On the origin of speeches

Genetics is about natural selection, say the English. No. it's about language, the pioneering Italian, Luigi Luca Cavalli-Sforza tells Steve Connor

Something like 99 per cent of human history occurred before the invention of writing and the start of our cultural heritage. For tens of thousands of years, our ancestors lived, gave birth and died during the long period when history went unrecorded. It is thanks to Luigi Luca Cavalli-Sforza, and the other pioneers of human genetics, that scientists now have a powerful tool with which to open a revealing window on our earliest existence.

Something like 99 per cent of human history occurred before the invention of writing and the start of our cultural heritage. For tens of thousands of years, our ancestors lived, gave birth and died during the long period when history went unrecorded. It is thanks to Luigi Luca Cavalli-Sforza, and the other pioneers of human genetics, that scientists now have a powerful tool with which to open a revealing window on our earliest existence.

Cavalli-Sforza, Professor Emeritus at Stanford University in California, was among the first to demonstrate that the early chapters in the story of mankind can be told by reading the messages written in the indelible language of our genes.

He is one of the founding fathers of population genetics, the mathematical analysis of how and why the frequencies of genes change over time. The science has been an important influence on what has become known as "biological anthropology", which attempts to understand human prehistory by analysing the DNA of present-day people as well as the preserved tissue of those who are long dead.

One of the earliest attempts to understand human evolution using population genetics was carried out by Cavalli-Sforza in Italy in the the early 1950s, after he had studied in Cambridge under the great RA Fisher, the British geneticist who established a statistical link between mutation and evolution. Cavalli-Sforza says he owes this early insight to help from the Roman Catholic Church, which owned a unique, 300-year archive of births, marriages and deaths in Italian villages, as well as records on the family relationships of those intending to get married.

"I had access to all these records and I decided to use them for a problem that had not been solved at the time," Cavalli-Sforza explained during a visit this week to London to mark the publication of his latest book ( Genes, Peoples and Languages, Penguin, £18.99). Fifty years ago, he wanted to know whether natural selection was really the sole factor driving human evolution or whether another, random element, called "genetic drift", was also important.

At the time, there was something of a schism among geneticists studying human evolution, with those on this side of the Atlantic, under Fisher's influence, suggesting that only natural selection was involved. Many in America, led by Sewall Wright, the other leading geneticist of the day, thought that the frequency of human genes could drift up and down within a population purely on chance alone.

Cavalli-Sforza had an opportunity to test both ideas using his privileged access to the Catholic Church's records of nearly 100 Italian villages spanning three centuries.

He first set about collecting samples of blood from each village to record the frequency of the three main blood groups - groups A, B and O. Since the early part of the 20th century, geneticists had known that different populations from around the world had different proportions of the three blood groups for no apparent or obvious reason. Cavalli-Sforza thought this difference might, at least in part, be the result of random drift, something that Fisher, in his brilliant mathematical analysis, had dismissed.

"The question was, 'is drift really important to Man?' In England there was a tendency to believe that natural selection was the only thing that mattered," says Cavalli-Sforza. In trying to explain why human genes are what they are today, random genetic drift was always a theoretical possibility, but Fisher and others had calculated that human populations, even many thousands of years ago, were just too large for such a random event to have any importance. "I studied the blood groups of people so I could measure the genetic variation between villages. I could also make a prediction of what it should be from the records of the parish books, which allowed me to construct the demography of the people. I found my predictions were satisfied, so I did find the kind of drift that I expected," Cavalli-Sforza says.

It is not that the research downgraded the importance of natural selection, just that it demonstrated there were other elements to human evolution which had a more random, non-selective basis. "I have now reanalysed the data and I am convinced that, for one of the three blood groups we studied, there was some variation on top that was due to natural selection. We now know that the ABO blood groups are subject to selection of various kinds," he says.

Drift turned out to be more important for the small, isolated villages of the Italian mountains. Natural selection was more of an influence in the larger villages built on the fertile land of the Italian lowlands.

For many population geneticists today, this simple experiment seems blindingly obvious. At the time, however, it was revolutionary of Cavalli-Sforza to use completely different sources - historical records and a genetic analysis of blood groups - to try to predict and test the course of human evolution. Since then there have been many similar approaches that have helped to explain how early humans migrated from an African homeland to travel across the world over many tens of thousands of years, spreading their culture and their genes as far apart as Alaska and Australia.

The relative frequency of the genes determining the blood groups of modern Europeans clearly shows, for instance, a pattern of early migration from the Middle East to north-west Europe. Most scientists now accept that this must represent the movement of the earliest neolithic farmers about 10,000 years ago. From the physical remains of early human settlements, anthropologists know that it took several thousand years for farming technology to spread from the fertile crescent of the Middle East to the most westerly margins of Spain, the British Isles and Scandinavia. The language of the genes tells this story.

Cavalli-Sforza can lay claim to another radical insight into human prehistory in the way that he has combined genetics with linguistics. He saw very early on that genes and languages have many things in common. They pass from one generation to the next, they suffer "mutations" that change them over time, and that a small, isolated population is likely to share many of the same genes, as well as the same mother tongue.

In addition to natural selection and genetic drift, there is a third important force for change in terms of human evolution: migration. An invading group can change the gene pool of a "founder population". It can also change the language, either replacing it entirely or adding new words and phrases. Cavalli-Sforza saw that linguistics offers another route to cross-checking the predictions made from population genetics. Both could be used to unravel the story of the many mass migrations that must have taken place in human prehistory.

"Language is the central thing that makes us human. It would not be possible to have modern language without thinking in the way that we do. We all think in the same way and we can all learn to speak with equal skills in any language," Cavalli-Sforza says.

"I believe that language was very important for two reasons. By establishing communication you can spread knowledge... and you can learn more easily."

Each of the many thousands of modern languages owes its existence to the development of an early proto-modern tongue some 50,000 years ago, he says.

It was this, and the further development of stone technology and the invention of boat-building, that led to the most important migration in history - from our African homeland and across the four continents to become the most widely distributed species on the planet.

What we see today as racial differences are purely superficial, he says. Genetically we are surprisingly uniform, more so than our different skin colours and facial features may suggest. Race, says Cavalli-Sforza, has no biological meaning. Humans are not divided into racial groups but exist as a genetic continuum.

And all because we share a common history and a common set of genes.