NewsNEOLITHIC AGRICULTURE

Reading the Signs of Ancient Animal Domestication

See allHide authors and affiliations

Science  20 Nov 1998:
Vol. 282, Issue 5393, pp. 1448
DOI: 10.1126/science.282.5393.1448

This article has a correction. Please see:

Over the millennia, humans seeking a steady source of food, hides and wool, and companionship have tamed everything from wolves to turkeys to guinea pigs. Learning when—and why—each of the more than two dozen domesticated animals was brought under human rule has been a continuing quest for archaeologists. Now researchers are shaking up their old conclusions by using more sensitive techniques, such as tracing demographic patterns in bone assemblages, to tease out the signature of human handling. So far such methods are pushing back the dates of domestication of one animal—pigs—revealing animal husbandry in what is now southeastern Turkey long before cultivation began there. More examples may follow. The findings are “causing quite a stir,” says Bruce Smith, an archaeologist at the National Museum of Natural History in Washington, D.C. “People are now going back and looking at other animal species.”

Traditionally, the first farm animals were thought to be wild goats and sheep, tamed in southwest Asia around 10,000 years ago* by sedentary cereal farmers who had wiped out the local wild game and needed new sources of meat and hides. Domestic pigs and cattle followed around 9000 years ago. And the earliest firm evidence of dairy farming, from art and written texts, isn't until about 6000 years ago, although new dates could come from a new method for identifying milk fat residues on pottery sherds, reported on page 1478 of this issue.

Most archaeologists rely on a more mundane characteristic to identify domestic herds: size. Researchers assume that early domestic goats, sheep, pigs, and cattle were smaller than their wild cousins. Early pastoralists, the theory goes, kept their animals in worse conditions than in the wild and selected for smaller, more easily subdued males. “Who would you choose?” asks Melinda Zeder, curator of Old World archaeology and zooarchaeology at the Smithsonian Institution. “The nerdy goat with the glasses or the bully on the playground?”

But in a controversial new study, as yet unpublished, Zeder tests both the size idea and a newer indicator—a distinctive pattern of mortality that distinguishes herds from hunter's prey. Brian Hesse, a zooarchaeologist at the University of Alabama, Birmingham, reasoned that ancient pastoralists, like modern ones, probably tried to get as much meat as they could from their goats while still ensuring the herd's survival. The obvious strategy is the one still used around the world today for managing livestock: raising females to maturity and keeping them until they quit producing offspring, while butchering most males young and keeping only a few older males as breeding stock. “So there should be a very distinctive marker in the demography,” says Zeder, “and that should be instantaneous with the early period of managing.”

To hone her strategy for spotting this transition, Zeder examined nine different bones of a control group of 40 modern-day wild and domestic goats of known sex from Iran and Iraq, where goats are thought to have been first domesticated. She could reliably distinguish goats from sheep and determine the animals' age at death, based on the sequence of bone fusion from 10 to 36 months. She could also determine the animals' sex, because the bones of the males were consistently bigger than those of similar-aged females.

Encouraged, she turned to tens of thousands of goat bones from eight sites in Iran and Iraq, ranging from Paleolithic hunter-gatherer caves to two Neolithic villages. She found little size difference between goats at the 9800-year-old Neolithic village of Ganj Dareh and goats hunted by Middle and Upper Paleolithic bands more than 40,000 years earlier.

But she did find a significant difference in mortality patterns. In the early sites, almost all the male goats were 36 months old or older at the time of death; the less numerous females were younger, suggesting that hunters had targeted male goats in their prime. But at Ganj Dareh, few billies lived past 24 months, while almost all nannies survived to 36 months or more. This suggests that “they are allowing the females to live as breeding stock,” says Zeder.

The only evidence Zeder found for size reduction came at the Neolithic village of Ali Kosh, which new radiocarbon dates place at 9000 years ago. Zeder suggests that the animals were smaller there because it lies south of wild goats' natural range, so the animals were kept in hotter, harsher conditions—and females could no longer be bred with big wild males. Thus size reduction, rather than being the first sign of domestication, might instead indicate that animals had been transported beyond their original range or were no longer being bred with the wild type, Zeder suggests.

Not everyone is persuaded. For example, Harvard University faunal analyst Richard Meadow argues that some of the bones Zeder used, in particular the toe bones, don't accurately reflect an animal's size; he's not ready to give up on size reduction as an indicator of domestication. But other researchers, such as Curtis Marean, a zooarchaeologist at the State University of New York, Stony Brook, say Zeder's analysis is an important step forward. “It shows that the old idea that body size of the animals is directly related to domestication really doesn't fit the evidence,” says Marean.

Zooarchaeologist Richard Redding of the University of Michigan, Ann Arbor, agrees, and indeed Zeder's demographic patterns fit well with his recently published study of ancient pig bones at the site of Hallan Chemi in southeastern Turkey. For years, faunal analysts pointed to the declining size of pigs' second and third molars as a key trait for some reason associated with pig domestication, and they traced the earliest domestic pigs to a 9000-year-old village in Turkey. But Redding now believes he has found earlier evidence, by applying demographic criteria like Zeder's.

He analyzed animal remains found in layers at Hallan Chemi and noted that in early layers dating to about 11,500 years ago, pig bones made up just 10% to 15% of the fauna and were almost evenly split between male and female. But in the later layers, dating from 11,000 to 10,500 years ago, pig bones climbed to 20%. “They also become very heavily biased toward female, and they become very young. So the inhabitants are killing suckling pigs,” Redding says. All this happened before domestic cereal grains appear at the site, indicating that the people at Hallan Chemi were herding pigs before they began to farm grain. If Redding is right, the inhabitants of Hallan Chemi are the world's first known herders—and pigs, not goats or sheep, were the first farmyard animals to start on the long road to full domestication.

  • * All dates are calendar years.

Navigate This Article