Civilization lost?

On a desert outcropping known simply as NN2, archaeologist Ruth Shady Solís is kneeling over the remains of a clay wall, sweeping away dust with a small whisk broom. Then she stands up, baffled.

"Levels 1 through 3 are straightforward," she says, pointing to three separate tiers of dirt flooring at this site, some 120 miles north of Lima, Peru. But the next tier proves complicated because it's built of two different kinds of fill, one light gray, the other a reddish gray-brown studded with straw.

Were both sections built at the same time? If so, why the change in material? Remodeling was complicated, it seems, even 4,000 years ago.

The two-tone floor remains one of the small puzzles of the much bigger mystery known as Caral. Confirmed last year as the oldest city in the Americas, Caral has shattered the myth that civilization got a late start in the New World. Nearly 5,000 years ago, around the time that Sumerians developed writing and before Egyptians built the Great Pyramid at Giza, people here in the Supe River Valley began building a city.

They knew nothing about writing and had no knowledge of ceramics. But they planned and built huge public works, evolved a specialized and stratified society, and developed a sophisticated and diversified economy. The findings at Caral have added another millennium to the age of civilization in the Americas.

But here in Peru, their discovery evokes mixed emotions from the archaeologists who work the site and the rural people who live around it. There's pride, certainly, but also puzzlement.

"The campesinos always ask: Why did our ancestors have the capacity to build such an important city, and we live so poorly and don't have the ability to do similar things?" says Dr. Shady, the Peruvian archaeologist who recognized the importance of Caral five years ago. The answer "is very difficult for me."

It involves the rise and fall of civilizations.

If ever there were a spot commemorating the shifting fortunes of history, Caral is it. Set in a mountainous desert not unlike southern Nevada, bordered by a long, narrow stretch of green fields fed by the Supe River, Caral has spent millennia covered by dust and debris. Early in the 1900s, archaeologists realized that its six large dunes were too regularly shaped to be natural. But it took decades before excavation began, and until recently, archaeologists believed the site was relatively modern. In 1996, when Shady began working at Caral, she quickly guessed that it dated from the preceramic era but still had very complex architecture. Her excavations began to prove her theory.

For example, two partially excavated pyramids reveal adjacent, circular sunken plazas - a combination of square and round that would come to characterize later structures throughout Peru. The presence of plazas suggests two things. First, that the early society had evolved a need for large ceremonial gathering places. (Shady's team also unearthed 32 flutes made of condor and pelican bones, suggesting a knowledge of music and, perhaps, public ceremonies.) Second, the labor required to build such large public works needed some kind of hierarchy to plan the development and organize the workers.

The pyramid builders had unique building methods. They would tie up rocks in fiber bags, called "shicras," and then transport them to the construction site and lay them, bag and all, as fill to build up the pyramid. Shady's team has uncovered enough of these shicras to notice differences in their quality: some knotted expertly, others less so. This suggested Caral had developed a division of labor with people specializing in different trades.

A civilization arises because it controls something important. Mesopotamia prospered once it irrigated the desert and produced an abundance of food. Caral diverted water from the Supe River to irrigate fields, growing staples such as squash and beans. But its secret weapon may have been cotton. By growing cotton, used to make fishing nets, the people of Caral could trade for fish with the communities on the Pacific coast 12 miles away. Archaeologists have unearthed thousands of fish bones.

The community also traded with communities in the jungle farther inland and, apparently, with people from the mountains. Shady has found the remains of jungle plants at Caral as well as aspects of mountain architecture in the buildings of Caral. The Supe Valley hosts other communities, some of them much older and some within view of the city itself, but none of them approaches the scale and sophistication of this city.

"Caral is a fabulous complex of a site," says Michael Moseley, an anthropology professor at the University of Florida in Gainesville. Its sheer size and the scale of its pyramids suggest to some experts that its inhabitants were developing an economy different from maritime communities on the coast, he says, although the point remains controversial.

Caral's one-time splendor makes its current condition all the more troubling. Shady and her archaeologists have barely scratched the surface of this vast area. The central zone itself stretches out over 160 acres. Her university in Lima, La Universidad Nacional Mayor de San Marcos, supports her and her team of archaeologists. The Peruvian government has provided a new van and 25 soldiers during the work week to help with the digging. But it's not enough, Shady says.

The soldiers have no training in excavation. Because of its own financial woes, her university has cut her team of on-site archaeologists from six down to three. To add insult to injury, the site itself remains unprotected and unguarded. So private cars and even tour buses show up unexpectedly throughout the day. To keep errant tourists from trampling the site, archaeologists leave their own work and give guided tours.

"It's very difficult because in Peru, there is no political culture that favors archaeological investigation," Shady says. "Archaeologists find themselves isolated."

To raise money, Shady agreed to work with Jonathan Haas, curator of anthropology at the Field Museum in Chicago, and his wife, Winifred Creamer, anthropologist at Northern Illinois University in DeKalb. The pair helped Shady get several Caral samples radiocarbon-dated in the United States, which proved the site dates back to at least 2600 BC, as Shady suspected. (The city probably is older, she contends, because the dated samples didn't come from the oldest parts of the excavations.) The three then coauthored an article on Caral.

But relations cooled after the article appeared last April in Science magazine. The American press quoted Drs. Haas and Creamer extensively, making it appear they were leading the team even though their work at the site was limited to collecting the samples for dating. And US funds never materialized.

Haas did propose $50,000 in support if Shady would agree to let him and his wife pursue their research in the area. She refused.

"I think it's an ... unequal relationship," she says. "There are many benefits for the professionals abroad." Little, if any, trickles down to local archaeologists. Haas points out that the US government will only fund archaeological research abroad if an American plays a lead role.

"There are always problems with this kind of arrangement," says Betty Meggers, a research associate and anthropologist at the Smithsonian Institution who has worked for years with Shady and other Latin American archaeologists. "North Americans are always going to be dominant."

But "it's at a crossroads now," she adds. "As you're getting more well-trained people down there, they're saying: 'We've had enough of this.' "

Instead of looking for funds abroad, Shady is trying to build local support from the ground up. She has convinced a nearby village to make T-shirts and caps with Caral logos, which her museum will sell.

After a full day of digging on one recent weekend trek to Caral, she traveled to a nearby village for an hour-long meeting. By the glow of kerosene lamps (the village still has no electricity), she tried to convince local leaders to open a small inn and restaurant to accommodate tourists and visiting archaeologists. Tourism, she hopes, will convince the government that her site is important enough to receive more support. Village leaders, however, remain skeptical.

"There's a problem of self-identification in the country," Shady answers when locals ask her why Peru is so backward today. When Caral flourished, "the society was organized with a population that worked to do things collectively for the collective good. But with the rupture from the arrival of the Spaniards [3,500 years later], there was no more interest in the country except as a source of minerals to be exported to Spain."

Even after the colonizers were thrown out, she says, "our leaders, generally because of problems of identity and self-esteem, believed that everything from abroad was good. Never again did they try to understand the country from its geography, from its history, from its social problems."