Criticism of Sellar's "Family Origins in Cowal & Knapdale"


  Fergus Gillespie succinctly summed up the two prevailing opinions on the Anradan kindred in his article on the "Gaelic Families of county Donegal."1

Mac Suibhne; MacSweeney

  "The story of the descent of Clann tSuibhne from the O Neill kings of Cenel Eogain may have been a convenient fabrication in order to give the family a Milesian pedigree and thus an accepted place in medieval Tir Conaill.  W.D.H. Sellar, however, has argued most convincingly that the story is not unlikely and that Clann tSuibhne and other Scottish families such as Lamont may well be descended from the O Neill family."

  A major drawback to Sellar's "convincing" argument however is the fact that he completely ignored the possibility that the pedigree was a fabrication as described above.   In his article he states that the "main argument which has been advanced against the authenticity of the Cowal and Knapdale pedigree is that it is too short by several generations and it is likely to be a medieval fabrication."  That was the argument used by Skene and other Scottish writers; but Irish historians have pressed a more formidable argument:  that the pedigree was a fabrication designed to provide a suitable pedigree for the Scottish gallowglass family of MacSweeney who settled in Donegal in the early years of the 14th century.2   
  Prior to the introduction of gallowglass families in Ireland, the Scottish clans were completely ignored by the Irish scribes, with the notable exception of the Dalriadic line of the Kings of Scotland.  No Scottish pedigrees are to be found in any early genealogical manuscript, including Laud 610, Rawlinson B. 502 or the Book of Leinster, except for a pedigree of the King of Scotland.  This changed quickly with the introduction of Scottish gallowglasses in the 14th century.  A pedigree of the MacDonalds in descent from Colla Uais makes its first appearance in the National Library G2 MS (c. 1345); followed shortly thereafter by notices for MacDonalds and MacSweeneys in the Books of Ballymote and Lecan.  Once these families were permanently settled in Ireland (as the MacDonalds and MacSweeneys both were) with well-defined roles and important positions in local society, the Irish scribes finally took notice of them and included pedigrees for them in all later genealogical manuscripts.
  In the case of the MacDonalds, their pedigree linking them to Colla Uais and the Ui Meicc Uais of northern Ireland is much too short to be historical, a defect cleverly explained away by W.D.H. Sellar.3  His opinion is accepted by many modern writers, although recent DNA studies by Bryan Sykes show the family to be Norse in origin.4.    The MacSweeney pedigree is much more believable, at least in the sense that it is not historically impossible, as Sellar goes to great pains to show in his article.  
   There are no antecedents for either pedigree in Irish or Scottish manuscript that predate the middle of the 14th century (G2).  The pedigrees simply appear, fully formed, and are untraceable in either the annals or any historic sources.  The dates of their sudden appearance in the genealogical collections just happen to coincide with their settlement in Ireland as gallowglasses; a coincidence too remarkable to be skimmed over without comment as Sellar did in his article.

  A second criticism of Sellar's article is that he fails to deal with an anomaly in the Gaelic MS. 1467 of Skene that he mentions frequently.  A second pedigree is contained in this manuscript for the same string of names (Ferchar/Duinnsleibhe/Buircc/Anradan) linked in the Irish manuscripts to Aodh Athlaman; but in this pedigree Anradan is made a son of Gilleabeirt rig eilan Sidir (king of the Sudreys).  The pedigree terminates in Nialgusa of Lochaber, an ancestor of the MacDonalds.  As Sellar notes in his article the transcription of the manuscript given in the Collectanea De Rebus Albanicis is deficient.  Skene apparently transcribed the manuscript himself and was at best an eager amateur barely able to decipher Gaelic script.   But Sellar only acknowledges this odd pedigree in a footnote to a reference to the Lamonts by saying "MS. 1467 omits 'Aodh Athlamhan."
  Indeed it does!  It not only omits Aodh Athlaman but it also omits Flaithbertach an Trostain and any link to the O'Neills of Ireland.  Why Sellar chose to ignore this pedigree is a mystery.   But even Skene, amateur though he may have been, could not have butchered his transcription so badly that the pedigree simply omitted Aodh Athlaman.

  A third criticism (but one not likely to be widely accepted) of Sellar's article is that he failed to realize there were problems with the O'Neill pedigree itself; and these problems centered on any possible descendants of Aodh Athlaman.   O Ceallaigh pointed out these problems years ago5 and it is perhaps no coincidence that the first full pedigree of the O'Neills6 also made its debut in the Books of Ballymote and Lecan, in each manuscript followed by that of the MacSweeneys.  The two pedigrees are bookends, composed at the same date, sharing the same fictitious ancestor, Aodh Athlaman.   Rather than repeat the documentation here, see "The Descent of the O'Neills" and O Ceallaighs "Gleanings from Ulster history.


1, Donegal History & Society; Interdisciplinary Essays on the History of an Irish County; pp. 778-788; 1995; "Gaelic Families of County Donegal."

2.  See F.J. Brynes; A New History of Ireland; Medieval Ireland 1169-1534
Chapt. 1 "The Trembling sod"

"By the end of the thirteenth century the galloclaig, fighting
men from Inne Gall, 'the islands of the Foreigners', were being
endowed with lands in Ireland by Gaelic lords and given genealogical
legitmization with pedigrees derived from the Airgialla, or even,
in the case of the MacSweeneys, Ua Domnaill's most trusted vassals,
with Ui Neill ancestry."

3.  W.D.H. Sellar;  The Origins and Ancestry of Somerled;  Scottish Historical Review; Vol. XLV, 2 No. 140. Oct. 1966.

4.  Bryan Sykes;  Adam's Curse; Chapt. 16;  THE Y-CHROMOSOME OF SOMHAIRLE MOR.

5. Seamus O Ceallaigh; Gleanings from Ulster History; 1994;  The following enigmatic phrase appears in his section on the O'Neills::  "For six generations in sequence the family depends upon a single individual to maintain the succession; no brothers or uncles are mentioned, and L.E. expressly states that Aodh was the last of his race."  The Aodh referred to here is Aodh Athlaman.

6. Rawlinson B. 502 has a pedigree including Flaithbertach an Trostain in which the O'Neill ancestors are conspicuously absent:  "Aed m. Neill m. Flaithbertaich m. Murchertaich m. Domnaill m. Murchertaich m. Neill glunduib m. Aeda Findleith;  The earliest full pedigree of the O'Neills is found in the mid-fourteenth century G2 MS (National Library).   It is strikingly different from the later versions found in Ballymote and Lecan.  See "The Descent of the O'Neills.