It had been understood from the beginning that there could be counting errors and issues of sampling and contamination, but what soon became evident was that there were also built-in errors in the method itself. One of Libby’s key assumptions had been that the quantity of carbon-14 in the atmosphere was constant, but this, as it turned out, was not true. Changes in the earth’s magnetic field, sunspot activity, even human activities like burning fossil fuels, exploding atom bombs, and testing nuclear weapons, altered the atmospheric concentration of radiocarbon. Unless radiocarbon dates were corrected using a known formula (i.e., calibrated), they could be off by significant amounts (as much as seven hundred years in the case of dates around 3000 B.C.). Gradually, other complexities revealed themselves, including the marine reservoir effect, which lowers the proportion of radiocarbon (and therefore increases the age) of marine organisms, and the “old wood problem”—the fact that wood from a long-lived species of tree could actually be hundreds of years older than a house that was built from it or than charcoal from a fire in which it was burnt.
OF COURSE, EMORY was not the only one requesting radiocarbon dates for Polynesian materials. In New Zealand, Roger Duff sent samples of charcoal from an earth oven at Wairau Bar to two separate laboratories, one in Wellington and the other at Yale, obtaining two date ranges: A.D. 1015 ± 110 from the Yale laboratory, and A.D. 1225 ± 50 from the lab in Wellington. These correlated well with Emory’s date of A.D. 1004 ± 180 for the Kuli‘ou‘ou rock-shelter. But the most tantalizing early radiocarbon dates came from the Marquesas.
The archaeologist attached to the Bayard Dominick’s Marquesan team had reported in 1925 that the Marquesas offered “few opportunities for archaeological research.” But in 1956, a new expedition set out to reexamine the possibilities in these islands at the eastern edge of the Polynesian Triangle. An energetic Columbia University graduate student named Robert Suggs was sent ahead to reconnoiter, and he quickly discovered that the previous generation had gotten it all wrong. Everywhere he looked, he saw archaeological potential. “We were seldom out of sight of some relic of the ancient Marquesan culture,” he writes. “Through all the valleys were scattered clusters of ruined house platforms. . . . Overgrown with weeds, half tumbled down beneath the weight of toppled trees and the pressure of the inexorable palm roots, these ancient village sites were sources of stone axes, carved stone pestles, skulls, and other sundry curios.” There were ceremonial plazas “hundreds of feet long” and, high on the cliffs above the deep valleys, “burial caves containing the remains of the population of centuries past.”
The coup de grâce came when Suggs and his guide followed up on a report of a large number of “pig bones” in the dunes at a place called Ha‘atuatua. This windswept expanse of scrub and sand lies on the exposed eastern corner of Nuku Hiva. A decade earlier, in 1946, a tidal wave had cut away part of the beach, and since then bones and other artifacts had been washing out of the dunes. Not knowing quite what to expect, Suggs and his guide rode over on horseback. When they came out of the “hibiscus tangle” at the back of the beach and “caught sight of the debris washing down the slope,” he writes, “I nearly fell out of the saddle.”
The bones that were scattered all along the slope and on the beach below were not pig bones but human bones! Ribs, vertebrae, thigh bones, bits of skull vault, and innumerable hand and foot bones were everywhere. At the edge of the bank a bleached female skull rested upside down, almost entirely exposed.
Where the bank had been cut away, a dark horizontal band about two feet thick could be seen between layers of clean white sand. Embedded in this band were bits of charcoal and saucers of ash, fragments of pearl shell, stone and coral tools, and large fitted stones that appeared to be part of a buried pavement. They had discovered the remains of an entire village, complete with postholes, cooking pits, courtyards, and burials. The time was too short to explore the site fully, but the very next year, Suggs and his wife returned to examine it.
There were several interesting things about the objects that emerged during the excavation of Ha‘atuatua. Many were classically Polynesian: beautifully curved, nearly circular one-piece fishhooks; pearl-shell points and shanks of two-piece lures; saw-toothed coconut graters of nacreous shell. But some were surprising: a unique form of snail-shell vegetable peeler; “some very peculiar stone adzes” with “Melanesian” shapes; large ornamental pearl-shell disks with notched edges, from a kind of headdress that was common in the islands of Vanuatu but unknown in Polynesia. Like the reel-and-whale-tooth necklaces in the burials at Wairau Bar, in New Zealand, these artifacts seemed to hark back to an early period of Polynesian history and also to suggest a direct link with islands far to the west.
But what was really astonishing was the dates obtained for the samples of ash, charred bone, and charcoal from Ha‘atuatua that were sent for radiocarbon dating. They suggested that these Marquesan artifacts were far and away the most ancient cultural materials to have been discovered in eastern Polynesia, pushing the occupation of the islands back to the second century B.C. Even Marquesan traditions did not suggest such early settlement. “According to the most trustworthy genealogies,” writes Suggs, “the settlement of the archipelago by Nuku, the traditional first settler, occurred about A.D. 950,” whereas the carbon-14 dates suggested that at that point the islands had already been occupied for a thousand years.
But an even more startling discovery awaited Suggs in his second season of work at Ha‘atuatua. Returning in 1957, he began a systematic examination of the site, methodically trenching, digging, sifting, and recording what he found. At first this was comparatively little: “a few fragmentary pearl-shell fishhooks, and a coral file or two,” plus some postholes indicating buildings and a pit that might have contained a burial but didn’t. Then one day, as he was watching one of his workers sift a shovelful of earth, something among the rocks and sand caught his eye, “a flat fragment of some brick-red substance,” which appeared briefly and then disappeared. It looked, thought Suggs, exactly like a piece of pottery. But that couldn’t be—pottery was found throughout Island Southeast Asia and much of Melanesia, but it had never been seen east of Samoa. And yet, there it was: an unmistakable potsherd from the lowest level of the dark band of sand that indicated human habitation. Almost immediately, a second, larger fragment emerged, then a third: a piece of an ancient pot rim with a grooved and rounded lip and marks on the inner and outer surfaces, “from the hand of the potter who had smoothed this vessel in the dim past.”
In all, five fragments of pottery were discovered, belonging to just three vessels: a poorly fired, crumbly brown pot with a coarse temper; a well-fired reddish-brown bowl with a flared rim; and a fine-tempered fragment, also reddish brown, with marks that showed it had been polished using some kind of tool. Modest though they were, these ceramic tidbits changed “the complexion of Polynesian prehistory”—though, as was so often the case, it was not immediately clear in precisely what way. Had the ancient Marquesans made these pots, and if so, why were there not more? Or had they been in trading contact with islands thousands of miles away to the west where pottery was known to have been made? Or might these be fragments of pots that were brought by the very first settlers to the Marquesas? And if so, did that mean that the original settlers had come direct from the western edge of the Polynesian Triangle, bypassing the intermediate, apparently aceramic islands of the mid-Pacific—Tahiti, the Cooks, the Australs, and the Tuamotus?
It was impossible to say. But combined with the dramatic early dates from Ha‘atuatua, the discovery of pottery at the far eastern edge of the Polynesian Triangle was a sensation. Among the immediate implications was the possibility that the Marquesas, rather than pottery-less Tahiti, might be the Hawaiki of legend, the original dispersal point for all the islands of eastern Polynesia. And yet there was something counterintuitive about this idea. All one has to do is look at a map to see that sailing from Samoa or Tonga direct to the Marquesas—straight into the wind for more
than two thousand miles—and then doubling back to reach Tahiti and the other islands at the triangle’s center feels like an unlikely path.
But there it was. There was no getting around the fact that, based on the radiocarbon dates, Ha‘atuatua was the oldest known cultural site east of Samoa and the only one at which even the smallest sherd of pottery had ever been found. “The big question now,” writes Suggs, “was: Just how did this find alter our ideas and theories concerning the origins of the Marquesan people and the Polynesians in general?” Pottery, it turns out, was to be the key.
The Lapita People
A Key Piece of the Puzzle
Lapita patterns from Site 13, in “Archaeological report” by Christophe Sand, Journal of the Polynesian Society (1998).
COURTESY OF THE AUTHOR.
IN ORDER TO understand the significance of pottery for Polynesian history, we have to leave the Polynesian Triangle and travel west to the tiny Melanesian island of Watom. This speck of land, barely three miles wide, lies off the north coast of New Britain, which is itself an island off the north coast of Papua New Guinea, in what is known as the Bismarck Archipelago. One day in 1908, a German priest by the name of Father Otto Meyer was digging a foundation for his church when he uncovered a number of finely decorated potsherds. Father Meyer was not at all sure what exactly he had discovered: “I, poor hermit, what do I know of these scientific questions which are still so perplexing?” But it occurred to him that the sherds might be evidence of contact between the people of Watom and the inhabitants of South America, so he sent a few of the fragments to the Musée de l’Homme, in Paris.
A few years later, in 1920, the Bayard Dominick Expedition’s Tongan team reported the discovery of decorated pottery on the island of Tongatapu. Pottery was not considered to be part of the Polynesian “cultural complex,” but early European explorers had reported seeing small earthenware pots in the Tongan archipelago. Cook’s naturalists, who saw them there on the second voyage, thought they might be “memorials” of Tasman’s passage through the archipelago in 1642, while nineteenth-century Europeans who saw pottery in Tonga generally believed it to have come from nearby Fiji, where there was a robust and ongoing ceramic tradition. None of this prepared the Bayard Dominick team for the quantity of pottery they discovered once they started to dig. They collected no fewer than 1,577 pieces of a “highly porous, lightly fired, fragile ware,” ranging in color from red through brown to dark brown to black. Most of the sherds lacked decoration and were, in the words of one of the team members, “rude, indifferently made, and serviceable rather than ornamental.” But a few displayed a highly refined and distinctive type of ornamentation made up of small dots arranged geometrically in lines and curves, which appeared to have been stamped into the clay.
The senior member of the Tongan team was a man named Edward W. Gifford, a methodical, data-driven anthropologist with a curious claim to fame. An amateur ornithologist and conchologist whose first academic paper—on the mollusk Epiphragmorphora fidelis—had been published when he was only fourteen, Gifford took a job right out of high school at the California Academy of Sciences. He was then steadily promoted through the ranks until, at the age of fifty-eight, he reached the rank of full professor at the University of California, Berkeley without ever having attended college himself.
In 1947, Gifford returned to the Pacific on the first of a series of three archaeological expeditions to Fiji, New Caledonia, and Yap. He was hoping to establish a “succession of cultures,” following a path backward in time (and westward in space) from the western edge of Polynesia. On his first expedition, he discovered that what had been true for Tonga was even more true for Fiji: the “chief characteristic of Fijian archaeology” was “pottery, above all.” Potsherds were everywhere, in the dunes and rock-shelters and riverbanks, “conspicuous on the surface” and “abundant beneath.” Discovered in digs at depths of up to twelve feet, they were vastly more common than any other kind of artifact. Although Gifford also found tools and ornaments made of stone and shell, the ratio of these to potsherds was on the order of one to ten thousand. Present in all places and at all points in time, Fijian pottery came in a great variety of styles, including some with a distinctive geometric patterning made up of straight and wavy lines, arcs, dashes, and dots that appeared to have been stamped or pressed into the clay using some kind of tool.
Following his star westward, Gifford next embarked in 1952 for New Caledonia. New Caledonia, with its neighbor Vanuatu, occupies an intermediate position between the Polynesian Triangle and the Bismarck Archipelago. Like New Zealand, it is a piece of old Gondwana, and, also like New Zealand, it has a strange and unique natural history. It is home to an unusual number of curious animals and plants, including a “mistletoe-like” parasitic pine and a famous tool-using crow. Before the arrival of human beings, New Caledonia had one of the most diverse collections of reptiles anywhere on the planet, including a giant horned turtle with a spiked tail, a twenty-pound monitor lizard, and a rare pygmy land crocodile, all of which—along with a giant megapode, a flightless swamp hen, two falcons, a scrub fowl, and several other species of bird—are now extinct.
As soon as Gifford and his colleague Richard Shutler Jr. arrived in New Caledonia, they began touring the island in search of suitable archaeological sites, traveling around in a 1951 Chevy half-ton truck that had been shipped over from California. Setting out from the capital, Nouméa, they traveled up the wetter, more heavily forested east coast, crossed over the central mountain range, and made their way back down the drier, savanna-like western side, stopping at various points along the way. At the edge of an eroding beach on the Foué Peninsula, they picked up their first piece of decorated pottery. The spot, to which they would later return, was known as Site 13.
OVER THE COURSE of the next seven months, the two anthropologists, with their teams of local helpers, as well as their wives, who had come out from California to join them, surveyed fifty-three sites and excavated eleven. They put more than fourteen thousand miles on the truck—on an island just 250 miles long—and collected more than seventeen thousand artifacts, including sinkers, knives, peelers, bracelets, scrapers, choppers, grinders, and adzes. But, once again, the most common artifact was pottery. The majority of sites were what Gifford described as “coastal kitchen middens,” which could be recognized by surface scatterings of potsherds and shells. Almost all of these sites, he observed, shared three characteristics: they were close to the sea, they were near a source of fresh water, and there was good gardening land nearby.
It was during this expedition that Gifford received news of the first radiocarbon dates from his earlier excavations in Fiji. Anticipating the possibility as early as 1947—before the method of radiocarbon dating had even been experimentally proved—he had been careful to collect charcoal samples for testing but had been thwarted in his attempts to get them dated when he returned from the field. Libby’s lab at the University of Chicago was swamped, and requests made to both the California Institute of Technology and the Radiation Laboratory at Berkeley were turned down. “I am afraid,” wrote an assistant at the Rad Lab unhelpfully, “that our answer . . . will have to be in the negative. If, however, you are able to assign a student or technician to the problem of developing the apparatus and method for performing these determinations, we will be happy to consult and advise.”
Eventually, Gifford secured the assistance of the University of Michigan’s Memorial Phoenix Project, a research institution launched in 1948 to support “peaceful uses” of atomic energy. When tests of the first Fijian sample found it to be approximately 950 years old, he immediately dashed off a note to The Journal of the Polynesian Society. The date, Gifford wrote, was “of considerable significance, since it implies a much greater age for the deeper parts of the deposit.” And, indeed, subsequent analyses of charcoal from the same location later demonstrated that the site had first been occupied nearly a thousand years earlier, around the first century A.D.
At t
he end of six months, even as preparations for their departure were getting under way, Gifford and Shutler returned to Site 13 for the second-to-last dig of the expedition. They had plenty of workers and the site was easy to excavate; elsewhere there had been cave-ins and pits that continually filled with water. They marked out a number of rectangles along the sloping edge of the beach and across the isthmus in a low, overgrown area between the mudflats and the sea that showed signs of having recently been used as a yam field. Almost immediately it became clear that Site 13 was different from all the other sites they had examined. As in Tonga and Fiji, more than 90 percent of the pottery found in New Caledonia had been of a plain, undecorated kind, with only small amounts showing any form of decoration. But what emerged from the ground at Site 13 was not only a vastly greater percentage of decorated pottery—roughly a third of the total—but an entirely new class of ceramics that were “vastly superior” to anything the researchers had yet seen.
Richly and completely decorated—sometimes on both the inside and the outside of the pot—with delicate patterns of circles, semicircles, diamonds, squares, triangles, zigzags, palmettes, and “eye” designs (almonds with a circle inside), the pottery from Site 13 was reminiscent, some thought, of Corinthian vases from the seventh century B.C., or the black Etruscan pottery known as bucchero nero. Others described it as similar in style, and perhaps technique, to Polynesian tattooing. The patterns, which were both intricate and regular, seemed to have been imprinted on the clay using a tool with teeth and then filled with a white paste of coral lime—a technique strikingly similar to the traditional Polynesian method of tattooing using a sharp, toothed instrument to puncture the skin and rubbing dye or charcoal (or, in nineteenth-century New Zealand, gunpowder) into the wound.
On the last day of the dig, Gifford noted in his field book that “Lapita was name of village at this site,” thereby christening (unintentionally) not only the distinctive type of pottery that had been found at Site 13 but the as yet unknown people who had left it there. In his final report, he made several important connections, noting the stylistic continuity between the pottery from this site and sherds he had seen in Tonga and Fiji, and connecting the dots between his work and French discoveries on the nearby Isle of Pines, which, in turn, had been linked to Father Meyer’s Watom sherds in the Musée de l’Homme. Gifford and Shutler’s feeling that this was all quite significant was more than confirmed when the results of the radiocarbon tests came back, showing that by far the oldest New Caledonian samples were those from Site 13. But what was really surprising was how early the dates were: at 2,800 years before the present, they pushed the occupation of New Caledonia back to the end of the first millennium B.C.
Sea People Page 22