Schoolchildren can recite the story of the first Americans.

About 12,000 years ago, prehistoric humans walked out of Siberia, trekked across the Bering land bridge and down an ice-free corridor into inner North America, where they hunted Ice Age elephants and peopled the new world.

But mounting evidence is slowly turning that story to fiction, said Michael Collins, an archaeologist with the Texas Archaeological Research Laboratory at the University of Texas at Austin.

For more than 20 years, Collins and other scientists have been digging up artifacts from Chile to Texas that convince them the first Americans didn't walk here at all, but came by boat, and arrived much earlier than previously thought.

"This has been hotly debated," Collins said. "That theory has held sway for 70 years or so. But a few of us for the last 25 years have come to seriously doubt that theory."

Collins is in San Antonio today to talk about the shifting debate over the first Americans. Collins and archaeologist Robert Ricklis, who excavated a 7,000-year-old cemetery near Victoria, will speak at a Southern Texas Archaeological Association meeting at the University of the Incarnate Word. The meeting is open to the public.

For decades, the first Americans were thought to be the Clovis people, named after a site in Clovis, N.M., where 11,000-year-old fluted points were found in the 1930s. Since then, Collins said, other sites in Pennsylvania, Chile and Virginia have yielded older finds.

Collins first became convinced of "pre-Clovis" ancestors in 1967, after discovering burned mammal bones with butcher marks at a site called Cueva Quebrada in Val Verde County. Carbon dating of charcoal put the bones at 14,000 years old. To this day, most other scientists have ignored those findings, Collins said.

In the 1970s, Collins worked on a site in southern Chile called Monte Verde, which contained artifacts at least 1,000 years older than those at the Clovis sites. At first, many scientists attacked the validity of the evidence and clung to the theory that the Clovis people arrived first, Collins said. Over time, they began to accept the site and the tide of opinion turned, he said.

"I spent 20 years of my life being beat up over that project, as did everyone else," Collins said. "It has finally, begrudgingly, earned the support of a significant majority of archaeologists."

But if the Clovis people were not here first, who were the first Americans?

"It's really a case of stay tuned," Collins said. Theories have been proffered, but none universally accepted, he said.

Collins himself believes America was likely peopled on two fronts. Coastal communities in both Asia and Europe likely made their way to the New World on boats, sticking close to ice shelves to fish and hunt sea mammals. Though no ancient boats have been found, Collins points to evidence that Asians traveled to Australia 50,000 years ago, presumably in boats, since the island continent has never been connected to a land mass.

Collins also points to evidence from Japan that suggests prehistoric humans 30,000 years ago ate deep-sea fish and possessed obsidian found only on distant Japanese islands, which also suggests the use of boats.

Though this far-flung evidence interests Collins, his efforts to debunk the Clovis-first theory are closer to home.

For the past several years, he has led work at the Gault site, a large Clovis campsite midway between Georgetown and Fort Hood. A rich bounty of evidence at Gault suggests the Clovis people were not highly mobile hunters, as previously thought. It's more likely they were somewhat settled hunter-gatherers who occasionally felled a mammoth, but lived mostly on plants and smaller game such as frogs, turtles and birds.

"(Gault) is the poster child for Clovis not fitting the theoretical model," Collins said.