The Pacific Ocean covers roughly a third of the earth's surface, and it was navigated by people for thousands of years before any of the instruments that modern sailors take for granted became available. By the time European ships began to arrive in the sixteenth century, almost every habitable island in the Pacific had been settled, some of them several millennia earlier. How the ancestors of the modern Polynesian, Micronesian and Melanesian peoples crossed such distances, using only wooden canoes, has become one of the most active questions in archaeology and oceanography.

The spread was not a single event. Settlement of the western Pacific, particularly the islands near New Guinea and the Solomon Islands, dates back at least forty thousand years. The long-distance voyages into the remoter Pacific began later. Fiji, Samoa and Tonga were reached around three thousand years ago by people whose distinctive pottery, known as Lapita ware, allows archaeologists to follow their movements. The Hawaiian islands, Easter Island and New Zealand were all settled more recently, within the past twelve hundred years.

The canoes used for these voyages were not the simple hollowed logs sometimes imagined. They were double-hulled, lashed together with fibre cord, and driven by woven sails of palm leaves. A large example could carry thirty or more people, together with the livestock and plant material needed to found a new settlement. Experimental voyages in the last thirty years, including those of the Hokule'a, a replica Hawaiian canoe built in the 1970s, have shown that such craft could easily maintain a speed of seven or eight knots in favourable conditions.

The method of navigation, however, has been the more controversial subject. Early European explorers left records of conversations with Pacific navigators that seemed to imply remarkable knowledge of stars, winds and sea conditions, but the details were often fragmentary. By the twentieth century, much of the traditional knowledge had been lost in islands that had adopted Western navigational aids. The revival began when a few Micronesian navigators, in places such as the Caroline Islands, were found still to use and teach the older techniques. One of them, Mau Piailug of Satawal, agreed to teach Hawaiian sailors in the 1970s, and the subsequent voyages of the Hokule'a, including a navigated crossing from Hawaii to Tahiti, demonstrated that the traditional methods could still work over distances of thousands of kilometres.

Piailug's methods combined several sources of information. The positions of stars near the horizon at different times of night provided a kind of compass. The direction of ocean swells gave information about distant weather systems and, over weeks, about the bearing of distant land. The presence of particular birds at particular times indicated the direction of shore. Cloud shapes and sea colour carried further clues. None of these cues, taken alone, was sufficient; the skill lay in weighing them together. Anthropologist Dr. Mele Tuilagi has observed that a navigator's attention during a voyage was almost total, with sleep snatched in short intervals and frequent small adjustments to course based on changing conditions.

Academic debate has focused on whether the long voyages were deliberate exploration or accidental drift. The anthropologist Thor Heyerdahl famously proposed, in the 1940s, that Pacific islanders had arrived from South America on rafts drifting with the currents. His 1947 voyage on Kon-Tiki demonstrated that such a drift was possible but did not settle the question. Later archaeological, linguistic and genetic evidence has overwhelmingly supported a different picture: the main settlement of the Pacific came from the west, by people who deliberately sailed against the prevailing winds by waiting for seasonal shifts. Dr. Tuilagi, who has worked on canoe projects in Samoa and Tonga, argues that the very existence of a continuous tradition of celestial navigation, documented in multiple island cultures, is hard to explain if the voyages were accidental.

The revival of traditional methods has had effects beyond scholarship. Several Pacific communities now treat long-distance canoe voyages as a cultural practice in their own right, linking modern young people to the skills of their ancestors. These projects are more than pageantry; they are also a reminder that the open ocean, easily imagined as empty, has for a very long time been one of the most well-travelled parts of the planet.