Non-matching profiles and a non-paternity event
by Mike Spathaky Genealogical research has shown that the surname Cree does not have one single origin. The numerically most significant group of lines originated in Perthshire, Scotland. It is almost certain that the Cree surname in County Down, Ireland was brought there from Scottish Cree (or Crie) migrants in the early 17th century. Of the many lines in the USA, some came from County Down around 1770 or earlier settling in Pennsylvania, others came directly from Scotland, settling in Maine and other states. A Cree line that is traceable back to Sprotbrough, Yorkshire, England, in 1744 has long been thought to be of Scottish origin but with no clear documentary evidence. These are discussed in the second of these two articles. Another substantial line started in Bolsover, England, in 1644, when James, who later took the name Cree, was baptised as the son of a Mackree. He has Cree descendants world-wide today. A Cree line in Kent, England, derives from a Huguenot, Pierre Jacob Carré, who had migrated from France before 1735. A line from Ennis, Ireland, is thought to have originated from a change of name from Creagh to Cree around 1700. Some other lines, from Northumberland in England, Massachusetts and South Carolina in the USA, for example, may or may not be of independent origin. Genetic genealogyThe Cree YDNA Surname Project (co-ordinated by Gary Maher in New Jersey, USA) and the Cree One-Name Study (co-ordinated by the author in Oadby, England) were started independently but the two co-ordinators now work closely together and with other researchers, in particular Robert H Cree and Trevor Cree. For readers unfamiliar with genetic genealogy it should be pointed out that a match between the YDNA profiles of two males will normally indicate that they have a common ancestor in their male lines. The degree of matching gives a rough indication of how far back in their male-line ancestry this common ancestor lived. It is now generally recommended that a 37-marker test should be used as this gives enough points of comparison for meaningful deductions to be made. If two people's 37-marker YDNA profiles match exactly (37 out of 37 markers identical), there is a 50% probability that their most recent common ancestor (MRCA) is no more than three generations back. For a 36 out of 37 match the distance to their MRCA is eight generations (again at a 50% probability level). These calculations depend on the average mutation rates of the relevant genes and, because mutations are random events, are subject to a wide range of error. If on the other hand the profiles of two test subjects show a low degree of matching, say 31/37 markers or less, then it can be concluded that they do not share a common ancestor in the period of normal genealogical records. In this case, if the test subjects share the same surname and were thought from documentary evidence to be related, the possibility arises that there has been a “non-paternity event” at some point in the ancestry of one of them. A non-paternity event (NPE) is where a person takes or is given a surname that is not the surname of his biological father. This can happen where a person is adopted, or born of parents who were not married or has deliberately chosen to change his surname. Children of unmarried or widowed mothers were often obliged by law or social convention to take her surname rather than that of the father. NPEs result in the breaking of the link between the inheritance of a surname and the inheritance of Y-chromosome genes. So a mismatch of YDNA profiles should not be taken as disproving the genealogically established line of inheritance of a surname. The Cree YDNA ProjectYDNA testing of Cree males has been in progress for some years. However in December 2012 a campaign to recruit more test subject was carried out. This was done by trawling the author's email files for Cree males with whom there had been previous contact in the context of Cree genealogy. It was hypothesised that these were more likely to respond than people selected from telephone directories or electoral lists who may have no interest in genealogy and could well be suspicious of DNA testing. It was surmised that previous contacts would have both an interest in YDNA test results and a degree of trust in the author as broker of the testing project. This approach did indeed result in an increase in the number of tests undertaken by Cree males has increased our knowledge of the genetic links between Cree lines and has prompted this review. Tests were all carried out through the Family Tree DNA company (FTDNA). Including earlier test subject, fifteen Cree males have now taken YDNA tests with results analysed for 37 markers. Analysis of the YDNA profiles clearly indicates two groups of four and six results respectively, where the matching within groups is close, and a further five results which do not have any matches. The two matching groups are discussed in Part 2 of this article. Of the five results which do not have any matches, four are from Cree lines which are known to be separate from the Scottish, Irish and the USA lines. The fifth result (the 12-marker profile) is from a Cree male who was thought to be descended from one of the Pennsylvania lines. Our discussion here focuses on the four non-matching results. |
This is Part 1 of a two-part article which reviews the progress of the Cree YDNA Project testing as it stands in April 2013 and attempts to elucidate the implications of the test results for Cree genealogy.
Related articles:
Download file: |
The Kent England (Huguenot) lineA small group of Cree families in Kent was a mystery for some years. Eventually in 2007 a researcher sent me a genealogy indicating that this was a Huguenot family whose surname was originally Carré. We have one test result from the Huguenot Cree line. Since the this line is known to have separate origins to the other main Cree groups, no match with other Crees was expected. None was found, a result which is consistent with the line being separate from other Cree lines. Across the whole FTDNA database there are seven close matches of non-Crees with this Cree profile, all two steps away from it (35/37 match). They all have European roots: two each from England and Germany, and one each from Ireland, Austria and Russia. The spread of matches around Europe, with none in France, may reflect the dispersal of the French Protestants known as Huguenots from France to countries where Protestants were accepted. The Massachusetts USA lineThis line has been in Massachusetts since the 1723 marriage of Nicholas Cree but his origin is unknown. We have one result from the Massachusetts line, which does not have a significant match with any other Cree test subject. This lack of a match to the main Cree groups is consistent with the line being separate from other Cree lines. This has to be a provisional conclusion because, as always, we have to bear in mind the possibility of a non-paternity event in the ancestry of the test subject. It would be good to have two or three further test subjects from the Massachusetts line. The Bolsover Cree lineWe have two test subjects whose documented ancestry puts them in the Bolsover line, which started in the Derbyshire town with the 1643 marriage of Alexander Mackree and the subsequent adoption of the surname Cree by his son James. The two YDNA test results did not match with other Cree results. This is as expected and consistent with the line being separate from other Cree lines. Nor did they match with each other however. This was of course rather a surprise. The probability is 50% that their MRCA lived more than 4800 years ago! The preliminary result for one of these two participants showed a rare 12-marker profile which has only five close matches in the whole FTDNA database, four of whom are surnamed Humphrey or Humphreys. The 37-marker profile showed a close match with two USA test results, those of participants named Humphrey and Humphries. In both cases the match to the Cree result suggested that the probability that the most recent common ancestor (MRCA) was within 12 generations is about 90%, within eight generations about 70% and within 4 generations about 30%. As Gary Maher pointed out, “The common ancestor between them could be a Cree or a Humphrey(s). There's no way to know which based on the information we have before us.” Now that we have a second test subject from the Bolsover line however, showing no match with the first, we can say that it is more likely that the first Cree test subject of this line is descended from a Humphrey/ies than the other way round. Since the known genealogy of the Bolsover Cree line is entirely within England – within 30 miles of Bolsover in fact from 1643 to 1914 – we can suggest that a non-paternity event (NPE) occurred, probably around 8 generations ago, in the Bolsover area, by which a biological son of a Humphrey/ies acquired and passed on the surname Cree. A search for a Humphrey/ies male was made in Bolsover parish registers. The Family Search web site shows a Thomas and Anne Humphry who married in Bolsover in December 1779 and baptised two daughters in the parish church in January and August 1780. They subsequently had seven more children from 1784 to 1793 in nearby parishes. The date 1779 coincides exactly with a weak link in the chain of documentation of the relevant branch of the Bolsover Cree line. James and Mary Cree were married in 1777 but no children are recorded until 1788 after which three were born. The circumstantial evidence was thought convincing that three more known Cree children were born to this couple, two of them James and John during the gap period. One of these, John Cree, is the male-line ancestor five generations back from the Cree test subject who matches with Humphrey(s). One possible interpretation of the YDNA test results is that John Cree is the biological son of a Humphrey/ies who lived at that time, and Thomas Humphrey must be a likely candidate. This could have come about if John Cree, but not his wife Mary, was absent from Bolsover for the "gap" period 1777 to 1788 and that Mary had a relationship with Thomas Humphrey, resulting in the birth of two sons John and James who were then brought up by Mary as sons of her husband John Cree, alongside their documented progeny born after 1787. If this is the case, a likely reason for James Cree's absence from Bolsover would be service in the British Army in the American Revolutionary War (1775-1783). It looks as though the most fruitful line of research to test this hypothesis is to trace a present-day male-line descendant of Thomas Humphry of Bolsover. If such exists, a YDNA test might show a match between him and the tested Cree descendant of James and Mary Cree. This line of research is on-going and is described in the article A Cree-Humphrey relationship. The second part of this article will examine how the YDNA testing has thrown new light on the ancestry of the Scottish Cree lines and those other lines who are thought to be descended from them. |
Related articles:
Download file: |