In Search of the Ancient Polynesian Voyaging Canoe (1998)
Polynesia began with the voyaging canoe. More than three thousand years ago, the uninhabited islands of Samoa and Tonga were discovered by an ancient people. With them were plants, animals, and a language with origins in Southeast Asia; and along the way they had become a seafaring people. Arriving in probably a few small groups, and living in isolation for centuries, they evolved distinctive physical and cultural traits. Samoa and Tonga became the cradle of Polynesia, and the center of what is now Western Polynesia.
More than two thousand years ago, Polynesians exploring eastward, during times when winds shifted away from the prevailing easterlies, discovered the Tahitian and Marquesas Islands. From these "centers of diffusion" explorers reached outward as far as Hawai'i to the north, Easter Island to the east, and New Zealand to the southwest. Before European open ocean exploration began, Eastern Polynesia had been explored and settled.
Canoe Design Evolution
Because the exploration and settlement of Eastern Polynesia originated from the same centers, the design of the canoes must have been much the same throughout. But that design disappeared. Ships are as mortal as their makers. Except for fragments of ancient canoes excavated on New Zealand and pieces of a large canoe recently unearthed from a bog on Huahine, there is no hard evidence.1 Except for a petroglyph on Easter Island, and passing references in the old legends, there is no descriptive record.
The Easter Island canoe petroglyph found at Orongo and Herb's rendition of what the original canoe may have looked like.
Over the following centuries, this "archaic" form evolved into designs which became "classical" to each island group-specialized to meet the challenges of local winds and seas and timber resources. When Europeans arrived, they found pronounced differences in canoe designs from one island group to another.
One such design change was witnessed by Europeans. Schouten in 1619 saw only the tongiaki double canoe in Tongan waters. When the Cook expedition arrived in 1773, the drawings of double canoes by the artist Hodges depicted a transition from the tongiaki to the swift kalia--a borrowing of the Micronesian "double-ender" concept. When Cook visited again five years later, the artist Webber's drawings suggest that he saw only the new kalia.2
Design Strategy for Hokule'a3
What were the design features of the ancient double-hulled voyaging canoes (vaka taurua)? Applying the "age-distribution" method, we assume that similarities in hull shape, sail shape, and construction techniques which were widely distributed when Europeans arrived must have been carried outward from the centers of cultural diffusion during ancient eras of exploration and settlement, and may be accepted as features of the ancient canoes. By limiting the design of a voyaging canoe to these features, a performance-accurate replication of the ancient canoe is possible. Such features formed the design vocabulary of the replica Hokule'a.
Construction Drawings for Hokule'a
Kenneth Emory and I went through all designs of canoes recorded in early drawings and in other evidence and sifted out those features of hull design and sail plan which by their wide distribution may be taken to be most ancient. These I applied to the conceptual design. From my own experience with the Pacific swell and in consultation with more experienced sailors, I arrived at a waterline length of 55 to 60 feet as one that could handle the swells yet recover easily in the troughs, and Emory found that this could be taken as an average for the length of canoes used in the 18th and 19th centuries for long distance voyaging in the Tuamotus and Tahitian islands. Canoes of far greater length would put great stress on the lashings. The double-ended ndrua of Fiji and kalia of Tonga were of greater length, but these carried a smaller hull to windward involving less stress on lashings than two hulls of equal length, and were generally used on shorter voyages, which means during periods of predictable weather. (Click here for a photo of "Takitumu," a modern reconstruction of a Cook Islands kalia.)
Some classical Hawaiian features were included which did not affect performance, such as the styling of the bow and stern pieces (manu) and the arched cross-beams (not really ancient, invented by Kanuha several centuries ago).
Kino (Hull): Hulls were carved from logs wherever timber of sufficient size was found. The depth of a hull might be increased by adding one or two courses of boards (strakes) fitted and lashed above the hull's upper edges (gunwales). On atolls where large timber was not available for dugout hulls, the use of gunwale strakes was transformed into a method of building entire hulls "plank-built" over a dugout keel piece, with ribs and thwarts inserted to strengthen the planking.
All hull and sail design features must be compromises. Where paddling was the primary power mode with sail as auxiliary power, round-bottomed hulls were favored for their maneuverability; but where sailing was the primary purpose, hulls were deeper or had a greater amount of "V" shape along the keel for better tracking through the water. Such hulls are less maneuverable but offer lateral resistance to the water, reducing leeway (the sideways skidding of a boat hull away from the wind when sailing against or across the wind). Because the great distances covered by some ancient voyages could not have been accomplished by paddling, we may assume that the voyaging canoes were primarily sailing machines, with paddling being auxiliary. These were not the flat-sided "V" hulls of modern catamarans, but a rounded "V" by which the maximum floatation capacity could be carved from the natural shape of a log. A rounded "V" hull, with the sides swelling outward in convex curvature, is also stronger than a flat-sided "V" hull because it adds the strength of an arch against the impact of waves.
Where hulls were of unequal length, the smaller hull was carried on the left, and called the ama-the same term for the float outrigged from the left side of single-hull "outrigger" canoes. The one exception is the island of Tubuai in the Australs, where the ama is carried on the right; but today no canoe maker on Tubuai can explain why.
Below the waterline the curvature of all Polynesian hulls is convex, both in length and in section, with no cavities. Longitudinal curves below the waterline are smooth-flowing from bow to stern, creating a gentle entry at the bow and an equally gentle departure at the stern-features necessary for a "soft" ride and maximum hull speed. In these curves there are no abrupt breaks-no "chisel" bows to snag the water and make steering difficult, no abrupt departure at the stern which creates turbulence. For best speed the hull curves are faired out as much as possible (canoe builders knew how to use a flexible fairing strip to check their hull curves), with no hollows or flat areas to cause turbulence.
The volume of Polynesian hulls aft of the midsection is slightly greater than the volume forward of the midsection. This extra floatation aft offsets the tendency of canoes to "squat" at the stern when under a hard press of sail.
Because double canoes are held together by rope lashings, the hulls must be assembled closer together than the hulls of modern catamarans. This narrows the space through which water must pass between the hulls. To avoid excessive turbulence between the hulls, the greater volume aft of the midsection should be obtained by greater hull depth, rather than increasing hull width.
Pe'a (Sails): My first preliminary drawing for Hokule'a (1973) featured triangular sails carried with the peak of the triangle downward and mounted on straight spars, a design which by its simplicity and wide distribution seemed to be the most ancient form. This sail plan was modified later in 1975 and again in 1976 with a curved boom to more closely resemble the Hawaiian sails at the time of European contact. However, experiments in 1991 and subsequent voyages have demonstrated that the simple triangular sail carried on straight spars is no less efficient; moreover, it is easier to furl and handle on deck when the rig is dropped to ride out bad weather.
Sails were of pandanus matting except in New Zealand, where pandanus could not be naturalized and flax was substituted. Sails were cut from long rolls of matting seldom more than 18" wide, double plaited of strips 3/16" to 3/8" wide in a twill pattern, changing to a check pattern along the edges for strength.
The sail was built up by overlapping the edges of these strips and sewing them with a running stitch. The outer edges of the sail were hemmed over a rope. A line was fastened with a running hitch at intervals along the outer edges. This line was then tied to the spars with a spiral lashing or with many short lengths of line.
Curved booms, if desired, could be scarfed up from shorter poles to achieve the desired overall curvature and length. The long scarf joints were strengthed with splints and seized up with small line. Spars could also be strengthened at those places where sheets and stays were attached by seizing splints to them.
For a large voyaging canoe having no labor-saving winches, two sails are easier to handle than one large sail. The foresail should be the larger. By distributing the effort over two sails, the moment of capsize is lowered, imparting greater stability to a vessel which, being held together by lashings, is necessarily narrower than a modern multi-hull. If the vessel appears to be overpowered while sailing off the wind, sail area can be quickly reduced by dropping the aftersail.
'Iako (Connecting Cross-beams): A true replication of an ancient canoe should have crossbeams shaped from straight poles-the method most widely distributed. The arched crossbeam is a feature of the classical Hawaiian double canoe, invented only four centuries ago by the designer Kanuha in the time of Keawe.4
In a quartering sea the hulls of a double canoe will work against each other. By inter-connecting the crossbeams with diagonal bracings of strong rope, this motion can be restrained, adding very little weight to the vessel.
While most cross-beams were lashed to the gunwales, the connection of the two hulls could be strengthened by two or more lower crossbeams let through the hulls, as seen in the drawing of a beached Tahitian double hulled sailing canoe by Webber, with Cook.5
The flashing speed of modern catamarans results from their wide beam and rigidity of construction, made possible by steel fastenings. The vaka taurua is a slower sailer. Assembled with cordage, it lacks the rigidity of modern multihulls, and the hulls must be closer together to reduce stress on the cross-beams. Assembly by lashings seems to offer one advantage. As noted on the replica Hokule'a, the cross-beam lashings absorb much of the shock of waves that beat against the hulls, a pounding that is transmitted throughout a modern vessel.
Pola (Decking): Decking may be of light planks if these are supported by a webbing stretched between the crossbeams. Lighter planking means less weight. Deck planks should be spaced with gaps through which heaping waves can rise. Without such gaps to relieve wave pressure, strong surges can break the decking. In this compromise, it's better to be safe than dry.
Planks can be added over certain areas of the windward hull on long tacks, even out to the ends of the crossbeams, where they will deflect the splash of waves, and serve as hiking boards for the crew during gusts of wind.
Mast steps: Wind pressure on the sail drives the mast downward. Such pressure should not be borne by only one crossbeam. The masts may be stepped upon strong longitudinal beams (kua), each distributing the downward thrust over a least three crossbeams. Once the optimum center of effort is found by experimentally moving the masts forward or aft over these steps, additional crossbeams may be added under those points.
Manu (Bow and Stern pieces): As I discovered while sailing Hokule'a, end pieces have a practical function. Eastern Polynesian end pieces typically rise higher at the stern than at the bow. The sternpiece appears to break a following wave crest that might otherwise board the canoe. When the canoe surfs on a following wave, plunging forward, the bowpiece, in a burst of spray, helps prevent the bow from "boneyarding" into the back of the wave ahead.
As expressed in the carving of end pieces, symbolism associated with birds or bird-man (manaia) forms was widely distributed. The term manu for the abstract shape of the classical Hawaiian end piece suggests that the archaic form may have represented birds. A pre-classical Maori bowpiece unearthed on New Zealand has a long neck and the head of a bird-man figure. European drawings of some Marquesan canoes, and old Marquesan canoe models, have bird-like shapes when viewed in profile, with the head at the bow, the gunwale strakes resembling wings, and the stern-piece appearing as the tail. Feathers were widely used as pennants flown from the end of a spar (Tahiti, Hawaiçi), or black feathers hung from the stern piece (New Zealand); as bunches of feathers at the stern (Marquesas); and as feathers worked into the gunwale lashings (New Zealand, Marquesas).
Hale (Deck Shelter, pronounced "ha-lay"): This may be a construction of light poles and purlins covered by thatching and/or tightly plaited matting. The shelter should be easily moved. On long reaches or tacks it should be positioned over the windward hull.
Sailing the Vaka Taurua
Steering: The idea of steering a sixty-foot multihull without a rudder has intrigued conventional yachtsmen on their first sails aboard Hokule'a. On a downwind course the steering paddle is handled in the manner of a rudder, and long sweeps were used on some Polynesian canoes. On any other tack, however, the steering paddle is held against the lee side of the hull near the stern. The pressure of the water against the blade helps hold it fast, and very little effort is required to hold a heavy steering paddle in place. A slight twisting pressure to hold the leading edge of the blade firmly against the hull prevents the flow of water from getting under the blade and kicking it away. For this reason, steering paddles were often carved flat on the side held against the hull, and concave on the other.6
At a canoe's first sea trials, the masts should be experimentally shifted forward or aft until the center of effort is balanced with a slight weather helm, so that when the paddle is raised, decreasing its lateral resistance to the water at the stern, the stern will fall off the wind, turning the vessel into the wind. When the paddle is lowered, creating more lateral resistance at the stern than exists at the bow, the canoe will turn off the wind.
The Polynesian paddle creates less drag than the modern rudder, and is put in the water only when needed.
Steering with the Sails: On long reaches, steering paddles may not be needed at all; the canoe can be rigged to steer itself by sails alone. The aftersail is eased out slightly more than the foresail. As the canoe rounds up into the wind, the aftersail luffs and loses power. Pressure on the foresail now causes the vessel to turn off the wind a few degrees. The aftersail is again presented to the wind; it fills, and the vessel begins another slight turn to windward. Sawing slightly into the wind and off the wind, the canoe will steer itself on a close reach for hours.
Tacking: In light or moderate winds the double canoe will come about (turn into and through the eye of the wind) without stalling if the crew backs the foresail, harnessing the wind to push the bows over. In a strong breeze, however, it's difficult to come about without sailing. Then it is better to jibe (make the turn with the wind astern) by luffing the aftersail until the foresail powers the vessel well off the wind, then close-hauling both sails as the stern passes through the eye of the wind. Here, the blades of the steering paddles are held at full depth to grip the stern in the water. A double hulled vessel is slow to turn because its two hulls give it twice the waterline length of a sailboat of the same length.
Shortening the Sails: Sails and booms were brailed up to the mast while temporarily not in use.
In the path of a dangerous squall, the prudent act would be to release stays and drop spars and sails, lowering the center of capsize as much as possible. The canoe is brought into the wind, and the forestay (a running line) is eased out, lowering the mast aft. The shrouds will hold the mast in fore-and-aft alignment with the canoe as it comes down.
Lacing between sections of sail matting might be quickly removed to shorten sail.
Storm sails may be simple, small, strongly made sails lashed to short straight spars with stays and shrouds already attached. Rolled up and stored away, these can be raised to power a canoe in moderate gales.
Leeboards: The use of leeboards to diminish leeway and help a vessel without a keel go to windward was a Chinese invention which never got to Polynesia, but the same effect was accomplished, when required, by a row of men holding paddles against the lee side of a hull. This takes practice, but it can add ten degrees to a canoe's windward performance.
Storm Survival: An approaching storm meant getting down sails and spars, even jettisoning the deck shelter if necessary to reduce windage, and laying out a sea anchor on a very long line. Strong baskets are said to have been used in Hawai'i.
If necessary, the next step would be to deliberately swamp the canoe, a technique that modern minds find incomprehensible, but which is still commonly practiced in Micronesia. Wooden hulls provide sufficient floatation so that the crew can ride within the hulls with their heads and shoulders above water.
Being mostly under water, the canoe will not be buffeted about. Most important for navigation, it will not skate downwind, but will hold position fairly well.
After the storm has passed, the hulls can be bailed out and the voyage resumed. It's helpful to have additional positive flotation to raise the gunwales with enough freeboard to facilitate the bailing. On the old canoes, coconuts served as positive floatation as well as providing food and drink. These, and other cargo, were held down in the bottoms of the hulls under netting. Today, any inflatable devices secured under netting can be inflated to give the hulls more freeboard. Rigid foam or empty containers packed under the bow and stern covers will also add floatation.
1. "The main evidence that we have of what the voyaging canoes were like came from an island called Huahine, which is about 110 miles west of Tahiti. In an area called Maeva, a hotel called the Bali Hai was being built and when they were digging up the ground, they found some canoe bailers. They called Dr. Sinoto from the Bishop Museum and he went down and conducted an excavation. He thinks that six hundred to a thousand years ago there was a canoe under construction here and the work place was hit by a tsunami which buried the canoe under mud and sand and preserved it by cutting off the oxygen that causes wood to rot.
"Sinoto unearthed planks of the canoe with coconut fiber (aha) still holding them together. There was a knot in the plank and what the builders did was to put wood in from behind and lashed the two pieces of wood together to make a sandwich. Dr. Sinoto guesses that this canoe was 72 feet long, ten feet longer than Hokule'a" (Nainoa Thompson, Speech at Kamehameha Schools, April 1998).
2. [Haddon & Hornell explain, "The principal features wherein the tongiaki differed from the kalia were: (1) in the approximate equality of the two hulls; (2) in sailing, the same ends were always directed forward and in consequence the manipulation of the sail was entirely different; (3) the mast was much shorter and had a forked head in which the yard rested, with the tack of the sail confined by ropes between the prows and not stepped at the fore end of one hull; (4) the presence of an outrigger balance spar; (5) the deck platform was relatively larger and extended considerably farther aft than in the kalia; (6) the deck shelter was a tunnel-shaped hut without a platform above the roof. (Canoes of Oceania, Bishop Museum, 1936, reprinted 1975, 271-272).] (Click here for a photo of "Takitumu," a modern reconstruction of a Cook Island kalia.)
3. Kane lists the following people as important contributors to the design and building of Hokule'a:
Kenneth Emory: consultation on ancient design features.
Herb Kawainui Kane: conceptual design and establishing design parameters in a general drawing.
Rudy Choy: consultation and advice on hull design.
Vince Bartelone: the hull-lines drawing
Kim Thompson: the lines drawings for the end pieces (manu)
Warren Seaman: lofting the hulls and laying up sections and stringers on the strongbacks.
Curt Ashford and Malcolm Waldron: chief shipwrights.
Jim Ebersold, Calvin Coito, and others: boat carpenters.
Wright Bowman, Sr. and Wright Bowman, Jr.: crossbeam construction.
Keola Sequeira: carver of the koa mast heads (he donated the koa).
Bob Fortier: protective fiber-glassing of the wooden hulls.
Kane recalls: "These were most of the hands-on guys. Additionally there were part-time carpenters and a host of volunteers. Most of the board members donated some time. Publisher Carl Lindquist brought his family down to help with the disagreeable task of poisoning the interiors of the hulls to prevent rot. Kawika Kapahulehua facilitated air transportation for materials and supplies. Slim's Power Tools donated the use of power tools. Many merchants helped procure supplies and materials at cost. Dillingham Corporation donated the use of the building premises. I don't remember the name of the trucking company that gave us a price break to haul the completed parts over the island to Kualoa Park (how I obtained that launching site from mayor Frank Fasi after the State gave me the run around for months on San Souci beach is another story). The U.S. Marines at Kane'ohe brought equipment to lift and set the hulls in perfect position for the lashing up, which was done over five weekends by many volunteers.
4. Malo, David. Hawaiian Antiquities. Honolulu: Bishop Museum. 130.
5. Beaglehole, The Voyage of the Resolution and Discovery, Cambridge, 1967, Pl. 25-b.
6. "In the earliest days of Hokule'a, when she was going through sea trials, the steering mechanism was like the kind we use on a six man canoe but the problem was that a six-man canoe weighs 600 pounds and Hokule'a weighs-fully loaded-about 24,000 pounds and guys who were trying to steer her were getting knocked out, ending up in the hospital, and it wasn't working and I was looking at that thing and saying, 'Oh, man, it is really going to be a long trip.' And so some of the beach boys said, 'We are not going to do it this way anymore. We have got to come up with a steering sweep.' And so they designed a steering sweep that would work-from their experience at Waikîkî-and it solved the steering problem. And the steering sweep that they pulled out of that swamp in Huahine [see note 1]-it was discovered after the beach boys figured out how to design a sweep for Hokule'a-and this sweep [excataed in Huahine] is very similar in design and only two feet shorter than the ones we use to steer Hokule'a" (Nainoa Thompson, Speech at Kamehameha Schools, April 1998).