by W.D.H. SELLAR
30th January 1998
[Transactions of the Gaelic Society of Inverness, Vol. LV, 1997-1998, The Gaelic Society of Inverness, 2000, pp. 233-258.]
[NOTE: Click on the Footnote Numbers to read the footnotes; Click on the number again to return to the main body of text.]
Even among Highland clans the origins of the MacLeods has proved notoriously controversial. The question has been debated at length in the standard histories of the clan, Canon R. C. MacLeods The MacLeods of Dunvegan, and I. F. Grants splendid The MacLeods:The History of a Clan. More recently two honorary chieftains of the Gaelic Society of Inverness, Alick Morrison and the late Rev. William Matheson, have written on the subject from strongly opposed points of view. The MacLeods have always gloried in a high Scandinavian ancestry. The nub of the problem concerns the further claim that they descend directly from the thirteenth century kings of Man and the Isles. For some time this claim won a measure of acceptance, both R. C. MacLeod and I. F. Grant holding that Leod, the eponym of the clan, was a son of Olaf the Black (d. 1237), king of Man and the Isles. In 1977 William Matheson challenged that claim, arguing that the MacLeods descended not from the kings of Man and the Isles, but from a Norseman named Olvir, whom he identified tentatively with a character in Orkneyinga Saga called Olvir Rosta. Ten years later Alick Morrison launched an attack on Matheson in turn, arguing once again for descent from Olav the Black. This paper follows Matheson for the most part, although differing from him in some respects. In particular, it suggests a different context for Olvir and reintroduces a relationship with the kings of Man.
Toward the end of the eleventh century the kingdom of Man and the Isles (or Innsegall) came under the rule of the dynasty of Godfrey (Gofraidh, Godred) Crovan. Godfrey, who was also for a time king of Dublin, is said by the Chronicle of Man to have fought at Stamford Bridge in 1066 with Harald Hardrada, and to have ruled over the Isles for about sixteen years before his death in Isla circa 1095. The Annals of Inisfallen record his death in that year as rí Innsi Gall. Godfreys official style as ruler of the Hebrides is not known. His son and eventual successor, Olaf the Red, however, styles himself in his charters rex insularum (king of the Isles), the Latin equivalent of the Gaelic rí Innsegall. As is well known, the rule of Olafs son, another Godfrey, was challenged by his sisters husband Somerled of Argyll, acting on behalf of his son (and Godfreys nephew) Dugald. As a result, the kingdom of the Isles was partitioned, with the descendants of Godfrey retaining the Isle of Man, together with the northern isles of Skye and Lewis, and the descendants of Somerled taking possession of the southern Hebrides and possibly also the Uists. Both dynasties continued to use the style rex insularum.
Godfrey, son of Olaf the Red, left two sons by different mothers, Ragnall and Olaf the Black, who vied with each other over many years for the kingship of Man and the northern Isles. Ragnall is the subject of a famous Gaelic praise-poem, and either he or his cousin and namesake, Ragnall son of Somerled, is praised in Orkneyinga Saga,, and famously described there as the greatest warrior then in the Western lands. The fullest account of the reigns and rivalry of Ragnall and Olaf the Black is given in the Chronicle of Man, written in the mid-thirteenth century, and very much the in-house record of the Manx kings. As many of the dramatis personae in this story figure in later MacLeod tradition, it is worth considering here in some detail. According to the Manx Chronicle Olaf was expelled by Ragnall from the kingdom after a succession dispute, and kept in custody by King William of Scotland. After Williams death in 1214, Olaf was released, and allowed by Ragnall to rule in Lewis. Olaf then married Lauon, or Lyanuon, of Kintyre, sister of the wife of Ragnall. The bishop of the Isles, however, objected to the marriage as being within the degrees of affinity prohibited by Canon law, on the ground that Olaf had previously had intercourse with a cousin of Lyauon. Olaf dismissed his wife accordingly, treating the marriage as a nullity, then married Christina, daughter of Ferchar, earl of Ross.
Ragnalls wife (whose name is not given) took considerable offense at this insult to her sister, and sent letters secretly in Ragnalls name to their son Godfrey Donn (the Brown) in Skye, directing him to seize and kill Olaf. Godfrey collected some men and set off for Lewis. Olaf, however, narrowly escaped and fled, according to the Chronicle, to his father-in-law the earl of Ross. The vicecomes of Skye, Paul son of Balki (pol filius boke), described as a vigorous and powerful man throughout the kingdom of the Isles (vir strenuus & potens in omni regno insularum) wished no part in the murder of Olaf. He too fled to the earl of Ross where he entered into an alliance with Olaf. The title vicecomes given to Paul Balkason by the Manx Chronicle has sometimes been translated as sheriff, vicecomes being the regular Latin form for sheriff in later medieval Scotland. However, it seems more likely that behind the vicecomes of the Chronicle lies some Scandinavian title, such as syslumadh or armadhr (Gaelic armann). In any case, Paul Balkason was probably the official representative, or governor, of the king of Man in Skye. As will be seen, he was long remembered in later Gaelic tradition.
Olaf and Paul came secretly to Skye and learnt that Godfrey was staying with only a few men on a certain island called the isle of St. Columba (in quadam insula que vocatur insula sancti columbe). Under cover of night they dragged five ships (tractis quinque navibus) from the nearest shore of the sea about two stadia distant, and encircled the island. When Godfrey and his men awoke in the morning, they were amazed to see that they were surrounded. Although outnumbered they put up a brave resistance but were eventually defeated. Olaf and Paul put to death everyone they found outside the bounds of the church. Godfrey Donn himself was seized, blinded and castrated. The Chronicle states that this was done against Olafs wishes, and at the bidding of Paul Balkason. The year is given as 1223. The contemporary Icelandic Annals confirm both deed and date, recording that in 1223, Olaf, Godfreys son, king of the Hebrides, caused to be blinded his brothers son Godfrey, son of Reginald [Ragnall], king of Man. Despite his injuries, Godfrey remained active, and was appointed king in the Hebrides by Haakon, king of Norway in 1230. He and Olaf subsequently agreed to divide Man and the Isles between them, the northern Hebrides falling to Godfreys share. Shortly afterwards Godfrey killed Paul Balkason in the Hebrides, only to meet his own death in Lewis. Godfreys son Harald was later briefly king of Man and the Isles.
The story of the blinding and castration of Godfrey Donn was long remembered in tradition. There is an extremely garbled account of it, displaced by eighty years or so, in Hugh Macdonalds History of the MacDonald, where he writes that Somerled and Olaf the Red ... killed Godfrey Du, or the Black, by putting out his eyes, which was done by the hermit MacPoke, because Godfrey Du had killed his father. For Hughs Olaf the Red, read Olaf the Black, for Godfrey Du read Godfrey Donn, and for the hermit MacPoke read Paul Balkason. Paul was remembered in later tradition as Pall mac Phaic. Some of his descendants settled in Bernera, Harris, where they adopted the surname MacKillop. Some remain there to this day. More surprisingly, Paul was still remembered this century in Kilmuir, Skye as Fear Caisteal Eilein Chaluim Chille (the man of the castle of Eilean Chaluim Chille). This raised the question of the location of the island of St. Columba (insula sancte columbe) of the Chronicle.
It has sometimes been suggested that the island in question is Iona, the most famous island of St. Columba; but, apart from the fact that Iona is more commonly referred to as I Chaluim Chille, rather than Eilean Chaluim Chille, a close reading of the Manx Chronicle makes it clear that the action took place somewhere on Skye. At least three separate locations in Skye have been put forward as the scene of the blinding. Alick Morrison has suggested an Island of St. Colm by Portree but gives no supporting reasons. A stronger candidate is the Island of St Columba which lies in the middle of the river Snizort, a few hundred yards from where the river meets the sea. This island was for a time the site of a medieval cathedral of the Isles, the ruins of which can still be seen. It is the traditional burial place of -- among others -- the Nicolsons (MacNeacail) of Scorrybreac, a family like the MacLeods long established in Skye, and like them of Norse descent. It seems to conform to the description in the Chronicle in that it lies only a few hundred yards away from the sea. However, it is not clear why Godfrey and his men should have been encamped on what was presumably even then ecclesiastical ground; nor does the description of five ships encircling this island ring true. If Godfrey had been encamped on the island at Snizort, the obvious tactic of his assailants would have been to anchor their ships at the head of the Loch and to proceed from there on foot, rather than to navigate their ships a few hundred yards up a fast flowing and treacherous river. In addition, this location does not explain the phrase tractis quinque navibus.
In fact, Olaf the Black and Paul Balkason almost certainly surprised Godfrey at Eilean Chaluim Chille in Kilmuir, which until the eighteenth century lay in the middle of the now drained Loch Chaluim Chille. This is the Eilean Chaluim Chille which was long associated in oral tradition with Pall mac Phaic, as noted above. Its situation fits the description in the Manx Chronicle perfectly, for the northern end of the former loch lies only two or three hundred yards from the small harbour of Bornaskitaig. Olaf and Paul could easily have landed at Bornaskitaig and dragged five ships the few hundred yards from the sea to Loch Chaluim Chille at dead of night, ready to surprise Godfrey Donn, encamped on an island in the loch, in the morning, just as the Chronicle relates. The Royal Commission report, compiled over seventy years ago, notes that there must once have been two island in Loch Chaluim Chille: a cashel, the wall of which enclosed the entire area of one of the islands, and seventy yards away, but probably once linked by a causeway, another island with on it the remains of a church. The presence of a church as well as a fortified site tallies will with the account in the Chronicle.
The ancestry of Godfrey Crovan, itself a subject of considerable speculation, is also relevant to the origins of the MacLeods. All are agreed that Godfrey is likely to have been of royal Scandinavian stock and related to other Scandinavian kings in Ireland and the Hebrides. The only hard evidence as to his immediate paternity, however, for long was thought to be the passage in the Chronicle of Man, in which he is described as filius haraldi nigri de ysland, that is, son of Harald the Black of (or from) Ysland. But where is Ysland? Both Iceland and Islay have been suggested, and sometimes even island. I have little doubt that the place in question is Iceland. It is certainly unlikely to be Islay, not only because the intrusive and unpronounced s in Islay belongs to a later period, but also because Islay is, in fact, referred to elsewhere in the Chronicle, as Yle. Thus the death of Godfrey Crovan is recorded in the island que vocatur yle. Why Harald the Black was associated with Iceland can only be a matter for speculation, although it is not difficult to imagine an appropriate scenario, Iceland being then an integral part of the wider Scandinavian world. A number of suggestions have been made as to Haralds possible affiliation. Some have derived him from the dynasty of Maccus and Godfrey, sons of Harald, rulers of the Hebrides in the tenth century. Others have suggested descent from the Scandinavian kings of Dublin. The various theories were reviewed by Canon R. C. MacLeod and I. F. Grant, among others. More recently, however, attention has been drawn to medieval Welsh genealogical tradition as throwing light on the origins of Godred Crovan. This is considered at the end of this paper. The last king of Godfrey Crovans dynasty was Magnus (d. 1265), son of Olaf the Black. The last known male of the dynasty as to whose relationship there can be no doubt was Godfrey (killed 1275), illegitimate son of Magnus.
The MacDonald Lords of the Isles eventually succeeded to the title rì Innsegall, previously borne by both the Manx dynasty and by the immediate descendants of Somerled, although Lewis and Skye, ruled from Man until 1266, were incorporated into the MacDonald territories relatively late. Prominent among the clans or families which followed the later MacDonald lords were the two main branches of the MacLeods, the MacLeods of Harris and Dunvegan (who later adopted the style MacLeods of MacLeod) and the MacLeods of Lewis, known respectively in Gaelic as the Siol Tormoid and the Siol Torcaill -- the descendants of Tormod (otherwise Norman) and the descendants of Torquil. The earliest contemporary references to MacLeod chiefs are two charters granted about 1343 by King David II to Malcolm MacLeod of the lands Glenelg, and to Torquil MacLeod of the lands of Assynt. Significantly both grants relate to land on the mainland of Scotland rather than in the Isles, Glenelg being held by the Harris MacLeods, and Assynt by the MacLeods of Lewis and their cadets. The Siol Tormoid and the Siol Torcaill functioned effectively as separate clans, each under its own chief, neither holding their principal lands of the other, with separate coats of arms and separate places of burial; they even, on occasion, took separate sides in island conflicts. It is a sign of the aspirations and importance of both these MacLeod branches that their arms, together with those of the Lords of the Isles, appear in the Armorial de Berry, a leading European mid-15th century armorial, being the only Hebridean arms given: the MacLeods of Harris bear a castle in their arms (presumably for Dunvegan), and the MacLeods of Lewis a mountain on fire.[24 ]
Traditionally the Siol Tormoid and the Siol Torcaill descend from and are named after two brothers, Tormod and Torquil, said to be sons of Leod, the eponym of the clan. However, in his paper to the Gaelic Society on The MacLeods of Lewis, William Matheson argued convincingly that the first Torquil of Lewis (from whom the Siol Torcaill take its name) was not a brother of the first Tormod, son of Leod, but rather his grandson by a son Murchadh (or Murdo[ch]) who had married a MacNeacail (Nicolson) heiress. An important plank in this argument is provided by two Irish genealogical manuscripts which give the Gaelic pedigree of the MacLeods of Lewis -- the only such Lewis pedigrees surviving -- starting with Ruairi, chief of Lewis c. 1400, and tracing his ancestry back to Leod and beyond, before moving into a Nicolson pedigree. These pedigrees will be considered in greater detail later. The Nicolson connection appears to reflect the fact that wherever MacLeods of Lewis are to be found in the later Middle Ages -- in Lewis itself, in Wester Ross and in Waternish in Skye, for example -- there are traditions of the Nicolsons having preceded them. I believe, with Matheson, that the MacLeods of Lewis stepped into the shoes of the Nicolsons through marriage with an heiress, and that this explains their identity as a clan separate from the MacLeods of Harris and Dunvegan.
Turning now more specifically to the ancestry of Leod, the MacLeod eponym, it is notorious that the MacLeods, in their poetry and traditions, have always celebrated and gloried in their Scandinavian descent. We may start with the old and oft-repeated statement, not confined to the MacLeods, and occurring in many variations, that Gunn, Leod and Leandres were the three sons of the king of Denmark. This statement is not to be taken literally; rather, it is a way of expressing the belief that prominent among Scottish clans or families of Scandinavian origin -- Denmark standing for Scandinavia generally --are the Gunns, the MacLeods and the descendants of Leandres. These last are the Clan Gillanders or Anrias, many of whom later adopted the surname of Ross. Their most prominent member was Paul MacTyre, who lived in the fourteenth century, and was well remembered in later tradition. Some of his descendants, incidentally, took the name of Fraser and became barons of Moniack; and from these Fraser barons in turn, I understand, descend some of the name of Barron in the Aird district, including the long serving secretary of this Society. There were, of course, other families in the Highlands who claimed Scandinavian descent, apart from the three groups mentioned: notably, in the context of this paper, the Nicolsons (MacNeacail) of Scorrybreac, who continued to use the distinctive Scandinavian forename Arnketil in the Gaelic form Armchul until this century.
An important witness to the ancestry of the MacLeods is Sir George MacKenzie, first earl of Cromarty (1630-1714), grandson of Roderick (Ruairi) MacKenzie of Coigeach, the dreaded tutor of Kintail, and of his wife, Margaret, the elder daughter and heiress of Torquil Cononach; Torquil in turn being the eldest son, if son he was, of Ruairi MacLeod, the last effective chief of the MacLeods of Lewis. At the end of his History of the Family of MacKenzie, written about 1669, Sir George appended a short history of the MacLeods. In it he declared Leod (Leodus) to have been a son of Harald, king of Man (1249-50), son of Godfrey Donn: To Harald succeidit Leodus his only son. Sir George also asserted that after Man came into Scottish possession, Alexander III granted Lewis, Harris, Waternish and Minginish to Harald. Leods wife, according to Sir George, was Adama, daughter of Ferchar, earl of Ross.
Sir George, however, is not a credible witness. His false ascription of a Geraldine origin to the MacKenzies, deriving them from Colin FitzGerald, earl of Kildare, rather than from Gilleoin of the Aird has long been known and exposed. As Matheson pointed out, Sir George was familiar with and used the Chronicle of Man, the text of which had recently become available through Camdens Britannia. Sir George knew, therefore, of the marriage said by the Chronicle to have taken place between Olaf the Black and Christina, daughter of Ferchar, earl of Ross -- indeed, he mentions it (although calling Christina a sister of Ferchar) -- but he does not derive Leod from this marriage, or affiliate him to Olaf. There is no earlier evidence for the existence of a son of Harald named Leod, or a daughter of Ferchar, earl of Ross named Adama, and none is likely to be found. Nor is there any trace of a grant of lands by Alexander III, as described. All are figments of Sir Georges imagination.
Sir George also seems to have been the first writer to declare Torquil and Tormod (Torkell and Dormeth) to have been brothers. Characteristically he suggests that Torquil, the ancestor of the MacLeods of Lewis, from whom he himself descended, and arguably represented, was the elder brother! His subsequent account of the MacLeods of Lewis is sketchy as regards the earlier generations, and partial as regards the later ones.
Later MacLeod accounts favour a descent from Olaf the Black, by his wife Christina, daughter of Ferchar, earl of Ross, an option clearly rejected by Sir George MacKenzie. Thus the Assynt MS (c. 1680) gives Leod as the son of Olaf; while a Memorial of 1767 narrates that Olave and Christina had three sons, named Gunn, Leod and Leandres! Canon R. C. MacLeod reviewed the various accounts, and accepted Olaf as the father of Leod. However, as he believed that Leod must have been born about 1200, too early for the marriage between Olaf and Christina, he suggest that Leod must have been a son of an earlier (and otherwise unknown) first marriage of Olaf the Black. This account, as can readily be appreciated, is at variance with all the sources, and would presumably make Leod older than the well attested sons of Olaf who later succeeded him as kings of Man and the Isles. One strand in Canon MacLeods argument, and indeed in the acceptance of Tormod and Torquil as brothers, is a supposed charter of Donald, king of the Isles, dated 1245, unquestionably a forgery, although believed by Canon MacLeod to be genuine, which names as witnesses both MacLeod of Lewis and MacLeod of Harris. I. F. Grant defers to Canon MacLeods opinion on the parentage of Leod, but notes that the charter of 1245 is of doubtful authenticity.
There are many reasons why these accounts of MacLeod origins are not to be believed. Most of them have already been put forward by Matheson in his two articles on the subject in the Societys Transactions. There is absolutely no contemporary evidence, or anything approaching contemporary evidence, in favour of Leod being a son of Olaf the Black, or of Harald, son of Godfrey Donn, and a good deal of circumstantial evidence against. One can, in fact, see the story building up, and becoming ever more elaborate, from the time of Sir George MacKenzies original account, so clearly inspired by a reading of the Chronicle of Man.
There are other arguments additional to those put forward by Matheson for not accepting these accounts. In the first place, if the MacLeods were so closely related to the last Manx kings, why did they not put forward a claim to Man? Other claimants are know. Magnus, last king of Man, and son of Olaf the Black, died in 1265, as has been noted, and his illegitimate son Godfrey died leading a revolt in 1275. At least two later claimants to Man, both female, are known. In 1293 Affrica, claiming to be the cousin and heir of Magnus, appealed her case from John Balliol, as kings of Scots, to Edward I. In the same year another claimant named Mary was referred by Edward to John Balliol. Her claim was later pursued by her grandson, John Waldeboef, in 1305. It is difficult to believe that Leod, or his son Tormod, would not also have claimed if they represented the continuing male line of the kings of man.
Another argument against the MacLeods descending directly from, or representing the kings of Man is the fact that they apparently did not adopt or difference the arms of Man until the later seventeenth century, that is, about the time Sir George MacKenzie was writing his account. These arms, the distinctive three legs of Man, were already born by the kings of Man in the thirteenth century. Yet the early arms of both branches of the MacLeods, recorded in the fifteenth century in the Armorial de Berry, were quite different, as has been seen. A further argument against the proposed descent is that the MacLeods did not adopt as forenames the royal Norwegian names favoured in the twelfth and thirteenth century Manx kings: names such as Godfrey, Ragnall, Olaf and Harald. By way of contrast, the MacDonald and MacDougall descendants of Somerled by Ragnhild, daughter of Olaf the Red, did adopt these names, or some of them, and have continued to bear them to the present day.
I am therefore, firmly of the same opinion as Matheson (and pace Morrison) that the claimed descent from the later kings of Man, more particularly from Olaf the Black, has been fabricated. There is, however, further evidence relevant to the origins of the MacLeods: the evidence of Gaelic praise poetry and genealogies. These too were considered in some detail by William Matheson. The praise poetry certainly claims a Scandinavian descent for the MacLeods and seems to point to a relationship with the kings of Man. The famous poetess, Mary MacLeod (Màiri nighean Alasdair Ruaidh), writing in the second half of the seventeenth century, refers to a notable ancestor name Olghair on several occasions. For example, she compliments her patron Sir Norman MacLeod of Bernera on his noble ancestry: Sir Tormod mo rùin Olgharach thu; and describes Norman chief of MacLeod (1699-1706) as de shliochd Olghair nan Iann. Mary MacLeod also refers to the MacLeods as being descended from kings who laid the Isle of Man under tribute: de sloineadh nan righrean leis no chiosaicheadh Manainn. The Rev. James Fraser, minister of Wardlaw (Kirkhill), writing his history of the Frasers about the same period, refers to the MacLeods, in Gaelic, as Sliochd Oliver, and to Leod, in Latin, as the son of Oliverius Norwegie. Another contemporary, the Rev. John Morison (Bragar(, referred rather extravagantly to Leod as Claudius the sone of Olipheous ... the king of Noruway his son.
A number of traditional Gaelic genealogies of the MacLeods are known. Four of the most significant are shown on Table 2, lettered A to D. Pedigree A comes from Duald MacFirbiss great collection of genealogies. Tacked on to an account of the early Macleans is the pedigree of Christina MacLeod, wife of Hector Maclean (Eachann Reaganach), who flourished in the second half of the fourteenth century, and founded the branch of Lochbuie. Christina is given as the great grand-daughter of Leod through his son Tormod and grandson Murchadh or Murdoch. Her son Murdoch was clearly named for his MacLeod grandfather, and the forename Murdoch (or Murdo) has been used by the MacLaines of Lochbuie ever since. W. F. Skene printed this genealogy in his Celtic Scotland, but unfortunately read the earliest name in the pedigree as Semmair, rather than Sen Imair (that is, Ivar the Old), thereby giving rise to quite false trail.
Pedigree B is the MacLeod of Lewis pedigree in the Royal Irish Academy manuscripts 23 H 22 and 23 G 4 which runs back from Ruairi MacLeod of Lewis, chief about 1400, and continues into a Nicolson (MacNeacail) of Scorrybreck pedigree. I brought 23 H 22 to the attention of William Matheson, and it is discussed by him in his article on the MacLeods of Lewis. In its current form it dates from the early nineteenth century. The Nicolson portion of the pedigree runs on directly from that of the MacLeods without a break. It probably begins with the name Angus which appears four generations above Leod, and certainly no later than one name beyond that. Although not separately identified in the manuscript, the Nicolson pedigree can be recognized as such because of it correspondence with the Nicolson genealogy given in the celebrated Highland MS 1467 (NLS MS 72.1.1), also printed by Skene in Celtic Scotland. Manuscript 23 G 4 was kindly brought to my attention by Kenneth Nicholls of University College, Cork, who informs me that all the material in 23 H 22 seems to be copied from it, or at least from its original, which Nicholls believes to date from about 1600. As the relevant portion of the pedigree in these two manuscripts is identical, save that 23 G 4 abbreviates two names. I have presented them as a single pedigree.
Pedigree C was again printed by Skene, under the name of the Kilbride MS, in Collectionea de Rebus Albanicis and again in Celtic Scotland. The original now seems to be lost. This pedigree is traced back to Iamhar Atacliadh, that is Ivar of Dublin, and for many names beyond. Pedigree D comes from a pedigree of, and presumably composed for, Sir Norman (or Tormod) MacLeod of Bernera (c. 1614-1705), perhaps by Cathal MacVurich. This pedigree, like the MacLeod/Nicolson pedigrees, gives the name Angus four generations above Leod, and may, therefore, belong to the same genealogical tradition.
A further pedigree, nominally MacLeod, not shown on Table 2, is also given by MacFirbis, and comes from the tract On the Fomorians and the Norsemen. This is a highly inventive genealogy, mostly gobbledygook, which includes Malcolm Canmore, Alpin and Loarn in the generations below Leod, and Scandlain Sgainde (who also appears in pedigree C as Sgoinne Sganlain); Arthur and Alexander, and man others, in the generations above Leod. It finishes in fine style with Old Ivar the Great of the Judgments, from whom descent the race of Old Ivar in Scotland, Ireland and Scandinavia. The reference to Old Ivar is significant, as emphasizing the general belief in the Scandinavian origins of the MacLeods.
These genealogies constitute the best evidence now available for the origins of the MacLeods. Pedigree D dates from the later seventeenth century. The fact that it is likely to have been composed for Sir Norman MacLeod of Bernera, a well known and discriminating patron of Gaelic poetry and tradition, gives it extra authority. Pedigree C, the Kilbride manuscript, was dated by Skene to about 1540. The first six names in the pedigree (that is, those later than Leod) have been carefully erased, probably by a partisan of the rival house. Some of the earlier names in this pedigree appear to be slightly garbled. The original behind B, the MacLeod/Nicolson pedigree, presumably dates from around 1400, the time of Ruairi, chief of Lewis, with whom the pedigree begins. The original of A, the pedigree of Christina, who married Hector Maclean, may reasonably be presumed to date from the time of that marriage, clearly regarded by the Macleans as a prestigious one, and thus from the mid-fourteenth century. Pedigree A, therefore, represents the earliest account of MacLeod origins we now possess, removed by only three generations from Leod himself.
In fact, the four pedigrees correspond closely, allowing for the garbled nature of C. Leods great grandfather Olbhar seems to be recognized as a particularly significant ancestor, and is given the by-name of Snoice, Snaige or Snaithe in three of the pedigrees. He can readily be identified, as Matheson underlined, with the Olghair of Mary MacLeod, and the Oliver of James Fraser and John Morison. His name also appears in a fragmentary pedigree which runs Suain o tainig clann Suain (from whom clan Suain) mic Olbhuir mic Iomhuir Oig mic Sgoinne Sgiathluinn mic Iomhuir Athaclaith. It is reasonable to assume that this represents the pedigree of the MacSwans of Roag, who claimed to belong to the same stock as the MacLeods. The run above Olbhar is similar to that in pedigree C. There can be little doubt, as Matheson sets out, that the name Olbhar represents the Scandinavian forename Olvir, later confused or assimilated with Olaf, as in the description in 1630 of John MacLeod as John MacOlaus of Dunvegan. Pedigrees B and D end with Olvir, if the name Angus beyond his is taken to belong to the Nicolson rather than the MacLeod portion of the pedigree. Pedigree C gives three names above Olbhar: Young Ivar, Old Ivar and Sgoinne Sganlain before reaching Ivar of Dublin. These seem rather formulaic. Only A, the oldest pedigree, appears to offer detailed information as to Olvirs antecedents.
Pedigree A is alone in giving the name of Leods father as Gillemuire. The others give another Olvir. Is Gillemuire then to be regarded as a mistranscription as Matheson believed? I would suggest not. Given that the Gaelic pronunciation of son of Gillemuire (mac (Gh)illemhuire) and son of Olvir (mac Olbhuir) is very similar, especially when spoken in the course of a long genealogical run, it is easy to see how Gillemuire could have been assimilated to Olvir, given the prominence of the latter name in MacLeod tradition. It is not so easy to explain a change of name in the other direction. In any case, pedigree A clearly does distinguish carefully between the two names, as it gives another Gillemuire as the father of Olvir Snoice. The name Gillemuire was certainly known early in the MacLeods lands, for the Harris Morrisons, to whom Alick Morrison belongs, have long been known as Clann MhicGillemhuire.
The name given above Gillemuire/Olvir as the grandfather of Leod is not immediately recognizable, and seems to have been unfamiliar to the later pedigree compilers. Pedigrees A, B and D give Raice, Raoige and Raisi respectively, clearly all versions of the same original name. The name given in C is Oib, which appears to be wholly garbled and may be ignored. Matheson made the ingenious suggestion that Raice should be emended to Paice, th Gaelicised form of the Scandinavian name Balki, as in Pall mac Phaic for Paul Balkason. He further suggested that the Paice or Balki in question was the father of Paul Balkason, and that Leod was, therefore, the nephew of Paul, as set out in Table 3. This emendation by Matheson was later criticized by Alick Morrison in his Chiefs of Clan MacLeod, and does, indeed, seem unlikely. It is not easy to see why the name Paice, reasonably well known in MacLeod tradition, should have been altered to the unknown Raice in Leods own pedigree. It is even more difficult to equate Paice with the forms Raoige or Raisi. In addition, the proposed linguistic shift from P to R would seem to be a substantial one. Mathesons proposed emendation stems largely from his trying to explain later MacLeod tradition which claimed that Paul Balkason was the foster-father of Leod, and made him his heir. As Matheson points out, similar traditions are to be found in other families, and often camouflage a more direct blood relationship. However, this is by no means the only possible explanation for claiming such a relationship of fosterage, and the tradition, as we now have it, dates only from the eighteenth century.
If Mathesons emendation is rejected, what explanation can be given for the name Raice? It is worth noting that the name Raingce appears in the pedigree of the Macleans, and was presumably pronounced not unlike Raice or Raoige, the more so if a suprascript dash indicating an n has at some stage been omitted from the i in Raice. From Raingce comes the surname Rankin (Clann mhic Raing) in Mull. Another possibility, suggested to me by Andrew MacLeod, is that Raoige represents the comparatively rare Norse name Ragi.
Beyond Raice comes Olvir Snoice (Snaige, Snáithe). He was remembered by the MacLeods as a significant ancestor, and is probably a historical character. William Matheson, among others, has made a strong, although speculative, case for identifying him with Olvir Rosta, a character in Orkneyinga Saga. At first sight this seems an attractive suggestion. Olvir Rosta is last heard of in exile in Lewis after a distinctly chequered career. His presumed dates fit will enough with a great grandfather of Leod. As a new man in the Hebrides he might seem a fitting ancestor for a new family. In addition, according to Orkneyinga Saga, his mothers father father bore the splendid name of Ljot (that is, Leod) Nithing. However, against this theory is the fact that the evidence is entirely circumstantial. There is no hard evidence in favour. Olvir was not such an unusual name as to suggest the identification in itself. In addition Olvir Rostas genealogy and family relationships, as given in Orkneyinga Saga, do not tally at all with the genealogy of the Olvir in the MacLeod pedigrees, Ljot Nithing notwithstanding. Above all there is the by-name. The Olvir in the MacLeod pedigrees is never called Rosta. Rather he is named Snoice or Snaige. On balance, therefore, the proposed identification must be rejected.
The by-name Snoice is unusual, and its meaning is unclear. Captain Thomas, writing last century, suggested that it might mean the Hewer? Matheson, however, argued convincingly against this derivation, but did not suggest another in its place. At the meeting in Inverness I suggested that the name might be Scandinavian rather than Gaelic, and possibly referred in some way to the nose, as in the phrase to cock a snook, or in the place name the Snook on the island of Lindisfarne. In the discussion after the meeting Mr. D. J. MacKay alerted me to the use of the term snaoicein in the dialect of Bernera, Harris, earlier this century, to refer to someone with a cleft palate. It would seem, therefore, that Olvir Snoice, ancestor of the MacLeods, acquired his by-name because of a cleft palate or other nasal deformity!
According to pedigree A, our best source, Olvir was the son of another Gillemuire whose fathers name has not been preserved, but whose mother is named as Ealga Foltalainn -- Helga of the beautiful hair -- daughter of Aralt (Harald) son of Sen Iamhar (Old Ivar). Matheson has suggested, on the strength of later MacLeod tradition that their ancestor Olvir acquired Dunvegan through marriage with a daughter of MacRailt Armuinn (which might be loosely translated as the lord MacHarald), that Helga of the beautiful hair was the wife of Olvir rather than his grandmother. However, it does not seem necessary or desirable to emend the pedigree in this way.
Ivar the Old was remembered as an important ancestor. As has been seen, he occurs also in other pedigrees, sometimes in the company of an even more remote forebear named Iamhar Athaclaith or Ivar of Dublin. As has often been noted, this suggests an association with the Scandinavian rulers of Dublin, who descend from a late ninth century Ivar, king of Dublin.
However, the matter can be taken further. The Welsh, like their fellow Celts in Ireland and Scotland, were great genealogists, and preserved the pedigrees of their ruling dynasties. The authenticity of a group of these pedigrees relating to Fruffudd ap Cynan, prince of Gwynedd (d. 1137), and including many Irish and Scandinavian connections, has recently been vindicated by Sean Duffy. Another Welsh pedigree gives the ancestry of Ragnall (killed 1229), king of Man and Isles, the father of Godfrey Donn, and presumably dates from his time. It is the only traditional pedigree of a member of the Manx dynasty known to have survived. It runs, Rhanallt m. Gwythryg ap Afloyd m. Gwrthryt mearch m. Harallt ddu m. Ifor gamle m. Afloyd m. Swtrig, that is, Ragnall son of Godfrey son of Olaf son of Godfrey Mearch [mearch or mearanach being another epithet sometimes attached to Godfrey Crovan] son of Harald Dubh son of Ivar Gamle son of Olaf son of Sihtric. The father of Harald the Black de Ysland is thus revealed as Ivar Gamle. Ivars father Olaf son of Sihtric is probably to be identified with the famous Olaf Cuaran (d. 980), of the ruling dynasty of Dublin, son of Sihtric Caoch (d. 927). However, P. C. Bartrum, the editor of Early Welsh Genealogical Tracts, suggested that there should be a break in the pedigree after Harallt ddu, in the mistaken belief that there was firm evidence that Godfrey Crovan descended, via his father Harald, form the earlier Godfrey, rí Innsegall (killed 989).[71 ] In fact, there seems to be no compelling reason for not taking the pedigree at face value and accepting it, as George Broderick does, as the best evidence for the origins of Godfrey Crovan and his dynasty. Indeed, the preservation of the Scandinavian epithet Gamle in a Welsh context argues for its authenticity. Broderick identifies the Ivar Gamle of the pedigree with Ivar, king of Dublin (d. 1054), a grandson -- rather than a son as in the pedigree -- of Olaf Cuaran by his son Harald (killed 999).
Whether Ivar Gamle is to be identified with the Ivar, king of Dublin who died in 1054, or with an otherwise unrecorded son of Olaf Cuaran named Ivar is probably now impossible to determine. However, the pedigree arguably provides the key which unlocks MacLeod genealogical tradition. The Scandinavian epithet Gamle attached to Ivar means old, and Ivar Gamle, therefore, translates in Gaelic as Sen Iamhar of MacLeod genealogical tradition are one and the same person. Indeed, provided that chronology fits, this seems a more likely explanation than pure coincidence, especially as in both cases the later descent was claimed through Harald son of Ivar. It would certainly explain the strongly held MacLeod belief in kinship with the kings of Man and the Isles. The chronology, in fact, fits perfectly, as can be seen from the reconstruction in Table 4. If the identification of Ivar Gamle and Sen Iamhar is correct, then the MacLeods descend, not from thirteenth century kings of Man and the Isles, but from Helga, sister of Godfrey Crovan. The scenario is entirely plausible. If accepted, it sheds some light not only on MacLeod origins, but also on a very dark period of Hebridean history.
Notes and References
I should like to express my gratitude to Aubrey MacLeod and to Andrew MacLeod both of whom have let me see material of their own on MacLeod origins. Andrew MacLeod, indeed, has been working on lines parallel to my own, and I am most grateful to him for commenting on an earlier version of this paper. I should also like to thank Kenneth Nicholls of University College, Cork and, above all, to acknowledge my debt to the late William Matheson from whom I learned so much about Hebridean genealogy.
1. R. C. MacLeod, The MacLeods of Dunvegan (Clan MacLeod Society, 1927); I. F. Grant, The MacLeods: The History of a Clan (London, 1959; 2nd. ed. Edinburgh, 1981)
2. W. Matheson, The Ancestry of the MacLeods and The MacLeods of Lewis, in Transactions of the Gaelic Society of Inverness TGSI vol. LI, 68-80 and 320-337 respectively; Alick Morrison, Chiefs of Clan MacLeod (Associated Clan MacLeod Societies, 1987), ch. 1, The Origin of Leod; also Morrison, The Kingdom of Man and the Isles TGSI vol. LVIII, 425-481.
3. I have used George Brodericks edition of the Manx Chronicle: Cronica Regum Mannie & Insularum [Cron.Man.] (Manx Museum and National Trust, 1979) for the Latin text, although I have generally preferred to supply my own translation.
4. For a discussion of these titles, see W. D. H. Sellar, Hebridean Sea-Kings: the Successors of Somerled, 1164-1316, in Alba: the Highland Dimension in Medieval Scotland, ed. E. J. Cowan (forthcoming).
5. It is clear that the descendants of Somerled controlled Isla and Mull and their surrounding islands. There is no contemporary evidence regarding the Uists, but their later history suggest that they may also have come under the rule of Somerleds line.
6. Or latterly, in the case of the dynasty based in Man, rex Mannie et insularum.
7. Brian O Cuiv, A Poem in Praise of Raghnall, King of Man, Eigse viii (1957) 283-301; Orkneyinga Saga, ch 114. For the argument that the description in the Saga applies to Ragnall son of Somerled, see Sellar, Hebridean Sea-Kings.
8. Cron. Man. f41v-43r.
9. Broderick gives the name as Lauon and Lyauon, rather than Jauon (Joan?) as sometimes suggested. Could Lyauon stand for Liamhain, or for Liobhan, both attested Gaelic names?
10. It is not clear when Ferchar was created earl of Ross: the Complete Peerage suggests c 1225-6.
11. Broderick's translation of this passage is misleading: tractis quinque navibus translates as having dragged five ships, rather than Brodericks They launched five boats. Andersons, following Goss, gives they brought five ships (Early Sources of Scottish History [ES]. ed. A. O. Anderson [2nd ed. M. O Anderson, Stamford, 1990] ii, 459).
12. Icelandic Annals in ES, ii, 454.
13. These events all seem to have taken place in 1230 or 1231, the sources being the Manx Chronicle, the Chronicle of Lanercost and Haakon Haakonsons Saga (ES, ii, 471, 472, 478). Haakon Haakonsons Saga refers to Godfred as the Black, rather than Donn, the Brown, but I have assumed that all the references are to the same man. Hugh MacDonald also refers to Godfrey Donn as the Black (see immediately below).
14. Hugh MacDonald, A History of the MacDonalds Highland PapersI, ed. J. R. N. Macphail (Scottish History Society, 1914) 7-8.
15. I Pall Mac Phaic, see Matheson, Ancestry of the MacLeods, 73-6. Andrew MacLeod suggests to to me that the description hermit may represent a memory, at one remove or another, of the title armadhr.
16. See, for example, Matheson, Ancestry of the MacLeods, 74-5.
17. Matheson, Ancestry of the MacLeods, note 38.
18. Morrison, Kingdom of Man and the Isles, 470.
19. For St. Columbas Isle see, for example, Alan MacQuarrie, Kings, Lords and Abbots TGSI vol. LIV 355-75 at 369-70. For Nicolsons, see W. D. H. Sellar and Alasdair Maclean, The Highland Clan MacNeacaill (MacNicol); A History of the Nicolsons of Scorrybreac ed. C. B. Harman Nicolson (Waternish, Skye, 1999). Dr. Alasdair Maclean has advanced the case for Snizort.
20. RCAHMS: The Outer Hebrides, Skye and the Small Isles (Edinburgh, 1928) no. 535. The suggestion that the location of the action was at Loch Chaluim Chille was strongly supported at the reading of this paper to the Gaelic Society.
21. Cron.Man., f. 32v.
22. Cron.Man., f.33v.
23. For these grants see Regesta Regum Scottorum VI: David II ed. Bruce Webster (Edinburgh, 1982) nos. 486 and 487.
24. See C. I Fraser of Reelig, Some Notes on the Heraldry of the MacLeods, in I. F. Grant, MacLeods, appendix ii.
25. Matheson, MacLeods of Lewis, 324.
26. The pedigrees are from Royal Irish Academy Manuscripts 23 H 22 and 23 G 4. Only the first of these was known to William Matheson. They are considered further below.
27. Matheson, MacLeods of Lewis; Sellar, History of the Clan in Highland Clan MacNeacail.
28. For example, the passage on Paul mac Tyre attached to Ane Breve Cronicle of the Earls of Ross, ed. W. R. B[aillie] (Edinburgh, 1850) 30, which begins, Thair was thrie sones of the King od Denmark, callit Gwine, Loid, and Leandres, quha came be sea out of Denmarke and landit in the north parts of Scotland, to conquest lands to themselffis be the sworde...
29. On this see, for example. W. Matheson, Traditions of the MacKenzies TGSI vol. XXXIX-XL, 193-228 at 214-6; W. Matheson, The Pape Riot and it Sequel in Lewis TGIS vol. XLVIII, 395-427 at 423-7; and W. David H. Sellar, Highland Family Origins -- Pedigree Making and Pedigree Faking in The Middle Ages in the Highlands ed. Loraine Maclean of Dochgarroch (Inverness, 1981) 103-116 at 111.
30. For Paul MacTyre see, for example, Matheson (as per note 29 above).
31. Duncan Warrand, Some Fraser Pedigrees (Inverness, 1934) 58-60. I am indebted to Hugh Barron for this reference.
32. Sellar, History of the Clan in Highland Clan MacNeacaill13-14.
33. The disputed paternity of Torquil Cononach was, of course, one of the reasons for the fall of the MacLeods of Lewis.
34. Sir George MacKenzies History of the family of MacKenzie is printed in Sir William Frasers Earls of Cromartie (Edinburgh, 1876) ii, 462-513.
35. See, for example, W. F. Skene, The Highlanders of Scotland (1836) ed. Alexander Macbain (Stirling, 1902) 325-6, and Matheson, Traditions of the MacKenzies.
36. Matheson, Traditions of the MacLeods, 71.
37. R. C. MacLeod MacLeods of Dunvegan, 4.
38. R. C. MacLeod MacLeods of Dunvegan, 20.
39. A transcript of the supposed charter is given in Chronicles of the Frasers, 55; see also E. C. Batten, The Charters of Beauly Priory (Grampian Club, London, 1877) 31-2.
40. I. F. Grant, MacLeods, 25, note 1.
41. Rotuli Scotiae i, 18b.
42. A. W. Moore, History of the Isle of Man (London, 192; Manx Museum and National Trust, 1977) i, 186-7.
43. C. I. Fraser of Reelig (Some Notes on the Heraldry of the MacLeods) notes that the three legs of Man appear in the arms of Iain Breac, chief from 1664-93.
44. Gaelic Songs of Mary MacLeod ed. J. Carmichael Watson (1934) (Scottish Gaelic Texts Society, 1965): Cronan an Taibh, lines 520-1; An Cronan, line 791. I do not know who Ochraidh is in the line Bhair sliochd Olghair is Ochraidh (Cumha do Mhac Leoid, line 698), and presume that the name is garbled (and see Matheson, Ancestry of the MacLeods, note 15). Morrison (Chiefs of Clan MacLeod) seems mistaken in talking of Olghair mac Ochraidh.
45. Cumha do Mac Leoid, lines 695-6.
46. James Fraser, Chronicles of the Frasers: The Wardlaw Manuscript (Scottish History Society, 1905) ed. William MacKay, 40.
47. John Morison, Description of the Lewis in MacFarlanes Geographical Collections ii, 214, cited in Matheson, Ancestry of the MacLeods, note 49.
48. Skene, Celtic Scotland, iii, 482; and see Matheson, History of the MacLeods, 76, note 46, where he thanks me, for suggesting a redistribution of the minims, that is, suggesting the reading Sen Imair, rather than Semmair. Thus Semmair, although accepted by R. C. MacLeod and Alick Morrison, is a phantom. Kenneth Nichols kindly informs me that MacFirbis autograph genealogies in University College, Dublin (at p. 406) have six undifferentiated minims between Se and air.
49. RIA MS 23 H 22, p. 48.
50. Matheson, MacLeods of Lewis, 324.
51. Celtic Scotland, ii, 461 (from MS 1467). Surprisingly, no MacLeod pedigree appears in MS 1467.
52. RIA MS 23 G 4, p. 396.
53. Celtic Scotland, iii, 460. The name of Leods father in this pedigree is transcribed as Oloir in Collectanea de Rebus Albanicis (Iona Club, Edinburgh, 1847), pp. 360-1.
54. RIA MS E i: see Angus Matheson, Poems from a Manuscript of Cathal MacMuireadhaigh Éigse 11 (1964-6) 1-17 at 17 (also Éigse 10 (1961-3) 270-8); Matheson, Ancestry of the MacLeods, 69.
55. The tract was published by Alexander Bugge, (Christiania (Oslo) 1905), with translation and notes. The pedigree had previously been published by John ODonovan in his The Fomorians and the Lochlanns. Pedigrees of the MacCabe of Ireland and MacLeod of Scotland Ulster Journal of Archaeology 1st. ser., ix (1861-2) 94-105; I am grateful to Andrew MacLeod for this reference.
56. Canon R. C. MacLeod discussed this pedigree in his MacLeods and gave it a great deal more credence than it deserves.
57. Skene, Celtic Scotland (iii, 460, footnote.)
58. RIA MS A V 2, fol. 80r. The pedigree continues beyond Iamhuir Atacliath (Ivar of Dublin) for many generations. I am most grateful to Kenneth Nicholls for checking this pedigree. He notes that the earlier generations are essentially the same as those in the Kilbride pedigree as printed by Skene in Celtic Scotland, save that they continue back to Adam!
59. Matheson, Ancestry of the MacLeods, 72 (citing TGSI vol. XXXVIII, 398). Matheson was not the first to equate Olbhar with Olvir; see, for example Alexander Macbains notes to Skenes Highlanders of Scotland, 420; and Captain F. W. I. Thomas, Traditions of the Macaulays of Lewis (note 64 below).
60. There are conflicting views as to the origins of the Morrisons: see Alick Morrison, Clan Morrison (Edinburgh and London, 1956); William Matheson, The Morisons of Ness TGSI vol. L (1976-78) 60-80.
61. Matheson, Ancestry of the MacLeods, 73-6; also MacLeods of Lewis, 337.
62. Morrison, Chiefs of Clan MacLeod, 7-8. Morrison is mistaken, however, in stating that a chieftain named Paul son of Balki is named by the Manx Chronicle in 1144 [recte 1154-6]; The Chronicle only mentions the name Paul (princeps Paulus nomine).
63. For Raingce in the Maclean pedigree see Celtic Scotland, iii, 481. For the Mull Rankins see N. R. Morrison, Clan Duiligh: Piobairean Chloinn Ghill-Eathain TGSI vol XXXVII, 49-79. Andrew MacLeod notes the tenth century Icelandic Lawspeaker, Thorarin Ragis brother (Njals Saga, ch. 13).
64. Matheson, History of the MacLeods, 71-72. This identification seems to have been first suggested by Captain F. W. I. Thomas, Traditions of the Macaulays of Lewis PSAS, xiv (1879-80) 363-431 at 364, note 2. I am grateful to Andrew MacLeod for this reference.
65. Matheson, History of the MacLeods, 78, note 12. Morrison (Chiefs of Clan MacLeod, 10) says that the epithet probably means white, without further explanation.
66. I am most grateful to Mr. D. J. MacKay, Teandalloch, Muir of Ord, for confirming this in writing. He also notes that the postman at Rhenigadale, Harris earlier this century, who also suffered from a nasal deformity, was commonly addressed as Staoicein. I note also Old Icelandic snaga a snag axe, snokr a snake, snyta to blow the nose, and snagi clothepeg; and Scots snoke to sniff.
67. Matheson, History of the MacLeods, 76-7.
68. As in pedigree C, the Kilbride MS.
69. P. C. Bartrum, Early Welsh Genealogical Tracts (Cardiff, 1966), Hanes Gruffudd ap Cynan; and Sean Duffy, Ostmen, Irish and Welsh Peritia 9, 378-96.
70. Bartrum, Early Welsh Genealogical Tracts: Accau Brenhinoedd a Thywysogion Cymru, 6 c (p. 99).
71. Bartrum, Tracts, 152.
72. George Broderick, Irish and Welsh Strands in the Genealogy of Godred Crovan, 1980 Journal of the Manx Museum, 32-8.