he archaeologist David V. Burley looked out on Fanga 'Uta Lagoon and tried to think like ancient seafarers. Here, at the head of the lagoon, is where the first of their outrigger canoes must have pulled in, concluding heroic voyages of hundreds, perhaps more than a thousand miles from the west. So here he decided to dig on the South Pacific island of Tongatapu in the kingdom of Tonga.

Although the site had been excavated before, Dr. Burley, of Simon Fraser University in British Columbia, made a new and revealing discovery concerning a fiercely debated issue in archaeology: the origin and migration routes of the Polynesians.

What Dr. Burley found were shards of the distinctively decorated pottery of the Lapita peoples, cultural ancestors to modern Polynesians. Radiocarbon dating of charcoal among the shards showed that adventurous seafarers had reached the Tonga islands between 850 B.C. and 900 B.C., making this the earliest known settlement in Polynesia.

Tongatapu, Dr. Burley concluded, "probably served as the initial staging point for population expansion" to other islands of Tonga and into Samoa. The place seemed to be what anthropologists and geneticists call a founding colony. These people, he said, must have "formed the gene pool for all the rest of Polynesia."

From this new frontier, scholars think, the ancient navigators perfected the double-hull outrigger sailing canoe and set out on their final expansion, venturing over even more immense stretches of open sea. Each of their bigger canoes probably carried tens of people with their pigs and cargo.

The seafarers made it all the way east to Tahiti and northeast to Hawaii. Hawaii is separated from Samoa and Tonga by more than 2,500 miles, and from Tahiti by 2,700 miles. Then they ranged south to New Zealand and even farther east to Easter Island. The whole of Polynesia, extending over almost one-fourth of the Pacific, thus became the last large area of the world to be settled by people.

Of particular interest, the pottery by the lagoon, which was excavated in 1999, held clues to where the seafarers who reached Tongatapu had originated. In an analysis of bits of the shards, Dr. William R. Dickinson, a University of Arizona geologist, found sandy minerals exotic to Tonga. Some of the pots had been brought there from elsewhere. Further study revealed that the artifacts were composed of minerals found only on the Santa Cruz Islands in Melanesia, some 1,200 miles to the west and closer to New Guinea and Australia.

In a report recently in The Proceedings of the National Academy of Sciences, Dr. Burley and Dr. Dickinson called the shards the first physical evidence that linked the voyages of the Lapita people between the western and eastern regions of the wide Pacific. This further established their capability for "geographically extensive interisland voyaging from or through island groups well to the west of Tonga during the earliest human presence in Polynesia."

The researchers also wrote that the "ties back to the Santa Cruz Islands may imply that Tonga was initially settled by voyagers traveling directly from central Melanesia, rather than through intermediate settlements in Fiji, as has commonly been assumed."

Other scholars of Pacific prehistory said, however, that the new research left unresolved many of the fundamental questions of Polynesian origins. Who were the people that produced the Lapita pottery as well as distinctive stone tools, beads, rings and shell ornaments? Were they an ethnically distinct society of newcomers, or one with diverse groups sharing a handicraft style? What was the relationship of the Lapita culture to descendants of the first settlers of the Pacific, beginning some 45,000 years ago, who occupied Australia, New Guinea and nearby islands?

Although hypotheses abound, no consensus answers have emerged. "We really don't have a good common-sense picture or story of what the migrations were really like," said Dr. John Edward Terrell, an anthropologist at the Field Museum of Natural History in Chicago.

The first of the ornate pots with geometric designs were excavated in 1952 and called Lapita after the discovery site in the Melanesian island of New Caledonia. Subsequent exploration has uncovered the earliest known Lapita artifacts, some 3,500 years old, in the Bismarck Archipelago, northeast of New Guinea. The trail of shards leads from there east to Polynesia. But did the art of making Lapita pottery originate with the indigenous Melanesian population there, or was it introduced by new arrivals?

This so-called "Polynesian problem" may be more than academic. Newly independent nations in the Pacific, anthropologists noted, are expected to look to history and archaeology in creating distinctive postcolonial cultures.

"We must be aware of the political content and implications of our findings," said Dr. Matthew Spriggs, an anthropologist at Australian National University in Canberra. "What has already been written and what is said in the future will be read by audiences we have only recently acknowledged."

Early European explorers seem to have been the first to speculate about the identity of the Pacific islanders. In the 18th century, Capt. James Cook was struck by the resemblance of the customs and appearance among the light-skinned Polynesians on islands several thousand miles apart. His theory was that they had originally come from Malaysia or Micronesia. French navigators were sure that they could not be related to the dark-skinned Melanesians in the vicinity of New Guinea.

The classification of three general groups of Pacific islanders — Polynesians ("many islands"), Melanesians ("dark islands") and Micronesians ("little islands") — was itself a European invention. Some of their differences, it now seems, may be only skin deep.

Until recently, several archaeologists and linguists supported a view that ancestors of the Polynesians left Taiwan and mainland China 3,600 to 6,000 years ago. They spread through the Pacific relatively swiftly, largely bypassing Melanesia, which would explain why the Polynesians are not dark-skinned but speak an Austronesian language, rooted in Taiwan, instead of a Papuan language in parts of Melanesia. This view became known as the "express train" model.

But the model soon ran into trouble. Nothing resembling prototypes of the Lapita pottery has been found in Taiwan or southern China. They first show up in the Melanesian Bismarcks. Although early genetic studies seemed to support the model, more recent ones reveal that the ancestors must have stopped off in Melanesia for considerable interbreeding, which has left clear genetic markers in today's Polynesians.

Dr. Mark Stoneking, an anthropologist at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, said that the express- train model in its simplest form failed the genetic test and that the debate now focused on where the intermarriage took place and how extensive it was.

Quoted in an article last year in the journal Science, Dr. Stoneking proposed that "the ancestors of Polynesians were Austronesians who moved out of Southeast Asia — not necessarily Taiwan — whose population expanded along the coast of New Guinea, intermingled and then moved out into Polynesia."

Or as Dr. Patrick Kirch, director of the Hearst Museum of Anthropology at the University of California at Berkeley, said, the migration "was a train that dropped off boxcars along the way, encountering other human groups, interacting with them, exchanging cultures and sharing genes." From his excavations in the Bismarck Islands, Dr. Kirch said he found 3,500-year-old evidence suggesting that the newcomers came from the islands of southeast Asia. They lived in houses built on stilts, similar to ones in southeast Asia. They brought agricultural plants with them and outrigger canoes.

Dr. Kirch said many archaeologists now subscribed to the "triple-I" model, for intrusion, innovation and integration, which recognizes that there was some fusion between the newcomers and the Melanesians. Out of this interaction apparently emerged, among other things, the Lapita pottery style. Archaeologists are not sure if the pottery was first developed where the oldest specimens have been discovered, in the Bismarcks, or if it was introduced there by the seafarers from the west.

In an article last year in Current Anthropology, Dr. Terrell of the Field Museum and colleagues wrote that current research had yet to account for the differences in appearance between Polynesians and Melanesians that made such an impression on European explorers.

"It is anyone's guess, therefore, how biologically homogeneous or diverse other people were in the Pacific at the time when the ancestors of the Polynesians set sail," the anthropologists said. "Some people living in the older settled parts of the Pacific must have had traits now seen also among modern Polynesians; some may have exhibited traits that were more like those currently seen among southeast Asians; some may have looked more like today's Austronesian-speaking Melanesians."

Dr. Terrell and two of his colleagues, Dr. Kevin M. Kelly of the University of Iowa and Dr. Paul Rainbird of the University of Wales, said an important reason that Polynesians did not resemble their presumed Melanesian cousins or any Asian forebears was that, in a sense, the Polynesians did not "come from" anywhere. They became Polynesians after their ancestors settled in the Fiji Islands, Tonga and Samoa.

That is, they passed through what is known as a genetic founder event, in which everyone is a direct ancestor of an extremely small number of forebears — perhaps a few canoeloads of people who landed at Fanga `Uta Lagoon.

Whatever their roots, the people whose pottery served as signposts of their eastward migrations eventually abandoned the elaborate Lapita style. After their arrival in Fiji, Samoa and Tonga, archaeologists found, the decorative ceramics disappear in a century or more in favor of plain, functional bowls, cups and storage vessels. The people apparently carried no Lapita pottery on their later voyages to the rest of Polynesia.

"The decorated pottery was probably connected with ritual, belief, religion," Dr. Terrell said. "When it got to Tonga, the ideas or politics behind the style became irrelevant. People no longer bothered to produce it and found some other way of expressing their relations to the universe."

By about 2,000 years ago, Dr. Kirch of Berkeley said, the people in the Tonga area had developed a new technology: the double-hull sailing canoe. Even though they could not see other islands on the distant horizon, as had their ancestors in the southwestern Pacific, the notion that the ocean was full of islands endured in tradition. So now that they could, they set forth. These were not all accidental voyages, as was once thought. They were tacking against the prevailing east-to-west winds, knowing that, if necessary, they could ride a following wind back home.

Anthropologists have also learned from oral tradition that these societies had a social structure with status ranked by birth order. "They were descendants of settlers who were junior siblings and their own explorations were very often conducted by junior siblings," Dr. Kirch said. "They were the ones with a reason to explore, to find new land and claim that for themselves."

A similar pattern often applied in the early settlement of British North America.

The "Polynesian problem" may not be solved, but it is attracting increased attention from archaeologists, linguists, geneticists and other scholars. After all, as Dr. Kirch said, the peopling of the Pacific was "a huge diaspora, one of the biggest that happened before European expansion" after the 15th century.