The Invisible History of the Human Race (25 page)

Surnames have the longest history in China, where they began around five thousand years ago. They are about seven hundred years old in England, where they became hereditary in the fifteenth century (and likely earlier). They became hereditary in Scotland a century after that. Surnames are older in Ireland, beginning around nine hundred years ago. In the Netherlands they’ve only been in use for the last two hundred years. In Turkey they were adopted in the early twentieth century. Iceland is the only European country to retain the older-style patronymic that many other countries abandoned, in which children are surnamed with the father’s forename appended to the word for “son” or “daughter,” so the surname changes every generation. For example, Axel Stefansson is Stefan’s son, but Axel’s son Baldur will be known as Baldur Axelsson. In English, names like Johnson, and even just
Johns, originated this way.

Many surnames were based on the occupation of their owners, such as Smith, Wright, or Sawyer. It was common for others to be based on nicknames or diminutives, like Redhead. Still others were inspired by local topography or a place name, like York, London, Lake, or Townsend. Some
types
of surnames are especially common to a particular region. In East Anglia, it is said, a number of surnames originated with the local pageant in medieval times. If an actor was well known for his role in the local annual play, such as Herod or the Egyptian pharaoh, he could end up with a surname like King or Farrar.

Even though much water has passed under the cultural bridge since surnames began, their use is still strongly shaped by their origins. For many surnames there’s a clear “geographic hearth”; that is to say, where the name originated is still the place where most people carry it. In Great Britain the surnames of Scotland and Wales demarcate distinct regions, and the border created between Scottish and English names is more absolute. Along the geographic Welsh border, the old Welsh surnames are starting to encroach into the English region.

When it became clear that the Y chromosome was passed down the male line, researchers became quite excited about the possibility that a relationship could be established between surnames and the Y. But many were skeptical that names could be treated like genes. After all, names are words and as such are surely affected by the tides of change that sweep through languages and dialects.

But are surnames words in the literal sense? They were originally drawn from language, and the words that gave rise to them (smith, lake, hill, etc.) behave like other words. But once they became names they were, in a sense, arrested, and with their existence as simple nouns over, they became subject to some unusual rules. Now, unlike any other words, surnames can be known by anyone, but they also, in an odd way,
belong
to a select group of people.

Indeed, while dialects and names have a strong connection to geography, they do not have the same relationship. The association has been likened to children’s shoe size and reading ability, which appear to be correlated but only because they both correlate strongly with a child’s age. One study from the Netherlands revealed that surnames are unaffected by changes in dialects and seem to change according to their own rules,
not the rules of language. What about the biological landscape?
Do
surnames track the Y?

 • • • 

In 902 a large Viking band led by a man named Ingimund settled in Wirral in northwest England after being expelled from Dublin. We know that the Viking presence in Wirral was a significant one, because a great deal of Viking jewelry, weapons, and tombstones have been unearthed locally. Many place names in that region are also of Norwegian origin, such as Tranmere (meaning “crane sandbank”) and Meols (meaning “sandbank”). Wirral even has a village called Thingwall, the word for a field where Viking men met and conducted their parliament. Neighboring West Lancashire has the same legacy of Norse artifacts and place names, but the story of how Vikings arrived there has been lost in time. In both areas the invaders clearly affected the culture, but did they reproduce with the locals or did they keep primarily to themselves?

As many of the incoming band were men, the most straightforward place to look for an answer was on the Y chromosome. But even though the Ingimund-led immigration was a large one, there have been other significant immigrations into both areas since, including waves of people from Ireland and other countries. How do you find a specific
Y in this population? If the researchers simply took fifty men off the streets of Wirral and West Lancashire and analyzed their Y chromosomes, it would likely tell them nothing—the men or their ancestors might have arrived in the region the previous year or more than a thousand years ago. If there were a signal from Ingimund’s day, it would have been drowned by the noise of all the different immigrations since.

Instead of using a random sample of men, the researchers decided to seek out those whose families had lived in the area for a long time. In addition, they sought out men whose surnames existed in the medieval era. Ideally the researchers could have combed through the Viking sagas, compiled a list of the invaders’ names, and then checked if anyone in Wirral or West Lancashire still carried them. At the time that the sagas were written, however, names were not passed down in families, and recall that English surnames only came into being seven hundred years ago. The best bet for plumbing the social makeup of Viking-era Wirral and West Lancashire was to go back as far as possible and hope that any family groups that had originated a few hundred years earlier might still be detectable when surnames were adopted.

The researchers found lists of surnames in documents dating from the 1300s and the 1500s that many Wirral and West Lancashire locals still carried, like Barker, Beck, Bushell, and Sherlock, as well as less familiar ones like Bilsborrow, Lunt, Tottey, and Crumblehome. For a control group that also had some local heritage, the researchers included men without medieval names but whose paternal grandfathers were born in the region.

The study found that there were indeed different types of Y chromosomes in both groups, and in the medieval name samples half of the Y chromosomes could be traced to Norse ancestry. Not only did these results reveal that there was mixing between the Vikings and the locals, but they also underscored just how greatly the invaders from more than a thousand years ago influenced the makeup of the local population today. There was a small Norse influence on the Y of the modern control group, and indeed, there’s a good chance many of them also had antecedents in the Viking invaders. But by focusing on the men with medieval names, the researchers were able to locate more precisely the descendants of Ingimund and his band and of their contemporaries in West Lancashire.

Another comparison was made with a group of men from nearby Mid Cheshire, which is physically close but does not have the same history of Viking artifacts or Norse-influenced place names. It was found that the Mid Cheshire Y chromosomes were also far less likely to be Scandinavian than those from Wirral or West Lancashire.

In a similar experiment researchers from Trinity College Dublin investigated the Y on the Emerald Isle and discovered that it had little diversity. Why do Irish men look so similar on the Y? One reason, they proposed, was that life in Ireland may have been a bit Genghis Khan–like before St. Patrick arrived. No doubt the Catholic Church has deeply shaped the contemporary culture and population of Ireland, yet a significant part of the nation’s genome still carries the stamp of a decidedly
non-Catholic fifth-century warlord.

Niall of the Nine Hostages was so named for his habit of kidnapping the kin of other chieftains to ensure their cooperation. He famously fathered many children to consolidate his power, as was the tradition of the day. In 2005 researchers at Trinity College Dublin investigated a Y that was shared by 17 percent of Irish men in the country’s northwest. Because the geography of the modern Y overlapped with the region once ruled by Niall, the researchers believed the Y might be his legacy. They narrowed the subject population further by testing fifty-nine men whose surnames originated in Niall’s clan (including Donnelly, O’Donnell, O’Gallagher, O’Doherty, Flynn, Egan, and O’Rourke) and found that the putative Niall Y was particularly frequent in this group. The Y also dates back to more than 1,700 years ago, which puts it in the Niall era.

As with the Genghis Khan Y, the frequency of the Niall Y was amplified by the social mores of the period. Divorce and polygyny were socially acceptable for hundreds of years in Ireland, even after it became Christian, which enabled Niall’s male descendants to spread his Y far and wide. In one famous example, Lord Turlough O’Donnell, a descendant of Niall who lived at the turn of the fifteenth century, had eighteen sons with ten different women. Those sons, in turn, went on to have fifty-nine sons of their own. Because the Irish diaspora of the nineteenth century was so large, and because the Niall Y was already so frequent in Ireland, it’s thought that two to three million men worldwide can trace their Y to Niall.

 • • • 

Researchers began to use surnames as an index of relatedness in the late nineteenth century. In the years since they have been utilized, with some success, to estimate how much inbreeding has occurred within a population. In the 1970s a doctor in Montreal established a connection between Y chromosomes and surnames when he tracked a novel mutation on the Y of a modern French Canadian man and his extended family back through generations to a pioneer of the same name. It took until the beginning of the twenty-first century to show that clusters of repeated segments could be tracked through time, along with surnames. In one of the earliest Y/surname experiments that tracked short repeats, researchers in 2000 found that a particular kind of Y was common among Irish men with Irish names but much less common in Irish males whose names were obviously English or otherwise non-Irish.

Since the basic link between the male chromosome and the male-line surname was established, it’s become clear that because history shapes names, and because history and biology affect the occurrence of Y chromosomes, there is no single straightforward relationship between them but rather many subtle and variable patterns. For example, if you want to investigate the old surname/Y connections, uninterrupted by migration in the modern era, you are better off looking in rural communities where people do not often move into the community. But that doesn’t mean you can’t learn about the character of a city from its names too. The centuries-old, worldwide immigration into London is underscored by the fact that it overwhelmingly has the most unique surnames of any place in Britain.

Most curious is the way that Y/surname patterns differ between countries. In Britain, on average, a man who has the same surname as another is significantly more likely to have a similar Y chromosome, and therefore a common ancestor, than he would with someone of a different surname. But there’s a twist: The Y similarity depends on the frequency of the surname within the population. If you are a Smith, for example,
the rule does not apply.

The name Smith was adopted many times over in England—all, of course, based on the occupation of the bearer. A man called Smith is no more likely to share a Y chromosome with another Smith than he is with anyone else not called Smith. While Smith is, unsurprisingly, the most common name in England, any English surname that is held by at least ten thousand people is effectively a Smith-type name. (This includes the Kings, the Brays, and the Steads, for example.) No doubt, if surnames were just coming into general use now, Smith would be one of the rarer names, and we would perhaps be encountering more John Analysts, Jack Realtors, and Susan Hackers.

Setting the Smiths aside (all 600,000 of them), there are many other names that are less common, and in fact, a large pool of them are distinctly unique. There may be all sorts of biological and cultural reasons why they are rare, but one of them is simply that the source name was unique to begin with. Another possibility is that the male line was simply not successful. Instead of having ten sons, like his medieval neighbor John Johnson, Bill Billson might have had only one. If Billson junior grew up to father ten daughters but no sons, that would mark the end of that Y and the good Billson name.

For the English names that have survived with very few holders, there is a strong link between the name and the bearer’s Y chromosome. In fact, the less common the surname is, the more likely it is that any two men with that name have a similar Y, as is the case within name groups like Werrett, Titchmarsh, and Attenborough. The Attenborough men are the most homogeneous of them all; 87 percent of them have descended from a single man and share the same Y chromosome, which is found in only 1 percent of the rest of the population.

So strong is the link between the surname and the Y in such cases that it’s been suggested that, for many of England’s unsolved cases of murder and rape in which a Y DNA sample has been taken from the scene but no suspect has been identified, the Y profile could be matched against a database of Y + surname matches, generating a list of possible names for the bearer of that Y and significantly narrowing the pool of suspects.

In Ireland, by contrast, a man is thirty times more likely to share the same short repeats on his Y with another man of the same name than with a man of a different name, regardless of how popular that surname is. The surname Ryan, for instance, is held by approximately 1 percent of the Irish population (Smith is held by 1.3 percent in England), yet there’s a forty-seven-times-greater chance that any two men who are called Ryan will have the same Y than any two men who are chosen at random.

Even though there are sixty thousand Irish Ryans, researchers believe that many of them have a Y in common because when surnames began, there were only a few, and perhaps even only one, Ryan. This could mean that most of the many Ryans of modern-day Ireland descend from this founder. The O’Sullivans, O’Neills, Byrnes, and Kennedys all appear to have had a dominant founding Y as well.