata Florida 0. 1202 (the so-called ‘Annals of Strata Florida’) and that additions to the text were made during transmission, possibly before and/or after 1202. Furthermore, while the C-text was indeed redacted at St David’s c. 1288, to what extent it represents a complete version of the ‘Annals of St Davids’ at that date is not clear.

5 The Eilaf named here is not, as some have sug— gested, Olaf Sihtricson of Dublin (and therefore a pos— sible Hiberno—Scandinavian ally of Reyn) but rather the Anglo—Scandinavian Eilifi‘ borgilsson, earl of part of Mercia under Cnut, who would become involved in the coalition against Cnut in 1026: see Simon Keynes, ‘Cnut’s earls’, in A. R. Rumble (ed), The Reign of Cnut: King of England, Denmark (”Ml Nam/(1y (Leicester, 1994), 43—88 (esp. pp. 58—60). Eilaf is also mentioned in the Welsh chronicles s.a. 1035, when he is said to have departed for Gei’iiitlxizia following the death of Cnut in that year.

6 The phrase se . . . essefilium Maredut and the use of the verbs expugno (overcome, defeat) and [mgno (fight, disagree) may indicate that the two texts are in some way related.

7 See Kathleen Hughes, Celtic Britain. in. the Early Middle Ages: Studies in. Scottish and Welsh Sources, ed. D. N. Dumville (VVoodl)ridge, 1980), and discussed in David N. Dumville, Histories and Pseudo-Histories of the Insular Middle Ages (Aldershot, 1990), essay 111. The work of Hughes and Dumville 0n the chronicles is currently undergoing some revision by julian Harrison. I am very grateful to Mr Harrison for dis-cussing his research with me, some of which was pre-sented in a paper, ‘The origins of Brut. y Tywysogyonfi delivered at Jesus College, Oxford, on 1 February 1996; for now, see Julian Harrison, ‘A note on Gerald of Wales and Avixnales Cambriue’, WHR, 17 (1994—5), 252~55 (pp. 25443).

133 Consequently, notable differences between the texts (such as those outlined above for 1022) may either represent additions made to the ‘Annals of St Davids’ during the process of transmission which culminated in the B—text, or may be due to losses incurred during the compilation of the C—text. These and other textual points are crucial to how we interpret of the events of 1022, and I shall return to them below.

Additional accounts of the events of 1022 are to be found in the two extant versions of Brut y Tywysogyon (=ByT) and the related Brenhinecld y Saesson. (=ByS).8 The accounts of 1022 in these three Middle Welsh chronicles are notable for their length (not only when compared with the Latin equivalents, but also with other annal—entries at this chronological point in the chronicles); and they possess what has been described as a more ‘literary’ character than the Latin versions of these events.9 (On account of their length, the three texts are therefore given at the end of this article.) Comparison of these vernacular texts demonstrates that, despite certain differences including the slightly shorter length of that in ByS,10 they must all derive ultimately from a single account of 1022. As the late Thomas Jones emphasized,11 despite their greater length, these three vernacular texts do not add much more detail to our knowledge of Reyn but rather embellish the briefer material supplied in the extant Latin chronicles with a eulogy on Llywelyn ap Seisyll (including a rather formulaic description of the prosperity under his rule) and a lengthened account of the battle (though, again, this is shorter in Bren/zinedtl y Saesson), both of which serve to praise Llywelyn and by the same token den— igrate Reyn. and his claims. Furthermore, if we compare the instances where the ver-nacular texts (lo relate the same basic information as the two versions of the Annales, it would seem that they are closer to the B~text2 thus, like the B-text (and tml’ihe C), the vernacular texts explicitly deny Reyn’s claim to be son of Maredudd and describe him as Irish; and, in addition, they mention Deheubarth (De/zen), which is equivalent to the Dextrales Britones of the B—text. Only in the case of Reyn’s opponent do the Brntiau, clear— 1y agree with the C— against the B-text of the Annales in giving Llywelyn ap Seisyll rather than the erroneous Seisyll who, I have argued, probably derives from a scribal error here.12 It seems likely therefore that the (lost) account of 1022 from which the three extant vernacular accounts derive was a longer and embellished version of a text close

8 Brnl y Tyugvsogyon. Peniarlh MS. 20, ed. T. Jones (Cardiff, 1941) and Brut y Tywysogyon. or The Chronicle of [he Princes. Peniarlh MS. 20 Version, tr. T. Jones (Cardiff, 1952), so. 1020 = 1022; Bruty Tywysogyon or The Chronicle of the Princes. Red Book of Hergest Version,

ed. and tr. T. Jones, 2nd edn (Cardiff, 1973), so.

[1022]; Bren/zinedd y Saesson or The Kings of the Saxons. BM Cotton MS Cleo/)alra Ba) and The Black Book of Basingwerh NLW MS. 7006, ed. and tr. T. Jones (Cardiff, 1971), so. 1020 = 1022.

9 Jones, ‘Historical writing in medieval Welsh’, 26;

on the exceptional length of the vernacular versions,

note also, Mauncl, Ireland, Wales, and England, 121.

‘0 In addition to the examples discussed below, only Bi'enlnrzedtl y Saesson makes the claim that Rhain also desired Gwynedd against Llywelyn ap Seisyll (y (lamunws ynlev Gwyned yn. erbyn Llywelyn. val) Seissyll).

Also, the Red Book Brut and marginally in the Pen.

20 version, the pillaging following Llywelyn’s victory is said to have extended hyt y Mars, ‘all the way to the March’, which is evidently a post—conquest addition.

‘1 Jones, ‘Historical writing in medieval Welsh’, 26.

12 There is a similar onomastic divergence in these five chronicles for the year 1018: the B-text states that in that year Aeddan ab Blegywryd and his four sons were killed a Grifino filio Lewelin rege Briton'lmz, where—

as the C-text has a Lewelino; and the vernacular

chron-icles, in partial onomastic agreement again with the

C—text, name Llywelyn ap Seisyll as the killer. Again,

the B~text looks to be in error (perhaps erroneously replacing Llywelyn ap Seisyll with his more famous son and successor Grutfudd ap Llywelyn) and again the Bruts follow AC (C) and have the ‘correct’ reading.

134 DAVID E. THORNTON

in detail to that preserved in AC (B). This conclusion for 1022 is supported by our cur-rent understanding of the relationship between these various chronicles overall:13 the common source of the three vernacular texts (of which, the Pen. 20 version is general-ly thought to be the best witness) seems to have been compiled at Strata Florida c. 1300 from the lost Latin ‘Annals of Strata Florida’ (from which, we have seen, the B—text of the Annales was also derived) with additions and embellishments (such as proverbs) being made to the material derived from those annals. Consequently, the longer accounts of 1022 in the vernacular chronicles should be regarded as secondary to the shorter Latin equivalents, and any points of similarity between the B—teXt and the Brutiau, against the C-text, should be treated as possible additions or alterations characteristic of the ‘Annals of Strata Florida’, and not necessarily integral to the original St David’s annals. Again, it is important not to forget these various textual points when employing the chronicles as sources for the events of 1022; and I shall return to this problem below. Having thus presented and discussed the sources for the deeds of Reyn Scotus, it is necessary to deter-mine what, if anything, they can tell us about this enigmatic figure.

The two Latin annals refer to the alleged ‘pretender’ of 1022 using the form Reyn, and the vernacular texts have Rein. These are perfectly acceptable Middle Welsh forms of Old Welsh personal name Regm, modernized as Rhain, and I will use this modern form henceforth.14 The Latin texts simply give this Welsh name without comment, as do the Red Book Brut (a’e heme 'Ult Rein, ‘and his name was Rhain’) and Bren/zinecld y Saesson (a Rein, oed y hemu, ‘and Rhain was his name’); but the Peniarth 20 version is a little more suspicious (a Rem y mymzazud a alw, ‘and he desired to be called Rhain’).15 However, doubts about Rhain’s claims are stronger when we examine his alleged genealogical antecedents. Of the two versions of the Amzales, the C—text states simply that Rhain ‘said himself to be the son of Maredudd’ (se dicebat essefili'zmz Maredut), with— out accepting or rejecting the truth of this claim, whereas the B—text explicitly states that the claim was false (mentitus est se esse filz'um Mareduc). The vernacular chronicles agree with AC (B) in denying this alleged parentage. Neither Latin text identifies the Maredudd in question, and of the Brutiau, only Bren/zmedd. y Saesson calls him Maredudd ab Owain (the other two texts having simply ‘King Maredudd’). The most obvious can-didate for Rhain’s alleged father is Maredudd ab Owain, king of Deheubarth and, for some time at least, of Gwynedd too, who died in 999.16 In the light of the textual

‘3 Again, elements of the standard interpretation of the relationship of the three vernacular chronicles, proposed by Lloyd and endorsed by Jones, that they represent three independent translations of different copies of a lost [min chronicle from Strata Florida describing events down to 1282 is undergoing an amount of revision by Julian Harrison: J. E. Lloyd, ‘The Welsh chronicles’, Proceedings of the British Academy, 14 (1928), 369~91; Brut y Tywysogyon. Peniarth MS. 20 Version, tr. Jones, pp. xxxvi—xxxvii; Jones, ‘Historical writing in medieval Welsh’, 21. For Harrison’s work, see above 11. 7.

14 For example, compare AC (A) 808 Begin. with the form Reyn. given in the C~text for the same year; or the two figures called Regin in §§ 13~14 of the Harleian

genealogies with the equivalent form Rein. in the Jesus College collection (§ 8) and the Book of Llandaff: see Early Welsh Genealogical Tracts, ed. P. C. Bartrum (Cardiff, 1966), 9—11, 45; The Text of the Book of Lian. Dav reproduced from the Gwysa'ney Manuscript, ed. J. G. Evans and J. Rhys (Oxford, 1893), 118, 125, 167.

15 For the equivalent phrase in the Red Book ac y mynnawd y alw ehzm ylt men/tin (and he desired to have himself called king), Jones suggested an earlier scribe had misread regin as regent, and therefore that this ele-ment of doubt was in the underlying Latin text and not simply in the Pen. 20 version.

‘6 For thefigure of Maredudd ab Owain, see David

E. Thornton, ‘Maredudd ab Owain (d. 999): the most famous king ofthe Welsh’, WHR, 18 (1996—7), 567—91.

135 discussion above, it seems safe to assume that such a negative View of Rhain’s parental claims formed part of the annal for 1022 in the lost ‘Annals of Strata Florida’, from which both the B-text and the common source of the Brutiau derive, though it is more dif_{ficult to decide whether it had been added into these annals C. 1202, or had occurred} in the ‘Annals of St Davids’ but was subsequently omitted from the version underlying in the C-teXt.

What then of the claim in the chronicles that, despite his evidently Welsh personal name, Rhain was of Irish origin? Comparison of the Latin texts is revealing here. The C-text (which, we have seen, does not explicitly deny Rhain’s alleged parentage) makes no mention of the supposed Irish connection whatsoever. Thus, only the B-text (which does openly deny the parentage) attributes the epithet Scouts to Rhain; and again, the vernacular texts follow and characteristically elaborate this in calling him nebzm Yscott (a certain Irishman) as well as more specifically Rem Yscott. Thus, the so—called ‘Annals of Strata Florida’ would seem to have contained the epithet. Furthermore, incorporated into the extended and ‘more literary’ account of the battle in the Brutmu is the description of how this Rein. Yscott boastfully incited his troops, promising them victory ‘as is the custom among Irishmen’ (y iizegys y mae moes gtm Ysgotyeid), but subsequently himself retreated at the crucial moment. This is certainly a colourful episode, but must surely be considered one of what Jones termed the ‘studied embellishments’ in the Brutmu — in this case prob— ably inspired by Rhain’s epithet Scams — and should not therefore be treated as additional or independent verification of an Irish origin. Again, however, it is more difficult to deter— mine whether the epithet was part of the ‘Annals of St Davids’ but came to be omitted in the C—text, or whether it was one of the additions characteristic of the ‘Annals of Strata Florida’. While this textual analysis does not prove that the Irish connection must be untrue, it does remind us that we should not automatically assume that it is true simply on account of weight of textual evidence (four texts against one).

The_{final point relating to Rhain is his eventual fate. The various texts appear to agree} about the location of the battle: the Latin accounts place it at Ostz'mn. Guilt, that is, ‘the confluence or mouth of the river Gwili’, which is a translation of the place—name Aber Gwyli. given in the vernacular chronicles. Based on the reference to the kingdom of Deheubarth, then this was perhaps Abergwili, near Carmarthen. The two Latin texts disagree as to Rhain’s fate at the battle: whether he was slain (B—text) or merely defeat~ ed (C—text). As Thomas Jones pointed out, the vernacular accounts would seem to reflect the consequent uncertainty in giving the compromise statement that ‘he was never seen again’.17 If so, this represents one instance for the annal of 1022 when the vernacular Brutiau do not follow the reading of AC (B) against that of the C-text, though of course they do not exactly follow the latter either. In fact, Jones suggested that his ‘compiler’ — that is, of the common text underlying the three vernacular chronicles — had before him versions of bot/1. possibilities (represented by our extant B— and C—texts), and being unable to decide which one to follow, chose the middle way.18 However, in the light of

17 Jones, ‘Historical writing in medieval Welsh’, modern historian, he arrived at a happy compromise

26—7. with the ambiguous statement that Rhain “was never

‘8 ‘One can almost see the compiler pause and pon— seen again” ’ (Jones, ‘Historical writing in medieval

I 36 DAVID E. THORNTON

our current understanding of the medieval \Melsh chronicles, it would be safer to reject Jones’s attractive scenario and instead suggest that the divergence between the B—text and the vernacular texts is to be explained either as one of the additions and em— bellishments introduced into the common source of the Brutiaru. or as a change to the ‘Annals of Strata Florida’ as eventually redacted at Neath, or perhaps both!19 It is not possible to argue convincingly either way: on the one hand, the ‘more literary’ form of the annal for 1022 in the vernacular texts may suggest we should favour the reading of the B-text (thus, that Rhain was indeed slain at Abergwili); but, if we simply regard the phrase ‘and he was not seen again’ as an addition for literary effect, then the vern— acular texts are closer to the C-text (Rhain simply defeated) and perhaps his death was added to the B-text at Neath. However, given the derivative and literary character of the vernacular texts here, it would no doubt be safer to opt for the former possi-bility and regard the B-text as the better witness to the ‘Annals of Strata Florida’ at this point.

From the foregoing consideration of Rhain in the light of an understanding of the relevant five texts, a number of probable conclusions emerge. First, there seems no rea— son to doubt that he bore a genuine Welsh personal name, though whether this was from birth or was subsequently adopted is less apparent. Secondly, he said he was the son of a Maredudd (seemingly Maredudd ab Owain of Deheubarth), and only one strand in the chronicles (though, by the accident of textual survival, that represented by most of the surviving texts) explicitly casts any doubt upon this claim. Thirdly, Rhain’s sup-posed Irish connections are limited to one (the same) strand in the chronicles; and, final-ly, his death at Abergwili is also limited to the same strand (or rather in this case, one substrand) of the chronicles. Analysis of the inter—relationship of the five chronicles thus reveals that we are essentially faced with two different annalistic views of Rhain: on the one hand, that, judging from his epithet Seams, Rhain was an Irishman (or, at least, was regarded as an Irishman) who falsely claimed to be son of Maredudd, thus acquired con— trol of Deheubarth, and was subsequently killed at Abergwili by Llywelyn ap Seisyll; or that, judging from his name, he was a Welshman who said he was (and therefore may very well have been) a son of Maredudd and was defeated (but not necessarily killed) at that battle by Llywelyn. Furthermore, despite the textual balance favouring the first interpretation (both in terms of the number of extant texts, four against one, and the greater length of the three vernacular accounts), analysis of the chronicles indicates that both possibilities should be treated with equal weight by the historian. In attempting to gain a fuller understanding of the significance of ‘Rhain the Irishman’, I will consider both possibilities below, starting with the strand represented by AC (B) and the Brutiau. As stated above, those historians who have previously commented on the events of 1022 have generally favoured the interpretation that Rhain Seams was indeed Irish or at least partly so. Furthermore, this interpretation has recently been strengthened by Sean Duffy’s suggestion that he may be identifiable in the Irish sources as Roen (or Raen), king of Meath, who died in 1027: ‘One cannot, of course, say for certain whether 19 A possible explanation may be that the vel‘HaC- co (to blind, make invisible, conceal): that is, if Rhain

ular texts’ wording reflects a scribal misreading of the was hidden or concealed, then he would not have phrase oeeisus est (was killed) of AC (B) as oceecus est been seen thereafter.

the “Rein” in Wales and the “Raen” in Mide are one and the same person, but it must be a possibility.’20 On the face of it, this identification certainly has its strengths: the two figures were more or less contemporary and they bore very similar personal names. In order to give this identification appropriate consideration, I shall now examine the fig-ure of Roen, beginning with the onomastic and genealogical details relating to him before proceeding to discuss his political career and its background in the politics of the king— dom of Meath during the years following the death of Mael Sechnaill Mér, king of Tara, in 1022.‘21

There is some variation in Roen’s personal name as given in the Irish chronicles and other sources (which are of course of differing dates of redaction) and some texts give more than one form: by far the most common is R0911 (including at least one case of the genitive form 110611122), followed by the orthographic variant Ram, then the modernized form Room, and finally one case each of the ‘corrupt’ forms R0111 and 1361111. Henceforth, I will use the form Roen when referring to this figure. The Irish annalists were not enti1ely suie what to make of this obviously extremely rare name: for example, in (11110111611111 510110111711 the form Roen is glossed 1. Rao11, seemingly by way of explanation; 23 and (aas Duffy notes) one copyist of the Annals of Ulster wrote i11te1lineally the phrase 111 tl1111g1111 so (I do not understand this!) above the erroneous form R0111. Despite this apparent confusion and its evident rarity, however, the name could indeed have a pure— ly Gaelic etymology as a monothematic name based on the Old Ilish noun 1oe11 (Mod. Ir. 1(1011), meaning path, way, t1 all and defeat, 1out; act of falling, collapsing’. 24

There 1s 11101e ag1eement over Roen s political status. He occuls in the 1egnal lists of Ushnagh (that 1s, of "Mide or Meath)25 and 1s va1iously desc11bed 111 the ch1onicles as 11’ M1de (king of Meath);111 Mal/11 og115 11gdc11111111 E1e1m (king of Meath and p1 111ce of Iieland); 1‘17gda11111t1 Temm (prince of Ta1a); 111111.06 of Meat/1 (AClon.); and 111 two 1elated texts he 13 termed 1'1 1(11‘1/11111‘ M1116 (king of Westmeath)?“ Of these titles, those of king of Meath and 1‘1’gda11111a of Tara are perfectly compatible, and the latter was often held by the kings of Meath in this period. However, the 111111.06 of Meat/1 in the Annals of Clonmacnoise prob— ably suggests the lower status title 1‘1’gda11111a Mide; and, the use of ‘king of Westmeath’ (admittedly only at the same point in the related AT and CS) perhaps implies that there

2” Duffy, ‘Ostmen, Irish and VVelsh', 383. 21 For the Clann Cholmain kings ofMeath, see Paul Walsh, ‘The Ua Maeleachlainn kings of Meath’, Iris/1 Ecclesiastical Record, 57 (1941), 165—855; F. .l- Byrne in T. W. Moody at al. (eds), A New Hislmy of Ireland, 9 vols. (Oxford, 1982—), ix, 130—1, 195~7; Liam Cox, ‘The U3 Maeleachlainn kings of Meath’, R’aclit 1111 111111716, 5 (1972), 22—53. F01 my own (impeifect) attempt to 1ec0nst1uct the genealogy of the dynasty, see David E. Thor,nton ‘Kings, chl onicles and genealogies: 011 1econst1ucting mediaeval Celtic dynasties, in K. S. B Keats—Rohan (ed), Family T111111 malt/113 Roots of Politics Woodbiidge, 1997), 23—40 (pp 32—37).

22 For the genitive form, see below, p. 140. 23 This occurs at CS [1024]; in the next annal, the scribe uses the modern form.

24 The name has not, to my knowledge, been dis«

cussed by specialists, but for the relevant noun, see Carl j. S. Marstrander et (11., Dictionary of the 1115/1 Language Based Mainly on. Old 111111 111M111; Iris/1 Materials

(Dublin, 1913—76), ‘1'6e11 (raen)’; Patrick S. Dinneen, Focloir Gaea'lzilge agus Bérz'rla: A11 Irish—English Dictianmy (Dublin, 1927), 5.1). ‘raon’.

25 For the regnal list of the kings of Usnagh, see The Book of Leiiistei‘, formerly Lebor 11a Niiaoliongbzila, ed.

R. I. Best 61 al., 6 vols. (Dublin, 195483), 1, 198, iv, 808; and john [=Eoin] MacNeill, ‘Poems by Flann Mainistrech on the dynasties of Ailech, Mide and Brega’, Arc/1111111111Hibarnica, 2 (1913), 37~99 (pp. 85,

91).

2“ I use ‘VVestineath’ literally here, though it should

be stressed that the medieval ia1‘tl1ar Midr,’ was not nec— essarily co—terminous with the modern county of that

1410 DAVID E. THORNTON

in the chronicles indicate that the period following the death of Mael Sechnaill Mor in 1022 was rather turbulent politically for Meath, and a number of intra-dynastic killings are recorded in the chronicles, including that of Muirchertach ua Carraig Chalma (a can— didate for Roen’s father, as suggested above) by the same Mael Sechnaill Got.” Therefore, Roen’s battle against the men of Meath in 1026 may have been part of the ongoing intra— dynastic conflict which began in 1022.

However, 1027 does not necessarily mean the end of Roen. For example, both the Annals of the Four Masters and those of Clonmacnoise record for 1041 the death of a ‘Mael Ruanaid mac Roein’, called rigclamna of Tara, who was presumably therefore our Roen’s son. In addition, although I cannot trace this particular line of descent beyond Mael Ruanaid, it is perhaps worth noting here that among the numerous grave—slabs at Clonmacnoise, there was once a slab bearing the simple inscription ROEIN (according to the transcription, with lines over the letters R and N, perhaps indicating abbrevia-tion).39 R. A. S. Macalister, being unable to verify the reading of the lost slab, was ini-tially dubious of it (stating ‘which can hardly be right’), but he did subsequently index it without comment as Roam (that is, without any abbreviation lines). This name—form could arguably represent Roam, the genitive of Roen.: the inscription therefore meaning ‘[The grave] of Roen’. This name, as we have seen, was very rare (if not otherwise unique) in medieval Ireland. Furthermore, according to a series of later poems, the cemetery at Clonmacnoise housed the graves of numerous Irish kings and, while no Roen or Raen is explicitly named among their number, the dynasty of Clann Cholmain is indeed represented.“ It is tempting (but admittedly beyond proof) to suggest that this lost slab once marked the final resting place of Roen mac Muirchertaig meic Mail Shechnaill, king of Meath.

What then are the grounds for identifying this genealogically and onomastically obscure Roen, king of Meath, with ‘Rhain the Irishman’ of the Welsh chronicles? In addition to Rhain’s alleged epithet Scams, the case for identifying him with Roen rests primarily on the fact that the two men bore strikingly similar names and were active at more or less the same time. Do these facts bear up to the foregoing analysis of Rhain in the light of the \Nelsh chronicles? Duffy states that both ‘Rein’ and ‘Roen’ had ‘the same unusual name’ and further that ‘the Mide dynast bore the Welsh name Rhain’ but, as stated above, there is no difficulty in regarding the Middle Welsh forms Reyn and Rein, as the modernized Rhain. However, while the phonologically not dissimilar name Roen and its variants seem to have troubled the Irish scribes, it may indeed have been a purely Gaelic name in etymology — which is not surprising as he was a member of an Irish dynasty. Christian. Inscriplions in the Irish Language, ed. M.

Stokes, 2 vols. (Dublin, 1872), i, 4~7, 76—82; Caitlin

3“ Other assassinations include that in 1023 of Conchobor mac Oengusa meic Charraig Chalma by ‘the Gots’ (probably including Mael Sechnaill there—

fore).

3" R. A. S. Macalister, The Memorial Slabs of

Clonmmrnois, King’s County: With an. Appendiv 0n the A/Iateriuls for a History of the Monasteny (Dublin, 1909),

43; idem, Cor/ms lnscriplionem I'lzsularzmz Celticm’zun, 2

vols. (Dublin, 19454)), ii, 60 (=no. 755) plus 'Index’. “" R. I. Best, ‘The graves of the kings at

Clonmacnois’, Erin, 2 (1905), 163—71; George Petrie,

NI Mhaol-Chréin, ‘Cathair Chiarain', An Sagart, 7 (1964), 37—9; John O’Donovau, ‘The Registry of Clonmacnoise: with notes and introductory remarks’,

journal of [he Killcenny and South-East Ire/11ml Archaeological Society, 2nd series, 1 (1958), 444—60 (pp.

449—51); j. G. O’Keeffe, ‘The kings buried in

Clonmacnois’, in j. Fraser, P. Grosjean and J. G.

O’Keeffe (eds), Iris/I Texts, 5 vols. (London, 1931—4), iv, 44—6.

From a chronological standpoint, Duffy notes the importance of the year 1022 for both men.41 As discussed above, it was in this year that the death of Mael Sechnaill Mor sparked off a succession struggle between various segments of the ruling dynasty of Meath, and Roen (who was probably a member of one of these segments) may well have felt that Wales was a safer prospect in the light of the (possible) dominance by the rival Mael Sechnaill Got and his kinsmen. It is not inconceivable therefore that he could have turned up in “7ales in 1022 as Reyn, Scotus, seeking to further political ambitions thwart-ed by current developments in Ireland, and that following his defeat at the hands of Llywelyn ap Seisyll, he may have returned to Ireland to await a later opportunity to claim the Meath kingship. However, there is a major difficulty in this scenario: accord— ing to what must be regarded as the ‘primary’ text of the relevant annalistic strand (that is, the B-text of the Ammles) ‘Rhain the Irishman’ was not only defeated but killed by Llywelyn ap Seisyll at Abergwili in 1022. This would rule out any identification with Roen, whose own death is firmly dated to 1027 and placed in Ireland (and whose grave stone may even be identifiable). Of the Latin texts, it is the C-text (which, taken alone, permits 720 necessary Irish connection whosoever) which states that Rhain was defeated, without any reference to his death. Admittedly, the statement in the vernacular texts that ‘he was not seen again’ could be taken to mean defeat only; but, as discussed above, these embellished accounts are certainly derivative and their details accordingly suspect. Thomas Jones himself was not prepared here to give the Brutiau precedence over the Latin versions and regarded the phrase in question as an attempt to resolve the appar— ent contradiction between these two copies of the Ammles. Therefore, if this bipartite textual analysis of the different Welsh accounts holds, then the case for identifying Rhain ‘the Irishman’ (d. 1022) with Roen king of Meath (d. 1027) must surely collapse.

Of course, to rule out the case for Roen does not in itself preclude the possibility that Rhain was a different Irishman. The idea of Rhain’s Irishness depends primarily upon our interpretation of the epithet Scot/us, since I have argued the further references in the vernacular chronicles to his Irish character are probably embellishments inspired by the epithet in their common Latin source. It is not always easy to penetrate the signif-icance of epithets in medieval sources since they are often posthumous coinages, there— by reflecting subsequent — and not necessarily contemporary — Views of an individual’s character. Indeed, as later coinages, they are accordingly rather rare in annals which were (roughly) contemporaneously recorded: for example, the earliest version of the Ammles (the so-called A-text, which terminates in the mid—tenth century) invariably men— tions Welsh kings by their personal name (and perhaps their patronymic) only but, for certain notable early kings, the two later versions have added the nicknames by which they came to be known.42 In the case of 1022, therefore, the epithet Scams may indicate that the pretender was indeed an Irishman or was regarded as such by a later annalist,

*1 Duffy, ‘Ostmen, Irish and Welsh’, 383.

42 The best example is that of the ‘obit’ ofthe great

Hywel Dda, ‘Hywel the Good’ (950): the A—text sim—

ply calls him Higucl my Britton'zn but the B—text adds cognomenlo Bonus and the C-text scilicet Bonus. In this particular instance, it is probable that the epithet did not emerge until the twelfth century, possibly as part

of the cult of the king among his descendants in Deheubarth and possibly in connection with his

alleged association with Welsh law: see David Kirby, ‘Hywel Dda: AnglopliilP’, WHR, 8 (1976—7), 113

(p. 9); and Huw Pryce, ‘The prologues to the Welsh

lawbooks’, Bulletin. of the Board of Celtic Studies, 33

142 DAVID E. THORNTON

or that he was a non—Gael to whom the epithet had become attached in the chronicles at some point. The fact that he bore a Welsh personal name may lend weight to the last interpretation.43 As such, specifically ‘racial’ epithets like Scams borne by Welshmen are rare in the chronicles. The best (and far more common) parallel is that of the epithet Sais (Englishman) which recent scholarship has understood variously to mean a \Velshman who could speak English, or had spent some time in England, or had an English parent, or who ‘admired and imitated’ English ways.44 By analogy, might Rhain ‘the Irishman’ thereby have been a Welshman who could speak Irish, or had spent some time in Ireland, or had an Irish parent, or who even ‘admired and imitated’ Irish ways (or, of course, two or more of these options)? While there is no means of proving this suggestion with any certainty, such an association between a Welsh ruler (actual or aspir-ing) and Ireland would not have been without parallel in the eleventh century.45 On the other hand, we should also note the fact that the words Gwyddel and Sais could have derogatory connotations when employed by the medieval Welsh.46 In the light of this fact, could the epithet Scams have been used to denigrate Rhain in some way? This is clearly possible since the relevant Latin text also explicitly denies Rhain’s alleged parent-age (where the ‘non-Irish’ C—text is simply non—committal) and the related vernacular texts are even more open in attacking his claims and character, including outlining his supposedly cowardly and ‘Irish’ behaviour during the battle. The alleged Irish origin of Rhain could simply be an element in the negative tone of these four texts derived from the lost ‘Annals of Strata Florida’.

If we reject the common elements of the accounts of 1022 in the B-text of Avmales C(mzbriae and the three vernacular chronicles as later additions in the so-called ‘Annals of Strata Florida’, then we are left with that given in the C—text: in 1022, Llywelyn ap Seisyll, king of Gwynedd, fought against Rhain, who said he was the son of Maredudd, and Rhain was defeated (by Llywelyn?) at Abergwili. Taken on its own, this account need not imply that these events were anything other than a purely internal Welsh affair, nor does it necessarily suggest that Rhain’s claims were false since, in effect, we only have his word for it. How do the claims face up to more detailed scrutiny?

On the face of it, the extant Welsh genealogies would seem to support AC (B) and the Brutiau since they record no ‘Rhain’ as a son of Maredudd ab Owain. For example,

*3 Duffy argues that Rhain ‘was not simply a Welshman with the epithet Gwyddel’, but he bases this argument primarily on the additional material in the vernacular chronicles (especially the Pen. 20 Brill) to

which, I have argued, we should not attach any inde—

pendent value.

44 See Melville Richards, ‘Gwyr, gwragedd a gwe— helyth’, Transactions of the Honourable Socieiy 0f

Cymmmdarion, 1965, 27—45 (p. 41); and, more recent

ly, the ongoing research into this epithet by Fred

Suppe: see, for example, his ‘Who was Rhys Sais?

Some comments on Anglo-Welsh relations before

1066’, The Has/(ins Society/ourmll, 7 (1995), 63—73 (pp.

64—5). I must also acknowledge Prof. Suppe’s paper as the inspiration for the title of this present article.

45 A number of eleventlrcentury Welsh dynasts are

known to have journeyed to Ireland, usually in reac-tion to politically unfavourable developments in

Wales, and subsequently returned seeking to re—estab-lish themselves in Wales, often with Irish or Hiberno—

Scandinavian help: Moore, ‘Gruffudd ap Cynan and the medieval Welsh polity’, 26—7. Most famous of these was Gruffudd ap Cynan, who was probably born in Ireland and was himself of part Hiberno-Scandinavian descent.

*6 See Patrick Sims—VVilliams, ‘Irish elements in late medieval Welsh literature: the problem of Cuhelyn and *Nyf’, in M. J. Ball at al. (eds), Celtic Ll'nglu'sties/ Ieit/lyddiael/z Gelz‘aidd: Readings in. Biyz‘lzonia Languages, Feslsc/zrift for T. Anuyn. Watkins (Amsterdam, 1990), 277,95 (pp. 278, 287, n. 7); idem, ‘Celtomania and

WHO WAS RHAIN THE IRISHMAN?

the later medieval Welsh genealogies now collectively known as ‘Achau Brehinoedd a Thywysogion Cymru’ (=ABT) list Maredudd’s offspring as comprising one son, Cadwallon (who we know predeceased his father by seven years), and one daughter, the thrice—married Angharad.47 However, the earlier jesus College collection, which also mentions Angharad, names another daughter, Lleucu, described as wife of Llawr (of unknown origin) and thereby ancestress of Hywel ap Goronwy (d. 1106) which, if reli-able, indicates that ABT does not contain a complete list of Maredudd’s children. Finally, some late and unreliable genealogies mention another son of Maredudd called Rhys, father of ‘Aron Fraich Hir’, though in this case I am very dubious of its historical reli* ability.48 Elsewhere, I have discussed how the Welsh (and Irish) genealogies by no means name every member of a dynasty and often neglect whole dynastic segments,49 and it is perfectly possible that political or dynastic reasons (as well as simple scribal error or mere ignorance) can often underlie such omissions. Consequently, the absence of Rhain from the relevant genealogical tracts need not in itself automatically confirm the state-ments in the chronicles that his claim to be son of Maredudd must be false. A possible support for Rhain’s claim comes from his very name. An analysis of what I have termed the ‘naming strategy’ of the ‘Second Dynasty’ during the ninth and tenth centuries demonstrates that the different branches of the dynasty tended to use names associat-ed with the earlier Welsh dynasties which they had replacassociat-ed, perhaps in order to lend anthroponymic continuity to their political claims.50 When understood in these terms, the name ‘Rhain’ would have been entirely appropriate for a dynast of Deheubarth dur-ing this period: for example, the name had been borne by two kdur-ings of Dyfed in the eighth and ninth centuries, though originally it had been transmitted to this Dimetian dynasty following a marriage union with the line of Brycheiniog; and, in the tenth cen— tury Hywel Dda had named one of his sons Rhain, who was thus an uncle of Maredudd ab Owain.51 According to this analysis, Rhain’s name was not only Welsh, but also respectably ‘Dimetian’ in its dynastic associations, and therefore it was not unreasonable for a son of Maredudd ab Owain to have been called Rhain. This does not in itself prove that Rhain was Maredudd’s son but, I think, it should encourage us to take his claim more seriously than both modern historians and medieval chroniclers have previously been prepared to do.

If Rhain’s claims were indeed true, do they tally with the politics of Deheubarth in the early eleventh century? Reconstruction of the political situation in Deheubarth

*7 For the significance of Angharad, see below,

p. 146.

48 P. C. Bartum, ‘Pedigrees of the Welsh tribal patriarchs’, Nalional Librmy of Wales journal, 18 (1963—4), 93—146 (pp. 113, 134); also see Thornton, ‘Maredudd ab Owain', 570.

“‘9 Thornton, ‘Kings, chronicles and genealogies’,

33—8. For the two case-studies, only about half of the

members of the Irish example occur in the relevant dynastic genealogies, and while the figure is in fact much higher for the Welsh example, there is still

an evident amount of omission. Roen of Meath, we

might recall, was omitted from the Clann Cholmain genealogies.

5" David E. Thornton, ‘Predatory nomenclature

and dynastic expansion in early medieval Wales’, Alledieval Prosopogmphy, 20 (1999), 1—22, and more

briefly ‘Kings, chronicles and genealogies’, 39—40.

51 As such the name Rhain would sit comfortably

alongside names such as Hyfaidd, Owain, Maredudd, Tewdwr and Llywarch which had been employed by

the earlier kings of Dyfed and were also used subse— quently by the ruling line of Deheubarth. Some tracts

refer to Rhain ap Hywel Dda as ‘Rhun’, which is prob— ably an error.

144 _{DAVID E. THORNTON}

during the period after the death of Maredudd ab Owain in 999 and before the battle of Abergwili in 1022 is clif_{ficult because no dynast is explicitly linked to the kingdom in} the chronicles until Rhain makes his appearance in 1022. In 1015, a figure called Owain ap Dyfnwal is said to have been slain, though his genealogical and political affiliations are not apparent. This silence in the chronicles need not, of course, indicate necessar-ily that Deheubarth was kingless during these two decades nor, as Lloyd suggested, that it was experiencing ‘the beginning of a period of anarchy unexampled even in that turbulent age’, but rather the exact opposite: a period of internal stability and a relative lack of external attack.52 Indeed, kings are normally named in the relatively terse chronicles when they die, attack or are attacked by neighbours or rivals, or (very occa-sionally) when they perform some particularly notable deed, such as going on pilgrim-age to Rome. If a king of Deheubarth did none of the above during the early eleventh century, then he may well have escaped the attention of even the local St David’s annal— ist. It is at least possible that the_{figure of Rhain named in the C—text of the Annales may} represent just such a king, who had remained anonymous from the time of his father’s death in 999 until he was defeated by his Llywelyn ap Seisyll in 1022. Of course, the silence of the chronicles means that it is impossible to prove this assertion, but equally it prevents us from certainty in the opposite interpretation.

If the situation in Deheubarth is of limited use, then perhaps our understanding of this \Velsh Rhain may be facilitated by an examination of his opponent Llywelyn ap Seisyll. Llywelyn is _{first mentioned in the chronicles 3a. 1018 when he is said to have} killed one Aeddan ab Blegywryd and his four sons. Next, his encounter with Rhain at Abergwili is recorded for 1022;53 and finally, notices of his death in 1023 are preserved in the Irish as well as the Welsh chronicles.“ A few annals accord Llywelyn the title ‘king of Gwynedd’; and in addition, his killing of Aeddan in 1018 has been interpret~ ed by some as marking the point of intrusion into Gwynedd, though this is certainly not necessary.55 The notices of his death in 1023 in the Welsh annals are followed by the statement that Rhydderch ab Iestyn took the kingdom of Deheubarth — a fact which could indicate that Llywelyn had ruled that kingdom as well, perhaps after defeating Rhain in the previous year,56 though again there is no explicit statement to that effect. The Irish chronicles call him 177 B'retan, ‘king of the Britons’, which probably indicates that his power extended over a large part of Wales, if not necessarily the whole. Judging from the rather extensive panegyric on the benefits of his rule ‘from one sea to the other’ recounted in the vernacular chronicles,57 the compiler of the lost Latin source for these texts also thought Llywelyn’s rule covered much of Wales. There is good evidence therefore that Llywelyn ap Seisyll was a figure of considerable, though seemingly

short-52 There are records of two external attacks by non- _{non with the battle: EWGT, ed. Bartrum, p. 141;}

Welsh during this period: Dyfed was raided by Vikings

in 1001, and St David’s by the English in 1012.

53 Llywelyn’s involvement at the battle of Abergwili is also mentioned in a pedigree of Ellelw, wife of

Llywelyn ap Cadwgon (d. 1099), where Ellelw’s

ances-tor non ‘Ceneu Menrud’ is said to have been killed by Llywelyn ap Seisyll: see JC § 33. The chronology of this pedigree is evidently erroneous, and histori-ans have generally rejected the synchronization of

Maund, Ireland, Wales and England, 46, 60.

5“ AU s.a. 1023.7; AT [1023]; CS s.a. [1021=A.D.

1023]; AClon. 3.11. 1023; ALC s.a. [1023]. See Lloyd, History, i, 347, n. 92; Maund, Ireland, Wales and

England, 59.

55 Maund, Ireland, Wales and England, 90—2. 55 Ibid., 60.

WHO WAS RHAIN THE IRISHMAN?

lived, influence on the Welsh political scene. Yet, despite this importance, his genealog— ical antecedents are somewhat difficult to unravel. According to later medieval tracts, Llywelyn’s mother was Prawst ferch Elise, grand—daughter of Rhodri Mawr, and his wife was Angharad, daughter of Maredudd ab Owain (or at least he fathered his famous son Gruffudd by her).58 Of these two connections, while the reliability of the former has been questioned, the latter is generally accepted as true by historians.59 This link to Maredudd could have been important for his possible interests in Deheubarth. However, of Llywelyn’s paternal ancestry nothing is known and his father Seisyll ‘appears to have been entirely obscure’.60

This genealogical obscurity means that any kinship—based claim which Llywelyn may have made to justify his rule, whether in north or south Wales, would seemingly have rested on his apparent marital connection through Maredudd’s daughter. Such a situ-ation would not have been without precedent,61 and of course it is a moot point as to what extent political reality depended upon genealogical background when military might was sufficient. Yet, judging from the text of AC (C), kinship was indeed an impor-tant issue in the events of 1022: Rhain, we are told, se (lieebat esse fill/um Marednt; while, as noted above, his opponent Llywelyn was allegedly married to Angharad ferch Maredudd ab Owain. Both protagonists of 1022 therefore had or claimed to have kin— ship ties to Maredudd around which the right to rule Deheubarth would have partly hinged, whether at the time or retrospectively. Since Llywelyn won the battle at Abergwili it is perhaps not surprising that Rhain’s claim to kinship was reduced to mere hearsay. The text of AC (C) therefore takes on the appearance of the Victor’s ‘version of events’; and if, as is possible, the victory established Llywelyn’s power in Deheubarth, then the local St David’s annalist would have had little option but to toe the line in this regard. Following the strand represented by the C-text of the Annales, it could be argued therefore that What we have is an account of the defeat of a Welshman called Rhain, who may very well have been the son of Maredudd and who may even have ruled Deheubarth during the preceeding decades, until he was overthrown by his ‘intrusive’ brother—in-law Llywelyn ap Seisyll. The supposed Irish origin of Rhain would thus become part of a later version of the annal for 1022, characteristic of the ‘Annals of Strata Florida’, and seemingly intended to denigrate Rhain and his claims while, by the same token, praising Llywelyn. Thus, the B-text of the Annales includes the (possibly) derogatory epithet Scams and explicitly states that Rhain lied about his parentage. Furthermore, the additional material in the three vernacular accounts of 1022 effectively amounts to a eulogy on Llywelyn and the prosperity which obtained during his rule,” plus a lengthened description of the battle at Abergwili. For example, the description

58 See jC § 27; ABT §§ 1e, 2f, 7f and k; ByT (RB), 51 For example, the famous Hywel Dda gained con— s.a. 1113 = 1116; ByT (Pen. 20), 5.11.. 1113 = 1116; trol of Dyfed in the early tenth century though his ByS, 5.11.. 1113 = 1116; P. C. Bartrum, ‘Rhandiroedd only (known) connection with its ruling line was Powys’, National Library of Wales journal, 18 (1973—4), through his wife Elen; and it seems that his ancestor

231—7 (p. 232). Maund, Ireland, Wales and England, Merfyn Frych had established the ‘Second Dynasty of

60~2; Thornton, ‘Maredudd ab Owain’, 569. Gwynedd’ in the previous century under similar cir— 5“ Lloyd, Hislmy, i, 347; Maund, Ireland, Wales, and cumstances.

England, 61. 62 Lloyd, Histmy, i, 347, who suggests we compare

‘5" Lloyd, History, i, 347; Maund, Ireland, Wales and this with The Text of the Book of Llan Dav, ed. Evans

146 _{DAVID E. THORNTON}

of Llywelyn as gm‘yclzaf vrenhm Gwyned a [)lzenaf a c/zloluomssaf urea/1m 0’7" hall Viymnyeit (supreme king of Gwynedd and foremost and most praiseworthy king of all the Britons) or y bran/2m clotuorussaf a wydit 0’7' mar [)zuy gilid (the most praiseworthy king who was

known from one sea to the other) clearly echoes the earlier vernacular descriptions of Llywelyn’s father-in—law Maredudd ab Owain as y clotuomssaf wen/M1 y Biylmzyelzf (the most praiseworthy king of the Britons).63 The use of similar phrasing for the two kings may have been intended to emphasize Llywelyn as appropriate successor of Maredudd

against Rhain, who claimed to be his son.

The reasons for this later manipulation of the annal for 1022 in order to denigrate Rhain are not immediately apparent, though a detailed analysis of the different Welsh chronicles may facilitate our understanding of the texts, their composition and the inten— tions of their compilers in this and other respects. Pending such a study, a few brief tentative comments might be ventured. At the probable time of the compilation of the ‘Annals of Strata Florida’ (6. 1202), the abbey of Strata Florida was under the patron-age of the dynasty of Rhys ap Gruffudd, the Lord Rhys (d. 1197); and it is not unrea-sonable to expect the monks thereof to support their powerful benefactors. Now, among the jesus College collection of genealogies is a series of pedigrees of Rhys Gryg (d. 1233), son of Rhys ap Gruffudd:64 these describe Rhys’s paternal descent from Rhodri Mawr both through his great-grandfather Rhys ap Tewdwr (d. 1093) and through his grand-mother Gwenllian, daughter of Gruffudd ap Cynan (d. 1137), and also his maternal descent from Angharad ferch Maredudd ab Owain through her Powysian son Bleddyn ap Cynfyn (d. 1075). In addition to Bleddyn, the relevant text (IC § 27) also names Angharad’s other sons, that is Gruffudd ap Llywelyn (d. 1063) and Trahaearn ap Caradog (d. 1081). Thus, the text has the effect, whether reliably or not,65 of linking Rhys Gryg (in various ways) to many of the important Welsh rulers of the second half of the eleventh century, including Gruffudd son of Llywelyn ap Seisyll. Though not a direct lineal ancestor of Rhys Gryg, Gruffudd was clearly regarded as sufficiently impor— tant to be included in the genealogy. Thus, given this apparent importance as a politi— cal predecessor for the family of the Lord Rhys in the early thirteenth century, it would not be surprising if the matter of Gruffudd’s ancestry was felt in need of attention. All that would have been available were the few annalistic accounts of his father Llywelyn ap Seisyll — including the annal for 1022. While it remains a tentative hypothesis, pend-ing further analysis of the chronicles and their backgrounds, it might be conjectured that the account of 1022 in the ‘Annals of Strata Florida’, in which Rhain’s claims were explicitly denounced and he himself branded an ‘Irishman’, represents just such a retro-spective attempt to strengthen the legitimacy of Gruffudd ap Llywelyn by eulogizing his father and attacking Llywelyn’s opponent at Abergwili, thus reinforcing the genea-logical and political claims of the patrons of the abbey almost 200 years later.

As such, this material demonstrates the need for caution when consulting the medieval Welsh chronicles, and further underlines the need for a thorough study of these texts “3 The phrase occurs following the notice of _{Leinster, as cloluorussaf a (:hadarnaf vreizlzz'n y Gywydyl,} Maredudd’s death in 999. For a discussion of which, _{‘most praiseworthy and bravest king of the Irish’.}

see Thornton, ‘Maredudd ab Owain’, 567—8, 572—3. “4 JC §§ 24—7.

We might also note that the vernacular chroniclesma. “5 For some comments about its content, see 1072, refer to Diarmait mac Mail na mBo, king of Maund, Ireland, Wales and England, 67.

and their history. For the year 1022, a short annal-entry of about thirty words was aug— mented to a longer version of up to 270 words and, I have argued, we can detect the manipulation of the information contained therein during this process. Admittedly, this is an extreme case, but it does serve to illustrate the danger of using these annalistic texts Without a consideration of how and when they came into being. I have suggested that it is only by means of analysing the relevant texts that we can achieve a fuller under— standing (as far as is possible) of the position and significance of ‘Rhain the Irishman’.

APPENDIX:

The Three Vernacular Welsh Chronicles’ Accounts 0f1022

ByT (Pen. 20): Vgein mlyned a mil oed oed Krist pan gelwydawd nebun Y scott o dywed— ud y uod yn vab y Varedud vrenhin, a Rein y mynnawd y alw. A chymeredic vv gan wyr y Deheu, ac a gynnhelis gyfoeth. Ac yn y erbyn y kyuodes Llywelyn vab Seissyll, brenhin Gwyned a goruchaf a chloduorussaf vrenhin holl Brydein. Ac yn y amser ef megys y gnotaei yr henwyr dywedud frwythlawn oed yr holl dayar o’r mor bwy gilyd o dynyon ac 0 bob ryw da, hyd nad oed neb essywedic na neb yn reidus yn y gyfoeth, ac nyd oed vndref wac na diffeith. A Rein yn wann ac yn llesc a gynnullawd lu, ac y megys y mae moes gan Ysgotyeid ef a annoges yn vocsachus y wyr ac ef a edewis vdunt oruod ohonaw ef, ac ef a erbynnyawd y elynyon yn ymdiredus. Ac wynteu yn wastad dier— grynnedigyawn a aroassant y trahaus annogwr hwnnw. Ac ynteu a gyrchawd y vrwydyr yn lew diargysswr. A gwedy bod 0 bob tu diruawr ladua yn gyttuhun, a’r Gwyndyd yn gwastad ymlad ef a orchvygwyd Rein Yscott a’y lu, kanys megys y dywedir yn y diareb Gymraec, ‘Annho dy gi ac na Cherda ganthaW’, velly ynteu yn lew yn kyrchu ac o lwynog-awl defawcl yn ymchwelud ar ffo. A’r Gwyndyd gan eu hymlid yn greulawn lidyawc a’y lladassant ac a diffeithassant yr holl wlad, ac a gribdeilassant yr holl da. Ac nys gweled ynteu etwa. Y vrwydyr honno a VV yn nrws auon Wyli yn Aber Gwyli. Ac odyna y doeth Eilaf ynys Brydein ac y diffeithyawd ef Dyued, ac y torr(r)ed Mynyw.

ByT (Herg.): Ac yna y dechymygawd nebun Yscot yn gelwyd y vot yn vab y Veredud vrenhin ac y mynawd y alw ehun yn vrenhin. Ac y kymerth gwyr y Deheu ef yn arglwyd ar y teyrnas; a’e henw vu Rein. Ac yn y erbyn y ryfelawd Llywelyn ap Seisyll, goruchaf vrenhin Gwyned a phenaf a chlotuorussaf vrenhin o’r holl Vrytantyeit. Yn y amser ef a gnotae henaflyeit] y teynras dywedut bot y gyuoeth ef o’r mor py gilyd yn gyflawn o amylder da a dynyon, hyt na thybygit bot na thlawt nac eissiwedic yn y holl wladoed, na thref wac na chyfle diffyc. Ac yna y duc Rein Yscot lu yn dilesc; a herwyd defawt yr Yscotteit yn valch syberw annoc a wnaeth y wyr y ymlad; ac yn ymiredus adaw a wnaeth vdunt mae ef a oruyde[i]. Ac ymgyuaruot a oruc yn ehofyn a’e elynyon. Ac wynteu yn wastat diofyn a oryssant y chlwydedic drahaus anogwr. Ac ynteu yn hy diofyn a gyr-chawd y vrwydwr; a gwedy gweithaw y vrwydyr a gweuthur kyffredin aerau 0 pop tu, a gwastat ymlad clrwy lewdwr o’r Gwyndyt, yna y goruuwyt Rein Yscot a’r lu. A her-Wyd y dywedir yn y diaereb, ‘Anoc dy gi ac nac erlit’, ef a gyrchawd yn lew ehofyn ac a gillawd yn waratwydus o lwynogawl defawt. A’r Gwyndyt yn llidyawc a’e hymlynawd

View publication stats View publication stats

148 _{DAVID E. THORNTON}

drwy lad y In a diffeithaw y wlat ac yspeilaw pop man a’e distryw hy y Mars. Ac nyt ymdangosses ynteu byth 0 hyny allan. A’r vrwydyr 110110 a Vu yn Aber Gwyli. A gwedy hyny y deuth Eilad ynys Prydein. Ac y diffeithwyt Dyfet ac y torret Myniw.

ByS: Y doeth nebvn Yscot kelwydauc a dywedut y vot yn vab y Moredud \rab Oweyn; a Rein 06d y henw. Ac y kymyrth gwyr y Deheu ef yn bennaf arnadunt. Ac )7 damunws yntev Gwyned yn el‘byn Llywelyn vab Seissyll, y brenhin Clotuorussaf a wydit 0’1“ mor pwy gilid. Ac yn y 065 cf ny bu eissieu da yn y gyfoeth 11a neb gouudus nac vn dref wac na diffeith. A gwedy dyuot y lluoed hyt yn Aber Gweili, y kyrchws Rein yr ymlad yn valch bocsachus gan annoc y wyr. Ac yn hynny )7 goruuwyt arnaw ef ac y ffoas yn llwyn— ogeid ffyrnic, ac y Has y wyr yn Olo_{fi‘ud, ac yd anreithwyt yr holl wlat. Gwedy hynny y} doeth Eilafy dir Kymmre a thorri Myniw a diffeithiaw Dyfed.