Jump to ContentJump to Main Navigation
Sailors and TradersA Maritime History of the Pacific Peoples$

Alastair Couper

Print publication date: 2008

Print ISBN-13: 9780824832391

Published to Hawaii Scholarship Online: November 2016

DOI: 10.21313/hawaii/9780824832391.001.0001

Show Summary Details
Page of

PRINTED FROM HAWAII SCHOLARSHIP ONLINE (www.hawaii.universitypressscholarship.com). (c) Copyright Hawaii University Press, 2022. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in HSO for personal use.date: 29 June 2022

The First Pacific Seafarers

The First Pacific Seafarers

Chapter Two The First Pacific Seafarers
Sailors and Traders

Alastair Couper

University of Hawai'i Press

Abstract and Keywords

This chapter describes the characteristics of the first Pacific seafarers. The peoples of the Pacific Islands have a history of early long-distance seafaring that is unrivaled anywhere in the world; their ancient ancestors were the first ever to make use of the open sea for large-scale migrations. Although there are many differences, overlaps, and transitional zones between islands and peoples, regions, and somewhat tenuous groupings of related human characteristics, have been geographically classified into the divisions and boundaries of Polynesia, Melanesia, and Micronesia. This chapter first provides an overview of the actual vessels used on the ocean voyages of exploration, migration, or early long-distance trade, with particular emphasis on the construction of oceangoing ships. It then examines the influence of weather and climate variables on Pacific voyages and migrations before discussing prehistoric seamanship and the process of navigation. It also considers the nutrition of Pacific seafarers and concludes by assessing the challenges faced by seafarers during arrival at the beach.

Keywords:   seafarers, Pacific Islands, seafaring, vessels, voyages, ships, weather, prehistoric seamanship, navigation, nutrition

THE PEOPLES OF the Pacific have a history of early long-distance seafaring unequaled anywhere in the world. As far as can be determined, their ancient ancestors were the first ever to make use of the open sea for large-scale migrations. Sometime before 40,000 BC they entered the western region of the Pacific from Southeast Asia. Sea levels were rising in this period of the late Pleistocene ice age, but still stood about fifty meters below those of today. This exposed dry areas of continental shelf, reefs, and islands, interlaced by waterways.1 The migratory Asia-Pacific huntergatherers followed these stepping-stones and channels and settled in New Guinea and possibly northern Australia. Some moved farther eastward over successive generations, following the coast of New Guinea until they reached open water and sailed to the near offshore islands of New Ireland about 40,000 BC, and the Solomons around 30,000 BC.2

The islands of the late Pleistocene were larger and had a different coastal morphology from their equivalents today. Sea passages to many offshore potential human habitats could be achieved by sailing between what were intervisible high landforms. This involved crossings of less than fifty miles, although on reaching the Admiralty group and the northern Solomons, people would have lost sight of land for a short time. It is not certain what the weather conditions were like during these Pleistocene voyages. Geoffrey Irwin agrees with other researchers that at least in this equatorial zone the winds would have been little different from what they are today.3

The Pleistocene-era migratory hunter-gatherers may be the most direct ancestors of the New Guinea highlanders and the Australian aborigines. Little is known about their material culture; their vessels were probably dugouts, bark boats, and bamboo rafts. The eastward migrations of these first seafarers apparently ceased when they encountered the open-sea horizons extending outward from the Bismarck Archipelago and the Solomon Islands. They established interisland trading systems in this area of “near (p.23) Oceania,” which included the carriage of obsidian by sea. Patrick Kirch points to the importance of the archaeological evidence of oceanic trading seafaring some 18,000 to 20,000 years ago that covered over 350 kilometers by boat, much of it beyond visual landforms.4

Open-Ocean Migrations

Many centuries after the initial settlements of near Oceania, a more technically advanced seafaring society moved from Southeast Asia and emerged in the region of the Bismarcks. These were Neolithic cultivators, fishers, and rearers of livestock. Already by 4000 BC they appear to have possessed large sailing craft with affinities to those of Indonesia. From possibly 3500 BC they sailed by interisland passages to Santa Cruz, then through Vanuatu. From there it was more than 450 miles to Fiji, and from Fiji 650 miles to Samoa (ca. 1000 BC), with possible en route islands. These routes are marked by their distinctive Lapita pottery.

Straight-line distances are not very meaningful, since actual miles covered could be three times greater, due to set and drift of currents, leeway, and tacking. Simply as a basis for comparison, the Polynesian descendants of the Lapita people in Samoa later sailed much greater passages of 1,000 miles of open sea to Tahiti (arrival about 1000 BC) and 700 miles to some of the Cook Islands. Furthermore, in this easterly sector of the Pacific the southeast trades are strongest, and unless El Niño was prevalent, chances of westerlies would be minimal but not impossible. The onward voyages from Tahiti to the Marquesas through the islands of Tuamotu would entail only 300 miles of open sea, but from there to Hawai‘i, nearly 2,000 miles (ca. AD 400), and from Mangareva possibly to Henderson Island, 380 miles, and from there to Rapa Nui, 1,000 miles (after AD 1000). Eastern Polynesia clearly presented formidable challenges.

It was several centuries following the colonization of the eastern Pacific that the southern periphery of what became the Polynesian triangle was completed, with voyages of over 1,600 miles assumed from Rarotonga to Aotearoa.5 These passages appear to have been undertaken about AD 1000 to 1300.6 There is no evidence of the great Polynesian voyages reaching Australia, although vessels did return northwestward from Western Polynesia into Melanesia and Micronesia, reaching Nukuoro, Tikopea, Anuta, and other islands, which are known as the Polynesian outliers. Polynesian seafarers also reached Norfolk Island around AD 1200; from there Australia is about 800 miles, well within their capability.7 Similarly, there is no firm evidence of the Lapita descendants sailing (p.24) from the remote east Pacific islands to South America. This would have been a more difficult but not impossible voyage of 2,000 miles from Rapa Nui. The sweet potato has origins in South America and was subsequently diffused throughout the Pacific.8 There were obviously some early links between the American continent and the Pacific islands, but the extent of those links is still uncertain.

The Micronesian island groups of the northwest Pacific received seafarers direct from the Philippines into the Marianas. Others came from the Lapita of the Solomon-Vanuatu region. Yet other related groups came from the Bismarcks to Yap Islands. The probable dates of various island settlements in Micronesia, including the Marshalls and Kiribati, range between 1500 BC and AD 500. All the dates so far alluded to may be subject to future changes as DNA research progresses.

As it stands, DNA indicates that the migrants who entered the remote Pacific were not genetically homogenous. They had progressed through Melanesia and mixed with preexisting non-Austronesian populations. The research also confirms Fiji as pivotal in human dispersions. These voyagers carried commensal animals, including pigs and rats. Pig bones found in middens reveal a single genetic inheritance over several routes. The bones of the Pacific rat (Rattus exulans) are unique and are proving valuable in tracing the origins and routes of their seafaring hosts. So far the DNA analysis show general west-to-east movements. Of fundamental interest are the bones of the Pacific rats as proxy for human migrations in Rapa Nui, the very extreme eastward position. These rats do not have the complexities of human genetic data, which are open to interpretations. Excavations by University of Hawai‘i teams in 2004 at Anakena beach on Rapa Nui have revealed rat bones in early arrival strata at about AD 1200. These rats are not native to North or South America, and this confirms a founding Polynesian population.9

The great Pacific migratory voyages were the basis for identifying the ethnographic stamp which the diaspora of 2000 BC to AD 1300 has put on the Eastern Pacific in particular. During this period, generations of seafarers visited every island in the vast Pacific Ocean, and they transferred thousands of people, plants, and animals to occupy most of the island world. Much of the flora, fauna, cultural, and human physical inheritances that are there today were the products of these voyages. In the course of time more locally distinctive cultural characteristics of people emerged in various island regions according to environmental challenges, levels of insularity, and later colonial impacts. Although there are many differences, overlaps, and transitional zones between islands and peoples, there (p.25) is a broad basis for the accepted geographical classifications of regions, and somewhat tenuous groupings of related human characteristics, into the divisions and boundaries of Polynesia, Melanesia, and Micronesia. This geographical classification was initially made by Dumont d’Urville in 1831 and remains a useful generalization for some Pacific divisions, as indicated by map 2.1.10

The Ships

The great migrations from west to east and subsequent voyages were made on a variety of vessels. Unfortunately the European misnomer “canoe” for these seagoing ships of the Pacific is misleading but has persisted over centuries and would be difficult to displace. There were, and still are to a much reduced extent, many types of vessels indigenous to these islands. They range from single dugout canoes—some with outriggers, like the Fijian takia, which is still extant—to the substantial ships that made the great oceanic voyages of the past but have long since disappeared.

The generic term for island vessels is “wa” and its variants. It is frequently a prefix indicating the type and function of Pacific craft, as in the Fijian “waqa ni koro” (boat of the village) and “waqa tabu” (sacred, or chief’s, vessel). In Samoa it is expressed as “va‘a” (often indicating a large vessel) and “va‘aalo” (small outrigger). Likewise in Tahiti and the Marquesas “va‘a” is used as “va‘a motu” (small vessel), as well as the term “vaka.” The latter word is common in Tonga and the Cook Islands, and “waka” is also used in the Cooks. In Hawai‘i it is “wa‘a,” in Aotearoa “waka” (as in “waka taua,” or war vessel). This is also a boat word among the northwest Pacific Polynesian outliers. In Kiribati the term “wa ririk” is still in use for small outriggers,11 and in the Marshalls it is “wa lap.” There are many other descriptive terms for vessels, but the occurrence of “wa” over a vast area of the Pacific is one of the many indicators of a common maritime inheritance.

Little has remained of the actual vessels used on the ocean voyages of exploration, migration, or early long-distance trade. The teredo worm and the hot, humid conditions of the Pacific effectively put an end to most of the ancient littoral and underwater cultural heritage. Excavations on Huahine in the Society Islands have uncovered some ship artifacts circa AD 800, including a carved steering paddle 3.8 meters long and two adzed planks that are “believed to have been part of the platform of a double canoe about 24 meters long.” Among other small but vital finds

The First Pacific Seafarers

Map 2.1. Cultural areas and prevailing winds.

(p.26) (p.27) were hand clubs similar to those being produced in Aotearoa at the time of Cook.12

The Pacific oceangoing ships took many years to build, often on an island that was within a group with suitable timber and specialized in shipbuilding. The whole community might be involved in divisions of labor, including carpentry, caulking, sail making, and cordage manufacturing. Food to provide for this specialized labor force would sometimes be brought from other islands.13 Such ships were undoubtedly the most advanced achievement of the Neolithic Pacific.14 They were highly valued and prestigious, had spiritual connotations, and were venerated and protected by ceremony and taboos. Many were also objects of art and had names that were both functional and poetic, such as the warship Rusaivanua (Fijian: destroyer of the land).

In summary, Micronesian vessels were built from planks and had single hulls, outriggers, and lateen sails. As the outrigger had to be kept to windward, coming about required the sail to be shifted fore to aft. In Fiji and Western Polynesia, long-distance vessels were mainly double hulled, with lateen sails, which likewise had to be shifted to come about. In Eastern Polynesia, including Tahiti and Hawai‘i there were double hulls, but with sprit sails on fixed masts, so that tacking was achieved by bringing the bow round to the wind. Melanesia is characterized by both doublehulled outriggers (east coast of Papua) and lateen and sprit sails of various forms. These and other differences in vessel technology were responses to available materials and wind, sea and coastal conditions, and to diffusion and adaptations of successful solutions for sailing and survival—all reinforced over generations by taught skills and memorized rules and rituals.

The ship construction techniques, all without metal, relied on the use of fire, wooden mallets, and adzes of basalt and obsidian, either local or obtained by trade. Other tools included adzes from the shells of giant clams, shell scrapers, shark-tooth- and coral-tipped drills, and sharkskin and sometimes pumice for smoothing the hulls. The vessels had to be big enough to carry viable family groups and broad for stability, with minimal draught for reefs and lagoons. They needed space for the crew, passengers, livestock, tools, weapons, plants, seeds, fishing gear, spare sails and spars, and possibly an alternative mast, as well as food and water for the voyage.

A vital first stage in ship construction was to choose the correct wood for various components of the vessel. All the parts had to respond intrinsically to incessant wave motions, sudden winds, and shocks of periodic heavy pounding. The natural shapes of trunks and branches were vital. (p.28) Mifflin Thomas describes the long process in Hawai‘i of selecting trees for hull timbers. The party would be guided by the kahuna kalai wa‘a, the specialized builder, who would follow the elepaio (a flycatcher species) in the forest over many days and observe its tree-pecking patterns until its behavior signaled a tree in the best condition. The forest god would then be placated by a pig sacrifice before the wood was transferred from a land to sea domain.15

John Twyning of the whaling brig Minerva (captained by Thomas Lewis), which sailed from Sydney in 1829 and was wrecked in Fiji, describes the process of building a large ship at Lakeba in the Lau Islands, where he and others in the crew were given refuge. After a tree of the hardwood vesi was felled under the directions of the Matia shipbuilder, it was split in the middle and each half reduced to the proper thickness. It was, Twyning wrote, “surprising to see the accuracy with which these planks are joined to each other; a piece of very thin tapa cloth with a kind of gum, made from the bread-fruit tree and laid between the two pieces to be joined, after which they are sewn together with threads of sennit, and made perfectly tight by small wedges driven between the threads and the wood.” He went on to say that the design and building of the ship would have received “the admiration of even the most skillful and scientific naval architect in Europe.”16

The ship seen by Twyning was most probably the Fijian drua, the building of which was usually directed by Tongans. Over one hundred feet in length, the drua had two hulls with length differences for hydrodynamic purposes, set about seven feet apart, over which an athwart-ship deck was laid. It had hatches for entry to the hulls, a deckhouse, and a platform erected above this from which the captain gave orders to the crew handling the sail and the massive steering oar. The single sloping mast of the drua, made from strong, flexible wood, stood about fifty feet above the deck and supported a massive lateen-type sail, contained between two flexible yards extending over the whole length of the ship. J. Glen Wilson of HMS Herald painted a scene of a fleet of drua sailing off Levuka in 1855.17

The Tongans and Samoans had vessels somewhat similar to the drua, including the tongiaki, which had two hulls of equal length, a sail like the drua’s, and two massive steering oars but was considered more difficult to handle in bad weather. A painting from the Schouten expedition shows a double-hulled vessel west of the Tuamotus in 1616 (figure 2.1). An impressive Tongan craft for speed was the kalia. Thomas West describes his 1865 trip on a kalia:

(p.29) Up went the huge sail, down went the great steer-oars splashing into the sea, and away we shot like a race-horse. … Every timber of the canoe creaked again; while the mast bent like a reed, and cracked in its socket as if it would split the deck in two. … [T]he sea was like a hissing cauldron on either side of our course, and the kalia, instead of having to mount over the smaller waves, cut its way right through them.18

In Hawai‘i the wa‘a kaulua, a particularly fast and elegant ship, was double hulled and had crab-claw sprit sails and steering paddles.19 The pahi of the Society Islands was unlike the other indigenous vessels in Polynesia. The pahi had a near-European type of keel, ribs, and knees and was full bellied, and its stem and stern posts were high and elaborately carved. It had two masts and unique sails. Robert Langdon, in his Lost Caravel, sees this vessel as possibly an adaptation of technology from sixteenth-and seventeenth-century Spanish ships.20

The First Pacific Seafarers

Figure 2.1. The Le Maire and Schouten voyage provided inspiration for the first image of a double-hulled canoe. The incident depicted in this painting took place off the Tuamotu Islands between 9 and 13 May 1616. The Dutch fired a warning shot from the boat of their ship, the Eendracht, but the crew of the canoe refused to stop. The Dutch then fired on the vessel.

(Courtesy of the Pepys Library, Magdalene College, Cambridge)

(p.30) In Micronesia the Kiribati baurua was a large, fast, slender vessel with an outrigger. The hull was more V-shaped than other ships in the Pacific, reducing the effects of leeway, and was also asymmetrically curved to compensate for the drag of the large outrigger.21 People of the Marshalls and Carolines also built outrigger craft with lateen-type sails. These were characterized by a lee platform on which the captain stood to direct sailing and steering; in some atolls these vessels were kept small because of timber shortages and difficult reef passages. The vessels of western Melanesia such as the puka of Santa Cruz were outriggers and used the crab-claw sail.22 On the New Guinea coast the Motu trading lakatoi was double hulled and had similar sails.23

Only a few types of large sailing craft were noted in Aotearoa by the first Europeans.24 A general view is that after possibly multiple voyages from and to the original homeland of Hawaiki, the Maori Polynesians started to build mainly single-hulled vessels from giant kauri logs.25 These waka tana were ornately carved, powered mainly by paddles, and used for coastal trade, and the waka taua, with its high stern and elaborate carvings, were used for warfare and were propelled by a hundred paddlers.

Weather and Climate Variables

The planetary wind systems of the Pacific are generally predictable in direction and force. These northeast and southeast trades blow from about 30° N and 30° S toward the equator, vary a few degrees in latitude over winter and summer, and are most consistent in the remote easterly zones. In higher latitudes beyond a zone of variables, the westerlies are on average fairly constant (see map 2.1), and hurricanes in the tropical latitudes have well-defined seasons and prevail in the westerly zones. In detail there are periodic and reasonably predictable departures from the prevailing planetary wind systems in several areas, which vessels on west-to-east voyages across the Pacific could take advantage of. The ocean currents in the main migratory zones would, however, often be adverse (map 2.2). Navigators in the eastward voyage from Samoa to Tahiti in the Hokule‘a wanted to test this replica of an ancient Polynesian voyaging canoe by driving close-hauled in the southeast trade winds. The result surprised them: “Instead of the anticipated hard struggle to reach Tahiti, an embarrassment of favourable northwesterly winds sped the Hokule‘a eastwards.” In his account of the passage, Ben Finney says of the old voyagers, “Seafarers had to learn how to exploit spells of westerly winds to

The First Pacific Seafarers

Map 2.2. Main ocean currents.

(p.31) (p.32) keep pressing eastwards against the direction from which the trade winds often but not always blow.”26

In Kiribati, predictions of winds were guided by the star calendar. Stars rise four minutes earlier each day, so that a rising morning star becomes an evening setting star in six months. When Antares (Rimwimata) appeared over the eastern horizon just after sunset, it marked the beginning of the best distant-sailing season (June). When it was replaced by the rising of the Pleiades (Nei Auti), dangerous westerlies could be expected (November–December).27 In latitudes farther south the hurricane season was also often avoided. Hocart notes that canoes in Fiji were beached during December to March.28 But this period of summer was, and is, often preferred for local voyages. It is called draki vinaka, when there are long periods of fine weather more comfortable than the boisterous winter season, which can be cold and wet on deck before dawn. Decisions about voyaging were a matter of keeping a weather eye open for signs of hurricanes.

The weather was not the only variable. There were changes in climate during the long periods of human migrations. These are recognizable in detail in the latter part of the migratory era.29 During the final migrations of the Polynesians, which took them to Aotearoa, there is evidence that the weather was favorable. This period from AD 800 to 1250 is identified as the “little climatic optimum,” with clear skies and fewer storms, which may have allowed seafarers from around the Cook Islands to extend their exploratory voyages beyond the trade wind zones, sail through the zones of variable winds, and work through the barrier of what are now strong westerlies to reach Aotearoa. Similarly this more benign climatic period extending into the thirteenth century may have allowed long-distance two-way voyages to take place. Later climatic change into the fifteenth century may account for deteriorations in weather, especially in the higher latitudes, that seem to have brought two-way links between Aotearoa and the original homelands to a close.30


The designs and materials of hull, sails, rigging, and steering gear were the strengths and maneuvering assets of Pacific vessels. This system was successful only when competent seafarers were in the loop. The ship had to bear variable loads and respond to a wide range of conditions at sea and on the coast. She would roll and pitch in a seaway and be subject to squalls and sudden shifts in wind force and direction. The vessel had to (p.33) survive severe storms without capsizing, foundering, or suffering irreparable damage. This meant expert seamanship was required.

The captain of an oceangoing Pacific vessel would undoubtedly be an experienced seaman and have status and authority. He would probably have selected the crew and decided on the time of sailing from weather and astronomical observations. As the principal navigator, the captain would determine positions and courses during the voyage and maintain discipline and morale. As with all seagoing vessels, a captain of a Pacific craft would for safety reasons have been regarded as the equivalent of “master under God,” with no one reigning above him while at sea.

The seamanship tasks would start from ensuring correct loading of the vessel for its stability and trim. Passengers would be distributed, livestock penned, and ship-handling areas kept clear. Among those on board a range of skills would be found in addition to those of the sailors. Possibly they would include fishermen, sailmakers, carpenters, divers, and specialists in traditional medicines, who could all contribute to safety.

Long west-to-east voyages, frequently into the prevailing trades, would inevitably involve extended periods of exhausting tacking or reversing lateens, as these vessels could sail no closer than about 75 degrees to the wind. When running free with a strong following wind and sea, a heavily laden double-hulled vessel required competent handling in adjusting to combinations of wind and sea. Twyning observes that the sea has to be kept on the quarter, for if a high wave was allowed to run between the two hulls, it could part them.31 Scudding before a gale could also make any of these craft difficult to control, and there was the danger of being overtaken by waves and pooping a heavy sea, which could wash over the length of the vessel. It is likely that experienced seamen would in such weather pay out a long line with buoyant material, such as wooden spars attached to act as a sea anchor, and ride out the storm or drag it astern to slow down, but there is no evidence of this in the Pacific. Some vessels carried a heavy stone on a rope for anchoring. This could be hung in the water over the bow, then weights shifted aft and the vessel kept nearly head-on to the sea, assisted by expert use of a steering oar.32 Morrison also observed in Tonga that “when taken by a squall they luff head on to it and shake it out—if long they jump overboard and hang her head to windward till the squall is over.” He added that bringing the sail down on very big vessels could be dangerous, but they carried plenty of cordage and masts to repair damage.33

Calm weather could bring other problems, particularly where strong (p.34) ocean currents were encountered (see map 2.2). The equatorial current can set in a westerly direction at thirty to forty miles per day, and under fresh trade winds at about three and a half knots. The easterly countercurrent has a rate that reaches over one knot. These currents vary seasonally, with the equatorial countercurrent extending just south of the equator in June and July. Captain G. H. Heyen, who commanded the brigantine Alexa, the last sailing vessel to operate regularly out of Sydney to the Pacific Islands in 1929, recalls becoming becalmed twenty miles west of Tarawa on the fifty-fifth day out of Sydney and drifting away; the Alexa did not reach the Gilberts again for another one hundred days.34


The systems of determining direction and position finding varied in detail in different parts of the Pacific. There were many common features, but most of the local knowledge was closely guarded and passed on through hereditary channels. This process also included formal training onshore and long periods at sea under the instruction of leading navigators. Father Ernest Sabatier describes how the maneaba, the great social meeting house in the islands of Kiribati, was built north-south and was used for teaching astronomy and navigation. The ridgepole of the maneaba represented the meridian, the thatch between the parallel rafters conceptually contained various constellations and positions of individual stars in their diurnal transits throughout a year, and the eaves marked the horizon.35 Grimble describes how outside under the changing night sky, the uma ni borau (roof of voyaging) would be the main school where names and passages of stars were identified and memorized, visualizing them within lines of rafters. Of particular significance would be the succession of “guiding stars,” which rose and set in the direction of known islands. The navigator initiates also learned those zenith stars that reached their maximum altitude in the vicinity of the home island and other islands known to them.36

The Jesuit missionary Father Juan Antonio Cantova, writing in 1722, says of navigational training in the Caroline Islands that “young men received lessons in practical astronomy and navigation”;37 and Gladwin, writing of the same islands in 1970, records: “Formal instruction begins on land. It demands that great masses of factual information be committed to memory. This information is detailed, specific, and potentially of life-or-death importance. … Often they sit together in the canoe house. … The pebbles usually represent stars, but they are also used to illustrate (p.35) islands.”38 The Marshall Islands stick charts are yet another type of teaching device. These simulate how swells bend and interact with patterns of distant islands—represented by cowry shells—below the horizon, thereby allowing navigators to establish approximate positions by the nodes of swells and courses to be adjusted.39 It was not until the development of satellite imagery that these intricate long-distance swell patterns used by Pacific island voyagers (and to a more limited extent Shetland fishermen) could be fully appreciated.40

The process of navigating would begin on departure by setting a course for a known group of islands on the appropriate wind. At the low island of Arorae in Kiribati there are transit stones providing initial departure directions to specific islands. Transit stones have been identified also on Atiu, giving the direction to Rarotonga. By keeping the leading stones in sight from stern observations, a navigator could maintain a course and estimate the current set and drift. When the marks went below the horizon and the guiding stars appeared, the course could be adjusted to those stars, offset from the destination to allow for current and leeway. The sky was now one big compass, and the feel of the wind could also act in this way.41 The departure from a high island is similarly described by Raymond Firth. The mountain peak known as Te Uru Asia, at one thousand feet above sea level on Tikopia, is the highest of five marks that, by relating vertical and horizontal scales, allow a series of first estimates of distance run.42

Gladwin details a voyage process of departure to arrival using the concept of etak. The disappearance of the island shore marks would show the end of the “etak of sight,” the second zone would be marked by the “etak of birds.” The third would be a series of etak distances in the open ocean. Then the navigator has guiding stars ahead in line with the destination, while another reference island, real or imaginary, was offset from the course. He would now visualize himself as stationary, and as successive stars came in transit with the reference island, the island of destination would move closer. Gladwin analyzes this system in detail and describes as a system for conceptually bringing together raw information and converting it into a solution of the essential navigational question, how far is our destination?43

Guiding stars—the sun and those stars close to the cardinal points of north (polestar), south (Southern Cross), east and west (Orion’s belt and Altair), the latter two on rising and setting—plus allowances for currents and leeway, would enable skilled navigators to keep courses with some certainty throughout a voyage. There is also significant evidence that the (p.36) Pacific navigators could establish approximate latitudes in relation to the wide spread of an island group to which they were heading, providing they knew the zenith star of or close to these islands. When, for example (in the present period), Sirius is directly overhead, its approximate declination of 17° S corresponds to the latitude of a navigator at 17° S. This includes the south of Vanua Levu in Fiji. Similarly, Arcturus (declination 19° N), when reaching its zenith directly above the navigator, indicates that the vessel was in the latitude of the southernmost island of Hawai‘i. The navigator of old would not know with certainty if he was east or west of the destination, because he had no way of establishing longitude. However, having made a directional allowance on departure and during the voyage for a windward approach to the destination, he could, when in the latitude of the islands, run downwind to pick up islands and would be guided in this by seabirds and other seamarks.

These indigenous systems of navigation have been well tested by David Lewis.44 They were undoubtedly no less accurate than the European methods used before the mid-eighteenth century, when longitude at sea could not be established with certainty. This problem came to a head in 1707 when four British warships were wrecked on the Isles of Scilly, with the loss of two thousand men. The ships were sailing in reduced visibility with an estimated longitude by dead reckoning, which put them farther west than they actually were.45 One of the purposes of Cook’s second voyage to the Pacific in 1772–1775 was to test the accuracy of Harrison’s timepiece for determining longitude by chronometer.

By this period of the eighteenth century, Europeans had some appreciation of the possible competent navigational skills of the Pacific islanders, primarily on the basis of their detailed geographical knowledge. Tupaia, the Ra‘iatean priest-navigator who came from a family of famous seamen, proved invaluable to Cook.46 In 1769 Tupaia named and gave sailing directions from Tahiti (the center of his world) to Fiji and to virtually every island group of Polynesia for more than one thousand miles north-south and three thousand miles east-west, with the exception of Hawai‘i, Rapa Nui, and Aotearoa. Similarly, Cantova in 1722 was able to draw a map of Micronesia, based on information from local informants, that, Hezel observes, “reveals a surprisingly complete knowledge of an island world that, stretching over fifteen hundred miles of ocean as it did, must be considered large even for a seafaring people.”47 This map remained the best available of the region into the nineteenth century. The intellectual process and remarkable feats of memory involved in navigation are all (p.37) the more impressive when one recalls that Pacific mariners had no written language and no mechanical timepieces.

Cook, during his first voyage, expressed admiration for these Pacific navigators, although he probably could not find the time to understand their systems in detail. On his return to the eighteenth-century society of land-based astronomers in England, this very practical navigator deferred to their views; as G. S. Parsonson expresses it, the English astronomers found it difficult to accept “that ignorant savages had long ago achieved the mastery of the sea which navigators in the old world had scarcely yet attained.”48

This account of navigational skills has referred only to men. There are very few references in the Pacific (or other regions) of women in charge of navigating ships. Women have led trading parties on voyages, and often canoes with female crews can be seen on short passages in archipelagos. But, as indicated in chapter 1, on deep-sea vessels there were, and some-times still are, prejudices and taboos. An oral account of a woman navigator was given to Father Sabatier in Kiribati during the 1930s.49 It is also an interesting recollection of one of many minor wars between islands in the 1780s, and how the conflict between Abemama and Tarawa was amicably settled for a time in the traditional way. The account contains some ambiguities, as well as conflicts between gender, status, and skills. He entitled the account, “La vengeance d’une femme,” which can be summarized as follows:

A Tarawa fleet attacked Abemama. It was beaten, and an Abemama fleet sailed to Tarawa to reinforce its victory. Tarawa paid tribute of feasts and gifts to Abemama. The Abemama fleet was led by a “general,” a “soothsayer,” and a “navigator.” The latter was Baintabu, a daughter from a navigator family without sons. At the feast she was ignored and did not receive a share of the gifts. On the seventy-mile return voyage to Abemama, Baintabu “sulked” and lay at the bottom of the leading boat, giving no instructions for tacking. She was eventually rolled up in her mat and thrown overboard. Fortunately for Baintabu, the last vessel picked her up. She navigated the boat to Abemama, and the other vessels were never seen again.

The questions is, was she “sulking,” or was she actually on her back, feeling the swell or looking for the zenith star? At the Tarawa feast and on board the vessel, however, she certainly did not have the prestige ordinarily (p.38) accorded to a navigator. There is another rare example of a woman navigator recorded about 1900. She was called Libe and was a teacher in the Ebon school of navigation in the Marshall Islands.50 This is quite possible, for in Britain during the early nineteenth century some wives of captains ran private navigation courses.


There is no doubt that skilled navigation, while prestigious, was not hard to come by on the long voyages by Pacific seafarers. Lewis concludes, “The special problems of the longer journey concern such factors as food supply, manpower, motivation, and strength of the vessel—not navigation.”51 Passage could be extended in time, and with fifty or so people on board, food supplies could run out. The voyagers would then have to depend on food from the sea. This would not have been a major problem for Pacific island seafarers, with their detailed knowledge of the marine environment.

In contrast to Pacific people, some of the European sailors who came to the Pacific were too far removed in perception and suitable marine technology to exploit the diversity of food resources available from the ocean. When, for example, stores ran out on Magellan’s ship Trinidad in 1521, the crew arrived at Guam ill with scurvy and starving, having been reduced to chewing cowhide and eating the last of the rats during the ninety-nine day crossing.52 On the 1740–1744 voyage to the Pacific led by Commander George Anson, some 1,400 seamen out of 1,900 on his squadron of six ships died mainly of “disease and starvation.” Unfortunately for Anson, who was a competent and caring naval officer, the seafarers that the British Admiralty provided him were of poor quality, and the northeast trade winds he had relied on to carry them across the Pacific proved to be very weak in the low latitudes he followed in May, thereby greatly increasing the length of passage.53

William Bligh was a little more successful in obtaining food in 1789, when he made a remarkable voyage of 3,500 miles in forty-eight days while in charge of eighteen men on a seven-meter launch with little free-board. These castaways from the Bounty set off with food and water for about only five days in normal conditions. They managed to make a brief stop at nearby Tofua in the Tongan group and obtained some breadfruit and coconuts before one sailor was killed and they were driven off by hostile islanders. Bligh then safely navigated to the northern barrier reef of Australia en route to Timor, and the men were able to obtain large quantities (p.39) of mollusks. Otherwise the boatload was able to collect rainwater and live on boobies, seabirds that they caught by hand. Bligh reported seeing plenty of fish, and although they had hand fishing lines, they did not have the skills.54

Even less successful in terms of living off the sea were the survivors of the whaleship Essex in 1821. The ship sank, and the three whaleboats eventually lost company with one another. When their meager supplies were exhausted, the castaways did what seafarers in lifeboats have done before and since: they agreed to cast lots and kill and eat one another in turn, in the hope that at least some would survive to be rescued or reach land. In the event that only one sacrifice was needed before they were rescued, the other who was eaten died of general weakness.55

If Pacific islanders had to resort to cannibalism at sea, which they did on other occasions on land, they would have been more conserving of related people by using only parts of someone who died, or of a voluntary victim, as bait to catch fish and birds. In practice, they would have lived off the sea. There are many contemporary accounts of islanders surviving very long voyages after losing their vessel or being carried away on small craft by the equatorial current. In June 2001, for example, four Samoan seamen lost their fishing craft northwest of Samoa. They took to the ship’s boat without adequate provisions for a long voyage. Two of them survived and arrived at Normanby Island on the northeast of mainland New Guinea five months later, after drifting about 2,500 miles. The survivors said that they lived by catching fish and birds and drinking rainwater.56

The Pacific Ocean beyond the island fringes is not in fact very nutrient rich. However, all island mariners would recognize certain signs at sea that meant sustenance. Flurries of flying fish might indicate schools of pursuing skipjack tuna. Behavior of tuna might signal the existence of predatory sharks. The flight paths of certain birds would be known, and their behavior would indicate shoals of specific types of fish. It was also possible to use floating baited lines to catch birds that spend most of their lives at sea. The expert fishermen on the ships would be well equipped with spears, fishhooks (of shell, bone, and wood), dip nets, and trolling and deep-water lines for catching a wide range of fish, as well as torches for attracting squid and flying fish at night. They possessed, as they still do, a profound practical knowledge of marine ecology and fish behavior. They were aware of what is now termed fish aggregation devices (FADs). Their stationary wooden vessel would act as a FAD, and fish would congregate around and under it. They also knew how to use sound and vibration to attract fish from long distances. By beating rhythmically on the sides of (p.40) the hull or using submerged coconut-shell rattles, they could attract certain species, including sharks. Pacific fishermen were using such methods centuries before modern hydrophone experimentation confirmed that fish communicate by sound and vibration over very long distances.

As fishermen, most Pacific islanders do not share the European fear and neurosis about sharks. They respect the shark, and they understand the habits of different shark species. They know when to remove them-selves from aggressive or mobbing sharks but also recognize other forms of shark behavior that mean the islanders can safely share the sea with them. After killing a shark, fishermen in some of the islands of west Melanesia still release the spirit of the shark by sounding notes on a conch shell. Sharks would have been difficult to catch in early times with coconut sennit lines, which could be bitten through, resulting in the loss of precious shell hooks. R. E. Johannes has demonstrated a system whereby a noose is slipped over the head and gills of a shark, and then the big fish is speared and clubbed as it is held alongside the canoe. Johannes notes, “The Palauan fisherman had a special tattoo on his wrist. When holding out a flying fish to entice a shark to swim through the noose, he was not supposed to let go of the bait until the tip of the shark’s snout reached the tattoo.”57

The Pacific island migrants would not have starved on voyages led by experienced mariners. Fresh water might have been more of a problem, and on every crowded vessel during a long passage there would have been, as the British Admiralty terms them, the “sick and hurt.” Greg Dening is realistically cautious of the view that “before the coming of the white man they [the islanders] had died only of old age, war and sorcery, that all the diseases that killed them came with the ships.”58 Undoubtedly the major epidemics and sexually transmitted diseases came from visiting ships, but on long voyages, as elsewhere, islanders may have suffered from intestinal, respiratory, muscular, skin, and skeletal problems, as well as accidents.

Ships would have carried stocks of island medicines, including those based on plants, possibly kava (Piper methysticum), seaweeds, coconut toddy and milk, and fermented pandanus and breadfruit would be taken or applied, as would the healing and cleansing properties of seawater and the ministrations of shaman healers.59 Fresh water would no doubt be rationed; however, the vast pandanus sails of the ships could collect rain from periodic heavy squalls to supplement that carried in gourds and pottery containers. Other foods on board would be similar to those customarily stored on many islands in Kiribati: green nuts, pandanus flour (p.41) (te kabubu), preserved pandanus fruit (te tuae), coconut toddy syrup (te kamaima), pounded breadfruit and taro (kept in leaves), dried shellfish, live fowls, pigs, and rats. The rapid reproductive capacities of the rats would generate many young, providing a regular source of protein.

Arrival at the Beach

Having survived the ocean leg of a voyage, the flotilla faced other dangers before reaching the beach. All seafarers become more vigilant when approaching even a well-charted coast. When they sailed into an unknown but very likely reef-strewn archipelagic region, the dangers could be immense. Signs of the approach to land would be observed long before it was sighted. These would include the types and directions of the flight of birds, the characteristics of ocean swells and clouds, and the occurrence of floating vegetation. Night sailing would now become particularly hazardous. Most high islands have areas of fringing reefs along their coast, and several are characterized by offshore barrier reefs enclosing lateral lagoons, although some, like the Marquesas, present primarily beaches and cliffs direct to the sea. Atolls comprise a series of low islands (motus) standing on wide reefs on their weather side. To leeward they have a large, broad reef flat enclosed by barrier reef, within which is the lagoon. Low reef islands have only coastal fringing reefs and no lagoons. Dangers are encountered in the vicinity of all these islands.

The weather side of atolls is steep-to, and breakers can be seen and heard a mile or so offshore. Likewise, the reflection of lagoon waters on the base of cumulus clouds can sometimes be seen before the barrier reefs are sighted. If a vessel finds itself too close in on the weather side, and there are strong onshore winds and currents, it may be difficult to clear the long line of the barrier reef and avoid stranding. On the lee side of atolls the waters can be even more dangerous, with numerous sunken reefs extending seaward for many miles from the shore. The prudent navigator would lay off an island until the sun was up and behind him if possible. Then, with a masthead lookout, he would avoid the yellow and brown areas and follow the deep blue water toward the reef.

On the final part of the approach, deep and wide reef passes would have to be found to enter lagoons. On high islands, breaks in the reef occur on any side of the island where rivers carry fresh water and sediments, which inhibit coral growth. On low islands, reef passes into lagoons lie on the lee sides; through which water driven across the weather reefs circulates in the lagoon and flows outward through these passes. The currents in such (p.42) passes vary in strength with the state of tide and wind. Depths also vary, but there can be shallow patches and reef spurs. Cook in 1773 nearly lost HMS Resolution in a reef pass,60 and Tupaia in 1769 sent island divers down to determine clearance under the rudder on several occasions.

Clearly, in meeting the challenges of finding their way over vast distances to and from small islands, the Pacific islanders of the remote past were in practice quite close to European mariners in the application of some basic principles of seamanship. In navigation the latitude by zenith star was not far short of a latitude by meridian altitude, which was the method of the European seamen using a simple measuring instrument by the sixteenth century. The bearing of celestial bodies for course setting is still in a sense the ultimate safeguard of the modern navigator. He relies on a gyro compass, but the prudent navigator checks this mechanical device against the magnetic compass. In turn, errors of the magnetic compass are established by observing the azimuth of a star. (p.43)


(1.) Peter Bellwood, “Footsteps from Asia: The Peopling of the Pacific,” in The Pacific Islands: An Encyclopedia, ed. Brij V. Lal and Kate Fortune (Honolulu: University of Hawai‘i Press, 2000), 53–58.

(2.) Geoffrey Irwin, The Prehistoric Exploration and Colonisation of the Pacific (Cambridge: Cambridge University Press, 1996), 19–30.

(3.) Ibid., 23.

(4.) Patrick Vinton Kirch, On the Road of the Winds (Berkeley: University of California Press, 2002).

(5.) Ben R. Finney, Voyages of Rediscovery (Berkeley: University of California Press, 1994), 182. About two-thirds of the way on the voyage from Rarotonga to Aotearoa, he sighted the high Kermadec Islands, where there is evidence of early Polynesian occupation.

(p.213) (6.) Kirch, On the Road of the Winds, 207, puts the date of colonization as AD 1000 to 1200.

(8.) Robert Langdon, “The Bamboo Raft as a Key to the Introduction of the Sweet Potato in Prehistoric Polynesia,” Journal of Pacific History 36 (2001): 32–58. See also Thor Heyerdahl, The Kon-Tiki Expedition (London: George Allen and Unwin, 1951). For Micronesia, see Kirch, On the Road of the Winds, 165–178.

(9.) S. S. Barnes, E. Matisoo-Smith, and T. L. Hunt, “Ancient DNA in the Pacific Rat (Rattus exulans) from Rapa Nui,” Journal of Archeological Science 33 (2006): 1536–1540.

(10.) Dumont d’Urville, French scientist and naval officer, sailed widely in the Pacific from 1822 to 1825 and returned in 1826 until 1829. The name “Polynesia” was first used by the French explorer Charles de Brosses in 1756, and the three-way division by D’Urville in 1831. See Lal and Fortune, Pacific Islands, 63–159.

(11.) A. Grimble, “Canoes of the Gilbert Islands,” Journal of the Royal Anthropological Institute 54 (1924): 101–139.

(12.) Peter Bellwood, The Polynesians: Prehistory of an Island People (London: Thames and Hudson, 1987), 20.

(13.) John Young, “The Response of Lau to Foreign Contact,” Journal of Pacific History 17 (1982): 29–50.

(14.) A. C. Haddon and James Hornell, Canoes of Oceania, 3 vols., Bernice P. Bishop Museum Special Publications 27–29 (Honolulu, 1936–1938).

(16.) John P. Twyning, An Account of the Life and Adventures of John Payer Twyning: Comprising the Wreck of the Minerva and the Author’s Years in Fiji and the Friendly Isles, 2nd ed. (Bristol: for the benefit of the author, 1850), 161.

(17.) The painting is in the collection of the Fiji Museum, Suva.

(18.) G. S. Parsonson, “The Settlement of Oceania,” in Polynesian Navigation, ed. Jack Golson (Wellington, New Zealand: Polynesian Society, 1965), 38.

(19.) Mifflin Thomas, in Schooner from Windward, describes Hawaiian vessels at the time of Cook, although not specifically the wa‘a kaulua.

(20.) Robert Langdon, The Lost Caravel (Sydney: Pacific Publications, 1975), 108–110.

(22.) Haddon and Hornell, Canoes of Melanesia, 2:8.

(23.) Clive Moore, “Hiri Trading Voyages,” in Lal and Fortune, Pacific Islands, 131.

(24.) Philip Edwards, ed., The Journals of Captain James Cook (1769; reprint, London: Penguin Classics, 1999), 88. See also Anne Salmond (p.214) Two Worlds: First Meeting between Maori and Europeans 1642–1772 (Honolulu: University of Hawai‘i Press, 1998).

(26.) Finney, Voyages of Rediscovery, 159–161. The Hokule‘a in 2004 was still sailing to each of the islands of Hawai‘i, traveling more than four thousand miles in support of coral reef conservation (Honolulu Advertiser, 3 May 2004).

(27.) Ernest Sabatier, Sous l’equateur du Pacifique: Les iles Gilbert et la Mission Catholique (Paris: Edition Dillen, 1939); and Grimble, “Gilbertese Astronomy,” 200.

(29.) Patrick D. Nunn, “Illuminating Sea-Level Fall around AD 1200–1510 in the Pacific Islands: Implications for Environmental Change and Cultural Transformation,” New Zealand Geographer 56 (1993): 46–53.

(30.) Patrick D. Nunn, “Facts, Fallacies and the Future in the Island Pacific,” in A New Oceania: Rediscovering Our Sea of Islands, ed. Eric Waddell, Vijay Naidu, and Epeli Hau‘ofa (Suva: University of the South Pacific, 1993), 114. See also David Lewis, We, the Navigators (Canberra: Australian National University Press, 1972), 89–90.

(32.) Paul Johnstone, The Sea Craft of Prehistory (London: Routledge, 1980), 210.

(34.) G. H. Heyen, pers. comm., 1963.

(37.) Francis X. Hezel, The First Taint of Civilization (Honolulu: University of Hawai‘i Press, 1983), 53.

(38.) Thomas Gladwin, East Is a Big Bird: Navigation and Logic in Puluwat Atoll (Cambridge, MA: Harvard University Press, 1970), 128.

(40.) A. D. Couper, “Seasat Images,” in Times Atlas and Encyclopedia of the Sea, ed. A. D. Couper (London: Times Books, 1983), 205.

(42.) Raymond Firth, We, the Tikopia (Boston: Beacon Press, 1936), 27.

(44.) A. D. Couper, “Pacific Seafarers in Trade and Navigation,” in Localization and Orientation in Biology and Engineering, ed. D. Varjú and H. U. Schnitzler (Berlin: Springer-Verlag, 1984), 227–243.

(45.) Dava Sobel, Longitude (London: Fourth Estate, 1995), 11–13.

(46.) J. C. Beaglehole, The Life of Captain James Cook (London: Adam and Charles Black, 1774), 192.

(47.) Francis X. Hezel and Maria Teresa del Valle, “Early European Contact with the Western Carolines 1525–1750,” Journal of Pacific History 7 (1972): 26–44.

(51.) David Lewis, “Polynesian and Micronesian Navigation Techniques,” Journal of the Institute of Navigation 23, no. 4 (1970): 432–447.

(52.) O. H. K. Spate, The Spanish Lake (Canberra: Australian National University Press, 1979), 47. See also William Bligh’s journal, The Mutiny on Board HMS Bounty (reprint, New York: Airmont, 1965).

(53.) Glyn Williams, preface to The Prize of All the Oceans (London: Harper Collins, 1999), and 134.

(56.) Kathy Marks, “Fishermen Rescued after Five Months Lost at Sea,” Independent, (London), 13 November 2001.

(59.) Kate Fortune, “Traditional Healing Practices,” in Lal and Fortune, Pacific Islands, 443.