Sunday, 7 December 2014

Were Richard III's parents second cousins as expected?

Isabella of Castille:
 Unfaithful wife of Edmund,
Duke of York?
As I wrote previously, the investigation of Richard III's skeleton revealed that he does not share the same Y-DNA as his putative paternal-line relative Henry Somerset, 5th Duke of Beaufort and there are 18 possible birth events where non-paternity may have occurred.

There are four possible breaks between Edward III and Richard's own birth. One of the more probable false paternity events would be the birth of his grandfather, Richard of Conisburgh. Another is the birth of John of Gaunt. Provided the researchers were able to extract sufficient autosomal DNA, there is a simple, free, test which may reveal that there is a false paternity within these five steps of the broken line (John of Gaunt-Edward III-Edmund of York-Richard of Conisburgh-Richard Plantagenet-Richard III). has a tool which can take autosomal DNA testing results and compare one side of a chromosome pair against the other, looking for similarities in regions which are expected to vary. It can then estimate which percentage of the DNA on each side is likely to have come from a common ancestor and how many generations back (basically based on amounts shared). Everyone gets one side of each chromosome pair from their father and one from their mother and if the parents are not related, there will be the expected variety and little similarity. The test evaluates how related an individual's parents are without having to test anyone but the individual.

According to the genealogy, Richard's parents are supposed to be, at their closest relationship, 2nd cousins (generations to most recent common ancestor would be about 4). His maternal grandmother's father should be the brother of his paternal grandfather's father. These brothers are John of Gaunt and Edmund of York and both paternity events are key interconnected steps on the broken line.

Additionally, Richard's father is supposed to be the 2nd cousin twice removed of his mother as his father's mother, Ann Mortimer, was also descended from Edward III, as the granddaughter of one of his granddaughters. Finally, his parents should have an additional more distant cousinship from common ancestry that is not related to Edward III.

Therefore, five steps in the paternity line happen to be uniquely tied to this supposed 2nd cousin relationship. Running Richard III's autosomal DNA results through the parental consanguinity calculator should indicate his parents have a common ancestor about 4 generations back. If the result is that the ancestor is farther back than 4 generations (or that his parents are unrelated), there is a break in the paternity line in the five steps between John of Gaunt and Richard III.

The key question is, have the researchers recovered enough autosomal DNA?

Friday, 5 December 2014

Infidelity possibilities in the line between Richard III and Henry Somerset

As demonstrated in Nature Communications, a multidisciplinary investigation of Richard III's skeleton revealed at least one non-paternity event in the 19 generations separating Richard III from his putative paternal-line relative Henry Somerset, 5th Duke of Beaufort. Looking at the events surrounding the conception and birth of all 19 individuals may help illustrate where the breakage (or breakages) most likely occurred.

As Ed West pointed out, "in primates, including humans, high-status males are rarely cuckolded; on top of everything else a king could have his wife beheaded if caught, something your average man couldn't get away with." Therefore, relative status of the father within each relationship needs to be considered, along with whether the child was an heir presumptive at time of conception. Births of sons to reigning male monarchs or males first in line to the throne are underlined below. These would represent both treasonous infidelities and counter the known improbability for a high status (in such a case, arguably the highest status) male to be cuckolded, and therefore are much less likely than other scenarios. Other birth scenarios and event characteristics which may increase the probability for infidelity are in red.

Four potential breaks: The births from Edward III to Richard III

  • 5 June 1341 birth at  Kings Langley Palace in Hertfordshire of Edmund, Duke of York (died 1402): He was the fifth son and seventh child of the marriage of a reigning monarch, Edward III to Philippa of Hainault. He was conceived in 1340 when his mother was 26 years old and 12 years after his mother was crowned queen. All of her children were born one to three years apart over a 25 year period 1330-1355, when she was aged 15 to 41. They had five children after Edmund. She was known for travelling with her husband throughout their marriage. No time apart between the two in 1340 has been identified by preliminary Internet research.
  • Approximately 20 July 1375 birth at Conisbrough Castle, Yorkshire, of Richard, Earl of Cambridge (died 1415). He was conceived as the second son and third child of the marriage of Edmund and Isabella of Castille, the youngest daughter of Peter the Cruel, King of Castille. At the time he was conceived, Edmund was Earl of Cambridge, the fifth son of the King and no closer than sixth in line to the throne (making Richard, if born male, 8th in line). Edmund was not well regarded, described as a man without the abilities suited to his station, while Isabella was the daughter of a reigning monarch known for despotism. Richard was his mother's third and last child, born when she was about 20, although she lived for another 17 years and Edmund outlived her. She had married Edmund when she was 17 and had a firstborn son, a daughter, and Richard within three years. In a 1988 book, modern scholar T.B. Pugh concluded "they were an ill-matched pair."  There are significant problems with the paternity of Richard, Earl of Cambridge that greatly increase the probability this is a non-paternity event:
    • The Yorkist claim to the throne made by his grandsons ran through the line of his wife, Ann Mortimer, and not through him.
    •  Wikipedia notes: "As a result of her indiscretions, including an affair with King Richard II's half-brother, John Holland, 1st Duke of Exeter (d.1400), whom Pugh terms 'violent and lawless', Isabella left behind a tarnished reputation, her loose morals being noted by the chronicler Thomas Walsingham. According to Pugh, the possibility that Holland was the father of Isabella's favourite son, Richard of Conisburgh, 3rd Earl of Cambridge, 'cannot be ignored'."
    • "according to Harriss, since he received no lands from his father, Edmund ... Duke of York, and was not mentioned in either his father's or his brother's wills, he may have been the child of an illicit liaison between his mother and John Holland, 1st Duke of Exeter."
      • The Y-DNA line of John Holland goes back to Sir Robert de Holland of Upholland, Lancashire, born about 1250. Haplogroup G could be expected to have a 1-2% prevalence in Lancashire at that time, with an unknown but possible presence for subclade G2. John Holland's male line is not known to have persisted at this time.
  • 21 September 1411 birth of Richard, Duke of York. He was the third child and second son, but probably only surviving son of the marriage of Richard, Earl of Cambridge and Ann Mortimer. Ann died shortly after he was born, aged about 20. Ann was for much of her life the daughter of the heir presumptive to the throne, and briefly the sister of the heir presumptive until Richard II was deposed by the Lancastrians.
  • 2 October 1452 birth of Richard III at Fotheringhay Castle, the eighth son (probably fifth or sixth surviving) and twelfth child of Richard Plantagenet, 3rd Duke of York and Cecily Neville. When Richard was conceived in 1451, his father, Richard Plantagenet was the wealthiest and most powerful noble in England, other than the king, and had the leading Yorkist claim to the throne. Cecily Neville was noted for her pride, temper, and piety. Her children were all born 1-3 years apart over a 17 year period when she was aged 22 to 40. No time apart between the two early in 1452 is noted. The couple had one other child after Richard.

Fourteen potential breaks: The births from Edward III to Henry Somerset

  • 6 March 1340 birth in Ghent of John of Gaunt. He was the fourth son and sixth child of the marriage of a reigning monarch, Edward III to Philippa of Hainault. He was conceived in 1339 when his mother was 25 years old and 11 years after his mother was crowned queen. All of her children were born one to three years apart over a 25 year period 1330-1355, when she was aged 15 to 41. They had six children after John. She was known for travelling with her husband throughout their marriage. Rumours circulated that he was the son of a Ghent butcher but only later in his life after he became unpopular and were based on a spurious conclusion that Edward III’s absence from his birth was significant (investigating the timing of these rumours, given the 1374 conception and 1375 birth of Richard of Conisburgh which seems to have the highest probability of being a non-paternity event, now has greater priority given the DNA results).
  • 1373 birth of John Beaufort. He was the first of four illegitimate children of the relationship of John of Gaunt, Duke of Lancaster and son of the king at the time John was conceived, and Katherine de Roet Swynford. While not married, the power differential between the couple is more in favor of actual paternity than not. John and Katherine had three sons and a daughter (1373, 1375, 1377, and 1379). Katherine had previously married Hugh Swynford and had children with him (about 1369 and 1367, and possibly two more for which details are unavailable) before he died in 1371/2. John’s second wife, Constance of Castille (sister of Isabella who married Edmund, Duke of York and may be responsible for the non-paternity event on the other s) died in 1394 and John married Katherine in 1396. The pope legitimized their children shortly after, as did Richard II by 1399.
  • 1406 birth of Edmund Beaufort. He was the third son and fifth child of six children of the marriage of John Beaufort, legitimated grandson of a sovereign and Earl of Somerset at the time of Edmund’s conception and Margaret Holland. Margaret appears to have had the six children between 1401 and 1409 when she would have been aged 16-24. Edmund died in 1410. Margaret then married Edmund’s nephew, Thomas of Lancaster, 1st Duke of Clarence, but the couple had no children. She died thirty years later in 1439.
  • 26 January 1436 birth of Henry Beaufort. He was the third child and eldest son of Edmund Beaufort, heir to the Duke of Somerset at the time Henry was conceived, and Eleanor Beauchamp. While Edmund “was the head of one of the greatest families in England, his inheritance was worth only 300 pounds.” Edmund and Eleanor were probably married about 1425 and had two daughters before Henry and then four more sons and three more daughters. Eleanor seems to have had her last child about 1455 when she would have been 47 and which was the year Edmund died. There may or may not have been a gap between the children in the 1440s as evidence providing detailed birth dates is scant. Eleanor lived until 1467.
  • Circa 1460 birth of Charles Somerset. He is the only known son of the relationship of Henry Beaufort, Duke of Somerset at the time of his conception, and his mistress Joan Hill. Born illegitimate, he was legitimized later in life by his father. Little is known of Joan Hill, but the relative status difference in the couple was in favor of actual paternity rather than not.
  • Circa 1496 birth of Henry Somerset [full disclosure: my 13th great grandfather, genealogically]. He was the only son of two children of the marriage of Charles Somerset, a Knight of the Garter (and illegitimate son of Henry Beaufort, 3rd Duke of Somerset) at the time Henry was conceived, and Elizabeth Herbert, 3rd Baroness Herbert. Charles and Elizabeth married 2 June 1492 and only two children are known. Charles was created Lord Herbert in right of his wife. Elizabeth is known to have died before 1513 but in which year is not clear.
  • Circa 1526/7 birth of William Somerset. He was the eldest son of Henry Somerset, Earl of Worcester at the time (or shortly after) William was conceived, and Elizabeth Browne. William and Elizabeth were married before 1526/7 and are known to have had three more sons and four daughters after William. Elizabeth was lady-in-waiting to Anne Boleyn and her main accuser. Updated: Her accusations were apparently a reaction to her own adultery coming to light in 1536.
  • Approximately 1550 birth of Edward Somerset. He was the only son of three children born to William Somerset, Earl of Worcester at the time (or shortly after) Edward was conceived, and Christian North. William and Christian are known to have had two daughters in addition to Edward, but birth dates and additional details are scant.
  • 1577 birth of Henry Somerset. He was the eldest surviving son of Edward Somerset, heir to the Earl of Worcester at the time of his conception, and Elizabeth Hastings. Edward and Elizabeth married in December 1571 and are known to have had fifteen children. Edward was an important advisor to King James I and served as Lord Privy Seal.
  • 9 March 1602 or 1603 birth of Edward Somerset. He was the firstborn of nine sons and four daughters of the marriage of Henry Somerset, son and heir to the Earl of Worcester at the time of his conception, and Anne Russell. Henry was considered an “outstandingly wealthy peer.” Henry and Ann married 16 June 1600 and had thirteen children after Edward.
  • 1629 birth at Raglan Castle of Henry Somerset. He was the firstborn and only son of Edward Somerset, Lord Herbert of Raglan at the time of his conception, and Elizabeth Dormer. Edward and Elizabeth married in 1628, shortly after Edward graduated from Cambridge University with a Master of Arts degree and when he was still one of the richest lords in England. Elizabeth had a daughter about 1631 and another before 1635 when she died.
  • 25 December 1660 birth of Charles Somerset. He was the second but eldest surviving son of Henry Somerset, Lord Herbert at the time of his conception, and Mary Capell. Henry and Mary had married, using the “republican” form of marriage before a justice of the peace (Wikipedia) in 1657 and Henry made himself acceptable to Cromwell and survived the interregnum. However, he was involved in the Royalist plot of July 1659 and was committed to the Tower of London. A letter he wrote to his wife from the Tower in August of 1659 survives. Henry was released 1 November 1659 and Charles was conceived about five months later. Mary had previously been married to Henry Seymour and had two children by him before he died in the Tower in 1654. As outlined above, she married again in 1657, had a son before 1660 who died as an infant, had Charles in 1660 when she was aged 30, and then likely had four daughters and one more son (1663, 1664, a gap until 1670, 1671, and 1673).
  • 2 April 1684 birth at Monmouth Castle of Henry Somerset. He was the eldest child and only son of the marriage of Charles Somerset, the Marquess of Worcester and heir presumptive to the dukedom at the time of his conception, and Rebecca Child. Rebecca married Henry in 1682, had Henry in 1684 and a daughter eight years later in 1690. It is not known if other pregnancies occurred or children who did not survive were born in the interval. The couple are not known to have had further pregnancies or children between 1690 and Charles’ untimely death in a coach accident in 1698, aged 37. The dukedom skipped to Henry when Charles’ father died in 1700.
  • 12 September 1709 birth of Charles Noel Somerset. He was the second son of Henry Somerset, 2nd Duke of Beaufort at the time of his conception, and Henry’s second wife, Rachel Noel. Rachel married Henry in 1706 and bore him a son Henry in 1707 and then Charles in 1709, shortly after which she died. Henry himself died in 1714, at the young age of 30, after marrying a third time (no issue) in late 1711.
  • 16 October 1744 birth of Henry Somerset. He was the third child and firstborn son (ultimately only son) of Charles Noel Somerset (who was the heir presumptive his childless brother’s dukedom at the time of his conception in 1744) and Elizabeth Berkeley. Elizabeth was 31 years old when he was born, having married Charles Noel Somerset four years prior and born two daughters earlier. The couple would go on to have three more daughters, one in 1746, one in 1748, and one after a gap of 8 years in 1756 (the last arrived when Elizabeth was 43 years old).  

Appendix: Understanding where Edward III's Y-DNA came from and haplotype probabilities

According to Wikipedia, Edward the III's Y-DNA line traces to Fulcois, the Count of Perche, a French noble living in the 900-1000s who was probably of the family of viscounts from Châteaudun in northern France.

Fulcois (if there were no non-paternity events along the descent to Edward III), was Edward's 9th great grandparent via:
  1. Hugues du Perche, born northern France (probably Châteaudun), the youngest son of Fulcois (the count of Perche at the time of his conception)
  2. Geoffrey II, born northern France, probably firstborn son (of the count of Perche at the time of his conception), became the Count of Gâtinais and married a Duchess of Burgundy
  3. Fulk IV, born in northern France, the younger son of the above (who was Count of Gâtinais at the time of his conception).
  4. Fulk V of Anjou, born between 1089 and 1092 at Angers, a city in western France, as the second son, and the only known child of his parents, Fulk IV (Count of Anjou at the time of his conception) and Bertrade de Montfort (Bertrade was Fulk's fifth wife). In 1092, aged about 22, Bertrade left Fulk IV and bigamously married Philip I, King of France.
  5. Geoffrey V, aka Geoffrey Plantagenet, born 24 August 1113, presumably northwestern France,  the firstborn son and probably third child of Fulk V (Count of Anjou at the time of his conception)
  6. Henry II, born 5 March 1133 at Le Mans, France, the eldest child of the above and Empress Matilda
  7. John, King of England, born in Anjou or Aquitaine, France, 24 December 1166, the sixth son and ninth child of reigning monarch King Henry II and Eleanor of Aquitaine
  8. Henry III, born at Winchester Castle, Hampshire, England1 October 1207, the firstborn son of the king 
  9. Edward I, born at the Palace of Westminster, London, England, the night of 17-18 June 1239, firstborn son of the king
  10. Edward II, born 25 April 1284 in Caernarfon Castle in North Wales, the fourth son of the king (two older brothers died before his birth and a third died four months after he was born, so a potential heir presumptive at time of conception).
  11. Edward III, born 13 November 1312 at Windsor Castle, Windsor, England, firstborn son of the king
Two Y-DNA haplotypes were identified in the Richard III investigation as the possible Plantagenet type.  The two types are are G2 (P287) for the Richard III skeleton, and R1b1b (U152) for Henry Somerset, 5th Duke of Beaufort (as reliably triangulated from descendants).

However, there may be multiple non-paternity events, in which case it is possible neither is the true Plantagenet type and/or Edward III's type.

Both haplotypes are found in modern populations in Britain and France, but R1b1b (U152) is substantially more common than G2 (P287):
  • According to Eupedia, R1b1b (U152):
    • has a prevalence of 10-15% in north western France and reaches 15-20% in northern France. A peak of 20-30% is found on the far eastern border of France and in a pocket centered on Poitou, south of Anjou.
    • has a prevalence of 10-15% in pockets of south eastern England fading as it moves westward to a prevalence of less than 1% in Wales.
  • While Richard's type is G2, according to Eupedia, most Europeans who are G belong to subclade G2a:
    • and G haplogroup overall is found at a small prevalence of 1-2% in both northern France and England, outside of Wales where the prevalence increases slightly to 3-5% and an area near Calais that can show a prevalence as high as 6-8%. The prevalence of subclade G2 is significantly less than that for G as a whole and it is considered a very rare type.
Therefore, assuming the modern maps do not fundamentally misrepresent a basic availability of both groups in France and England in the 900s, either haplogroup makes sense as a Plantagenet type, based on the known origin of the line at Châteaudun, or as a type introduced to the line by a European or Middle Eastern interloper.

Saturday, 20 September 2014

Diagram of how 4th cousin matches work in autosomal DNA testing

Simulated relationship between two DNA testers who show on each other's match lists as 4th cousins. The ancestry of one tester is represented on the left (starting with her parent's generation) and a second person who matches with 1 segment, 14 cM or 0.19% is represented on the right (also starting at parents generation). Note the complex inheritance of haplotypes, which are not shared between the matches.