Volume 41 1932 > Volume 41, No. 162 > Was the double-outrigger known in Polynesia and Micronesia?, by James Hornell, p 131-143
                                                                                             Previous | Next   

- 131
A Critical Study

IN his thoughtful and stimulating work The Building of Cultures (5), Professor Roland B. Dixon cites the supposedly wide though sporadic occurrence and marginal distribution of the double-outrigger canoe as indicating its antiquity and precedence in time to the single-outrigger canoe. Unfortunately the data when critically analyzed do not support this conclusion. If it be correct, then it must be proved by other evidence than that afforded by the distribution of the double-outrigger in marginal areas in Oceania.

Dixon (5, p. 77) states that the double-outrigger “is, or was, found in three other small, isolated and widely scattered regions [in addition to Madagascar and East Africa (1), Indonesia (2), and the Torres Strait's region (3) ], namely, (4) Easter Island, and the Marquesas group; (5) New Zealand; and (6) the Pelew Islands, Ponapé in the Carolines, Samoa and Nissan … In the … fourth area the double-outrigger co-exists or co-existed with the single; in the fifth it had gone out of use in historic times and was known only by tradition; and in the sixth it occurs as a doubtful local usage in regions where otherwise only the single form is in use.”

The inference he comes to is—“Its wide extension, its marginal position, and the scattered form of its distribution all combine to suggest that as a cultural trait the double-outrigger is older and more primitive than the single form” (p. 78).

Apart from Nissan, which instance is not called in question and which is merely an extension from the New - 132 Guinea or the Indonesian area, the localities cited by Dixon will now be dealt with in turn, adducing the evidence for and against the statement of distribution above quoted. It is but fair, however, to add that Dixon had authority behind him in his argument, for the instances in question have been accepted for some time without adequate scrutiny of the individual evidence.

Taking the last group first as it is the most easily disposed of, the facts are as follow:—


In this group the evidence of the former presence there of double-outriggers rests solely upon a deduction made by Mueller-Wismar (12, p. 245) that certain stout thwart-like cross-bars with shaped heads, called bákat, which project outboard a short distance on the port side of the hull of the larger canoes, represent the thwartship tension-bars or boom braces seen in the double-outriggers of Mindanao, Sulu and Sangir (described by Hornell, 10, p. 78, plates 12 and 13). Mueller also suggests that the rod connecting the ends of these bákat on the starboard side represents the stringer connecting these tension-bars in the Philippine and Sangir canoes.

These points are the sole foundation for Mueller's inference; they do not justify Dixon's statement (p. 81) that historical data show that in the Pelew group the double-outrigger “had been replaced by the single within traditional times.” There is certainly no local tradition to this effect.

On the other hand the proximity of the southern Philippines and the Sangir Islands is greatly in favour of the inference; the Pelew Islands are more likely than any of the other groups involved in this discussion to have possessed the double-outrigger in former times. This, however, does not help Dixon's argument, for the Pelews are not a marginal group and make contact directly with Indonesia. It merely tends to prove that the double-outrigger, even if it be evolved from the single type, may have reverted to the original form, in these islands, through subsequent contact with Caroline Islanders who use single-outriggers exclusively.

- 133
PONAPE. (In the Eastern Carolines.)

The evidence of the occurrence of double-outriggers in this locality is contained in the following passage in Captain Fraser's account (7, pp. 74-75) of a flying proa seen in the vicinity of the Andema Islands, near Ponapé, in 1832. These proas, he says, “are formed the same at each end, thus having no occasion to tack. They have a large mat sail, easily shifted on changing the vessel's course, and by which they may sail within six points of the wind. The sail being large and the proas narrow, they are obliged to use double-outriggers to give them sufficient stability. Two of the crew were at this time sitting on the weather outrigger.”

No other writer makes mention of any Caroline Islands' canoe having a double-outrigger, but all agree in describing a lee-platform on the side opposite to the outrigger-frame, which also has a large platform of poles laid athwart its booms, whereon men are stationed to counterbalance any excessive heeling over toward the lee side when the wind increases in force. Of the lee platform Fraser says nothing. From his omission of what is an essential feature of Caroline Islands' canoes, it is certain that his “double-outriggers” were in reality the double-outboard platforms, one on each side, mentioned by all other voyagers.

In this account we have an excellent instance of the confusion caused by the loose use of technical terms, parallelled unfortunately in the case next to be considered.


Again one writer only is the authority for the statement that some of the canoes of the Samoan Group were fitted with double-outriggers. G. Brown (3, p. 350) states that “all these canoes had outriggers, and in the case of the sailing canoes they had one on both sides. If the breeze was strong, one of the crew stood on the windward side, going out farther and farther, as the wind increased, and especially when the canoe was struck by a squall, and stepping inboard as it slackened.”

Like so many other of the older observers, Brown uses ‘outrigger’ in an unexact and over-wide sense, meaning - 134 merely an outboard counterpoise—the ‘balance-board’ well known to have been in use in Samoa and the Society Islands, described and figured by numerous writers. This fitting was a plank projecting outboard on the starboard side, the outrigger being upon the opposite or port side as usual in Polynesian canoes. These canoes as rigged were unable to sail either end first as is the case with Fijian canoes which have the mast stepped amidships, and pivotted so as to rake toward either end according to the tack they may be on. Because of the disability inherent in the Samoan rig, the outrigger was sometimes on the lee side; had it not been for the counter-balance on the weather side provided by a board projected outboard from that side, on which some of the crew could be stationed to function as a counterpoise of live ballast sufficient to prevent the outrigger float from burying itself in the water, imminent danger would have been incurred of a dangerous capsize and the breaking up of the outrigger-frame.


The reference to double canoes in New Zealand occurs in a traditional account of the fitting out of the vessel Takitumu for the voyage from Tahiti to New Zealand about 500 years ago; as will be seen much turns upon the accuracy or otherwise of the translation of a single word.

Elsdon Best (2, pp. 203-5) gives the text of the Maori version and then the following translation thereof:—

“The two korere were then manipulated, one for each side of the vessel. The korewa timbers were composed of extremely light wood. The use of the korewa is to safeguard the vessel, lest it capsize, and, if it does chance to capsize, that it may continue to float, so that it may be turned on to the keel by the swimming men when it so capsizes. As to the names of the outriggers (korewa) I did not hear them mentioned in the whare wananga.”

Best comments on this, “The speaker meant the proper names here, not the descriptive names …. It is not certain whether korere is another name for an outrigger or not, but it looks like it. Herein we have a plain statement that Takitumu was a double-outrigger vessel, and the office of the outriggers is clearly explained. This is taken from the superior version of the voyage, as preserved by the - 135 trained experts of the whare wananga or tapu school of learning. In another version of this tradition certain men are instructed to launch the vessel and to attach the oceangoing outriggers. (Ka whakamau ai nya korewa moana ki runga). Again the plural form.”

In translating from the Maori version Best gives “outriggers, ” presumably meaning outrigger-floats, as the translation of korewa. On page 202 he describes these as being “launched out, ” and in another place as “thrust out.” As we have noted above these korewa are described by the Maori expert as being of extremely light wood, and intended to keep the vessel afloat if it were to capsize. But so far as I know this latter is not a function of the outrigger-float in any modern canoes although it may contribute slightly to keep a canoe afloat if it capsizes.

That the translation of korewa as ‘outrigger’ is open to grave doubt is increased when we find Best saying on page 210, “The word korewa seems to denote buoyancy and drifting, hence it was applied to both an outrigger and to what may be termed a drift anchor, or sea anchor.”

In his Maori Canoe (2a, pp. 296-7) Best gives the terms for outrigger as ama, amatiatia and korewa, but makes no reference to ‘outrigger-float’ and uses the term ‘outrigger’ vaguely and indiscriminately both for outrigger-booms and outrigger-floats. Now I doubt greatly his inference or identification that korewa means outrigger either as boom or float, though more likely the latter, as he says it connotes lightness. Korewa is not used elsewhere in Polynesia so far as I am aware for either outrigger-boom or float; ama in some form is universal for outrigger-float, and kiato for boom. I cannot think that the Maori would not have used ama if the thing meant be an outrigger float. Ama is given by Best in his glossary as a synonym of korewa, but in all the quotations given korewa alone is the word used and not the ordinary Polynesian term ama. The Maori must certainly have used kiato generally for outrigger-boom, for this word remains to the present day as a word for thwart, the nearest approach to an outrigger-boom that is to be found in modern canoes and boats in New Zealand.

Besides the linguistic objection, there are serious technical difficulties to the acceptance of korewa as an outrigger- - 136 frame or float. On pages 201-202, Best translates a passage from a very old tradition as follows—“Should it be known that a gale was approaching then immediately the vessel was prepared to meet it. The tokotu and whiti were set up, the covering mats were drawn over and stretched taut, so that, when the storm broke, all would be well. Then the outriggers of the vessel were thrust out [Katahi ka kokiri i nga korewa o te waka ki waho], the hokai and huapae were secured, the water fenders were drawn into place, and the pare arai wai secured in place.”

If outrigger-booms and floats were put out on both sides in stormy weather as we are here informed, this involves the conclusion that the vessel was sailing without outriggers until the storm threatened. This is most improbable. Nowhere in Polynesia is the outrigger ever other than a permanent fitting when afloat; we know that the Maori people who related and sought to explain this tradition had never seen an outrigger-canoe and I am convinced that a misunderstanding of facts is involved, leading to a misrendering of the technical term. From the use to which the korewa were put according to the accounts quoted, I am inclined to infer that they were light timbers tied to the outside of the hull to give additional buoyancy rather than that they were true outrigger-booms of which the canoe must already have had her complement when leaving shore. Even in Best's account, korewa is interpreted both as outrigger and sea-anchor. 1

- 137

As regards Friederici's inference (8, p. 30) that Tahitian canoes were anciently provided with a double-outrigger because drawings made by early voyagers show the outrigger-booms (kiato) as projecting outboard to two-thirds their length on the side opposite to the single-outrigger which they possessed, I am satisfied that the explanation is otherwise. Actually it is only one of the two booms which projects on the off side as stated; this, the fore boom, is prolonged outboard to a distance of several feet solely to afford an easy means of securing a balance-board securely in position, the two being lashed together to effect this purpose. Sailing-canoes fitted with this form of balance-board still exist in the group, where I have had opportunity to study their construction in detail.


This island was discovered by Roggeveen in 1722; he describes the canoes but fails to mention the kind of outrigger used. Of others who visited the island in the same century, we find that all those who make mention of the islanders having outrigger-canoes, agree that this was a single-outrigger.

Thus one of Gonzales’ officers writes in 1770 (9, p. 121): “Their canoes are constructed of five extremely narrow boards (on account of there being no thick timber in the country) about a cuarta in width; they are consequently so crank that they are provided with an outrigger to prevent them from capsizing.” Note the use of the singular in the reference to the outrigger.

- 138

Forster, who accompanied Cook in his visit in 1774, describes (6, vol. 1, p. 558) a canoe as having the “head and stern raised considerably, but its middle was very low. It had an outrigger or balancer made of three slender poles” (two booms and a float?).

La Perouse was the next to visit the island (1786) and although he makes no mention of an outrigger in the text, an engraving is given (13, vol. 1, p. 76) of a small two-ended canoe with upturned ends as described by Roggeveen and Forster. A single outrigger is present, formed of two straight thin rods as booms, with a third, representing a float, resting upon it but showing no visible attachment, and too thin to be of any utility.

Beechey, whose ship, H.M.S. Blossom, called at Easter Island in 1825, contents himself with saying (1, vol. 1, p. 40) that the three canoes seen, “resembled the drawing in Perouse's Voyage.”

The only other reference to the outrigger of these canoes is a puzzling one by Choris who was aboard the Rurik when she touched at Easter Island in 1816. As his party remained “scarcely five minutes on the island” owing to the hostile attitude of the inhabitants, he had scant opportunity to note clearly the form of the canoes. He describes the only two seen as “chétives, pourvues de balanciers, et portant chacune deux hommes.” Sketches are given of these canoes (4, pl. 10, figs. 1 and 2); one is provided with a double-outrigger with two short, slender, straight booms, to which is tied on each side an equally slender rod, the homologue of a float; the other canoe is depicted with a curious bow-shaped fitting at each end which is certainly not an outrigger. Whatever value may be attached to these sketches—Choris’ figures are by no means accurate in details, as instance his clumsy drawings of Hawaiian canoes—there is no doubt that from the time of Gonzales to that of Perouse (1770 to 1786), the canoes of Easter Island had but a single-outrigger, a feature also certified by Beechey in 1825. It is therefore clear that the type of outrigger was single during the period of early European contact.

Even if we were to admit that Choris' figures are accurate, all they would prove is that many years after the discovery of the island, a modification took place which - 139 we cannot characterize as other than degenerative. It is only one of his two figures that is fitted with a double-outrigger, and that of a flimsy functionally useless type, the other with a fitting which cannot be considered in any way as an outrigger. Nor must we forget that the opportunity which Choris had to observe accurately was of the slightest—“scarcely five minutes.”


The statements regarding the former existence of double-outriggers in this group rest on records left by De Quiros, the Portuguese pilot of Mendaña's expedition (1595) and by Commodore Porter in the early years of the nineteenth century.

With respect to the former, the primary difficulty is the lack of an accurate translation of the original wording. Both Burney and Dalrymple include translations in their collections of Voyages, and Sir Clements Markham (11) has also provided an English version of the standard Spanish text by Don Justo Zaragoza (17). As none of these gives a sufficiently critical rendering of the passage which concerns the present problem, Zaragoza's text will be given, with a fresh and more literal translation.

No sooner had the Spanish ships been sighted off the island of Fatuhiva, named Magdalena by Mendaña, than a multitude of natives put off in canoes to meet them. Zaragoza's text (17, vol. 1, cap. VI, p. 36) reads as follows:

“De un puerto que está junto á un cerro ó picacho que queda á la parte del Leste, salieron setenta canoas pequeñas, no todas iguales, hechas de un palo, con unos contrapesos de cañas por cada bordo, al modo de postigos de galeras, que llegan hasta el agua en que escoran para no trastornarse, y bogando todas sus canaletes.”

Of this the following is offered as a critical translation of the essential technical matter:—

“From a harbour under a peaked hill which lies on the eastern side, there came out seventy small canoes, not all the same size, each made out of a tree trunk, with counterpoises of canes on each side, after the manner of the outboard rowing galleries (postigos) of galleys, which reached down to the water on which they rested to prevent capsizing; all rowing with their paddles.”

- 140

The significance of this passage lies in its reference to what authorities have hitherto considered to indicate canoes fitted with outriggers, from the phrase “contrapesos de cañas, ” translated by Markham as “outriggers of cane on each side after the manner of the gunwales of galleys.” The rendering of “postigos de galeras” as “gunwales of galleys” is a faulty and meaningless phrase. What De Quiros meant is clear enough; in the larger galleys of his day in the Mediterranean, a horizontal platform running the whole length of the waist of the vessel projected outboard on each side, to give greater leverage for the galley slaves to ply the great sweeps which formed the primary means of propulsion. This surely signifies something different from the double-outrigger as known in Indonesia. Had writers not been obsessed with the idea of outriggers, the description would convey rather the vision of a bundle of reeds or canes lashed outboard below the gunwale of the canoe on each side. It is important to note that the craft mentioned were specifically referred to by De Quiros as very small (canoas pequeñas); in small dugouts such as these were, the freeboard would be only a few inches; a bundle of canes or two or three bamboos lashed to the sides, would reach down to the water as described by De Quiros and would be exactly parallel in position to postigos de galeras, the outboard galleries of galleys. This is actually a method in use for preventing the capsizing of narrow, unstable canoes in Burma, the Philippine Islands, and the Solomons (Guppy, 1887, p. 149); the same principle has also been adopted on the Colombian coast of South America, probably introduced by the Spaniards from a Philippine Islands' model (Hornell, 10a). The employment of the word cañas, literally ‘canes’ or ‘reeds, ’ is particularly significant, for this material, even in the form of bamboos, is not employed in Polynesia to form outrigger-floats, which is the equivalent of contrapesos if applied to an outrigger canoe. The use of the plural form is also significant as the float is invariably a single log of wood in Polynesia.

There is also the possibility that an actual mistake was made in the original writing of De Quiros' account. Linton (18, p. 347) in commenting on a statement by De Quiros that the Marquesans hewed out their canoes “with adzes made of thick fish-bones or shells, ” suggests that - 141 De Quiros in writing up his experiences confused the Marquesas with some of the low coral islands which he had visited earlier in the voyage, where indeed they did use such tools. All evidence is against this statement by De Quiros; if lie is manifestly wrong here, this lends weight to the possibility that he is wrong also in his description of the outriggers of Marquesan canoes.

A passing expression used by Porter who visited the islands in 1813, is held by some to furnish corroboration of the former existence there of the double-outrigger. Porter in describing the method of rowing in Nukuhiva (14, vol. 11, p. 73) says that the fishing canoes were “managed with paddles more resembling an oar, and are in some measure used as such, but in a perpendicular position, the fulcrum resting on the outriggers projecting from each side.”

Every other voyager who touched at the Marquesas from Cook (1775) onward, describes and figures single-outriggers only; the explanation of the seeming contradiction in Porter's account is that the single Marquesan outrigger was and is characterized by the proximal ends of the outrigger-booms projecting some distance outboard on the side opposite to that on which the float is attached; this is clearly seen in Hodges’ figure in Cook's Second Voyage (1777, vol. 1, pl. 33) and in the larger outrigger-canoes (vaka poti) still in use in the islands. This projection on the off side is sufficient to permit of its use as a fulcrum for an oar, equally with the full length of the boom on the float side. Besides, unless the fulcrum be against or on the side of the canoe, the oar could not be used in an upright position.

From these considerations we may conclude that neither can De Quiros' passage be considered as conclusive of the former occurrence of double-outriggers in the Marquesas, nor can Porter's reference be taken as having a similar import or as confirming it.

Coming to 1775 when Cook visited the southern group, and to the period covered by the next few decades, we find constant reference to the use by the islanders both of single-outriggers and double-canoes, (but not of double-outriggers), and of them being paddled in the usual fashion by short paddles used sitting; what Porter saw was - 142 probably some exceptional form of sculling akin to that practised by Fijians.


There can now be little doubt that the Polynesians and Micronesians, save probably the Pelew Islanders, never possessed double-outriggers. This overcomes the difficulty—the “pretty problem” as he calls it—which Dixon experienced in explaining how the older form of outrigger, which he considers to be the double form, can monopolize a very extensive nuclear area and yet have a marginal distribution.

With the elimination of the Marquesas Islands, Easter Island, Samoa, the Society Islands, and New Zealand from the geographical range of the double-outrigger, it becomes clear that the single-outrigger has true and unequivocal marginal distribution as it is found throughout Polynesia and Micronesia in the East, and in India, Ceylon, the Maldives, the Andamans and Nicobars in the West. This leaves the double-outrigger to occupy a nuclear area in Indonesia with an extension to Madagascar and part of the east coast of Africa; this extension cannot be considered as primarily marginal, for it occurred at a comparatively recent period and may be considered as Colonial Indonesian just as the dominant culture of America is Colonial European, of a modern type.

The further consideration and analysis as above set forth of the geographical distribution of the double-outrigger and the single, confirm the view I put forward in 1920 that the single-outrigger is the older and the double-outrigger the later in time (10, pp. 104-111).

- 143
  • 1. BEECHEY, CAPT. F. W., Narrative of a Voyage to the Pacific and Beering's Strait in H.M's ship Blossom in the years 1825-28.” London, 1831.
  • 2. BEST, ELSDON, “Did Polynesian Voyagers know the Double Outrigger?” Journal of the Polynesian Society, vol. 32, pp. 200-214, 1923.
  • 2a. BEST, ELSDON, “The Maori Canoe, ” Dominion Museum Bulletin, No. 7, Wellington, N.Z., 1925.
  • 3. BROWN, GEO., Melanesians and Polynesians, London, 1910.
  • 4. CHORIS, LOUIS, Voyage pittoresque autour du monde, Paris, 1822-26.
  • 5. DIXON, ROLAND B., The Building of Cultures, New York and London, 1928.
  • 6. FORSTER, GEO., A Voyage round the World in H.B.M's sloop Resolution during the years 1772-5.” London, 1777.
  • 7. FRASER, CAPT., Nautical Magazine, vol. 3, pp. 74-75, 1834.
  • 8. FRIEDERICI, G., “Die vorkolumbischen Verbindungen der Südsee-Volker mit America, ” Mitt. aus den deutschen. Schutzgebieten, Band 36, Heft 1, Berlin, 1928.
  • 9. GONZALES, DON FELIPE, “The Voyage of Captain Felipe Gonzales in the ship San Lorenzo to Easter Island in 1770-1, ” Hakluyt Society, Ser. 2, vol. 13, Cambridge, 1908.
  • 10. HORNELL, J., “The Outrigger Canoes of Indonesia, ” Madras Fisheries Department Bulletin, No. 12, pp. 1-114. Madras, 1921.
  • 10a. HORNELL, J., “South American Balanced Canoes, ” Man, No. 102, 1928.
  • 11. MARKHAM, SIR CLEMENTS, “The Voyages of Pedro Fern. de Quiros, 1595-1606.” Hakluyt Society, Ser. 2, vols. 14 and 15, 1904.
  • 12. MUELLER-WISMAR, W., “Austroinsulare Kanus als Kult-und Kriegs-Symbole, ” Baessler-Archiv, Band 2, Heft 5/6, pp. 229-249, 1912.
  • 13. PEROUSE, J. F. G. de la, The Voyage of La Perouse round the World in the years 1785-1788 (Translated). London, 1798.
  • 14. PORTER, CAPT. DAVID, Journal of a Cruise made to the Pacific Ocean in the U.S. frigate Essex in the years 1812, 1813, and 1814. Philadelphia, 1815.
  • 15. QUIROS, PEDRO FERNANDEZ de, See under Markham and Zaragoza.
  • 16. SMITH, S. PERCY, Hawaiki, the Original Home of the Maori. London, 1910.
  • 17. ZARAGOZA, DON JUSTO, Historia del descubrimento de las regiones Austriales hecho por el General Pedro Fernandez de Quiros, 3 vols. Madrid, 1876-1882.
  • 18. LINTON, RALPH, “The Material Culture of the Marquesas Islands, ” Bernice P. Bishop Museum. Memoirs, vol. 8, No. 5. Honolulu, 1923.
1   The passage quoted by Best is somewhat obscure, and its value remains in doubt while its source is withheld. No light is available on korere. Best is hardly justified in claiming the passage as “a plain statement that Takitimu was a double-outrigger vessel.” It is said that there were korere on each side of the vessel; but while korewa are spoken of in the plural it is not said that they were on each side. Two expressions used of the korewa raise difficulty. Ka whakamau ai nga korewa moana ki unga, implies that the oceangoing korewa were fastened upon something, possibly upon the gunwales: in this position they might serve to give bouyancy to the canoe when capsized, and would not be any hindrance to the swimming crew in their efforts to right her; effective double-outriggers would be an almost insuperable bar to righting. The second expression, katahi ka kokiri i nga korewa o te waka ki waho, is even more difficult as kokiri means strictly to propel or thrust longitudinally, which would be applicable neither to outrigger-floats nor to floats on the gunwale. Best is speaking loosely when he says (p. 210), that korewa was applied to a drift-anchor or sea-anchor; the expression used for sea-anchor in punga korewa, where punga is anchor and korewa is an epithet implying bouyancy.
In the first passage quoted from Best it is said, ko te take o te korewa hei tiaki i te waka which can hardly have any meaning other than that given by Best: “the use of the korewa is to safeguard the vessel.” But the meaning of tiaki is to guard, watch over, keep, and it is very doubtful whether an old Maori would have used the word in the sense of safeguard, which is needed here: this adds force to Hornell's contention that the account was given in relatively modern times by a Maori who had not seen the contrivances he endeavoured to describe.—H. W. W.