Ethnology of Tongareva
Kinds and Uses
The houses (hare) now in use in Tongareva differ from the old pattern in that they are made of sawn timber, erected on high piles, and roofed with corrugated iron. No perfect ancient house was seen in either of the two modern villages, but from the rough houses seen in the food plantations and the verbal description obtained, they were evidently built on the same pattern as the common rectangular house of the Cook Islands. They differ in the treatment of the walls and in the names of some of the parts of the framework. The guests of a family were lodged in one of the dwelling houses. Distinguished visitors were housed at the hare nui of the local chief, if he had such a building. A large party received at a social center was lodged among different members of the local community. The local people could always crowd together and free some huts for the use of visitors. No great inconvenience was caused, as the only pieces of furniture needed were a few mats for the floor. To mark the occasion a fresh layer of coral gravel was sometimes spread over the floor. Large parties of visitors were not housed but were given a camping ground. If doubt existed as to the visitors' intentions the camping ground would be removed from the local center, or on another island. The rough shelters of coconut leaves were quickly made by the visitors themselves. Lamont (15, p. 211) remarked that while he was at Omoka, “the people of Matunga came over to see me, and encamped along the water's edge, where they remained for some time, having thrown up little temporary huts very quickly.”
The clear graveled space before the dwelling houses formed the family meeting places. The hare nui of a chief and a clear space at the social center formed the rallying places not only for the members of the community but for the official reception of guests. For religious purposes and the exercise of customs, the stone-inclosed marae was the assembly place.
Common Dwelling House
The ordinary house was small. Many of the house sites measured are 9 feet long by 6 feet wide; some are even smaller. However, as public congregations took place on the marae and people sat out on the graveled open spaces in front of their huts, the small size is accounted for by the use of the huts as bedrooms. The chiefs had larger houses. Lamont noticed the larger houses of the principal chiefs he visited at Omoka, Motu-unga and other islands. Turua's house at Motuunga he describes (15, p. 185) as hav- page 94 ing a framed roof about 10 or 12 feet square with the eaves about 5 feet from the ground. The house platform of the house of the high chief, Turua, in Motukohiti was, roughly, 38 feet long by 18 feet 6 inches wide, and though the house may not have covered the platform entirely, it was large.
Figure 3. House framework, a, side view; b, end view: 1, supporting posts of ridgepole (pou); 2, ridgepole (tahuhu); 3, wall posts (pou); 4, wall plates (hapai); 5, principal rafters (oka); 6, upper ridgepole (tamaiti); 7, purlins (torotoro); 8, thatch rafters (rau oka); 9, eaves rod (torotoro hiohio).
The house framework is shown in figure 3. A supporting post (pou) of coconut (niu) or hala (Pandanus; hara) was erected in the middle of either end of the rectangular ground plan. The ridgepole (tahuhu) was placed in position on the supporting posts. Side wall posts, also called pou, were used, and the wall plates were lashed to them. The principal rafters (oka) were stretched over the wall plates and ridgepole, with the upper ends crossing to form a support for the upper ridgepole, called tamaiti (child) to convey the idea of its being the child of the main ridgepole. About three horizontal purlins (torotoro) were used on either side on the outer side of the principal rafters. The thatch rafters (rau oka) were placed on the outer side of the purlins. The eaves rods (torotoro hiohio) were tied to the outer side of the lower ends of the thatch rafters. They supported the lowest tier of thatch and derived their name from torotoro (purlin) and hiohio (small), being thinner than the main purlins. The names which differ from those used in the Cook Islands are hapai for rape, torotoro for tarava, tamaiti for taorangi, rau oka for kaho, and torotoro hiohio for manuae or iniki. The thatch rafter name of rau oka (rau, thatch; oka, rafter) although appropriate, has displaced the widespread Polynesian term kaho, which occurs in some regions as kaho, 'aho, and 'aso. The term kaso is present in the Tongarevan dialect, but has been applied to the rod over which lauhala (leaves of Pandanus; Tongarevan, rauhara) is bent to form roof sheets.page 95
The thatching material was made of lauhala, which, like that of the lower Cook Islands, was first soaked in the salt water of the lagoon to kill the bugs and beetles. After being dried in the sun the leaves were rubbed against a stake stuck in the ground in the same manner as in Aitutaki (28, p. 13). The rubbing process is termed sahu (Aitutaki, oro). On the completion of the rubbing the tip end (siku) of each leaf was bent across the butt end (pu) and kept down with the foot at the bottom of the rubbing stake. When several leaves had accumulated around the stake they were tied together at the leaf crossings and formed a tupe bundle.
For joining the leaves into roof sheets, aerial roots (kaihara) of hala (Pandanus; Tongarevan, hara) were used. The roots were cut into three or four foot lengths, and the bark peeled off (ka sosore te kiri). The lengths were split (vavasi) into two kinds of rod, a thicker rod (kaso) and a thinner rod (haniu). The term kaso has been transferred from the original thatch rafter. The term haniu in Tongareva is also applied to the midrib of the coconut leaflet and is so used in the Cook Islands.
A wooden needle (tui) made of ngangie is used to sew the roof sheets of lauhala together. The implement resembles the au tui of the Cook Islands, except that it is wider and that most of them are sharpened at both ends. (See fig. 4.)
After material has been assembled, the sewing of the sheet commences (kua sausau te tui), as shown in figure 5. The sheets are three to four feet wide, according to the length of the kaso rod which stiffens the doubled-over edge of the sheet. The haniu rod keeps the leaves together. A large number of sheets are required and when they have been made the thatching commences.
In the Cook Islands the roof sheets are sewn together in exactly the same manner (28, p. 17). Here and in Samoa coconut leaflet midribs are used for pinning the leaves together. The use of thin strips of hala rootlet under the name of haniu shows that the leaflet midrib was also in use in Tongareva, and the name haniu was applied to the rootlet material after it had displaced the coconut leaflet midrib.page 96
The covering of the roof with lauhala is termed tapoki (to cover), but the actual thatching process is atohanga. Two-ply twisted sennit cord (hauato) is used for tying (sere) the sheets to the thatch rafters. In commencing, two sheets (parua) are laid together across the thatch rafters with their rod ends just above the eaves rod. The cord is tied to the lower end of the thatch rafters. The thatcher, from the inside of the house, carries the cord over the upper edge of the sheet on the right of the thatch rafter, crosses it over the thatch rafter on the outside, and after puncturing a hole through the sheet below the kaso rod on the left of the rafter, brings the cord through the hole to the inside. The cord is drawn taut with a half hitch around its standing part or with a single overhand knot. Single sheets are then added two or more inches apart, each being knotted in position with the continuous cord. The thatching works upward until the upper ridgepole is reached, when another double sheet (parua) finishes off the section. The roof is thatched in sections of the width of the roof sheets. The completion of the thatching is expressed by the phrase “kua tuku te raumanu,” raumanu meaning all the rau (leaf) sheets. In the Cook Islands a wooden needle with a hooked point is used in thatching (28, p. 22), but in Tongareva no mention was made of such an implement.
Figure 5.—Sewing hala (Pandanus) thatch sheet. a, technique: leaf (1) held with butt end toward worker and with back surface up; a kaso rod (4) laid across leaf about 10 inches from its butt end and leaf butt turned forward over rods; a second leaf (2) added on left in same way with side edge overlapping that of first by about 0.5 inches; the two leaves held together with left hand; right hand pushes the needle (5) up through both layers of first leaf beyond rod and to right of midrib; long diameter of needle held in same direction as length of leaves in order that it may split more readily through longitudinal fibres of leaf; point carried across midrib and passed back through overlapping edges of both leaves and up again through both layers of second leaf (2) on near side of its midrib; needle given half twist to open holes with its wide diameter; thinner haniu rod (6) is passed through the three holes as far as point of needle; a third leaf (3) added with overlap; needle removed; downward and upward piercing on third leaf (3) at parts marked by dotted lines repeated; and rod (6) pushed on through holes. b, section of completed sheet: other leaves have been added in manner indicated and needle removed; kaso (4) and haniu (6) rods in position.
Roof sheets of split coconut leaves with the leaflets plaited in check are used in the rougher houses. It is probable that lauhala was formerly not commonly used, for Lamont (15, p. 155) states that “their houses or sheds are made of the cocoa-nut branches.” The coconut leaf sheets are easier to make, as the leaflets are fixed naturally to the midrib strips and require no artificial kaso rods. In thatching, the midrib strip edges are placed above, and the tying with a continuous cord proceeds. The details of making the coconut leaf sheets are the same as in the Cook Islands (28, p. 7).
Ridge sheets (takaumisau) of plaited coconut leaf are laid longitudinally to overlap the upper edge of the roof thatching on either side. Pointed stakes of ngangie, termed soka, are pinned through below the upper ridgepole to keep the ridge sheets in position. Pinning on the ridge sheets is “sokahia te takaumisau.” The ridge sheets are also similar to those in the Cook Islands (28, p. 26).
The sides of the house are termed tua, and the ends, tara. Lamont (15, p. 185) states that the eaves of a large house on Motu-unga were about five feet from the ground. In the smaller houses the walls were much lower. The ends were thatched to about the level of the side eaves, and below that the house was open on the four sides. Shelter from wind or rain was provided by plaited coconut leaf screens.
There were two wall screens, a narrow screen (pataro) formed by plaiting the leaflets on one side of the midrib, and a wider one (pataro mangarua) in which the extra width was obtained by plaiting two sides of the midrib. Lamont (15, p. 185) noted that
… slight stakes, of about a foot high, are driven into the ground around the house, immediately under the eaves, and at night long narrow mats of cocoa-nut leaf are fastened to, or laid against, these pegs, forming a shelter from the blast. Otherwise the house is entirely open all round.
The mat was probably the long narrow pataro. The wider pataro-niangarua was attached by one of its split midrib edges to the wall plate above and hung down as a screen. To secure it during a strong wind, the lower midrib edge could be fastened to pegs below. To permit the circulation of air through the house, the wider screens were propped upward and outward by lengths of coconut leaf midribs. The wider wall screen is the one in common use.
The pasaka house was built without wall posts and wall plates. The lower ends of the principal rafters rested directly on the ground. The roof was page 98 made of lauhala sheets. (See fig. 6.) The pasaka was a recognized type of good dwelling house built for strong winds and hurricanes. It was not as likely to have the roof blown off as was the house with the roof raised above the ground. Since the people spent most of the daylight hours out of doors and sat out on the cleared space in the evenings, the houses were used for sleeping or resting in the recumbent position. Consequently, the low roof was not especially inconvenient, and any slight inconvenience was compensated for by the added security in windy weather.
Besides these dwelling houses, rougher houses were built on the food lands, where the people stayed at odd times and did not need well-constructed dwellings. The pasaka house involved less work as the essential uprights, ridgepole, and rafters which rested on the ground furnished sufficient framework for a roof of plaited coconut leaves. It was evidently a modified type of pasaka that Lamont (15, p. 113) described:
In recrossing I observed some low sheds, which, as I saw no other dwelling places, were doubtless the residences of the natives, though the most miserable shanties I had ever beheld for human beings, consisting merely of four inclined sticks, about five feet in height, with two uprights and crossbeam, forming a light frame for a small roof. At the base the house is some six feet wide by eight long, the whole covered by a thatch of the cocoa-nut tree, formed by splitting the bough and platting [plaiting] the leaves, till enough are linked together to reach about half way up the frame, on which it hangs so loose that it can be lifted or dropped at pleasure. Other boughs are then fastened on in the same manner, but secured to the frame till all is covered in, the ends being closed by platted boughs secured to the uprights. Those I saw on the present occasion had the leaves supported by a stick.
By “bough” Lamont means the coconut leaf with its midrib, and “leaves” are the leaflets. Clearly, the plaited roof sheets did not reach to the lower ends of the rafters, and the hanging plaited leaf that reached halfway up the frame was the wall screen (pataro) which was propped up by a stick, leaving an aperture through which Lamont crawled.
The Hare Pou
The hare pou was described by my informants as another type of house, but they knew nothing about its architecture. Their only knowledge was derived from a song (pese).
Poupou hare no Pou-toru,
Ka hanake koe mei Savaiki,
Ka ma tua koe ia Taritoa,
Ko hakaaroha, pinga te poi.
Posts of the house from Pou-toru,
You have come up from Hawaiki,
You are the senior generation through Taritoa,
Love wells up, all is good!
The significance of this song could not be explained in detail, and the rough translation may not adequately convey the meaning. The phrase, pinga te poi, was said to be short for pinga te ipoipo (very good).
Coconut leaf shelters were quickly made on the plantations from coconut leaves arranged roughly in the form of a bell tent. (See fig. 7.) A few of these shelters were seen on the islands not permanently inhabited, but the demonstration of the technique was, unfortunately, crowded out by the rush of other work. This type of shelter was probably made by visitors when they were given a camping ground away from the settlements. It required no wooden framework, a vast consideration with the tools in use.
A few mats were the essential house furnishings in Tongareva. Such personal property as sennit cord, fishing nets, and weapons was also kept in the dwelling.
Sleeping mats of lauhala were in common use, and two kinds of coconut leaf mat were described to me. The coconut leaf mats (pakere rei) were page 100 used as floor coverings, and shorter mats (tapakau) were used for seats. When the people sat outside for their meals or to rest, the tapakau were brought out of the house and spread on the coral gravel covering the open space.
Back rests (tokotua, from toko, to prop, and tua, the back) dubbed out of tou wood and propped up at an angle by a short stake, were described to me as of old invention. The sitting mat upon which the person sat with the back rest adjusted to a comfortable angle, was spread on the ground. (See fig. 8.) Lamont crawled into one of the small huts to observe the furniture, and he says (15, p. 114):
This was not of the most costly description. A roll of sinnet and a coarse bag-net were suspended from the ceiling, a rough mat of pandanus leaf partially covered the sandy floor, and another was thrown over what seemed a bundle at the far end. Curiosity tempted me to lift the latter, when I beheld an urchin whose little black eyes seemed fairly to start out of their sockets, as they stared at me; while his mouth, which was about as broad as his face, emitted the wildest screams of terror I had ever heard. Nearly as much frightened as the child, I backed out of his den and at the same moment the little imp, throwing up the opposite side of the house, darted into the woods; and, though his yells still were heard, modulated by distance, he was out of sight in an instant.