First author

Cattle, sheep and goats were domesticated more than 10,000 years ago, but when and where secondary animal products — such as milk or wool — were first used remains a subject of debate. Richard Evershed, a biogeochemist at the University of Bristol, UK, invented a stable-isotope technique to detect milk residues and has now used it to analyse an unprecedented collection of more than 2,200 pottery fragments in a search for direct evidence of milk usage. On page 528, he and an international group of collaborators trace the earliest known use of milk to residues on pottery from southeast Europe and the Near East. Evershed tells Nature that milk may have trumped meat as a Stone Age food source.

What inspired this project?

One of our archaeologist co-authors, the late Andrew Sherratt, developed a theory that secondary animal products were first used some 7,000 years ago. Ten years ago, I developed a compound-specific carbon isotope method to identify the fats produced by domesticated animals. This allowed us to test the dairying aspect of Sherratt's hypothesis on a macro-regional scale. Our findings show that milk use occurred 2,000 years earlier than he had suggested.

Why is milk such an important archaeological topic?

Animals were probably domesticated first for meat. These data provide the first hints of humans using cows for large-scale milk production. Slaughtering an animal, especially a cow, is a dramatic final act. The alternative — a sustainable, daily supply of milk as a nutritious source of carbohydrate and protein — helped people to meet their dietary needs more efficiently.

Where did the use of milk emerge?

Our data show that milking was particularly important in what is now northwest Turkey, in part because the environment was suitable for cattle grazing and herding. It seems that milk was produced there at a surplus level, and its widespread use by humans followed. These data are from some of the earliest pottery in Europe and the Near East, raising the interesting possibility that containers were needed to hold milk.

Did you gain unexpected insight from your data?

Yes. We set out to determine where and when milk use emerged. We then wondered how people at the time, believed to be mainly lactose intolerant, could consume milk. But the residues in pottery suggest milk was processed — for example, into butter or yogurt, which contain less lactose — getting around the lactose-intolerance problem.