The Root of the Kennedy family tree in Scotland
authored by Iain Kennedy 27th April 2006
Copyright © 2007 Iain Kennedy
Although this One-Name Study recognises Roland de Carryck as 'the first Kennedy', Roland does not stand at the root of the Kennedy tree as we can trace his ancestry with a degree of reliability. For our purposes we will consider the root of the Kennedy tree to be Fergus, Lord of Galloway. Some researchers have made claims of having traced back beyond Fergus but I am not at this stage convinced. In this position I have been influenced by the writing of R. Andrew McDonald, now at Brock University. Whilst in a previous post Andrew wrote the book 'Alba - Celtic Scotland in the Middle Ages' with Edward J. Cowan. With Andrew's permission I quote two paragraphs from a chapter he wrote in which he surveys these attempts to trace the lineage of Fergus:
'The most pressing of the questions relating to the origins and ancestry of Fergus of Galloway is, from which of the various peoples, Galwegian, Norse, Gall-Gaidhil, or Norman, did he come? What was his relation to King David I of Scotland, and to King Henry I of England? In attempting to answer these questions the lack of evidence is keenly felt. Unlike Somerled, upon whose background the genealogies shed some light, there is no direct evidence bearing upon Fergus's ancestry, and the matter is further complicated by the inventions and romantic speculations of some nineteenth-century historians...'
'If Fergus was not merely a protege of David I, then the question remains of just who he was. This problem is accentuated by the lack of reliable genealogical information concerning his ancestors. He was styled simply Fergus de Galweia in those charters which he attested, and the pedigrees of his descendants were carried back only to him. Notwithstanding his belief that Fergus grew up with David at the court of Henry I, Sir Herbert Maxwell suggested that Fergus was 'of the line of Galloway princes or native rulers', meaning presumably Strathclyde-Welsh blood. While there might be much to commend such a view, some recent studies have argued that the last of the native British line of rulers in Strathclyde had died out by the middle of the ninth century, making it unlikely that Fergus was related to them. Other historians have seen him as springing from Norse or Gall-Gaidhil stock. His association with a man named Somerled (not Somerled of Argyll) in the thirteenth-century Roman de Fergus lends some credibility to the belief that by the early twelfth century Scandinavian elements had indeed intruded into the ruling house of Galloway. Some investigators would link him to Suibne, Kenneth's son, called king of the Galwegians, who died in 1034, and it has even been argued that Fergus was the son of that Somhairle, styled ri Innse Gall, who died in 1083. Whatever the case, the most commonly held consensus among modern historians who have examined the problem is that Fergus's background was mixed Celtic-Norse, not unlike that of Somerled of Argyll.' [see bottom of page for McDonald's references].
What these extracts do not take account of (the book was written in 2000) are the recent results of genetic analysis of descendants of Somerled and Fergus, namely the McDonalds, Kennedys and McDowalls.
Clan Donald USA Historian Mark McDonald was first to announce the result of his determination of the Y-chromosome profile of Somerled as part of his Clan Donald DNA project. The project keeps its data private (of course, testing clan chiefs is a delicate matter) but the topic has been discussed from time to time on mailing lists like this one
http://archiver.rootsweb.com/th/read/GENEALOGY-DNA/2005-04/1114547896
The work was independently verified by Professor Brian Sykes of Oxford University, founder of the theory of the 'Seven Daughters of Eve' (a maternal DNA study of modern Europeans and their ancestry) and author of the book of the same name. In the sequel 'Adam's Curse' Sykes writes a chapter 'The Y-chromosome of Somhairle Mor' in which he describes his testing of all five of the McDonald clan chiefs and thus verifying the findings of Mark McDonald. The old question as to whether Somerled is of 'mixed Celtic-Norse' origin is now settled; his male line has been scientifically proven as Norse. [It is important to stress though that the Y-DNA test only tests one line of his ancestry, the all-paternal one. This line is Norse but his other ancestral lines could still be Celtic and this can only be determined by following all-paternal lines of descent forward. In this respect then, Andrew McDonald's assertion of Somerled's background still stands].
Following the success of his first book, and his equally famous genetic study of the Sykes surname in Yorkshire, Sykes created the DNA testing company Oxford Ancestors. He will even sell you an 'Are you descended from Somerled, King of the Isles?' test!
I have been inspired to carry out a similar exercise with Fergus, prompted in part by the discovery that along with other Kennedys, I started to find increasing numbers of McDowalls matching my own Y-DNA test results from DNA Heritage. At first I assumed this was due to both families originating from the Galloway region in South West Scotland. When I went back to take a second look though I found something that rarely seems to get a mention; the McDowall and Kennedy clan histories both claim descent from Fergus, Lord of Galloway - and here I am seeing multiple DNA matches with both these names!!
My current plan is to expand my study into the DNA of Kennedys in Scotland and join forces with those researching the ancestry and genetics of the McDowalls to see if we can produce the same level of proof that McDonald did for Somerled.
Book extract references (abridged).
Maxwell, History of Dumfries and Galloway, 47
A.P. Smyth, Warlords and Holy Men; Scotland AD80-1000, 218-9
J.A. McGill, A Genealogical survey of the ancient lords of Galloway, Scottish Genealogist 1955, 3-6
Truckell, Proto-history of Galloway, 50
Oram, Fergus, Galloway and the Scots, 121-2
I have a full list of the references if anyone wants them.