The Genetic Genealogy Revolution
Berry Pomeroy Castle, Devon, England.
The speed and enthusiasm with which the American genealogy community has embraced genetics has been truly astounding. I vividly remember addressing a meeting of the New England Historical Genealogy Society, America’s oldest, in 2001 when only a tiny minority in the audience knew much about DNA and hardly anyone had heard of mitochondria or Y chromosomes. Now knowledge is detailed and extensive, and genuine advances are being made, fueled by the curiosity of members of the public about their roots rather than by academics.
The very first scientific paper to feature Y chromosomes in any sort of genealogical connection concerned a man who has already made an appearance in this book, and will do so again. He is Thomas Jefferson, third president of the United States and principal author of the Declaration of Independence. Jefferson’s wife, Martha, died in 1782 following the birth of her sixth child, Lucy Elizabeth. Some years later Jefferson and one of his slaves, Sally Hemings, who was the half-sister of his late wife, became lovers and he may have fathered a further six children by her including her last son, Thomas Eston, born in 1808. The story is a fascinating one for many reasons, not least because, given his political prominence and the unforgiving nature of the times, Jefferson denied it. The scandal, for that was what it was, rumbled on for the next two hundred years until 1998, when genetic evidence proved, beyond reasonable doubt, that it was true. The study that clinched the verdict compared the combinations of markers in the Y-chromosome signature from a direct patrilineal descendant of Thomas Eston Hemings with equivalent relatives of the president (Figure 1).1 Thomas Jefferson did not have any surviving legitimate sons, so his Y chromosome had to be identified through a patrilineal relative, who in this case was the president’s paternal uncle Field Jefferson. Both Field and Peter Jefferson, the president’s father, had inherited the same Y chromosome from their father, Thomas. Therefore any patrilineal descendant of Field Jefferson would carry the same Y chromosome as the president. When the Y chromosomes of five such descendants of Field Jefferson (A–E) were compared with a direct descendant of Thomas Eston Hemings (F), they matched exactly. Moreover, the precise Y-chromosome signature was rare in the general population, making the match extremely significant. This conclusion has not gone down well with some Jefferson descendants, but it was the result of a decisive piece of work.
Figure 1. Patrilinear relatives and descendants of President Thomas Jefferson with matching Y-chromosome signatures. Lengths of vertical links are approximately proportional to the number of generations.
Although the Jefferson/Hemings case demonstrated the power of genetics to prove a genuine patrilineal connection, it did not make a general case for the use of the Y chromosome to follow surnames. In fact, quite the opposite. Because Thomas Eston Hemings was illegitimate, the son of a slave and her master, and carried his mother’s name rather than the president’s, it was a prime example of what geneticists call a nonpaternity event, where a surname does not follow the same line of descent as a Y chromosome. This could be because of illegitimacy, as here; adoption; a deliberate name change; or infidelity by the mother. Whenever this happens, the link between a surname and its Y chromosome is broken forever. It was the disruptive effect of nonpaternity on surname/Y-chromosome associations that persuaded those few geneticists who thought about such things at the time that the Y chromosome was unlikely to prove useful on a larger scale. I think we were swayed by the generally high rate of nonpaternity, sometimes as high as 10 percent, uncovered by conventional genetic fingerprinting among the modern population, whereas it was the historical rate that was more relevant. Only when I was curious to test another man with my surname, Sir Richard Sykes (at the time the chairman of the pharmaceutical giant Glaxo), to see if we were related, did the surprisingly high general correlation between surnames and Y chromosomes in men start to come to light.
After I found that our Y chromosomes matched and so we were likely descended from a common ancestor, I randomly recruited around one hundred other male Sykeses from the phone books of West Yorkshire, where the name is concentrated, and found that 60 percent of us also shared this same Y chromosome. I have written about this extensively in Adam’s Curse, so I won’t repeat myself here. Suffice it to say that the association must have lasted for eight hundred years, since, like many other English names, Sykes became hereditary during the thirteenth century. For that to be so, the nonpaternity rate must have been much lower than I or my colleagues had, rather lazily and certainly mistakenly, assumed. For the Sykes name this rate worked out at just over 1 percent per generation. For other surnames, like Dyson from the same Yorkshire valley, it was even less, with 95 percent of Dysons sharing the same Y-chromosome fingerprint. I was fortunate in that the shared Sykes Y chromosome was rare, rather like Jefferson’s in that respect, as this made the matches very significant from a statistical point of view because they were very unlikely to have happened by chance. I think the effect was so striking, though, that even if the Sykes Y chromosome had been a common type, the fact that 60 percent of us shared it would have shown up as unusual.
Not being a genealogist, I was not really aware of the practical importance of this discovery, but that was to change very fast. Within days of the publication of the paper announcing the Sykes results, in April 2000, the lab was swamped by requests from members of the public asking us to help with their own genealogical research.2 The scientific paper had a fair amount of press coverage in Britain, including an appearance on breakfast television where I remember sharing the sofa with the world chess champion, Garry Kasparov (and in complete contrast and basically setting the tone of the show, a man who could balance on a coconut using only his thumbs). In the United States there was an article in the New York Times. The following week the London Times carried a feature on the seven clan mothers I had identified and named, following the mitochondrial DNA research in Europe. That also triggered a huge response from the public, with requests for genetic tests. So we were very soon faced with a choice in the lab, and I called a meeting in the conference room to come to a decision.
We agreed that we could not, and should not, use our research time and money to test individual families purely for their own interest. Once the general principle of the surname/Y-chromosome link had been established with the twenty or so names we were then working on, we would not be able to justify much more work, except in special cases with their own intrinsic academic interest. Our choice was, then, whether to decline these requests from the public or to set up a proper mechanism for dealing with them. We debated this long and hard. One or two members of the lab were apprehensive and against any “commercialization,” as they saw it. But in the end the argument that won the day was that we had by chance stumbled onto something that evidently appealed to a large number of people. To do nothing about it seemed to me and most of my colleagues to be tantamount to saying that we were only doing this research for the small audience of our scientific peers, not for the public. We agreed that we should respond, and that it was right to do so. We set up Oxford Ancestors that same week, hired a technician, and began taking orders for both Y-chromosome and mitochondrial DNA tests, the first company in the world to do so.
It took another year to set up an operationally independent commercial arrangement with separate staff and premises in a local business park. The company is still going strong and has recently celebrated its tenth anniversary. I am sure I was mistaken in thinking that if we didn’t do something about the public demand, it would go unsatisfied—there are now plenty of other companies doing this sort of thing—but I don’t regret doing so for a minute. Oxford Ancestors has helped tens of thousands of people “explore their genetic roots,” as it says in the company logo. It has been a great experience seeing so many aspects of what people can discover from their DNA, and along the way I have encountered some absolutely fascinating reactions. As I write, a decade into the era of public accessibility, I estimate that almost a million people have had their DNA tested for either mitochondrial or Y-chromosome DNA, or both. A lot of this has been personally financed, as it should be, through private companies like Oxford Ancestors in the UK, Family Tree DNA and Relative Genetics in the United States, as well as through American companies specializing in African ancestry.
Though Oxford Ancestors began in my university genetics lab in response to an unanticipated demand, other companies had different primary motivations, even if all of them ended up doing much the same thing. In the case of Houston-based Family Tree DNA, it was the curiosity of its genealogist founder that provided the vital spark, as the ebullient Bennett Greenspan explained when I visited him at the company premises on the top floor of a commercial building on the outskirts of this thinly spread city. Greenspan had taken an interest in the work on the Y chromosomes of the hereditary Jewish priesthood, the Cohanim. Briefly, the male descendants of Aaron, the brother of Moses, were selected to serve as priests. Genetic corroboration of the Old Testament tradition came when a particular Y chromosome was found at much higher frequency among Ashkenazic Cohanim than in comparable non-Cohanim. Greenspan became aware of this when he read the Wall Street Journal report of the earlier Nature publication in July 1998.3 He knew nothing about DNA, as he readily admits, but was already an experienced genealogist and, as a Jew, the Cohanim story caught his eye. The news also came at a time when he had just sold his photographic-products business, having correctly anticipated the effect of the digital camera revolution on the demand for film. So he was on the lookout for something new into which to channel his energies. His first visit was to the University of Arizona and the laboratory of Dr. Michael Hammer, one of the early pioneers of uncovering the genetic variation in the Y chromosome on which all subsequent studies depended. Although Hammer was not one of the members of the Cohanim project, which was carried out in England and Israel, he was the obvious man to speak to. And, in American terms, it wasn’t that far from Houston to Tucson.
Bennett Greenspan had a particular project in mind when he called on Hammer. He was trying to see whether another man named Greenspan, from Argentina, was related to himself, and having been unable to find a paper trail connecting the two of them, he realized the potential of the Y chromosome to solve the issue. It was as he was leaving that Hammer said to him, “Someone really should turn this into a business”—which is exactly what Bennett Greenspan did. Thus, Family Tree DNA was born, opening for business within weeks of Oxford Ancestors’ debut in 2000. As it turned out, he and the Argentinean Greenspan were not genetically related, and before you ask, no, Bennett still doesn’t know if he is related to his namesake Alan, the former chairman of the Federal Reserve.
The third early genetic genealogy company also sprang from a university, but in a rather different way from Oxford Ancestors or Family Tree DNA. For many years Salt Lake City had been at the center of global efforts to map and sequence the human genome. The city’s location and its connection to the Mormon Church with its deep interest in tracing family connections and unequaled historical records made it a natural place to explore the potential for genealogy of the new genetic knowledge. I had gotten to know Dr. Scott Woodward from Brigham Young University when we had both been working in the field of ancient DNA, so it was no surprise to find him leading the initiative as head of the Sorenson Genome Institute. I had planned to visit him, but a bush fire in Yellowstone National Park intervened, and I had to ask my questions during a phone call from San Francisco. Since we had known each other a long time, this was almost as good as a face-to-face meeting, and I asked him to tell me how he had become involved.
Woodward told me of being woken in the early hours one morning by a phone call. It came from Norway, which explained the awkward timing, and the caller was James LeVoy Sorenson. Sorenson, who died in 2008 at the age of 86, was the richest man in Utah, and although Scott had never met him, he certainly knew who he was. Sorenson had made a fortune first in real estate and then as an inventor of medical devices, notably the first modern intravenous catheter and, more prosaically, the disposable surgical mask. He was also a Mormon, having served his obligatory mission for the Church of Jesus Christ of Latter-day Saints in New England in his early twenties. While in Norway, where his ancestors had their roots, and evidently aware of the growing application of genetics to the questions of human origins, he had come up with an idea.
“Scott,” he asked, “how much would it cost to do Norway?” It became clear to the by-now-wide-awake Woodward that “doing Norway” meant DNA testing the entire population of four million people. “Tens of millions of dollars, at least,” was his off-the-cuff response. There was no immediate answer from Sorenson, and Woodward imagined that the figure was way too high, much higher than Sorenson had in mind. The call ended, and Scott went back to sleep assuming he had heard the last if it. He was mistaken. When Sorenson returned to Salt Lake City he arranged a meeting with Woodward in order to come up with a plan for “doing Norway.” (When you are the richest man in Utah, with a fortune estimated at $4.5 billion, tens of millions must seem like small change.) To his credit, Woodward managed to persuade Sorenson that perhaps “doing Norway” was not the best way to go and managed to morph the ambitious yet geographically limited project into something with much greater promise. Why not collect DNA from volunteers all over the world along with their family histories? Sorenson quickly agreed, and Woodward left Brigham Young to lead the project at the eponymous research institute. The aim was to collect and store a hundred thousand DNA samples, an ambition realized in 2007. There was also a commercial arm, Relative Genetics, which like Oxford Ancestors and Family Tree DNA offered a service direct to the public.
Last, there has been a numerically impressive initiative with corporate backing. The Genographic Project, a joint venture of National Geographic and IBM, has tested DNA from 350,000 people from all over the world since it began in 2005. The project leader, Dr. Spencer Wells, who rejoices in the enviable title of a National Geographic “Explorer-in-Residence,” is himself a geneticist, and he and I once worked in the same research institute in Oxford, though not in the same laboratory. His expertise was in the genetics of the cellular immune system, a horrendously complicated natural defense arsenal that keeps our bodies from being overrun by pathogens. Unfortunately it can take its job a bit too seriously and turn its impressive destructive powers against our own bodies, leading to rheumatoid arthritis and other autoimmune conditions. It is also the system that causes tissue transplant rejection. However, the bewildering genetic variation that underpins the autoimmune system has been seized upon by geneticists as a more sophisticated equivalent of blood groups, and put to use in differentiating populations for the purposes of finding out where people came from. I have never particularly liked this approach to unraveling the past, for reasons I have written about in The Seven Daughters of Eve. It meant that Spencer and I never worked closely together in Oxford. While I cut my teeth in the coral-fringed islands of Polynesia, Spencer’s chosen region was the harsh and arid steppe land of Central Asia, in particular Uzbekistan and Kyrgyzstan. By the time he became the director of the Genographic Project, Spencer had wisely abandoned the genetics of the autoimmune system and switched to the stalwarts that are mitochondrial DNA and the Y chromosome.
Soon after finding the genetic link between the Y chromosome and a handful of English surnames, including my own, and while the whole field was firmly within the “research” phase—meaning that we weren’t going to get any objections from our main funders, the Wellcome Trust—an opportunity came my way to test, in depth, the usefulness of the surname/Y-chromosome association. With the Sykes study and the other names used to check that Sykes was not for some reason unique, there was no parallel genealogical research. I had basically picked the volunteers randomly. So long as they shared a surname, that was all that mattered. While I was on the lookout for a surname with a well-researched genealogy behind it, I was contacted by Chris Pomery. An interesting man in many ways, he had recently returned from a spell as a correspondent in the Berlin offices of the London Times as well as other assignments in Prague. He has since written two successful books on DNA and genealogy that have been very useful practical introductions for genealogists all over the world. When I first met Chris he exuded—whether deliberately or not—something of the air of an international man of mystery, which added spice to his visits to my lab in Oxford.
In between his (as we all liked to imagine) murky dealings in Eastern Europe, Chris had done a huge amount of genealogical work on his own surname. He had tracked down 825 living holders of the Pomeroy name and its variants and linked them through the records to one of fifty-one named ancestors. Some of these ancestors lived a long time ago, the oldest in the 1600s, while others were much more recent. There were also different spellings to be considered, like “Pomery,” “Pomroy,” and “Pummery,” that might or might not be genetically related. As so often in genealogical research, there were no reliable records with which to connect the different branches, and Chris approached me to see if genetics might provide the means of doing so. When he outlined the situation to us in Oxford, his project seemed to have all the ingredients we were looking for as a practical test of the Y chromosome, and we immediately decided to go ahead.
One of the advantages of working in Oxford is that undergraduate students in the biological sciences have to do a piece of original research during their final year and write a dissertation. Each year I took on at least one of these bright youngsters, and I could see that the Pomeroy project would be ideal. It would be bound to generate results for the dissertation, it was contained rather than open-ended, and above all it was original and interesting. The student in 2000 was David Campbell, and the three of us met in the coffee room at the Institute of Molecular Medicine, the location for many a planning session, to map out the details. Chris explained that the surname Pomeroy originated from the Norman adventurer Ralph de la Pomerai, who was granted a number of English manors by William the Conqueror as a reward for his loyalty and support during the Norman invasion of 1066. The manors were mainly in the southwestern county of Devon, with the family seat at Berry Pomeroy castle. The castle, now a ruin, is tucked away in a steep wooded valley a few miles outside the picturesque and historic town of Totnes. It was sold in 1547 to Edward Seymour, the first Duke of Somerset and brother of Jane Seymour, the third wife of King Henry VIII. Berry Pomeroy, now uninhabited except by the ghost of the White Lady (ghosts are considered a de rigeur accessory in all the best castles), still belongs to the Somerset family, now headed by John Seymour, the nineteenth duke.
We picked one member of each of the fifty-one groups that Chris had identified as being related through the records, and David began the (at that time) laborious process of generating their Y-chromosome signatures. By then we had expanded the number of markers we used from four to seven, still very limited by today’s standards. Even so, when the results were all in, they were very revealing indeed. Despite the fact that the name originated in a single individual, Ralph de la Pomerai, the genetics revealed that the fifty-one volunteers belonged to at least eight genetically unrelated branches. Chris expressed little surprise at this, but I was certainly taken aback. In the other names that I was researching at around the same time, the general rule had been that when a surname was rare and its geographical distribution localized, there was usually just one major branch descending from a single ancestor. Of course that would not be the case in a common occupational surname like Carpenter or Fletcher, or names that were clearly derived from a feature of the landscape such as Bush, Hill, or Greenwood. But for Pomeroy, confined as it was to Devon and Cornwall and having a very definite origin, I thought something else must be going on. But from a very practical point of view, the genetics had shown Chris and his fellow Pomeroy researchers which of the links between the fifty-one groups could be followed up in the records with confidence and which would be a waste of time. Since then Chris and his colleagues have expanded the Pomeroy project to become “probably the most advanced surname project in the world,” according to the project Web site. They have thoroughly investigated the genetic links between alternative spellings of the name: “Pomery,” “Pummery,” “Pomroy”; in Australia, “de Pomeroy” and “Pommeroy”; and, in the United States, “Pumroy.” Interestingly, men with some spellings, like “Pomery,” usually share the same Y chromosome, while with others the genetics bears little or no relationship to the spelling, a reminder to genealogists everywhere that names can mutate much faster than Y chromosomes.
In some cases we know the precise origin of a surname, and can be certain that it was unique. Take Dyson as an example, which careful research by Dr. George Redmonds has shown to be an unusual case of a matronymic name meaning “the son of Di.” The Di in this case was an unmarried cattle rustler called Dionissia of Linthwaite who named her son, John, born in 1316, after herself rather than his father. With this unusual single origin, it is far less surprising that the name Dyson has a very tight association with a particular Y-chromosome fingerprint, which also has the advantage of itself being quite rare in the general population. In the case of Pomeroy, where there is a named ancestor, Ralph de la Pomerai, what is the cause of this unexpected surplus of genetically different ancestors all with the same rare surname? Here we see for the first time that the rules that apply to the peasants, and I have to include the Sykes line here, do not apply to the nobility.
In feudal estates, like that granted to Ralph de la Pomerai and maintained by his descendants, there are two forces at work that can disengage the surname of the lord of the manor from his Y chromosome. There is no doubt that the privilege of the position was accompanied by increased mating possibilities. One of these was enshrined in the custom of droit du seigneur, when brides-to-be had to endure a night with the lord before beginning normal married life. When a son was born as a result, he would have the Y chromosome of the lord of the manor, but the surname of the woman’s husband. Sometimes the son would be given the lord’s surname in exchange for material support. That would not necessarily disengage the surname from the Y chromosome, unless the son was actually fathered by another man, the woman being happy to keep quiet and take the money.
The Pomeroy case also showed how genetics can reveal, or dissolve, links to other names with similar spellings. Alternative spellings of the same name abound, and pose a particular puzzle for genealogists. Sometimes surnames are deliberately changed, and there are abundant examples of European immigrants to the United States and to Britain who have anglicized their surnames. Gutmann to Goodman, Beckmann to Beckham, and so on. It is very common. Later the name may be reverted to the original. But mistakes, deliberate or otherwise, by officials who are either processing immigration papers or recording births, are a common historical cause of alternative surname spellings. In medieval Britain, when most of the population was illiterate, it was the job of the local parson to register the births in his parish. Inquiring what the surname of the baby should be, it was an easy mistake to mishear the parents and write down a slightly different spelling, which, being unable to read, the parents were not in a position to correct. Even now, whenever I am dictating my name for a form or booking a restaurant over the phone, more than half the time I become Mr. Skyes. The only place this never happened was when I was researching the Sykes heartland in West Yorkshire. There they got it right every time.
Alternative spellings are very common in the United States, either as deliberate adoptions to disguise a foreign origin or by the carelessness of clerks at Ellis Island, New York, or other entry points to the United States. Genetics has been helpful in linking American citizens with altered surnames back to their European origins. There are by now hundreds of examples, but one of the first in which I was involved, through Oxford Ancestors, was when we were approached by two American families, the Lehmans and the Bachmanns, with very similar requests. They were both looking to establish links back to presumed ancestors in Germany and Switzerland. Within each of the families, their own research in the records had uncovered an array of alternative spellings in the United States, so there were good reasons to turn to genetics to check whether the genealogical connections that had been made within each family were real or not.
To cut two very long stories short, the Y chromosomes in both families gave a very clear idea of the different branches and the range of spellings within them. Among the Lehman family, there were three clear branches defined by Y chromosome signatures, but the distribution of alternative spellings was more or less random between them. One branch contained mainly Laymans and Laymons, but another Layman clearly belonged to another branch with Lemons, Lemmons, and Lehmans while a third had a bit of everything; Layman, Laymon, Lemon, Lemmon, and even La Mance, all of them genetically related to one another. Among the Bachmann family, on the other hand, there were only two alternative spellings, Beckman and Baughman, but that was no help in defining the family structure as both were found within each of the four main branches defined by their Y-chromosome signatures. Once again genetics had shown the dangers of assuming that individuals with the alternative spelling of a surname were necessarily related. With the real branch structure now revealed in both families, they were able to link members of each branch to the correct relatives back in Europe.
In England it is surprising how many people claim to have ancestors who, like the Pomeroys, “came over” with William the Conqueror in 1066. This is a generally harmless boast, but I doubt if it is true in most cases. I hope it has not come as too much of a disappointment to the Pomeroys who have been proved to be genetically separated from Baron Ralph. I suspect, however, that the news will not have diminished their aspirations to a distinguished Norman ancestry.
The English are not alone in craving a noble ancestry, and genetics has reignited this desire by opening up the possibility of proving it. Again, an early case in which I became involved concerned the Cloughs of New England. The Clough Society of North America is one of hundreds of one-name groups that, very early on, saw the benefit of genetics in testing the links between their members. The society had the advantage of having an experienced genealogist, Sheila Andersen, who got in touch with me to test the link she had found, by working through the records, that the U.S. Cloughs were related to the Welsh nobility of the same name. Accordingly my lab tested Clough Society members and Sir William Clough, the head of the Welsh noble branch of the family. Like many American genealogy groups, the Cloughs enjoy coming to Britain to search out the locations of their ancestors, and Sheila is very good at organizing these tours. This one included the fairy-tale village of Portmerion, built by another ancestor, the late Clough Williams Ellis. The location for the denouement, where I was to reveal the results of the Y-chromosome tests, was to be in St. John’s College in Oxford. Not my own college, but considered appropriately medieval for visiting Americans. This is not the only time that visiting groups, especially television documentary makers, have transplanted my perfectly good study to more “authentic” surroundings. No substitution was more impressive than when I was engaged to help unravel the Welsh ancestry of the actress Susan Sarandon. The producer decided that the appropriate location for our scene together should be none other than the stately home of the Duke of Marlborough at Blenheim Palace, a few miles north of Oxford and known the world over as the birthplace of Winston Churchill.
Even the fabulously wealthy St. John’s College could not match the opulence of my “study” at Blenheim. Even so, the richly decorated college room with a fan-vaulted ceiling was a suitably historic venue, and after my introduction to the mechanics of the genetic tests, I came to the results. Most of the group of seven were women. They were Cloughs, all right; but, not having Y chromosomes themselves, had obtained the vital DNA samples from their male Clough relatives. In two cases this was from their husbands, so the women concerned did not have any strictly genetic Clough ancestry, but that did not diminish in the slightest their thirst for a touch of nobility in the family. I could sense the anticipation in the room when I flashed up the slide of the results, which was in the form of a color-coded table of the Y-chromosome fingerprints. Muffled yet audible sounds of delight filled the room as it became clear that the Clough Society Y-chromosome fingerprints matched the sample from Sir William. All except one, which was entirely unrelated. This chromosome, unfortunately, came from the holder of a high office in the society.
Sometimes the records are ambiguous and point in two different ancestral directions. This was the case with the Lockwood family in America. They had been researching their English origins for several years, but the majority had been unable to discover where their ancestor had lived. Only one had been able to follow a paper trail back to an Edmund Lockwood, born in the Suffolk village of Combs in 1574. The others did not know whether they too could claim Edmund as their ancestor or whether they were from somewhere else altogether. To try to untangle this, the family had spent many years combing the records in the two English locations where Lockwoods were concentrated. One was in Yorkshire, in northern England, while the other was in Suffolk, in East Anglia. But so far the effort had been in vain. Uncertainty always saps enthusiasm, so the Lockwoods had a very good reason to want to know which of these two locations was home to their ancestor.
This was a case ripe for genetics, and before long we had enrolled half a dozen male Lockwoods from either the old weaving town of Halifax, which our surname map soon showed to be the present-day epicenter of the Yorkshire branch, or from around Ipswich, the county town of Suffolk. Their Y chromosomes immediately revealed that the Yorkshire and Suffolk Lockwoods were unrelated to each other, but that within each region their chromosomes matched. When six American Lockwood chromosomes were tested, they were all identical, and not only did they match one another, they also matched the Suffolk Lockwoods but not their Yorkshire namesakes. At a stroke the ambiguity had been eliminated, and the American Lockwoods could concentrate their considerable energies on Suffolk, freed from the gnawing anxiety that they were wasting their time. Which, in Yorkshire, it turned out they had been.
The link I had discovered, almost by accident, between surnames and Y chromosomes has certainly found a use in testing the genealogical links between men with the same surname and illuminating the process of alternative spellings, as we have seen. This is all very interesting in what it says about surnames. But it also says a lot about men, and seemed to me to have the potential to explore the contrast in mating habits between the aristocracy and the peasantry, which my work on the genetic history of Britain and Ireland for Saxons, Vikings, and Celts had indicated was extreme. In the Pomeroy case there was circumstantial genetic evidence that the surname of the original Norman baron had been adopted by other men in the vicinity, and I have suggested how the behavior of the feudal lord and his descendants may have played a part in this. However, the Pomeroy case did not prove that the baron’s Y chromosome had been dispersed within the local population, as I suspect it very well might have been.
A few years later I had the opportunity to carry out a direct test of this phenomenon, not around Berry Pomeroy, but on another large estate, this time in Wiltshire, in southern England. The estate was Longleat, and I had gotten to know the owner, Lord Bath, when he had asked me to see if he was related to “Cheddar Man,” a nine-thousand-year-old fossil excavated from Cheddar Caves, on his estate. I had just published my DNA results from Cheddar Man, which showed that this ancient relic was related, through his mitochondrial DNA, to a history teacher in the local school. The story encouraged much mirth, especially in the United States, as it reinforced the bucolic, stay-at-home image of the English in that the descendants of Cheddar Man had taken nine thousand years years to travel three hundred yards down the road. It turned out that Lord Bath was not related to Cheddar Man, but his butler, Cuthbert, was. If I expand on this amusing anecdote much longer I will be guilty of repeating myself, as I described the episode in The Seven Daughters of Eve.
However, its relevance here is that when I was looking around for a test for what I came to think of as “aristocratic diffusion,” Longleat fitted the bill very well. The estate had been continuously in the hands of the Thynne family since the sixteenth century. More to the point, the majority of the estate workers, perhaps the most likely vectors of aristocratic diffusion, had come from the estate village of Horningsham a mile or so away. Lord Bath himself had become very enthusiastic about genetics and welcomed my inquiry. When I went to visit him in his penthouse at Longleat I was very pleased to see that he had mounted his mitochondrial family tree from the Cheddar Man case on the wall near his enormous chestnut desk. Lord Bath is famous as a confirmed polygamist, having at least fifty “wifelets” (his own description), who are rewarded with a cottage on the estate after sufficient years of service. Other than sexual pleasure, one objective of polygamy is reproduction, but, given that his offspring are still in single figures, I was not sure whether this has been as great a success as Lord Bath would have liked. However, the prospect of discovering that his ancestors had sired the population of Horningsham clearly had a certain appeal, and Lord Bath gave his permission for the project to go ahead. In fact he did a lot more than that, as we shall see.
As I was going home I stopped off at Horningsham parish church to have a look round the churchyard. It was November and getting dark, but I was still able, with the help of my flashlight, to make out the names on some of the crumbling headstones. I was looking for surnames, and sure enough, there were only a few—Trollope, Long, Carpenter—with several examples of each. This is what I had expected and hoped for. It is a sign of a stable and static population, where surnames are winnowed out over the generations as families have no sons to carry them on. Genealogists know this phenomenon very well as a surname becomes less and less common and then disappears. Rather cruelly in my opinion, the surname is said to have “daughtered out.” In the United States, Dearborn is a numerous and well-known name. Almost all of the 5,000–10,000 U.S. Dearborns are descendants of one man, Godfrey Dearebarne, who arrived in New England in 1639. But while Dearborns have thrived in the United States, the name Dearebarne has disappeared from England, where it originated, the hapless victim of “daughtering out.”
In the churchyard at Horningsham, daughtering out had reduced the variety of surnames to just a handful. This is very common in rural England, which I remember from my time as a postman in university Christmas vacations at my parents’ home in a village called Dedham on the Essex-Suffolk border. Being deep in the country, there were no house numbers. And there seemed to be only two surnames: Ablitt and Matthews. It was very frustrating trying to find the correct destinations for my consignment of Christmas cards, another unintended consequence of “daughtering out.” However, in Horningsham it worked to my advantage, and I set about contacting all the current residents with the few surviving surnames from the churchyard. Three months later, helped by the indomitable matriarch of the village, Vera Trollope, I had enough volunteers for the project to proceed.
Longleat is one of the most beautiful stately homes in the whole of Britain. Built during the reign of Elizabeth I in 1580, it has a symmetrical elegance not always found in houses of that period. Its position in a shallow bowl of land embraced by low rolling hills means that the first sight you have of the house is a distant one, looking down from the wooded rim set in parkland laid out long ago by the prince of landscape gardeners, Capability Brown. Longleat was built by Sir John Thynne, a man who was not born into great wealth but who worked his way up by administering the affairs of others, finally becoming the right-hand man, or steward, to Edward Seymour, Duke of Somerset, whom we have already met as the owner of Berry Pomeroy castle. At the same time Thynne began to build up his own property holdings, culminating in the purchase of Longleat in 1540. In 1549 he married into the wealthy Gresham family and began to plan the present house on the proceeds of his wife’s not inconsiderable dowry. Later the same year Edward Seymour lost not only his power base but also his head, but Thynne avoided execution and was imprisoned in the Tower of London. He was eventually released and, warned off politics, settled into country life at Longleat, where he and his wife, Christian, had nine children, with a further five, all sons, coming from a second marriage following Christian’s death. The Thynne dynasty continued to thrive in succeeding generations as their titles reflected their gradual ascent through the ranks of the peerage, first as baronets (1641), then viscounts (1682), and finally marquesses in 1789. Alexander Thynn, the current Lord Bath, who dropped the final e from his surname during the 1980s, is the seventh marquess, and a direct patrilineal descendant of Sir John. Thus he has inherited Sir John’s Y chromosome. The question was, could I also find the Thynne Y chromosome among the good people of Horningsham?
The day of the DNA collection had arrived, a brilliantly sunny Saturday in late February, and about fifty villagers had assembled in Longleat’s sumptuous Red Library. I arrived with two assistants: my son Richard who, at fifteen, was already a veteran of several sampling expeditions all over the world; and Charlotte, a Danish graduate student from Oxford who had answered my advertisement for research assistants. Before the sampling session began we had been invited to a drawing room to meet Lord Bath and his weekend guests. I remember feeling slightly nervous going into the room and being met by the eyes of a dozen or so people arranged on comfortable sofas and easy chairs. Lord Bath was there, of course, wearing a trademark embroidered skullcap, and he introduced us as the DNA collectors. Curled on the floor at his feet lay a cream Labrador. By his lordship’s side a strikingly elegant woman, her long blond hair interlaced with colored beads, reclined on a green velvet chaise longue. In an attempt to defuse my apprehension with bravado, I walked straight up to her and said, “We’ll start with you.” That is how I first met Ulla, my future wife.
The magnificent Red Library, with its leather-bound volumes filling shelf upon shelf of gilded oak, had been laid out with chairs that were filled with eager Horningsham residents. After a short speech by Lord Bath introducing the purpose of the day, Richard, Charlotte, and I set about collecting the DNA by means of cotton swabs rubbed on the inside of the cheek. I used to do this myself until the occasion, in the Shetland Isles, when an elderly lady’s false teeth came loose and clamped the swab fast. After that I asked people to do it for themselves. Before long everyone had given a sample and their consent. Even the dog, whose name was Boudicca, joined in, though I was obliged to sign the consent form on her behalf. This had been a very special day.
Returning to the scientific purpose of the visit, I was looking to match Lord Bath’s Y chromosome with men from Horningsham. If I found a match it would not necessarily mean that Lord Bath himself was the father of the man, though that was a formal possibility, but that the Thynne Y chromosome had escaped from Longleat House through the energies of one of his ancestors. The prospect appealed to Alexander, which is why he not only gave permission for the experiment but was wonderfully generous in making the arrangements, not only for the original collection but for the day when I returned to Longleat later that year to announce the results.
This meeting took place in the vaulted undercroft, where chairs and tables had been arranged around the stout pillars that supported the house. I left the announcement until lunch was over. First I went through the results for each family that had taken part. Within each surname the Y-chromosome signatures had been the same. I wasn’t surprised by this, so was taken aback when one of the Trollopes asked, “Are you saying that we are related to the Trollopes from past the church?” “Well, yes,” I replied. “No, that can’t be right, we are from completely different families,” came the response. I had clearly scratched a deep sore in the village.
Next I presented the results of the DNA tests from the other families, and, like the Trollopes, they were identical within each one. But did any share the distinctive Thynne chromosome? The time had come. It felt almost like an Oscar moment when the Academy Awards are announced. “And the winner is . . . George Long.” George Long, a gentleman about fifty years old, rose from his chair and came up to be photographed with his now-relative, Alexander, Lord Bath. Both men were descended from the same man, one of Lord Bath’s ancestors, although no one knows which one. Speaking entirely formally it could have been the other way around, but to me this was proof of aristocratic diffusion, from Longleat to Horningsham. But it was not overwhelming. The other families in the village did not have the Thynne Y chromosome, so had not sprung from the loins of Longleat. Attempting to compensate for any disappointment on both sides, I reminded the audience that, powerful though it no doubt is, genetics cannot detect unfruitful ancestral seductions.