Page 1

Report 2: Aaron Stark's (1608-1685) Ancestral Roots; A Theory

(Click on this Phrase to download a version formatted as PDF File)

Abstract

At the time this theory was prepared for publication, twenty-six men with the surname Stark or Starks have  — genetically and genealogically —  been proven to share Aaron Stark (1608-1685) as a common ancestor. Seven members were 37/37 matches to each other over their H37 Haplotypes (Markers 1 through 37).[1]  In Parts I & II,  the triangulation method devised by Charles F. Kerchner, Jr. was employed to define the H37 Ancestral Haplotype of Aaron Stark (1608-1685). In the genetic literature, a Modal Haplotype is composed of the most common Allele values observed at each marker over a specific haplotype. The H37 Ancestral Haplotype and the H37 Modal Haplotype of Aaron Stark's descendants has been observed to be identical. 

Two descendants of Aaron Stark (Kits #N17289 & #48711) have been confirmed to be members of Haplogroup R1b1a2a1a1a* — this result predicting all of the descendants of Aaron genetically tested will be members of the same Haplogroup (FTDNA's shorthand notation is R-U106).[2]  The purpose of this article will be to compare the Modal Haplotype of Aaron's descendants to persons in the FTDNA database  genetically confirmed to be members of Haplogroup R-U106 — none of whom will have the surname Stark. In particular, the focus will be on participants reporting the geographical location of their earliest known ancestor in the British Isles or Western Europe. The combination of geographical information and genetic comparisons will then be analyzed to attempt to determine Aaron's ancestral roots. [In the publication that follows, the H37 Modal Haplotype of Aaron's descendants will be designated as the Aaron Stark Modal Haplotype (shorthand notation will be ASMH).]

There has been much speculation about the ancestral roots of Aaron Stark which have not been satisfactorily answered. Part of the problem has been the belief of earlier researchers that Aaron Stark was a descendant of John Muirhead (alias John Stark). The account of John Muirhead saving King James IV from the charge of a bull and given the name Stark as a result of his bravery, was first presented in the 1903 Stark Family Association Yearbook.[3] On page 15 of the Yearbook, there is a poem entitled, “Story of the Origin of The Stark Name.” The events depicted in this poem came from an account by Sir George Mackenzie [1636-1691]; which tells the story of how John Muirhead was given the name John Stark by a grateful King James IV of Scotland.[4] On Page 21 of the 1903 Yearbook, there can be no doubt the membership believed they were all descendants of John Muirhead. In the preamble to the Association's Constitution, they stated:

“In order to form a more perfect union and to become more closely identified with each other, We, the descendants of one Muirhead, a Scotchman, to whom King James the Fourth gave the name of Stark (meaning strong) for his great bravery, do hereby adopt the following as the constitution of the Stark Family.”

Page 2

Historical Background Suggesting Aaron’s Old World Home

From their first arrival aboard the Mayflower in 1620, until 1629, only about 300 Puritans had survived in New England, scattered in small and isolated settlements. In 1630, their population was significantly increased when the ship Mary and John arrived in New England carrying 140 passengers from the English West Country counties of Dorset, Somerset, Devon and Cornwall. It was the first of the ships later called the Winthrop Fleet to land in Massachusetts.

While the passenger list related to this voyage is not well documented, we know the immigrants founded the First Parish Church of Dorchester in 1631, the place name of their new community taken from Dorchester, Devon County, England. In the 17th century this English town was at the center of the Puritan emigration to America, and the local rector, Rev. John White, was instrumental in organizing the voyage and supported the settlement of Dorchester, Massachusetts. For his efforts on behalf of Puritan dissenters, White has been called one of the unheralded founders of the Massachusetts Bay Colony.

In 1633, the Plymouth Trading Company established the first Connecticut settlement, a trading post at what would later become Dorchester, Connecticut in territory the Dutch claimed and in which they maintained a fort and trading post about seven miles downriver in what was later Hartford, Connecticut. In 1635, Puritan and Congregationalist members of Reverend Warham’s and Reverend Maverick's congregation, including, John Mason, Roger Ludlow, Henry Wolcott, and others, all prominent settlers in the new community, became dissatisfied with the rate of Anglican reforms. They sought permission from the Massachusetts General Court to establish a new ecclesiastical society subject to their own rules and regulations. About 60 individuals, totaling 23 heads of households, undertook a two-week's journey about 100 miles to the west. They founded a new town they initially also named Dorchester. Later, on February 21, 1636, the Connecticut General Count changed the name of the settlement from Dorchester to Windsor, believed to be named after the city of Windsor, England located on the River Thames. The new town was the first English settlement in the now state of Connecticut. All of the above suggests Aaron could have been one of these immigrants arriving in New England during this early historical period. Therefore, it would seem reasonable to begin a search for Aaron’s old world home in the southern part of England between 1608 to 1637 and his parents in the region before 1608. 

Aaron Stark's descendants are members of Group 1 while the legendary John Muirhead's descendants are in Group 2 (known as the Killermont Stark Families). 37 marker Y-DNA Comparisons of any member in Group 1 to any member in Group 2 resulted in a minimal genetic distance of 17 — obviously resulting in a common ancestor that lived well before the legendary John Muirhead. Further, Group 1 and Group 2 Members belong to different subclades of the R-M269 Haplogroup and could not have shared a common ancestor within the last 3,100 to 3,900 years. Assuming the Group 2 genealogy is correct and they are descendants of the Killermont Stark Families — who were descendants of the legendary John Muirhead — then their ancestors were members of the  Scottish Nobility. Obviously, if one has read the Introduction of the publication entitled, "The Life & Times of  Aaron Stark [1608-1689," one has to presume he was a commoner. Hypothesis I is suggested from the following quoted from, "The Life & Times of Aaron Stark [1608-1685],"

Page 3

Hypothesis 2 can be addressed by genetic analysis of others confirmed to be members of Haplogroup R-U106.[4]  Comparisons of the Aaron Stark Modal 37 Marker Haplotype (ASMH) to persons in the FTDNA database confirmed to be in Haplogroup R-U106 could determine if there was an early cluster of R-U106 in southern England at the approximate time of Aaron's birth. To test this theory, in the analysis that follows, persons will be selected for comparison to ASMH who report the year and place of birth of their earliest ancestor in England and Europe; and who also have been confirmed to be in Haplogroup R-U106. But, before the genetic analysis begins, some early Germanic Historical background.

^____________

Page 4

Hypothesis 2 is suggested because Group 1 is confirmed to belong to Haplogroup R-U106 found in higher concentrations in England (21.4%), Scandinavia (Denmark 17.7%), and reaches its maximum in the Netherlands (37.2%).

__________

Page 5

Construction of a Theoretical Historical & Genetic Timeline 

Suppose Aaron was a descendant of one of the Frisian tribes that migrated from the European Continent to southern England. Because Aaron's descendants belong to Haplogroup R-U106 — it is a given Aaron also had the U106 mutation. The progenitor of the U106 mutation could have been born between the calendar years 1900BC and 1100BC. Assuming  the time span of Aaron's ancestral generations consistently averaged 34 years, it is also a given that there were approximately 81 to 104 direct male ancestors from Aaron back to the U106 progenitor; all of whom would have tested positive for the mutation. 

An Infinite Alleles statistical mathematical model for estimating Time to the Most Recent Common Ancestor (TMRCA) was introduced in a publication by Bruce Walsh in 2001.[1]  On page 898 of this article, equation 3 introduced the "likelihood for the time t back to the MRCA given that we observe k out of n matches.." [n equals total # Markers compared. k equals total number of allele matches over the markers compared. t equals number of generations to TMRCA.] Rather than a genetic distance comparison, this will be a marker match ratio comparison. For example, when compared over 37 markers, if there is a match at 19 markers and a mismatch at 18 markers, the match ratio would be 19/37. The amount of difference in alleles values is not considered; only a mismatch in value at a specific marker is considered. This analysis will use the Infinite Alleles Model to genetically compare ASMH to other male participants confirmed to have the U-106 mutation.

For the hypothetical genetic comparison's that follow, an online calculator[2] using the infinite alleles model will be used to determine: "the two individuals genetically compared most likely shared a common paternal ancestor x number of generations ago." x in generations from this calculator will correspond to the peak posterior distribution (Likelihood in %) that occurs given the total number of Markers (n), the total number of marker matches (k), and the mutation rate.

 Using the Calculator, Table 1 is a representation of a theoretical historical and genetic timeline. The historical timeline focuses on historical events related to the Germanic Tribes; while the genetic timeline is based on the  TMRCA calculations producing results that fall within the historical events.  n  has the  constant value 37 Markers, k ranges in value from a low of 19 marker matches to a high of 35 marker matches. The total number of matches for each row in Table 1 is presented in parenthesis in the "Peak" Likelihood column.

 From the peak likelihood values calculated, those likelihood values that are ± .95 the peak likelihood value — provide a range of likely generations within which a Most Recent Common Ancestor may have lived given the number of Marker Matches k. The mutation rate used was .003, the approximate value (.00342 actual) calculated for Aaron's 37 Marker Haplotype Life Expectancy.

All of the Calendar dates were calculated using 34 years as the time span of a generation. The total number of generations calculated was multiplied by 34 Years; this resultant value then subtracted from the Calendar Year 1948; this resultant value being the approximate  average Calendar Year of births in that generation. Positive values are AD Calendar Years while negative values are BC Calendar years. [For example: 100 Generations Times 34 years equals 3,400 years. The Calendar Year 1948 minus 3,400 Years equals the value -1,452. Therefore the Calendar Year calculated will be 1452BC.]

Row 1 records the range of generations before the present in which the progenitor of the U106 mutation most likely lived. Using the TMRCA calculator, k/37 matches were entered until a ratio produced a range of generations and Calendar Time Span that approximated those in Row 1. The nearest match ratios that produced this result were 19/37 and 20/37 (recorded in Rows 2&3). For the remaining rows, the number of matches were increased by 1 and calculated until the a match ratio of 35/37 was achieved. In Column 2, observe the numbers in parenthesis begin with a 19/37 Match ratio in Row 2; the match value then increasing by 1 in each Row; ending with the Match ratio of 35/37 in the final row of Table 1. Historical and genetic comments related to the data presented appear in the last column. The BC Calendar Years appear on this page. Table 1 continues on the next Page and has the AD years. An analysis appears at the conclusion of Table 1.

Page 6

Discussion of the Theoretical Historical & Genetic Timeline 

Suppose Aaron was a descendant of one of the Frisian tribes that migrated from the European Continent to southern England. Because Aaron's descendants belong to Haplogroup R-U106, it is a given Aaron also had the U106 mutation. According to Table 1, the progenitor of the U106 mutation could have been born between the calendar years 1900BC and 1100BC. Assuming  the time span of Aaron's ancestral generations consistently averaged 34 years, it is also a given that there were approximately 81 to 104 direct male ancestors from Aaron back to the U106 progenitor. If the Infinite Allele Model is correct, Table 1  suggests genetic comparisons to ASMH with less than a 19/37 match ratio probably would not test positive for the U106 mutation.

Table 1 suggests that persons compared to ASMH having a 28/37 or 29/37 match ratio would share a Germanic common ancestor with Aaron's descendants that first arrived in Britain from northern Germany, the northern part of the Netherlands, or southern Scandinavia. This suggests comparisons to persons who are descendants of English ancestors having match ratios of 29/37 to 35/37 to ASMH are more likely to have early ancestors born in Great Britain after the migration period. If these persons are positive for the U-106 mutation and share a common ancestor who was a member of one of the Germanic Tribes, then, theoretically, one would expect a cluster of match ratios in southern England within 29/37 to 35/37. Not all persons in Western Europe with the U106 mutation migrated to England. Some families stayed behind and comparisons of ASMH to these persons who are descendants of early ancestors who lived in Main Land Europe, theoretically, would be expected to have match ratios from 19/37 to 28/37. We should also expect these ratios in England, for Aaron's common ancestor with these persons may have lived in Main land Europe before the migration period.

If Rannulf Stark was an ancestor of Aaron, he was alive 386 years before Aaron. If the record in which his name appears required Rannulf to be 21 years of age, then he was born in or before the year 1201AD. Using the same logic for William Stark;  he was alive 294 years before Aaron was born and his year of birth was in or before 1293AD. In England, surnames were most likely first introduced about 1200AD within the nobility and eventually established throughout England soon after the 1377 Poll Tax was introduced. If Rannulf was an ancestor of Aaron, he would have been born approximately 21 generations before Aaron's 10th generation descendants. Using the same logic for William Stark reported in 1314: he was born approximately 18 to 19 generations before Aaron's 10th generation descendants. Referring to Table 1 again, if Aaron was a descendant of one or both of these men, it would seem reasonable others compared to ASMH with different surnames would most likely have 33/37  to 37/37 matches ratios — but could not have a common ancestor who was born after the calendar year 1200AD.

Aaron Stark was born in 1608. Assuming persons not having the surname Stark could not share Aaron Stark as a common ancestor, then persons that match at 35/37 markers most likely share a common ancestor who lived 11 or more generations earlier than the present.

The comparison results data, selection criteria, and sources of the R-U106 comparisons to ASMH can be found in  Appendix 1.

Page 7

It would seem reasonable to suggest these theoretical ancestors of Aaron's descendants — genetically shared with those persons compared to ASMH — represent the beginnings of many separate surname descendant lines, one of which was Aaron's first ancestor with the surname Stark. It would also seem reasonable that Rannulf Stark, documented in the records of Suffolk County in 1222AD, was an ancestor of William; perhaps the first to have the surname Stark. If Rannulf Stark was an ancestor of Aaron, then he would have had the U106 mutation. According to Table 1, persons reporting their earliest known ancestor was a resident of Suffolk County and having 28/37 and 29/37 match ratios when compared to ASMH, could share a common ancestor born in England after the migration period. Perhaps these shared ancestors were earlier ancestors of Rannulf Stark — living in England before surname usage began.  Persons reporting their earliest known ancestor was a resident of Suffolk County and having  22/37 and 25/37 match ratios, when compared to ASMH, share a common ancestor born well before the Germanic migration to England. These common ancestors most likely lived in one of the Figure 2 Frisian Settlements.

The cluster centering in and around Lancashire County most likely share  common ancestors with Aaron's 10th generation descendants who lived in Continental Europe before the migration to England. The 24/37 and 25/37 match ratios suggests the common ancestors were born between 738BC and 58 BC, the later date approximately 400 years before the Germanic migration to England. Aaron's descendants and the descendants of these persons most likely do not share a common ancestor born in England. The earliest known ancestor of the 33/37 match ratio in Lancashire could have migrated from southern England.

The other comparisons south of the Thames River suggests common ancestors that could have been born as early as 466BC and as late as 1098AD. Table 1 suggests the common ancestors of the 29/37, 30/37 and 31/37 match ratios could have been born as early as 384AD and as late as 1098AD. Between 800AD and 900AD, the Kingdom of Wessex produced the first King to achieve direct rule over what is considered "England." This Kingdom was located south of the Thames River. It would not seem unreasonable to suggest some movement of the Suffolk families towards places of power could have occurred and the earliest known ancestor of those compared to ASMH were their descendants.

Comparisons to those reporting their earliest known ancestor lived in Western Europe had match ratios ranging from a minimum of 19/37 to a maximum of 28/37. It would seem reasonable the ancestors of those selected for comparison were descendants of members of the Germanic Tribes that did not migrate to England. In Table 1, the common ancestor of those compared lived between 180AD and 554AD. The historical comments for this time span are: 

"The Frisians return to the Clay District in 400AD. The Germanic migration period last from 350AD to 500AD. Germanic tribes migrate all over Western Europe after the collapse of the Roman Empire, forming new tribes in the newly conquered areas. Around 450AD, the Angles, Saxons, Jutes and a Frisian fraction cross the North Sea and establish the Anglo-Saxon empire (currently known as England). The Frisians colonized the county of Kent in southeast England. Around 480AD, Clovis establishes the Frankish Empire (currently known as France), originating from the Chaukians and Frisians.

Therefore, it is reasonable to expect mose match ratios will be below 29/37 in the European comparisons to ASMH. Further, as mentioned earlier in this publication, comparisons of persons with the U106 mutation were expected to result in a minimum 19/37 match ratio. However, there were two 18/37 match ratios observed — both being descendants of European Ancestors.

Conclusions

Observations presented in the above discussion certainly would appear to provide credible evidence Aaron Stark could have been born south of the Thames River in England. Based on the above, I would suggest the genealogical research for Aaron's ancestral home be directed into these regions, attempting to find links to others living in the region with the surname Stark at the time of Aaron's birth. Other derivations (Starks, Starke) of the surname should also be pursued. I believe Aaron's deep ancestral roots are connected to the Germanic Tribes discussed because he was positive for the U106 mutation. It now remains to search for his English Ancestors. 

Page 8 

Part III Appendix 1: Genetic Comparisons to Aaron Stark Modal Haplotype (ASMH)

Part III Appendix 1: Selection & Sources of Genetic Comparisons to ASMH

Tables 3 and 4 present the genetic matches of persons who have been tested and confirmed to be in Haplogroup  R-U106. The majority of this data was obtained from the web site entitled "Y-Search."[1]   Family Tree DNA created this web site as a free service to allow people that have tested with different companies to make their results available for comparison. Because all of the descendants of Aaron Stark were tested by FTDNA, only those persons tested by FTDNA were chosen to be compared to Aaron's 37 Marker Modal Haplotype (ASMH). Other criteria for selection for comparison were the following:

Y-Search offers a Haplogroup Search. An "All Regions Search" of those in the database reported to be in Haplogroup R1b1b2a1a (R-U106) was conducted. The results of various search criteria reported there were — at the time this article was published — 733 total persons in the database who tested positive for the U106 mutation; 249 persons in the British Isles search; 111 persons in the Western Europe search; and 17 persons in the Scandinavia Search. The remainder of the searches with results but not used in the comparisons were: 173 persons in the Americas; and 41 in Eastern Europe. Other search regions resulted in 0 persons. As presented in the above selection criteria, of the total potential subjects found, those persons and Haplotypes not meeting the above criteria were not compared to ASMH and were not presented in the genetic comparisons that follow. Some FTDNA surname projects were also searched using the above criteria for potential subjects for comparison. Persons found in these projects that were not listed in Y-Search were added to this Appendix. While there were more persons who met the above criteria than are presented in the Appendix, those presented were representative of the conclusions discussed in this publication.

There were two exceptions to the above criteria, both predicted to be in Haplogroup R1b1b2 (R-M269); but not tested for the U106 mutation. Having the surnames Halstead and Land, they had the highest number of matches when compared to ASMH and most likely would test positive for the U106 mutation.

Table 2: Aaron Stark Modal Haplotype

^____________

Page 9 

Table 3: Genetic Comparisons  to ASMH; Descendants of English Ancestors

^____________

Page 10

Table 3 Continued: Genetic Comparisons  to ASMH; Descendants of English Ancestors

^____________

Page 11

Table 4: Genetic Comparisons  to ASMH; Descendants of Western Europe Ancestors