Preface: This is an unpublished paper which I have given several times in different venues. It was originally written in about 2001 as a response to Ewen Campbell‟s „Were the Scots Irish?‟, Antiquity 75 (2001), 285-92. I have never got around to finally writing it up for publication and although I hope I will eventually do so I can‟t see myself getting it done anytime soon. Various people working in the field, such as James Fraser and Thomas Clancy, have seen it in draft and responded to it so I felt I should put it into the public sphere and I am posting it on Academia.edu on 9th April 2012. Feel free to cite it. It has not been significantly updated since 2005. Ancient Kindred? Dál Riata and the Cruthin Alex Woolf Until recently students of the history of Dál Riata, the Gaelic-speaking polity which in the early middle ages straddled the North Channel, linking Argyll and Antrim, were almost entirely dependent upon the works of John Bannerman, published originally as a series of articles in the nineteen sixties and then reprinted together with an additional analysis in 1974.1 Bannerman‟s ground breaking work remained the main stay of Dalriadan studies for thirty years, work by other scholars during this period merely tinkering with the detail of his analysis, and many of his hypotheses attained the status of factoids, most particularly his map of the division of Argyll between the cethri prímchenéla Dáil Riata. In the last few years this picture has begun to change radically. Richard Sharpe‟s contribution to the Festschrift for Marjorie Anderson, edited by Simon Taylor, presented the case for Dál Riata being a stronger and more centralised kingdom than its Irish equivalents, a case perhaps over-reliant upon the account of royal succession in the region appended to the later regnal lists of the kings of Alba, while David Dumville has re-edited the two tracts that we must now learn to refer to as Míniugud Senchusa Fher nAlban and Cethri Prímchenéla Dáil Riata.2 More revisionist than either of these readings, however, have been Ewan Campbell‟s claims that the Irish origins of Dál Riata are fictional and that the Scottish Gael are autochthonous. These views, first expressed live in Spring 1997, published in popular form in 1999 and in a scholarly article in Antiquity in 2001, apparently heretical as they may seem, have 1 J. Bannerman, „The Convention of Druim Cett‟, Scottish Gaelic Studies 11 (1966), „Notes on the Scottish Entries in the Early Irish Annals‟, ibid, „Senchus Fer nAlban‟, Celtica 7 (1966), 8 (1968) and 9 (1971) and collectively as Studies in the History of Dalriada, (Edinburgh, 1974). 2 R. Sharpe, „The Thriving of Dalriada‟, in S. Taylor, ed., Kings, Clerics and Chronicle in Scotland, 500-1297: essays in honour of Marjorie Ogilvie Anderson on the occasion of her ninetieth birthday, (Dublin, 2000), 47-61. D. N. Dumville, „Cethri Prímchenéla Dáil Riata‟, Scottish Gaelic Studies 20, (2000), 170-191, idem, „Ireland and North Britain in the Earlier Middle Ages: contexts for Míniugud Senchusa Fher nAlban‟, in C. Ó Baoill and N, R. McGuire, ed., Rannsachadh na Gàidhlig 2000: Papers read at the conference Scottish Gaelic Studies 2000 held at the University of Aberdeen 2- 4August 2000, (Aberdeen, 2002), 185-212. 1 been seized upon as a new truth by the British archaeological community, always eager for evidence that invasions never happened in the past.3 English anti-invasionism, or immobilism as it is technically termed, is an intellectual reflex of the same trend that turned Sax-Coburg Gothas into Windsors and von Battenbergs into Mountbattens and is doubtless linked to the current „Euro- scepticism‟, itself a „Windsor‟ to „Sax-Coburg‟ xenophobia. Although there are some signs that the tide is beginning to turn within the archaeological community, immobilism is now deeply entrenched in the British archaeological establishment, as anyone who saw Francis Pryor‟s recent, and very disturbing, television series on Britain BC will recognise. It would be a shame, however, if Campbell‟s views on the origins of Dál Riata were to be dismissed entirely as a by-product of anxieties about English origins. My intention in what follows is to take up Campbell‟s challenge to question the origin myths surrounding early Dál Riata in the light of both his own research and arguments and of the generally more sceptical approach to such material that has arisen amongst Insular historians since John Bannerman published his Studies in the History of Dalriada. Professor Dumville has interpreted Campbell‟s revisionist views as a reflection of the new vigour within Scottish nationalism but as someone who knows Campbell and has heard him speak on this topic on numerous occasions I remain completely unconvinced by this analysis.4 On the one hand the climate within British archaeology encourages immobilist interpretations regardless of the modern national reflexes such positions might impact upon, on the other Campbell has put together a cogent argument regarding the specifics of the case in hand which cannot be dismissed lightly. Campbell discusses the evidence for migration from Ireland to Scotland under three headings; archaeological, historical and linguistic. Under the archaeological heading he highlights the different morphology of settlement types found in Argyll and Ireland. The typical settlement type of early Christian Ireland, the rath or ringfort, is unknown in this part of Scotland, and crannóga, while native to Scotland are new in Ireland in Late Antiquity (superficially similar categories of site were common in the middle Bronze Age but had gone by the Iron Age). Campbell also compares the dress fastenings of the seventh and eighth centuries found in the two areas. Zoomorphic penanular brooches, common in Ireland are, to date, unknown from Argyll, the spiral- ringed ring-headed pin, the commonest dress fastening found in the north of Ireland is known from only one example in Argyll and whilst the type G penanular brooch was produced both at Dunadd in Argyll and at sites in Ireland, Dooey in Donegal and Moynagh Loch in Meath, Dunadd appears to have been producing them before the analogous Irish sites. Under the heading of historical evidence Campbell investigates the better known accounts of the migration of the Dál Riata to their Scottish territories. He points out that the chronicle record which states Feargus mor mac earca cum gente Dalriada partem britania tenuit et ibi mortus est reaches us via the Annals of Tigernach and has 3 E. Campbell, Saints and Sea-Kings: the First Kingdom of the Scots, (Edinburgh, 1999), 11-15, idem, „Were the Scots Irish?‟, Antiquity 75 (2001), 285-92. These views were first aired publicly at the Society for Medieval Archaeology conference held in Glasgow at Easter 1997. 4 Dumville, „Ireland and North Britain‟, 195. 2 all the indications that it is a post „Chronicle of Ireland‟ entry to that text, dating to the tenth century. Campbell also claims that the account in the Míniugud which describes six of Erc‟s sons taking Alba is part of a tenth century rewriting, before which “Alba was not used as a term for Scotland”.5 Finally Campbell points to the fact that Bede‟s account of the coming of the Scotti to Britain names the leader of the migration as „Reuda‟, the eponym of the Dalreudini or Dál Riata. Since this account is some two centuries earlier than that preserved in the Annals of Tigernach it demonstrates the constructed nature of the Fergus Mor story, a point also made by Bannerman. Demonstrating in this way that origins legends have been constructed Campbell argues that all such legends must be constructed, Bede‟s account he seems to be saying, is unhistorical because the Tigernach account is also false. Campbell‟s final category is that of linguistic evidence. He argues that although the preserved accounts of the migration have been recognised by most historians as erroneous in detail the fact that Argyll shared the same language as Ireland in Late Antiquity has been taken to demonstrate that a migration of some sort did take place. Campbell argues against this position on two grounds. Firstly the absence of any recognisable Pictish, that is to say Brittonic, stratum in the toponymy of Argyll is contrasted with the situation in Eastern Scotland where Gaelic conquest and language shift in favour of Gaelic still left a huge proportion of place names and place-name elements of Pictish origin on the map. If Gaelic invaders had encountered British aborigines in Argyll surely, he suggests, they would not have disappeared so completely from the onomastic record. Secondly he argues that while the divergence of the Goedelic and Brittonic branches of Celtic was facilitated by geographical separation, an uncontroversial position, the assumption that the Irish Sea was the defining barrier along the whole frontier may well be erroneous. Campbell rightly points out that in the north sea crossings between Britain and Ireland may well have been far less complicated than travel across the central highlands. This point is underlined by the distribution of imported fine-wares in Late Antiquity and indeed it seems likely that it was these distribution maps which first set Campbell on this course. If we take each of these categories in turn we can see that there are strengths and weaknesses to Campbell‟s position. Ironically, considering his training, his position is most easily assailed upon the archaeological front. The argument that the absence of ringforts is evidence fro the absence of Irish migration to Argyll is deeply flawed on chronological grounds. The ring-fort itself was an early medieval innovation within Ireland. To quote Matthew Stout‟s figures, half the absolute dates recovered from excavated examples fall within the period from AD 540 to 884 while two thirds of the one hundred and fourteen sites covered in his survey have the mid point of their date range falling between AD 600 and 900. Pre sixth-century dates can almost all be explained as the extremes of long, non-specific, C14 ranges or by reference to contamination from previous occupation of the sites. Earlier attempts to push the site type back into the Iron Age, or even the Bronze Age, were largely based on very lose areal association.6 It thus seems clear that had migration occurred in the period assigned to Fergus Mór, or at an earlier date we should not expect the ringfort to be part of the Irish „cultural package‟ taken to Argyll. Since most of the spiral-ringed 5 „Were the Scots‟, 288. 6 M. Stout, The Irish Ringfort, (Dublin, 1997), 22-31 esp. 24. 3 ring-headed pins, the commonest dress fastener from the north of Ireland, were recovered from ringfort sites the evidence they provide falls at the same fence. One other category of archaeological evidence sited by Campbell in his popular book but not made much of in his Antiquity article, is the ogam inscribed pillar stone. These whilst „common in Ireland‟, and perhaps chronologically better suited to the problem under discussion, are not found in Argyll. It has to be noted, however, that they are far from common in the far north of Ireland, the proposed place of origin of the Dál Riata. In short Campbell has tested Argyll‟s Irishness against a crude stereotype of Early Christian Ireland without taking the fine tuning of chronology or geography into account. In his analysis of the „historical‟ material Ewen Campbell appears to be on firmer ground. His observations on the Annals of Tigernach appear to be correct and few would argue that Bede‟s account could be relied upon as anything other than a report of a legend current in his time. If Bede‟s migration leader, Reuda, is to be equated with the Eochaid Riata (also appearing as Coirpre Rigfhota) of the pedigrees then he can be placed some thirteen generations above Fergus Mór giving him a notional floruit somewhere between AD 46 and 176.7 While these two accounts remain, respectively, the best known and the earliest, there are other versions of the story of the coming of Dál Riata to Britain and these have been neatly summarised by Professor Dumville in the introduction to his edition of the Míniugud. None of these origin legends is convincingly more than a legend and if nuggets of truth are hidden within them they are well hidden. One point in Campbell‟s „historical‟ analysis does require some attention, however, and that is his assertion that the appearance of the word Alba in section three of the Míniugud is in itself anachronistic. In his analysis of this passage Dumville has been careful to distinguish between the final Latin clause, in which the term Albania is used, and the rest of the section which is in Irish. The Latin clause is almost certainly a later gloss suggesting that the Irish text, if not certainly primary (a Latin ur-text does seem to lie behind it), is none the less early. Dumville further draws attention to the fact that this passage is noting the dispersal of the sons of Erc, some to Britain and some to Ireland, and argues that while it may be necessary to amend the text it is not the name „Alba‟ which is in need of attention, for it is balanced by „Hérind‟, Ireland, but the word gabsat which precedes it and implies a conquest of Alba rather than a settlement in it. On the note of the „anachronistic‟ use of Alba one might also point ot Adomnán‟s life of Columba which while it does not use the term „Alba‟ does consistently use „Britannia‟ for the land of Colum Cille‟s exile, even describing Conall as king in „Britannia‟. If Adomnán considered this passage in the vernacular then the word Alba must have been what he had in mind, and if, as we are always told, Dál Riata, straddled the narrow seas then it is highly likely that its two component parts were thought of as Ireland and Britain. We might compare the Roman usage of „Asia‟ for the tiny former kingdom of Pergamum, their first province on that Continent, of the Visigothic use of „Gallia‟ for the County of Narbonne their last province north of the Pyrenees. A closer parallel might be the Welsh use of Llidaw, probably originally meaning „The Continent‟, for Breton Armorica. If this interpretation of the usage is correct then even Dumville‟s suggested emendation need 7 Based upon the a generation length of between 25 and 35 years. 4 not be required. In a tract on Dál Riata composed prior to the Alpinid conquest of Pictavia the ambiguity that we find in the text may not have been conceivable. Campbell‟s final category of evidence, linguistic, is far more interesting. There is nothing at all implausible in his suggestion that the fundamental barrier between Goidelic and Brittonic dialects of Celtic need not always have been the sea. We are so used to seeing language families described through the Stammbaum, which is a very imperfect way of representing even genetic relationships between languages, that we forget that it is not arial divergence which requires explanation in language history but the emergence of hard linguistic frontiers. If we think of the two great Late Antique language dispersals in Europe, of Slavonic and Romance dialects, we must recognise that a high degree of mutual intelligibility remained within each group until very recent times. The division into a variety of standard languages rather than more or less similar dialects has been the result of the Romantic nationalism, the development of local dialects represented as regional acrolects and the production of literary forms of those acrolects. Until the eighteenth century, at least, one could easily have wandered across Romance or Slavonic Europe and never encountered two adjacent communities who were unable to speak to one another in their native tongue even if Macedonians and Sorbs, or Picards and Andalusians might have been at a loss to understand one another. What is most surprising about the division between British and Irish Celtic is that it happened at all. That is to say that two clusters of dialects with a high level of mutual intelligibility within them and a low level of mutual intelligibility between them ever developed. In modern Europe hard language frontiers have been produced by nationalist states, or state nationalism. In the Ancient World they must largely have been the product of the disappearance of Zwischenvolk or the intrusion of alien groups into the dialect continuum, as the Vlachs and Magyars separated the Yugoslavs from their northern Slavonic speech-kin. It may be that the collapse of Celtic across Europe as a whole and the fundamental division that arose between British and Irish was the result of Roman domination with Latin providing an interregional register that encouraged local dialects of Celtic to develop in isolation of one another, losing, eventually, their utility for anything other than local communication. Against such an argument, however, one could point to the apparent solidarity of British Celtic and Gaelic within their own dialect continua. In any case there is some supporting evidence for Campbell‟s assertion that the Highlands presented a greater barrier to free communication and interaction than did the North Channel. Birdwatchers will be aware that these islands play host to two varieties of the common crow Corvus corone. One, Corvus corone corone, the Carrion Crow, is entirely black in plumage whilst the other, Corvus corone cornix, the Hooded Crow, is partly black and partly pale grey. These two crows are simply races of the same species and where their ranges intersect they socialise and interbreed happily. Broadly-speaking the Hooded Crow is the Irish variant but it is also the most common variety found in Atlantic Scotland and occasionally turns up elsewhere in Britain‟s western extremity. The prevalence of the Carrion Crow in the greater Dublin area is perhaps further proof, were it required, of the veracity of a well known Irish prejudice. A distribution map of the ranges of these two types of bird would look very similar to Campbell‟s suggested distribution map of the original extent of Irish and British varieties of Celtic. Since the crows interbreed happily the maintenance of two distinct races is almost certainly to be attributed to geographical constraints and this suggests that Campbell may be correct when he suggests that 5 Atlantic Scotland is closer to Ireland than to the Eastern lowlands. At least, as the crow flies. Against this suggestion that Gaelic, or „Q‟-Celtic had an unbroken presence in Argyll is the evidence derived from Geographer Ptolemy who published his survey of the Roman Empire in about AD 150 but who was dependent largely on pre-existing written accounts. In Ptolemy‟s survey the two areas which were to form the Scottish portion of Dál Riata, the southern Hebrides and Argyll, are dealt with in the Irish section and the British section respectively. Ptolemy ascribes Argyll to the Epidii tribe. This tribal name seems to be a plural form of the personal name Epidios which is formed from epos „horse‟ plus a suffix –idios used to turn a common noun into a deity name or personal name. Epidios is thus a name meaning „Horse-god‟ or „Horsey-one‟ or some such. Now the presence of the „p‟ in this name marks it out, in the form Ptolemy preserves it, as a British name. Had the name Epidios existed in primitive Irish it would have had a form something like *Ekwidios. This hypothetical name is the Primitive Irish form of the name which would become Eochaid. It is interesting and perhaps not coincidence that the legendary ancestor of the Dál Riata, either as Eochaid Munremar in the Míniugud or Eochaid Riata in the pedigrees, shares the name of the tribe who inhabited Argyll in Ptolemy‟s day. W. J. Watson pointed out that the Mull of Kyntire, which Ptolemy designates Epidion Akron, is given the name Ard Echde in the saga Aided Cú Roí. Ptolemy‟s forms are definitely British and if they represent the original then it creates serious problems for Campbell‟s hypothesis. He turns, however, to the explanation put forward by Greg Toner and others for British tribal names in Ptolemy‟s Irish section. Irish and British were at this period, it is argued, so similar that it was easy for speakers of one language to nativize names in the other. Indeed the Irish chroniclers appear to have been able to do this with Pictish names as late as the ninth century so this should not surprise us. This argument, however, can be played both ways. How can we tell whether we are looking at Gaelicised British or Britified Gaelic? If we turn to the Irish section of Ptolemy‟s survey and to the account of the southern Hebrides we may find a clue. Ptolemy refers to five islands off the north coast of Ireland collectively as Ebudae Insulae. The word Ebudae survived into medieval Irish and appears in phrases such as Fir Ibdaig „Men of the Hebrides‟. This word looks very much like an Irish attempt to reproduce the word Epidii phonetically rather than by translating it. Intervocalic /p/ was to become /b/ in British eventually and it may be that this process had already begun in Ptolemy‟s time (his Irish data may be somewhat later than his British or the inability to vocalise /p/ may have encouraged precocious voicing of incipient lenition amongst his Irish informants). If „Ebudae‟ and its correlates do represent Irish attempts reproduce the name of the Epidii then it tells us two things. Firstly that the Epidii like the later Dál Riata had some claim to the islands as well as to the mainland of Argyll and secondly that they were indeed „p- Celtic‟ rather than „q-Celtic‟ speakers in the early Roman period. To summarise this analysis of Ewen Campbell‟s argument we can probably say that the archaeological evidence does not stand up to scrutiny and that the discussion of 6 origin legends is just that and has no bearing on the actually existing past. The linguistic arguments are more plausible but not without some potential problems regarding the Epidii, incontrovertible evidence probably being irrecoverable. Is there any value in pursuing the question further? Cruithni One of the most sensitive topics in the study of late prehistoric and early historical Ireland is that of the population group known as the Cruithin or Cruithni. Their name is the normal word used in medieval Irish for the Picts but it was also used for a group of túatha in the north of the country up until about AD 774. In origin this word is the Irish form of the British word for Britons, Pretani. In medieval Irish the Latin loan word Bretan was used for the Britons south of the Forth and Cruithni was reserved for the less Romanised peoples of the North who were termed Picti in Latin. At one time some historians, including the great Eoin MacNeill, believed that the Pretani were the original inhabitants of both Britain and Ireland and that the Gaels had arrived at a late stage in prehistory displacing them from most of Ireland. According to this argument the Cruithni of northern Ireland were the last remnant of the pre-Gaelic inhabitants of the island. It has now become clear that this view is not supported by linguistic, historical or archaeological evidence. If British-speaking Celts ever did settle in Ireland they must have done so subsequently to the development, in situ, of the Gaelic language. Unfortunately the idea that Northern Ireland was British ab origine has proved attractive to certain elements within the Unionist tradition during the political troubles of that province. As a result „Cruithni Studies‟, to coin a phrase, have become the preserve of Unionist apologists such as Ian Adamson whose most recent book on the Cruithni concludes with a chapter on the Scots-Irish experience in the Appalachians. Serious historians of early Ireland, tending as they do to have nationalist sympathies or to be politically neutral have tended, understandably, to steer clear of the topic. Jim Mallory is typical of most serious scholars when he summarises his brief discussion of the topic thus: “about the only thing the Cruthin hypothesis does emphasise are the continuous interactions between Ulster and Scotland. We might add that whatever their actual origins and ultimate fate, when the Cruthin emerge in our earliest texts they bear Irish names and there is not the slightest hint that they spoke anything other than Irish.” In typically provocative style Professor Dumville alluding to this kind of statement in his, so far unpublished, British Academy Rhŷs Lecture in Edinburgh a few years ago (1997?), asked what the evidence for such an assertion might be. I can only imagine that Dumville was questioning whether we had any texts of Cruthnian provenance and whether we could be certain that Gaelic writers, clearly able to Gaelicise Pictish personal and place names were not doing the same for the Irish Cruithni. Mallory is of course right that there is not the slightest hint that the Cruithni spoke anything other than Irish just as Dumville is correct, if I understood him, that absence of evidence is not evidence of absence, but is this really all that can be said? St Patrick aside, contemporary literary witness in Ireland begins only in the course of the century between AD 550 and AD 650 and it is true that our sources, the chronicles an hagiography, give us only the name of the Cruithni, which appears periodically between 446 and 774, to suggest their foreignness. At the beginning of the sixth century the western frontier of the Cruithni seems to have been in the neighbourhood of the Lough Foyle although by the 570s they had been pushed back beyond the river 7 Bann by the northern Uí Néill. In the East the boundary of the Cruithni seems to have been somewhere in the region of Belfast Lough. Crudely speaking their territory at the dawn of history was equivalent to the modern day counties of Antrim and Londonderry. To the West were the Uí Néill, to the South the Airgialla and to the East the Ulaid. In the middle of this territory, pushed up between the Bush valley and the north coast lay the lands usually assigned to the Dál Riata in Ireland by modern scholarship. This enclave was entirely surrounded by Cruithni túatha. Is it not odd that the most Irish people in Britain were, in their Irish territories, surrounded by those Irish people who were described by their countrymen as British? Can it be coincidence? The simplest explanation of this paradox would be to assume that, pace the later synthetic historians and genealogists the Dál Riata and Cruithni were in origin two parts of the same people, perhaps ultimately British in origin, who formed a political, cultural and linguistic bridge between the two islands. It is this proposition that I wish to explore in what remains of this paper. A key incident in the history of the north of Ireland, if we are to believe the chronicle record, was the battle of Móin Daire Lothair fought in AD 563, somewhere near the Bann in which, according to the annals of Ulster, the northern Uí Néill came to the aid of a Cruithnian king Baetán mac Cuinn who was leading two túatha of the Cruithni against the Cruithnian over-king Aed Brecc. Baetán and his allies won the bettle killing Aed and seven other kings of the Cruithni whilst at least one other, Eochaid Láeb, escaped the battle field. All this puts at least eleven Cruithnian kings at the battle, giving some idea of the scale of the federation at this time. The annal goes on to relate that as a result Lee and Ard Eolarg, roughly the land between Coleraine and Magilligan‟s Point, was handed over to the Uí Néill. In the same year, Adomnán tells us, Colum Cille, a cousin of the victorious Uí Néill kings, went to the court of king Conall in Britain. The possibility that, like the alliance with Baetán, the opening up of a channel to Conall was part of the northern Uí Néill Drang nach Osten cannot be discounted. Further hint of links between the Cruithni and the Dál Riata might perhaps be found in the career of Aedán mac Gabráin. Aedán‟s horizons, both literary and political, as they have come down to us, far outstripped those of any other ruler of Dál Riata, both in Britain and in Ireland. The chronicles record battles fought in Orkney, Manaw, the unidentified „Leithri‟ and perhaps two battles against the Angles of Bernicia. Adomnán adds a further battle in „Circinn‟, a Pictish province. Unless Leithri was fought in Ireland none of Aedán‟s recorded battles were. The eleventh century Senchus Sil hÍr contains, twice, an allusion to the king of the Ulaid, Baetán mac Cairill (c.572-581), taking submission from Aedan at Rinn Seimne on Island Magee. The location must be very close to the boundary between the Cruithni and the Ulaid and if it was indeed a meeting on a border the idea that it marked a full submission may be an optimistic gloss supplied by an Ulidian tradition or author. The meeting between Aedán and Baetán at Rinn Seimne (which must have taken place, if at all, between 574, the date of Aedán‟s accession, and 581, or possibly 587, the date of Baetán‟s death) should probably be considered alongside another far better attested and better known royal meeting that Aedán attended. Adomnán describes a condcitum regum – an agreement or conference of kings – “namely Aed mac Ainmirech and Aedán mac Gabráin” at a place called Druim Cett, a ridge by the river Roe near Limavady. Although the Annals of Ulster date this meeting to 575 Michael Meckler has argued that this entry is a twelfth century addition to the chronicle and Richard 8 Sharpe has noted that Aed mac Ainmirech did not become Uí Néill over-king until 586. Bart Jaski has urged caution in revising the dates but these details need not affect the argument here. Taken together these two meetings of kings put Aedan at the eastern and western boundary of the Cruithni territory meeting the over-kings whose imperia bordered that territory. Whilst Bannerman and Byrne interpreted these meetings a narrative sequence in which Aedán and the Uí Néill allied against the threat posed by Baetán I would rather suggest that both these meetings mark the recognition of Aedán‟s over-kingship of the Cruithni by his neighbouring peers. The gaps in the narrative can be filled in with reference to the Dál nAraide regnal-list preserved in the Book of Leinster and Adomnán‟s Vita Columbae. The Dál nAraide were the Cruithni túath who eventually solidified their over-kingship of the group and whose name came to replace that of the Cruithni in the chronicle record in the late eighth century. Their rise to power is a classic case of a loose federation being gradually welded into a dynastic kingdom and deserves further study in its own right. In the course of the Viking Age they were even, at times, able to extract submission from Dál Fiatach, who traditionally dominated the Ulaid, and the rewriting of the relationship between these two groups, that took place between the tenth and twelfth centuries, and which inserted the Cruithni into the history of the Ulaid has cast a druidical mist over the subject which, despite T. F. O‟Rahilly‟s best efforts, still obscures the vision of some of our clearest sighted colleagues. The Dál nAraide king-list is remarkable not least because almost none of the kings who are named in it are sons of previous kings in the period before the ninth century. It claims for Dál nAraide that Aed Brecc whom we saw as the over-king of the Cruithni who was slain at Móin Daire Lothair in 563. His successor is named as Aed Dub mac Subni. Aed‟s death is noted in AU at 588 but we know from Adomnán that he was expelled from the kingship long before this and spent time in exile as a monk and even a priest. His successor in the king list is named as Fiachna Lurgan who was killed by the Ulaid in 626. While it is not impossible that Fiachna‟s kingship might have extended from Aed Dub‟s expulsion this would be a remarkably long reign and Fiachna‟s career as related in the annals begins only in the late 590s. It is possible that the prequel to the royal meetings at Rinn Seimne and Druim Cett were the battles of Tola and Fertola in the territories of the Cruithin recorded in AU s.a. 574 and that this marked a Dál Riatan expulsion of Aed, and the assertion of Aedán‟s overlordship. Indeed the saga Compert Mongain seems to explicitly admit Fiachna Lurgan‟s tributary status in relation to Aedán. The interpretation of the meeting at Druim Cett put forward here also lends added weight to the story about Colum Cille and Comgall of Bangor‟s return journey from the council. The two saint stopped to drink at a spring near the fortress of Dún Cethirn, near Coleraine. Columba tells Comgall that the spring would one day be filled with blood when his kinsmen, the Uí Néill, fought against those of Comgall, the Cruithni, at that place. The battle he was prophesying would not take place until 629 but it is also likely that the connection with the condictum regum has been made by Adomnán because it was generally understood to have been a peace agreement made between the Cruithni and the Uí Néill, perhaps even brokered by the two abbots. Their sanctity is underpinned by the proposal that in their lifetimes the peace process could make progress whereas in modern times, as it were, lesser kings, bereft of the counsel of saints, have returned, like dogs, to their vomit. 9 Aedán mac Gabráin‟s military adventures in Britain can perhaps be accounted for by the twofold benefit of the situation in the north of Ireland. Firstly he seems to have reached an accommodation with the two major over-kings of the region, who were at this time more interested in conflict in the Irish midlands, which left that flank protected. Secondly he was able to call on a much larger military following than most subsequent kings of Dál Riata through his over-kingship of the Cruithni. This extended overlordship in Ireland, claimed by kings of Dál Riata may also be hinted at in the opening sections of the Míniugud. This text, surviving in four medieval manuscripts, and most recently edited by Professor Dumville, is, in its extant form, a product of heavy editing carried out between the tenth and the twelfth centuries with the aim of stressing certain links between Dál Riata of the seventh century and the actually existing kingdom of Alba. Nonetheless a genuine seventh century core seems to lie behind it, and the first part of the text is a typical genealogical framework which seeks to explain, in lineage terms, the political relationships of the seventh century. The apical figure of this genealogical framework is a man named Eochaid Munremar. He is given two sons, Erc and Ólchú. Each of these is given, in turn, about twelve sons (the precise figures vary slightly in the manuscripts but a symmetrical structure was probably intended in the original). E are then told that six of Erc‟s sons lived in Scotland but that all of Ólchú‟s and half of Erc‟s set up home in Ireland where their descendants still live. Eochaid Munremar is clearly some kind of doublet of Eochaid Riata, the original „Epidios‟ perhaps, and what we are looking at is a genealogical explanation of the over-kingship claimed by Aedán and his descendants. As David Dumville has pointed out eighteen out of the twenty-four grandsons of Eochaid established their kindreds in Ireland, three quarters of the whole, a fact which the surviving recensions of the text, and the Bannerman edition, it must be said, obscure by following only a handful of lines, those which remained active in Alpinid Scotland, beyond the first generation or two. This strongly suggests that the seventh century version of this genealogical tract was produced to bolster the claim to an overlordship in which the Irish portion of the population was perhaps, in crude terms, three times larger than the Scottish. Almost certainly we are not meant to imagine the little cluster of baronies on the north coast which are usually ascribed to Dál Riata in Ireland. Instead, I would suggest, despite the disagreement with later tracts, such as the Senchus Sil hÍr, that we have here a remnant of an early Cruithnian schema, covering all the túatha between the Lagan and the Foyle. Aedán‟s son Eochaid Buide seems to have inherited his father‟s overkingship for following the killing of Fiachna Lurgan, the Cruithnian king of Mag Line by his name sake Fiachna mac Demáin, king of the Ulaid, in 626 the Dál Riata rígdamnae Connad Cerr, perhaps Eochaid‟s son, was sent to avenge him and slew the Ulidian at Ard Corann, not so far from Rinn Seimne. The attribution of the title Rex Pictorum to Eochaid in the Book of Cuanu may, perhaps, reflect this later writers misunderstanding of the now obsolete term Cruithni. The supremacy of Dál Riata amongst the Cruithne was not, however, to survive Eochaid Buide and the Dál nAraide kings of Mag Line gradually came to assert their dominance, their own name eventually supplanting that of the Cruithni. 10 Conclusion I have suggested that the Dál Riata were part of the Cruithni and further that they in turn may have been in some sense derived from Ptolemy‟s Epidii. The use of the term Cruithni in Ireland but not in Scotland may relate to dichotomization. It is unlikely that Cruithni was ever a self identifying term for these people but rather the term used by their Irish neighbours to designate them. Mallory is probably correct that by the age of chronicles and hagiography there was little that was objectively British about them and the name may have been „residual‟. A relic from an earlier age. What they called themselves is lost to us, perhaps something like *Sil Echdach or some other Old Irish term referring to an apical figure. The frequent occurenceof the name Eochaid and the oft-confused Eochu in this region, the latter even preserved in the name of Loch nEchach/ Lough Neagh may reflect such an ancestry. The Gaelicization of the Epidii, if this is indeed what happened, is not hard to explain if we recognise that any migration from Britain to Ireland is likely to have occurred before the majority of the sound shifts that distinguish British and Irish had occurred. In linguistic terms the two varieties of Celtic were probably close enough to converge rather than to require formal language shift, in much the same way as the North Germanic Anglian and Jutish dialects converged with West Germanic Saxon to produce Old English, or as Scots and Standard English have converged to produce modern Scottish English.. One possibility which serious linguist may want to consider is whether the distinctive features which mark out Scottish Gaelic and the northernmost dialects of Irish, which David Greene and others suggested showed British influence might date to this period rather than to the later take over of the Picts which one might not have expected to have effected the surviving Scottish Gaelic dialects that much whatever it might have done to the now defunct eastern varieties. As for the origins of Gaelic Argyll I would suggest that a convergence dialect emerging in the north of Ireland may well have become the aspirational acrolect for the whole Epidian tribal grouping and that as this dialect went through the Neo-Celtic transition in partnership with Irish so speakers on both sides of the sea shared those changes. In short I am suggesting that the Irish language was brought to Scotland by people migrating from Scotland to Ireland. When might such a migration have taken place? A long time ago! The terminus post quem would be the gathering of the data that went into Ptolemy‟s survey, so perhaps Agricola‟s campaigns of the 70s and 80s of the First Millennium. The terminus anti quem would be the historical horizon of the earliest Irish chroniclers for they felt no need to mention it in their narrative of post-Patrician events. The earlier one places it the fewer the linguistic impediments, presumably? Archaeology may provide hints. As Ewen Campbell pointed out some common phenomena of Early Christian Irish archaeology, such as crannóga do have their origins in Scotland and have a deep history in Argyll – and maybe the ringfort originated as a beached crannog, which it resembles in plan. In more specific terms one cannot help but draw attention to the Severan coin hordes of County Antrim – almost certainly indicating some sort of relationship of clientage between recipients and Romans. Did the gift take place in Ireland or were the coins brought across from Britain after the transaction? Was the Epidian conquest sponsored by the Empire? Or did the Epidii turn there attention to Antrim because their traditional raiding grounds had found powerful protectors? Unless a remarkably explicit archaeological find comes to light it is unlikely we shall ever know the answers to these questions. 11