23. Rotuman Seafaring in Historical Perspective
[Published in Seafaring in the Contemporary Pacific Islands, edited by Richard Feinberg. DeKalb: Northern Illinois University Press, 1995.]
Rotuma is one of the more isolated islands in western Polynesia. It is located at 12º 30'S, 177ºE, some 465 km northwest of Cikobia, the northernmost island in the Fiji group (Woodhall 1987). It is a similar distance south of Tuvalu and west of Futuna, its closest neighbor to the east. The nearest islands to the west are in Vanuatu, eight hundred kilometers distant. Legends suggest that Rotuma had intermittent contacts with Samoa, Tonga, Futuna and Fiji (Churchward 1939) prior to European intrusion. Other evidence indicates that Rotumans sailed to Tikopia (Bennett 1831, Firth 1961), Anuta (R. Feinberg, personal communication), Kosrae (J. Tobin, personal communication) and possibly even to Raiatea in the Society Islands (Anonymous 1938). Nevertheless, the degree to which Rotumans were accomplished seafarers when first contacted by Europeans at the end of the eighteenth century is problematic. Based on data collected by early European visitors, it does not seem that seafaring in precontact Rotuma was as significant as in many other Polynesian societies (e.g., Nukumanu, see Feinberg this volume; for Sikiana see Donner this volume).
Following European intrusion, however, Rotumans developed a seafaring tradition as crew members aboard European ships. Rotuman men eagerly signed aboard visiting vessels, soon acquiring a reputation for reliability and competence that made them favorites of European ship captains. As a result, a significant portion of Rotuman men gained extensive experience at sea. This had, and continues to have, a distinct effect on the culture and economy of the island. Sailors have been a major source of information about the outside world, thereby introducing and legitimating cultural changes, and have contributed considerably to a standard of living that is extraordinarily high for an isolated Pacific island.
Since we are dealing with a heritage transformed from an earlier, precontact formation to one that has been shaped by European seafaring traditions, it is necessary to approach our subject historically. I begin by presenting background information on Rotuma and its environs, then I describe seafaring technology and traditions shortly after contact. Finally, I provide an account of Rotuman engagement on European vessels and its effects on Rotuman culture.
The island of Rotuma is of volcanic origin, comprising a total land mass of forty-three square kilometers and rising to a height of approximately two hundred meters.  It is surrounded by a fringing coral reef that attains a maximum width of 1.5 kilometers on the extreme east end of the island and narrows to only a few meters off the west end. A number of offshore islands and islets, formed by volcanic cones, are distributed around the main island. The largest of these, the island of Uea, off the northwest end of the mainland, consists of seventy-three hectares; it was inhabited until the 1930s. Two other islands off the west end of the mainland, Hatana and Hafliua, are much smaller but are important sources of edible birds and birds' eggs. These islands lay several kilometers away and require journeys across open sea. Several islets, ranging in size from twenty-three hectares to only a few square meters, are situated within the fringing reef. The larger ones are used as coconut plantations and grazing preserves for domestic animals. They can be reached by walking across the reef at low tide, or by canoe, punt, and launch when the tide is high.
Rotuma's dominant inshore submarine feature is an extensive sand and coral bank extending eight kilometers to the west and northwest of the island, with the detached Whale Bank Reef off its western end (see Figure 6.1). Whale Bank Reef is approximately 5 by 1.6 kilometers, with an average depth of seventeen to eighteen fathoms. A current, setting west-southwest at speeds of less than one knot, occurs most of the year (Fisheries Division n.d.:3).
Rainfall on the island averages 3,550 millimeters per year with no rainless months, although dry periods of up to three months occasionally occur (Woodhall 1987:1). The soil is very fertile, giving Rotuma a reputation for growing coconuts and other produce of exceptional size and quality. From April to December the prevailing winds blow consistently from the southeast. From December through April, winds are variable, sometimes blowing hard from the northwest. Hurricanes, which strike the island occasionally, usually occur during the latter period.
The island's resident population is approximately twenty-seven hundred individuals, but over six thousand Rotumans now live in Fiji, which has political jurisdiction over Rotuma, and perhaps an additional one thousand to two thousand Rotumans live in Australia, New Zealand, North America and Europe.
Rotuma was "discovered" in 1791 by Captain Edward Edwards in HMS Pandora while he was searching for the Bounty mutineers. Edwards remarked on the great number of paddling canoes coming out to meet his vessel. He wrote that although Rotumans knew of Tonga (the "Friendly Islands"), "their canoes were not so delicately formed nor so well finished as at the Friendly Islands, but more resemble those of the Duke of York's, the Duke of Clarence's and the Navigators' Islands" (Thompson 1915:65).
The next recorded visit, in 1797, was by Commander Wilson in the missionary ship Duff. Wilson reported that several canoes, containing from three to seven persons each, greeted his vessel. He commented, "Their single canoes (for we saw no double ones) were nearly the same in all respects as at the Friendly Islands, being of the same shape, sewed together on the inside, and decorated in the same manner seemed not so neat and well finished" (Wilson 1799:292).
In the next two decades, a great many whaling ships stopped at Rotuma for provisions. Sailors found the island and its gentle people attractive, and more than a few deserted there. Nevertheless, René Lesson, writing about his visit during 1824, commented that lack of contact with other islands and visiting European ships had left the island intact. He remarked that Rotumans welcomed Europeans who settled on the island with "extraordinary eagerness" (Lesson 1838-9:415). Lesson opined that the islanders must be skilled fishermen because of the huge nets they used, which he estimated at more than forty feet long. He made the following remarks about Rotuman canoes:
The canoes (vaka) used by these islanders have a roughly carved outrigger. They are enclosed and pointed fore and aft, and driven by oval paddles which are also carved without much taste. We saw only one small double canoe (aoe) which came in the evening. The mast was notched and set up on a piece of wood which linked the two canoes. It held up a sail of very coarse matting. The canoes were covered by a platform which prevented sea water from getting into the hull and which supported a shelter consisting of an awning of flexible branches. On the whole, it was a poorly designed vessel, and long-distance navigation was probably undertaken in larger canoes (Lesson 1838-9:431, translated from the French by Ella Wiswell).
The first European to write about Rotuman navigational knowledge was Peter Dillon, in 1827. He reported that Rotumans knew of several islands in their "neighborhood," including the islands of "Vythuboo" [Vaitupu] and "Newy" [Nui], in what is now Tuvalu. He stated that Vaitupu abounded with white shells much in demand on Rotuma and that Rotumans made frequent voyages there for the purpose of obtaining them. Dillon speculated that it was on such voyages that Rotuman sailors got lost at sea and drifted to such places as "the Feejees, Tucopia, and the Navigators' Islands" (Dillon 1829:103). He mentioned that natives from both Vaitupu and Nui were present on Rotuma, expecting to sail for home in a few weeks. He was unable to obtain information concerning wind, weather, and tides from his informantsÑtwo renegade sailorsÑbut assumed that westerly winds prevailed at certain times of the year, enabling Rotumans to sail to Tonga and the Navigators' Isles. He based his assumption on the fact that he met a Rotuman in Tonga whom he returned home (Dillon 1829:103-104).
Robert Jarman arrived on Rotuma in 1832 aboard the whaler Japan. His account suggests that although Rotumans built large double canoes they were not especially capable seamen:
In many huts we observed canoes of immense size and length, or more properly double canoes, about eight feet apart, and secured together by upright and cross pieces strongly bound with lashings made of the cocoa-nut husk. I should suppose these canoes capable of carrying from one hundred and fifty to two hundred men; and are from sixty to ninety feet in length. I was led to enquire for what purpose the natives had constructed them, apparently so unwieldy and useless, as they must have cost them with the tools they possess, such infinite time and labor.
George Cheever, who kept a log of the whaling ship Emerald, which visited Rotuma in 1835, also reported that Rotuman canoes were rough concerns, displaying little in the way of ornamentation. He mentioned one large double canoe, about 100 feet long, which was presumably the largest on the island (Cheever 1834-1835). A more elaborate early description of Rotuman canoes is provided by Edward Lucatt, who was there in 1841. He wrote:
The native canoes are of peculiar construction. They have no trees high enough of the proper wood to form the main body of the canoe, like the New Zealanders; they are therefore built out of several pieces, which are sewed together with a sort of twine, of their own manufacture, made from the husks of the cocoa-nut. They are deep and narrow, somewhat angularly formed; the thwarts for the pullers to sit upon are made fast to the gunwales, and, to prevent the canoe from capsizing, they have an outrigger attached. They are unsightly-looking things, wanting altogether the lightness and grace of a New Zealand canoe; their paddles, too, are clumsy and heavy, and lack the symmetry and grace of the New Zealanders. There are several large double canoes on the Island, connected together by a strong platform; and in former times, when the population of the country exceeded the means of support, or it was feared that it would do so, oracles were consulted, and at their instigation a party would start off in one of these canoes in search of fresh land: sometimes failing in their object, they would find their way back again in a most miserable plight; but the result of the generality of such expeditions was never known. Of late years, there have been no adventures of the kind, and these ship-canoes from their long disuse are fast falling to decay; there are seventeen or eighteen of them upon the Island, carefully built over to protect them from the weather. These ship-canoes are no two of the same length; the longer one will be from eighty to ninety feet, while the smaller, answering the purpose of an outrigger, would not exceed fifty or sixty feet: each canoe has from four to five feet beam, but they have no floor; and, looked at separately, without their stem and stern pieces, they would be taken for troughs. They are kept about six feet asunder by cross beams lashed and otherwise made fast to the gunwales of both canoes; the beams are planked over, which furnishes a deck of from fourteen to sixteen feet in breadth. Both canoes are entirely covered in, and there are small hatchways with sliding covers. When a party has determined upon an exploring expedition, they build a house upon the main deck and stow their provisions, &c. in the holds of the canoes. Their sails are made of a species of rush marled together: in form they resemble the New Zealanders, being when set like an inverted triangle (Lucatt 1851:177-178).
C.F. Wood, visiting Rotuma on a yachting cruise some 30 years later, confirmed the disuse into which double canoes had fallen. He reported seeing large double canoes, similar to those he had seen in Fiji and Tonga, lying in sheds on the beach. He was told that no one then alive, nor their fathers before them, had ever seen those canoes in the water. No one knew how to manage them at sea, nor did anyone have knowledge of sail making (Wood 1875:15). An old chief told Wood that in the past such canoes would sail off, loaded with men, never to be heard of again. Wood commented that he knew of canoes from Rotuma landing on the northern coast of Vanua Levu, where half-breed descendants could be found. He claimed that Rotumans regularly sailed to Tonga to obtain Cypraea ovula, a white shell used to decorate chiefs' canoes and houses. "They did not buy these from the natives of Tonga," Wood wrote, "but they themselves fished for them on outlying reefs" (Wood 1875:25-26). 
Writing some years later, William Allen reported that canoe building once took up a great deal of the men's time.  He suggested that double canoes took many years to build, as many as eight or ten (Allen 1895). During his visit in 1896, however, J. Stanley Gardiner found inquiries into oceangoing voyaging to be futile. Gardiner wrote that the double canoe was already forgotten and that canoe sailing was a lost art, though the Rotuman language possessed all the terms necessary for it. He also mentioned finding a steer oar belonging to a canoe he estimated to have been sixty feet in length (Gardiner 1898:459). Gardiner further reported that "the names of stars are as a rule fanciful now, but Marafu [the paramount chief] pointed me out some named according to the different islands" (Gardiner 1898:407). 
The most extensive information recorded about Rotuman canoes is contained in Gardiner's account and the field notes of Bishop Museum ethnologist Gordon MacGregor, who visited Rotuma in 1932. Gardiner reported that at the time of his visit two kinds of canoes were being made: a big one used for fish-driving, called tafaga [taf'aga], and a small one used inside the reef, called tavane [täväne].  Taf'aga varied from twenty-five to thirty-five feet in length, took from eight to twelve paddlers, and could carry up to twenty people. Regarding canoe construction, Gardiner wrote:
A suitable tree is selected, cut down, and roughly shaped. It is then properly allowed to lie for a few months, after which it is dragged down to the hanua noho (village) which is going to build it. It is then hollowed out to the desired shape, the ends being left solid and the walls up to 2 inches thick. In the centre the sides would not be strong enough to bear the strain, and so are removed, fresh planks being fitted into their place. These are fixed by sinnet, holes for the lashings being bored through the planks; wedges are then driven in between from the inside to make the whole watertight. The sinnet makes the holes watertight, but pieces of sponge from the reef are driven in to ensure it. There is a distinct bow and stern, the former sharp and pointed up, the latter blunter and curved downwards. The first three feet of the deck at each end is covered. The breadth along the whole centre is about the same: 1 1/2-2 feet. The side towards the outrigger, or sama, is slightly straighter [figure 6.2 C] than the other. The outrigger is about 5 feet or rather less away; it is not quite half as long as the whole canoe. It lies usually on the right, or starboard side, and consists of a post of light wood slightly pointed at one end. This is supported by two hard wood beams, driven into it, lashed across the canoe itself; the bend at right angles, which is necessary, is cut out, but can be, and was, frequently induced in the growth of the timber. Another beam runs just above the bend between these; to it rods of hard wood are lashed, previously driven into the post underneath [figure 6.2B]. A platform is generally made to take the paddles and carry the nets between the canoe and the outrigger; the paddle blade is of an oval form, 2 feet long by about 6 inches broad. The bailer is of the regular type, of one piece of wood with handle in the centre, and shaped to fit the canoe. The launch of one of these used to be the occasion of a feast. Kava was placed for the gods, after one of whom it was named and then supposed to be under his special protection.
FIGURE 6.2. Rotuman canoe features: A--side view of the tafaga, showing planks let in at the side, also bow and stern; B--section of the tafaga through one of the supporting beams of the sama, or outrigger; C--side view of the sama to show method of fixing; D--top view of the tavane (the thin lines at the sides show the bulge of the canoe); E--longitudinal section through the oie, or drum; F--transverse section of the oie. (After Gardiner 1898:458)
In a section of his notes labeled "canoes," MacGregor recorded the names of canoe types and parts. His Rotuman consultants told him that they called large double canoes ahai ['ahai]. Some of them thought the name came from the English word "ahoy," used by English sailors, a derivation favored by C.M. Churchward, who wrote the definitive dictionary of the Rotuman language (Churchward 1940:348). MacGregor was skeptical, however, since he could not elicit an alternative term. The traditional term, he wrote, would not have died out so quickly. Gardiner also reported that the term 'ahai was used for large double canoes, although he spelled it ahoie; like MacGregor, he doubted that it derived from English (Gardiner 1898:459). Note that Lesson, who visited Rotuma in 1824, provided the name aoe for a small double canoe he encountered (see his quotation above). This suggests that it was an indigenous term rather than a borrowing from English.
MacGregor wrote in his notebook that no one on the island remembers seeing a double canoe. Nevertheless, one of his Rotuman consultants told him that 'ahai were made from five or six planks to a side, according to the size of the tree, and that they were about five to five and one-half feet tall. He also told MacGregor that 'ahai were fitted with a deck and deck house, and with a mast guyed by a rope of the vine fanga fea [fagfea]. 
Churchward lists the Rotuman word samtutuki for the smaller double canoe (Churchward 1940:305), which MacGregor described as constructed of two hulls (vaka) joined by three crosspieces (kiata). The hulls were about eight feet high according to MacGregor's consultants, and were completely covered over to keep out water. The covering had a trapdoor to permit access to provisions. Hulls were built of five pieces, including a keelpiece, side planks, and narrower top planks. The planks had flanges and were sewn together. Stitches tying the planks together were held fast by driving small wooden wedges under the knots. Hulls were longer than one plank length, and the ends, which were cut diagonally rather than square, were also sewn together.
Poles were set up in the hulls and the deck (pupui) was held up on them about a fathom above the canoes' top planks. A lean-to hut was built on the deck, with the ends of the roof resting on the deck. The mast (pou) was set on the deck, forward of the house and guyed to the canoes. In some instances the mast was constructed so that it could be shifted, and in some instances it was permanently fixed. Sails (läe) were made of about twenty ordinary floor mats sewn together. They had two wings, or points, in the fashion of Fijian canoes. The steering oar (hose) was worked on the starboard hull. MacGregor's consultants told him that samtutuki carried about one hundred men, but it must be cautioned that they were reporting about the construction of a canoe type that had long since disappeared.
Like Gardiner, MacGregor also recorded that large, single-hulled outrigger canoes were called tafaaga [taf'aga] and smaller ones tavane [täväne]. Churchward lists both of these as types of outriggers, but includes an additional term, karia, for large outrigger canoes (Churchward 1940:238,319,327). MacGregor reported that täväne were made with the fore outrigger boom shorter than the aft one as an aid to steering. This gave the canoes an appearance of toeing in when viewed from above. "The stern is the wider part of the canoe because of the narrowness of the whole canoe for a man of any size to sit in and because it handles more easily," Macgregor wrote. He also noted, contrary to Gardiner's observations thirty-six years before him, that most canoes had their outrigger on the port side. This is confirmed by John Bennett, an American married to a Rotuman woman and an accomplished sailor. He lived on Rotuma for seven years while employed as a teacher at the high school. According to Bennett, outriggers today are commonly mounted on the port side, as most people are right-handed and find paddling on starboard more comfortable and unobstructed if the outrigger is mounted on the left. Turning with the paddle is also easier (J. Bennett, personal communication).
An account of the steps involved in building taf'aga was provided MacGregor by a craftsman from the village of Salvaka:
The Rotumans build two sorts of large outrigger canoes, the tafaka oro [possibly taf'aga 'i'oro 'shark fishing canoe'] and the tafaka fonfonu [taf'aga fonfono 'canoe with built up sides'].  The first is a three piece canoe with a deep hull, the second a five piece canoe with a shallow keel piece and built up sides, two planks to a side. When the whole body of the canoe is considered these canoes are actually five and seven piece canoes, but the last two pieces are deck pieces and not in the walls of the canoes... The difference in structure is determined by the material the carpenters have to work with. If a tree selected is not thick enough to give the proper depth to the hull, a five piece canoe is built.
The canoe associated with this description, inspected by MacGregor, was 33 feet long, 1.5 feet at the widest part, and twenty inches deep. MacGregor made a list of outrigger parts. Table 6.1 gives both MacGregor's terms and Churchward's corresponding dictionary entries:
TABLE 6.1. Rotuman canoe lexicon.
* MacGregor notes that the puka "is the part that runs from the end of the roa on one side around the stern or bow to the end of the roa on the other side. It should be of one piece." He adds that this implies it is not necessarily the whole stern piece. (MacGregor 1932)
** lit. 'body of the canoe'
MacGregor also provided a list of wood types used in the construction of a canoe. The ratau [ratu'a] tree was used for the hull of a canoe; the outrigger and paddle [hose] were made from penau wood (Thespesia populnea L., Pacific rosewood tree); the outrigger boom from hau (hibiscus) wood; and the outrigger brace from either hau wood or bamboo. Paddles were 4.5 to 5 feet long, with a blade from two to two feet four inches; MacGregor described them as flat on one side and rounded on the other (Macgregor 1932). This may suggest more uniformity of construction than was the case. Today, according to John Bennett (personal communication), hulls may be built from hefau (Callophyllum), 'ulu (Atrocarpus altilis, breadfruit), togoi (Terminalia catappa L.), 'ura (Morinda citrifolia L.), ratu'a (Erythrina variegata L.) or sa'a trees,  from which most paddles are now made. Bennett suggests that sa'a is used for paddles today because it is light, and much easier to shape than paddles made of penau. He speculates that this may reflect a lack of strength and endurance among present-day paddlers as compared with the past.
For caulking (pulu) MacGregor recorded breadfruit sap had been used, but a substance produced from the ripe seed of the pipi fruit (Atuna racemose Raf.) was a recent innovation that proved superior. The seed is pressed into a red watery fluid in the hand with a rock. It is worked into the seams where it becomes very hard and holds fast. Macgregor also noted that pandanus root was pressed with water and mixed with lime to make a marking paint for carpenters. His notes contain detailed drawings of lashings (Macgregor 1932).
Based on hull shape, outrigger attachment, and proximity, Hornell concluded that Rotuman canoes were an adaptation of a small canoe used on the northern Tuvalu island of Nanumea (Hornell 1936:281), but this conclusion is disputed by McQuarrie, who makes a strong case for their having a closer relationship to the canoes of Tonga and Samoa (McQuarrie 1980).
Gardiner and MacGregor also provide the most thorough accounts of Rotuman fishing techniques, including hook and line, fish traps of various types, and nets. Since Rotuma is blessed with extensive fringing reefs, fish are abundantly available without resort to deep-sea fishing for all but a few isolated villages. Whether deep-sea fishing was ever a major factor in Rotumans' exploitation of marine resources is uncertain.
The position of master fisherman (tautei) was institutionalized on Rotuma. Each district had a tautei who was in charge of fishing expeditions. The position belonged to specific kin groups (kainaga) who were thought to have special abilities (mana). Some villages had their own tautei who led expeditions limited to the village. Tautei titles derive from fishing activities. For example, according to Elisapeti Inia, retired Rotuman schoolteacher and authority on Rotuman language and customs, the title Reuas derives from räe ia'uas 'to see a shoal of fish'; Tokoar from tok 'o'oar ta 'the canoe used as a receptacle during fish drives'; Tokroa from tok roa 'long pole' (used to direct communal fishing activities); and Urakmat from ur'ak mat 'to respect + wet', suggesting the importance of following taboos associated with fishing (Elisapeti Inia, personal communication).
Gardiner reported that Rotuman fish-hooks were generally crude. He described them as follows:
The fe, or shark-hook, was made from a shrub, the tiere, which, when it reached the height of about 3 feet, was twisted into an open knot, with a diameter of about 5 inches; it was then allowed to grow for about two years before being cut. The hook was then shaped, and a piece of hard wood spliced on as a barb projecting inwards. The bait was tied on over the barb; the fish working at this, as the wood was springy, gradually got its jaw between the barb and the stem of the hook. On being struck the barb caught in the gills, and the fish was hauled up sideways. A similar hook, but smaller, the oiniafa, was used for catching a large species of rock cod, the roog. Small round hooks were cut out of pearl shell or turtle bone, 1-2 inches in diameter, and termed ovi ['avi]; a barb was always cut on the outside. Those of pearl shell for certain fish were not baited, nor towed behind the canoe. Proper spinning baits are termed pa [poa], and were of two kinds, the one large, of pearl shell fixed on bone 4 inches, or more, long, and the other small, 1-2 inches, of pearl shell alone. Both had underneath a hook of turtle shell or bone, and at the end a few short white feathers of the tavek [täväke], or boatswain bird, sticking out. Tjija [Jija], long fish with very narrow jaws [garfish], almost too small for any hook, are caught by a lump of spider's web at the end of a line on a long bamboo, when the tide is coming in. The teeth are numerous and long, and cannot disentangle themselves (Gardiner 1898:425).
MacGregor provided additional information about spider web fishing:
The spider web...is caught up from the corner of the house by twisting on the end of a stick. The web is made into a ball and colored with charcoal so that it will show up black. Charcoal of the leifau [hefau?] is used, called mahala. This ball is then tied on the end of the fish line.
Gardiner described several techniques for catching fish on the reef, mostly involving traps and nets. In three instances he mentioned the use of canoes: while turtle fishing, during fish drives, and when catching flying fish. For turtle fishing, Rotumans used sennit nets, according to Gardiner. A net was put down in a reef passage just before the tide began to ebb. Canoes were stationed at each end of the net, and when a turtle was seen to go into it, a man from each canoe dived after it, seized it by the front flippers and turned it over so that it was compelled to come to the surface (Gardiner 1898:426).  MacGregor confims this description, adding that the fishermen would frighten the turtles into the nets by drumming on the sides of their canoes. He also describes an alternative method of turtle fishing, involving taking canoes out at night on the high tide and searching for turtles with torches, then spearing them when they came to the light (Macgregor 1932). Today people dive for turtles using goggles or a diving mask and small spear made from a quarter-inch steel rod propelled by a flexible rubber tube (similar to surgical rubber), with a copper wire loop attached on one end and a tire-tube rubber loop, placed over the thumb, at the other. The spear is drawn like a bow and arrow (Bennett, personal communication).
Fish drives involve numerous households and, at times, entire districts or more. A large net is placed at a strategic location in a reef passage just before low tide, and fish are driven into it by lines of people beating the water. Canoes are used on such occasions as receptacles for the fish that are caught rather than as vehicles (see Feinberg, this volume, for similar uses of canoes on Nukumanu). Those picking the fish out of the net crush the skull between their teeth to dispatch them quickly and prevent them from flipping out of the canoes in which they are stored. Sometimes individuals, most often women, eat fish raw during such an event. Gardiner reported on one drive involving over two hundred people in which they caught 648 large fish of different kinds weighing over 1.5 tons (Gardiner 1898:427).
Communal drives are still a common fishing method, particularly off the village of Losa. Fish are driven out of the reef passage and, blinded by the setting sun, are caught in a net. This method is called furamasa 'to fish late in the afternoon' (Churchward 1940a:208), or more generically, hagoat vao 'net fishing'. The term jau 'to beat' (which in this instance refers to beating the water when driving fish to where they can be caught)  is commonly used for fish drives in which nets are not used. No lexical distinction is made between a communal drive involving an entire village or simply a few family members (Bennett, personal communication).
In the past, canoes were also used to catch "flying fish" using a special net [seu] about twelve feet long by six feet wide, fixed at the end of two bamboo poles. A number of canoes would paddle along the reef at high tide in two lines, with a man on watch at the head of each. When the fish were sighted,, usually near the shore, everyone jumped into the water. While some of the men surrounded the fish with a net, others got ready to catch them when they jumped over it (Gardiner 1898:428). From his account it seems that Gardiner was referring to mullet ('anasi) rather than true flying fish (säsäve), since the former feed on the reef while the latter inhabit the deep sea outside the reef. Mullet often jump out of the water, but do not have the anatomy to glide through the air. 
The tautei from Noa'tau district provided MacGregor with detailed information concerning seu expeditions:
The Seu is a fishing party of eleven canoes which are accompanied usually by a chief, for whom the fishing is done.  The fleet of canoes go out in two lines of five in file, and the eleventh canoe travels between the two lines opposite the last two in the files.
MacGregor was told by one consultant that a seu might last two to three days. A number of taboos were associated with seu. For example, on the night before an expedition, the fishermen were supposed to sleep alone; if any one of them had sex, fish would jump over the net he was holding. No one could mention the names of the men who went fishing. The houses of the fishermen were shut during a seu and no one was allowed to enter them while it was in progress. Making noise was taboo, and it was forbidden to make cooking fires. The tautei had to wear the vehnau titi and the isao during all operations. If any of these taboos were violated, it would bring bad luck. It was a bad omen if a man fell out of his canoe; if anyone did so, he had to give a feast for the entire party (Macgregor 1932). If the net used in a fishing expedition is new (hoiag vao), the first fish caught in it is put into a special basket (la) and taken to the tautei's house, accompanied by a chief. The doors of the house, kept closed until then, are opened and an announcement of the catch (ma hei'akia) is chanted (Elisapeti Inia, personal communication).
MacGregor mentions two locations in which hook and line fishing were used for catching snapper in deep water by submerged reefs: off the islets of Haua and off the village of Losa. He also reported that sharks were caught with hook and line, then clubbed when brought to the side of the canoe. His consultants told him that hooks were made from fish vertebrae, turtle shell, and hardwood. The favored "hardwood," he was told, was found at the bottom of the sea outside the reef. (According to John Bennett, the substance referred to is black coral, commonly found off the island of Uea, which has no fringing reef.) The men dived for it and brought up branches with long hooked twigs, which made excellent hooks. Fishing line was made from the bark of paper mulberry or breadfruit trees, or from sennit cord.
The practice of closing up a house when a man goes deep-sea fishing is still common today. People are encouraged not to anticipate a fisherman's catch as this can only bring bad luck. When one first takes out a new boat to fish, he is expected to acknowledge a good catch by providing fish to the chief and minister. Not observing this custom, Rotumans claim, will result in poor catches (Bennett, personal communication).
It is clear from the reports of nearly all early commentators that Rotuman men were eager to leave Rotuma aboard European vessels and took every opportunity to do so (e.g., see Bennett 1831:480). Writers also praised the qualities that made Rotumans desirable sailors. The remarks of Joseph Osborn, aboard the whaling ship Emerald when it stopped at Rotuma in 1835, are typical:
They love to visit foreign countries & great numbers of them ship on board the English whaleships... On board a ship they are as good or better than any of the South Sea natives: diligent, civil & quiet, 3 very necessary qualities. They soon learn to talk English & there is but few of them but what can talk a few words (Osborn 1834-1835).
John Eagleston, Captain of the Emerald, echoed Osborn's sentiments. "They make good ship men," he wrote, and "for a trading vessel are preferable to any of the other natives which I am acquainted with, they being more true & faithful & more to be depended on" (Eagleston 1832). He noted that he had had a number of Rotumans aboard as crewmen in the past, as well as other islanders, but found Rotumans to be the best.
Some forty years later Litton Forbes wrote:
The men of Rotumah make good sailors, and after a few years' service in sea-going vessels are worth the same wages as white men. Scarcely a man on the island but has been more or less of a traveller. It is no rare thing to find men who have visited Harve, or New York, or Calcutta, men who can discuss the relative merits of a sailors' home in London or Liverpool, and dilate on the advantages of steam over sailing vessels. Thus the average native of Rotumah is more than usually capable and intelligent (Forbes 1875:226).
Commenting in 1867 on the extent of emigration, Rev. William Fletcher, the first European Methodist missionary stationed on Rotuma, wrote that upwards of seven hundred young men were known to have left the island in recent memory (Fletcher 1870). Anxiety over the unimpeded emigration of young men was one of the first issues raised by the chiefs of Rotuma in their negotiations with British authorities prior to cession. Thus Arthur Gordon reported in 1879 that the chiefs desired regulations to check wholesale emigration. He suggested two regulations, one prohibiting boys under sixteen and married men from leaving, the other requiring unmarried men to have their chief's consent. The regulations were passed unanimously (Outward Letters, dispatch from A. Gordon to Colonial Secretary, 4 December, 1879 ).
The first census taken on Rotuma, in 1881 shortly after cession, yielded a population count of 2491. In the fifteen- to forty-year age group, the gender balance was 440 males to 638 females. The Resident Commissioner at the time, Charles Mitchell, attributed the surplus of females to the fact that so many young men had left the island (Outward Letters, dispatch from C. Mitchell to Colonial Secretary, 1 October, 1881). 
W. L. Allardyce, who was on Rotuma about this time, commented on the shift in traveling destinations resulting from the demise of the whaling industry as well as the social price homestayers had to pay:
Nearly all the men on the island have at one time or another been to sea, and while in the old whaling days Honolulu and Behring Straits formed the goal of their ideas, the sailors of the present day must needs visit New Zealand, Australia, China, and India, while others still more ambitious are not satisfied till they have rounded the Horn and passed the white cliffs of Dover. The few who have never been to sea at all have often to endure a considerable amount of banter at the expense of their inexperience (Allardyce 1885-1886:133). 
Allardyce also noted that the majority of Rotuman men abroad were engaged in the Torres Straits pearl-diving industry, which was flourishing at the time. Writing in 1884, William Gordon reported that over one hundred Rotuman men were employed in the industry, mostly in the management of boats. The boatmen earned two to three pounds a month. Divers earned much more, up to forty pounds a month according to Gordon, but tended to squander it during binges in the city ( Outward Letters, dispatch from Wm. Gordon to Colonial Secretary, 24 November, 1884).
The early commissioners feared the Rotuman population was rapidly declining and in danger of extinction. They saw emigration as a significant part of the problem. Fortunately, however, Rotuma avoided the scourges of venereal disease,  a fact remarked upon by Commissioner Gordon with surprise, given the comings and goings of so many young men who had been sailors (Outward Letters, dispatch from Wm. Gordon to Colonial Secretary, 24 November 1884). Attempts to dampen emigration were aided by the British decision to govern Rotuma as part of the colony of Fiji and to close Rotuma as a port of entry. This significantly reduced the number of vessels calling at Rotuma and meant that men had to go to Fiji to sign aboard as seamen.
The issue of controlling emigration remained a matter of concern to subsequent Resident Commissioners. Following an inquiry into the matter, A. R. Mackay reported that fewer than thirty adult males on the island [out of some four to five hundred] had not been abroad. He reiterated what others had said before, that "it is a cutting reproach to cast at a man that he has not been away from the island; hence, partly, the anxiety of the young men to accomplish their long cherished dream" (Outward Letters, dispatch from A.R. Mackay to Colonial Secretary, 10 January, 1887).
The concern for controlling emigration eventually led to the passage of Rotuma Regulation Number 3 in 1939, stating that "No native may leave Rotuma without the permission of the District Officer" (who replaced the Resident Commissioner as administrative officer in charge of Rotuma following a colonial governmental reorganization). It further specified that "No male adult responsible for the maintenance of his wife, children, or relatives may leave Rotuma without making adequate provision for the maintenance of said wife, children, or relatives to the satisfaction of the District Officer" (Rotuma Regulations 1939:457). 
Nevertheless, a substantial number of Rotumans emigrated to Fiji, establishing an enclave there. The outbreak of World War II accelerated Rotuman emigration to Fiji, and by 1946 approximately 17 percent of all Rotumans were residing there. Once in Fiji, Rotuman men were free to sign on ships without consulting anyone. The chiefs protested and demanded the right to choose who could go, but the governor of Fiji ruled that "the liberty of the individual must be respected and it is not for the Rotuman Chiefs to decide who shall and who shall not sign on ships" (Fiji Archives, Document F 24/44).
Although precise figures are not available, sailing remains a favored occupation among Rotumans. In a survey of 414 households (85 percent) on Rotuma during 1989,  I identified sixty-nine men engaged in sailing at the time. This accounted for 20.7 percentof all men abroad on whom I have occupational data. Five of them were ship captains, two were pilots. Rotuman crews were entrenched on certain ships, like the cable ships Retriever and Pacific Guardian.
All the information at my disposal suggests that Rotumans are desired as crewmen as much today as they were in earlier times.  The reasons given are similar: that Rotumans are exceptionally conscientious, learn fast, and are capable of taking responsibility.  I have elsewhere related these qualities, widely recognized by employers of Rotuman personnel, to socialization patterns and the social organization of the Rotuman community (Howard 1966, 1970).
With a few notable exceptions, Rotumans now residing on the island do not venture beyond the reef and have not developed seafaring skills. Almost all fishing is done on the reef, and canoes are used more as receptacles than as vehicles. The only type of canoe that has survived is the small outrigger, formerly called täväne, but currently known only by the generic term for canoe, vaka. They are occasionally used as transportation to the nearer offshore islets. Both men and women fish on the reef, and women are as likely to take canoes for this purpose as men. It is common to find husbands and wives exploiting the reef together.
The most usual techniques employed do not require canoes, however. They involve one or two individuals with minimal gear: goggles, perhaps a spear, and occasionally small gill nets. Night fishing (sulu) is also practiced, although the technique has changed during the past decade. Previously, benzine lanterns were used, and one would walk on the reef at low tide or spearfish from a canoe. Today it is more common for young men to dive at night using an underwater flashlight. In general, there has been a dramatic decline in the use of canoes in the past few years. Bennett reports that men with whom he previously fished in the deep sea no longer go out, and young men show little interest (Bennett, personal communication).
As the cash flow to Rotuma has increased, largely as a result of remittances (Rensel 1994), the islanders have exploited marine resources less and less, so that consumption of tinned fish now exceeds that of fresh fish. A few years ago a Fisheries report documented an average annual per capita consumption of 30 tins of mackerel, a total wet weight equivalent of more than 40 tons of fish. The authors comment on the unimportance of fishing in daily life, in stark contrast to the strong fishing traditions of other Polynesian peoples (Fisheries Division n.d.:5). 
The 1989 household survey (see note 18) included an inventory of selected items, including canoes and boats. Seventy-six households (18.4 percent) reported owning a canoe, and seven households (1.6 percent) reported owning a punt or launch with an outboard motor. All canoes were owned by individual households, and nearly all had been built by a member of the household. 
One man, with the title Sautiak, owned a small fishing boat with an inboard engine. Sautiak identified five fishing areas that he exploits. Three were off the west end of the island, including Whale Bank Reef; one was to the east, another to the southeast of Noa'tau. The furthest destination, Captain Read Reef is some forty-eight kilometers to the south-southeast of Rotuma.  It is approximately sixteen kilometers long and rises to within twelve fathoms of the surface. According to a report from the Fiji Ministry of Agriculture and Fisheries, this location "is likely to be very productive for both pelagic and bottom fish, and can be fished with a reasonable margin of safety during favourable weather in the SE trades, when prevailing winds would favour sail-assisted return of a disabled vessel to Rotuma" (Fisheries Division n.d.:18).
Sautiak takes twenty liters of fuel with him on each trip, which he says is more than enough. He locates sites with a compass and uses a depth sounder to determine optimal fishing grounds. When fishing close to Rotuma, he uses alignments of landmarks on the island and watches for sea birds feeding at the reef breaks. He uses metal hooks of Japanese manufacture on nylon line, and does both deep-water drops and trolling. He goes out once a week or so when his boat is in good repair. Most of his expeditions are extremely fruitful, and he distributes his catch to friends and family, although he usually sells a portion. Since demand far exceeds the supply, selling fresh fish is very easy. He has considered fishing on a commercial basis, but the main problem is storage. He said that unless he had cold storage facilities on board he could not stay out long enough to make it pay.
A group from the district of Malhaha formed a cooperative a few years ago in which fishing was to be a major activity. They bought two boats and a large, walk-in freezer to store their catch. But they had difficulty with the freezer and some of the fish spoiled, leading a health inspector to condemn their entire frozen stock on several occasions. As a result, they endured financial losses and their backers, a group of Rotumans in Fiji, withdrew financial support. After a while the freezer broke down completely. One of the boats was repossessed by the bank and sold to a group from Hapmak; the other developed a leak and was still in drydock on our last visit to the island. The Hapmak group uses the boat for commercial fishing on a sporadic basis, going out when there is a special need for fish and when sailing conditions are optimal. Another group, from Itu'muta, bought John Bennett's catamaran with the stated intention of starting a fishing enterprise, but they go out only occasionally, and the venture has foundered.
Despite a consistent demand for fish and recognition of commercial possibilities, little has been done to capitalize on the potential. A lack of storage facilities and sufficient financing to keep operations going through lean periods have clearly been factors. But the fact that many Rotumans in Fiji are engaged in commercial deep-sea fishing suggests an aversion to the rigors of the deep is not at issue.  While it is probably true that economic payoffs on Rotuma are insufficient to offset investment costs in time and money,  a more important factor may be a failure in leadership. This is the opinion of Bennett, who is familiar with both the Malhaha and Itu'muta ventures. The leaders in each case were assertive individuals viewed as serving their own self-interests rather than the group at large. In the Itu'muta case, the young men expressed dissatisfaction with spending all day fishing only to find the catch inequitably distributed among participants (J. Bennett, personal communication). Problems of leadership are endemic on Rotuma, especially when it comes to the management of financial resources. It may well be, therefore, that the kind of leadership required to sustain a fishing enterprise is absent on Rotuma, and that Rotumans engage in seabound ventures so much more readily abroad (and aboard European vessels), because they they find the leadership much more palatable.
Sailing as a Focus of Cultural Experience
The importance of sailing as a focus of cultural experience for Rotumans is evident in many ways. It is a source of solidarity between men who have sailed together and the subject of reminiscences when men congregate. Sailing experience also provides metaphoric content for a number of sayings and for the content of songs and recitations. In some instances it constitutes the core of personal identity, as in the case of a woman who changed her name to Al 'e sasi 'Died at sea' after her son was lost while serving aboard a ship. Travel abroad, in general, is metaphorically a sailing experience, and at its core remains a canoe journey. Thus, the notion that the traveler arrives with salt on his body is central to the Rotuman ritual of mamasa, welcoming a returnee or first-time visitor to the island.
The Mamasa 'Welcoming Ceremony'
The term mamasa means 'to be dry' or 'to become dry', and is used in reference to a ceremony performed when people return from a sea voyage or, in earlier times, from a dangerous deep-sea fishing expedition (Churchward 1940a:258). Presumably when people arrived by canoe, or by steamer prior to the building of a wharf at Oinafa in the 1970s, they landed wet and had to be dried and provided with clothes; hence the reference to drying out. The mamasa is one of several life crisis events that involve the central symbols of Rotuman ceremonial performances: sacrificial pigs, kava, sweet-smelling oil, garlands, and fine white pandanus mats (apei).
The recipient of a mamasa ceremony (forau) is seated in the place of honor upon a pile of mats (päega) topped with an apei, which symbolizes his or her elevation to a god-like status for the duration of the event. Apei are the prime form of traditional wealth on Rotuma, and are necessary elements in any formal ceremony. Each apei is consecrated by a koua 'sacrificial slaughter and cooking of a pig in an earth oven'; it symbolizes life and blessings from the gods.
While seated on the päega the forau is presented with a tefui 'garland', tied around his or her neck by a woman designated to play the welcoming role. The woman then performs the mamia, which in ordinary circumstances means 'to wash a person or fishing net in fresh water after having been in the sea' (Churchward 1940a:259), but in this instance is symbolized by anointing the honoree with sweet-smelling oil, which must be pure Rotuman oil. This highly symbolic act signifies cleansing the body of salt from the sea (Nilsen 1977:80). In earlier times, the woman would take off the forau's shirt and put a new one on him or her, then present him or her with a new lavalava (Marseu 1986:5). 
This act is followed by a kava ceremony and formal presentation of food. The forau is given the head of the largest pig (reserved for the person of highest status during feasts), which is placed in front of him or her on a ceremonial table ('umefe). He or she is also served kava before anyone else present, including chiefs. If the ceremony is an elaborate one, for a person of high status or someone whose achievements abroad brought honor to the community, dances and songs composed for the occasion may be performed, glorifying the person's adventures or accomplishments.
The mamasa ceremony has at least two clear functions. One is to reintegrate sojourners back into Rotuman life by communicating their importance to the community and elevating their status, if only for the duration of the event. Attention is called to their absence and return, to their bravery in leaving the island, and to their accomplishments. They are made to feel special. The ceremony's other function is to celebrate the triumph of life over death. Like all Rotuman ceremonies, the symbolism of the mamasa emphasizes the regeneration of life forces: the sacrificial pig is a gift to the gods, who are supposed to respond by giving life to the land and its people; kava symbolizes bodily fluids that give life; and fragrant oils and flowers (in the form of a garland) suggest the sweet smells of life as opposed to the stench of death.  Journeying beyond the reef for long periods was regarded as dangerous in the past; the mamasa was a way to offer thanks for a safe return. Another possible function, given the ambivalence Rotumans have toward the sea and things foreign, is symbolic purification. By washing away salt water residues with fresh water, they perhaps ritually cleanse sojourners of contamination not only by the sea, but by foreign influences of all kinds.
Rotuman sayings referring to sailing experiences can be lumped into four categories: (1) those calling attention to the special status of travellers; (2) those used metaphorically to call attention to hardships and fatigue; (3) those referring to chaos; and (4) those calling attention to undesirable behavior. The following examples are illustrative: 
1. The special status of travellers. 'Ou la mat la'mou 'Your legs (or feet) are still wet'. Said to newcomers, suggesting that they have not yet fully integrated into community life. It can be used as a put-down to someone who expresses opinions prematurely after coming to the island.
Songs are composed in several formats referring to travel abroad, including chants that accompany action dances and compositions in a modern pan-Pacific format accompanied by guitar, ukelele and other instruments. An example of such a song is presented by Mosese Kaurasi (1977:145-146):
Ritual presentations of kava occupy a central place in nearly all Rotuman ceremonies. A key part of the ritual is the recitation of a fakpeje, a text generally associated with the legendary coming of kava to Rotuma. Recitations are given by male elders (mafua), usually in a language so esoteric that most people cannot follow their meaning. Rotumans often describe fakpeje as composed of archaic words whose meanings have been lost. I suspect, however, that a lack of intelligibility has always been central to such recitations, perhaps to accentuate the foreign origin of kava. The prevailing myth concerning kava's origin has it coming from "Tonga," which in its generic sense is a place over or under the sea inhabited by potent beings or spirits.
That Rotumans associate experiences overseas and aboard ship with the foreignness (and potency) attributed to kava can be seen in a fakpeje recorded by Hocart on Rotuma in 1913. In this instance, intelligibility seems to be obscured by the seemingly random juxtaposition of English phrases, with Rotuman words interspersed:
Unpredictability and Ambivalent Emotions
Collectively, Rotuman attitudes toward the sea are ambivalent. It has been a source of wondrous adventures for some, of seasickness for many, and of grief over losing kinsmen for others. For all Rotumans, boats represent connections to the outside world. For those on the island, boats are bridges to relatives abroad, transporters of vital supplies, and the means by which money is earned through the shipping of copra. Days when boats arrive and leave are major occasions.
That the sea beyond the reef is dangerous comes home repeatedly. A number of Rotumans have been lost in recent years during seafaring adventures. Some have gone fishing and never returned. News of crewmembers aboard vessels that have sunk arrives periodically. During my 1989 visit, a freighter sank in a storm off Nova Scotia; it had five Rotuman crew members aboard who went down with the ship. Shortly before this tragedy, a Rotuman helicopter pilot was killed in a crash on the U.S. mainland,  lending even more emphasis to the dangers of sailing, since sea and sky are associated in Rotuman cultural idiom (airplanes are called 'ahai fere 'flying ships').
Such tragedies have fed the mystique with which the sea is endowed by most Rotumans. They conceive of the sea, and what happens in and on it, as dangerous and unpredictable on the one hand, enriching and vitalizing on the other. Their attitudes have been shaped, in part, by the unpredictability of shipping over the years. Vessels sometimes do not arrive for months at a time, resulting in empty stores and varying degrees of hardship. At other times several ships may arrive within days of each other. Despite numerous efforts on the part of well-placed officials, shipping to Rotuma has remained erratic since the initial arrival of Europen vessels. It has been the source of much black humor on the island. For example, when I heard on Fiji Radio that a boat was scheduled to sail for Rotuma in a week and mentioned it to my Rotuman friend, he laughed and told me always to multiply the time interval by two or three.
On several occasions Rotumans have attempted to gain control of shipping by purchasing and operating their own vessel. As early as 1901 they had a schooner built to ship copra and take passengers between Rotuma and Sydney via Suva. The fifty-ton vessel cost over 2,000 pounds and operated for eighteen months before sinking on a reef at Rotuma (Eason 1951:89-90). Most recently, in 1992, a Rotuman group purchased an interisland vessel, the Wairua, at a cost of Fiji $250,000. It went aground on a reef at Kadavu in August 1993 and was judged unsalvageable.
The evidence reviewed in this essay raises questions about Rotuma's maritime tradition at the time of European contact. Although there is no doubt that Rotumans occasionally voyaged to islands in their general vicinity, and perhaps beyond, their canoes were described by European observers as of poor quality, and their navigational knowledge was characterized as limited. Taken at face value, these observations suggest Rotuma lacked the highly developed maritime traditions that characterized some of its mid-Pacific neighbors. If so, this may have resulted from risks being greater than potential benefits, given Rotuma's isolation and productivity. Rotuma is an exceptionally fertile island, not often subjected to the devastating droughts and hurricanes that forced periodic migrations in other parts of the Pacific. There may have been little motivation to develop long-distance sailing skills.
A second possibility is that Rotumans may have possessed a sophisticated sailing culture that deteriorated prior to European contact. The only hope of verifying such a prior tradition would be through archaeological investigations yet to be done. A third possibility is that Rotuman seafaring skills were in fact highly developed, but that opportunities to sail aboard European vessels led to an extraordinarily rapid erosion of traditional maritime technology and knowledge. Perhaps, it might be argued, the most skillful sailors took advantage of the opportunities European ships offered, and they distinguished themselves because they already had the personal characteristics that made them good sailors, regardless of the craft involved. That sailing, and canoes in particular, have remained central symbols in Rotuman culture lends weight to such an argument. Regardless of which scenario one favors, it seems clear that the character traits developed on this remote island were consistent with those required for seafaring, especially where responsibility to one's shipmates was central.
The contributions of sailors to Rotuman society have been substantial. Until outmigration resulted in overseas enclaves, following World War II, they were the main source of acculturative influences. Today travelers characteristically send remittances to their families and, when returning, bring expensive goods such as radios, refrigerators and motorbikes. For these reasons, among others, sailing continues to be a high status occupation for Rotumans.