FLORENCE, TEX.—"Look at that—isn't it gorgeous?" Sandy Peck asks as she rinses dirt from a flaked stone about the length and width of a pinky finger. Peck runs a hose over soil on a fine-mesh screen, prodding at stubborn clods of clay with a muddy glove. "Look, there's another one."

View Slide Show of the Dig

Peck, sorting soil that had been disturbed by a recent thunderstorm, is a volunteer looking for artifacts in the Gault Valley in central Texas, some 40 miles (65 kilometers) north of Austin. The valley hasn't changed much over the last several thousand years: A spring-fed creek still runs among live oaks and pecan trees, jackrabbits and deer still live on the nearby uplands, and cobbles of chert, ideal for making stone tools, still bulge from the valley's limestone walls. Today, however, instead of working hides and shaping stones as they did 13,000 years ago, humans painstakingly sift the soil in search of ancient artifacts that will overturn long-held assumptions about the earliest Americans.

Since the 1930s textbooks have taught that the New World's first inhabitants, known for the town in New Mexico where their spear points were discovered, walked from Siberia to Alaska about 13,300 years ago. The Clovis people were believed to be highly mobile nomadic hunters, never settling in one place, instead surviving on massive mammoths, mastodons and ancient bison.

But in excavations starting in 1998 Gault has revealed that Clovis people lived at the site for extended periods over a span of 300 years, says Michael Collins, a research associate with the Texas Archeological Research Laboratory. The evidence? Scientists have found numerous tools manufactured from local stone, used until they were worn, then repaired repeatedly until they finally were discarded. In other words, Paleo-Indians were members of a settled community. "We're redefining Clovis," Collins says.

It's unusual to find a site that has both materials for stone tools and "enough resources that people could camp and live right there," says Dennis Stanford, head of theSmithsonian National Museum of Natural History anthropology department who has visited the Gault site several times. "Usually you don't get both, but at Gault you get the whole show. So it looks like people were living there for extended periods of time."

Gault also suggests previous assumptions about Clovis's diet were wrong. Sure, they ate mammoth and bison, but archaeologists are also finding bones from frogs, turtles, snakes and rabbits. "Coming home with three rabbits isn't as dramatic as the museum mural image of Clovis people sneaking up on a mammoth," says Collins's colleague, Andy Hemmings, but probably better reflects day-to-day life.

Not everyone is convinced that Clovis was such a homebody. "Gault is not completely rewriting what we know about Clovis," says Robert L. Kelly, head of the department of anthropology at the University of Wyoming. "[Collins] could be right that this particular population of Clovis was more settled than other peoples. But is Gault the pattern or is Gault an aberration? We don't know."

The Clovis finds would be dramatic enough, but Collins also claims to have found evidence of an earlier culture at Gault. In 2002 the team dug below Clovis layers and promptly found hundreds of stone flakes. The surrounding soil was dated to 350 years before Clovis. When they kept digging, the team dated other materials to even further back, although Collins used a technique that others have questioned, and that even he acknowledges is imprecise.

If it holds up, Collins's claim would add Gault to the list of proposed pre-Clovis sites, including Monte Verde in Chile, Meadowcroft Rockshelter in Pennsylvania and Cactus Hill in Virginia. They're all controversial, however, based on charges of contamination and other problems.

Collins hopes to bolster his case by digging further at Gault. He'll also use another method to test the soil's age. And some 800 miles (1,300 kilometers) from Gault, in the northeastern Gulf of Mexico, members of his team will soon tackle the perplexing question of how Paleo-Indians got here in the first place. Until an ice-free corridor opened in northwestern Canada about 13,300 years ago, providing a route to the interior, a dome of ice divided Asia from North America. Some archaeologists have proposed an ice-free route along the northwestern coast; others suggest the first Americans, like the first Australians, had boats and used them either to travel east from Asia or, a few daring archaeologists propose, west from Europe. Hard evidence is unavailable because the coastline moved several hundreds of meters inland when the ice sheets melted.

"The archaeological record is out there underwater," says Hemmings, "so that's the next frontier in this search." Hemmings plans to spend two weeks this summer in the northeastern Gulf of Mexico looking for Clovis and pre-Clovis sites. In an expedition funded by the National Oceanic and Atmospheric Administration, a team lead by Hemmings and James Adovasio of Mercyhurst College in Erie, Pa., will explore promising sites along ancient coastlines using remotely operated vehicles. They've identified prime targets by studying underwater maps for features Clovis people are known to have preferred, such as cliff faces near streams or rivers.

Back on dry land, Collins believes Gault has more to tell us about early Americans. "It's a special place," says Collins, "and it's been a special place for a long, long time."

View Slide Show of the Dig