Arts and Crafts of the Cook Islands
As quadrangular adz heads were in general use in the marginal areas of Polynesia, it may be assumed that the quadrangular type was the original form before dispersal took place. The fact that a few quadrangular adz heads were in use in the Cook Islands at the end of the second period indicates that the quadrangular form was introduced in the first period.
The form of adz heads influenced the shape of the foot of the haft to which they were lashed. Three types of hafts have been described (fig. 101). The toe haft and the medium haft are associated with quadrangular adz heads and with triangular adzes which have a flat back surface. The heel haft was used with the inverted triangular adz heads which have a median edge on the back.
The toe haft has a wide distribution in Melanesia, New Guinea, and Micronesia. Its occurrence in the three corners of the Polynesian triangle (fig. 269) supports the statement that the quadrangular adz head, with which it was associated, was the earliest form of adz head in Polynesia. Unfortunately, none of the quadrangular adz heads obtained in the Cook Islands were hafted page 444but we may reasonably assume that the toe haft accompanied the quadrangular adz heads into the Cook Islands in the first period but gave way to the heel haft when the inverted triangular adz head came into general use.
The lashing technique used with the toe haft was transverse turns around the toe and the butt of the adz head (fig. 101, c). This simple pattern was effective because the toe was long enough to accommodate the full length of the adz butt, and toe and butt formed an uninterrupted column for the continuous transverse turns of the lashing element. In Polynesia, no artistic pattern was developed with this form of lashing; but in Fiji, a multiple lozenge pattern was produced by a technique similar to that of the Mangaian inaere rafter pattern (fig. 18).
The commonest form of adz head in the Cook Islands was the reverse triangular, and it is evident that during the period of occupation the triangular form superseded the quadrangular form in general use. However, as reverse triangular adz heads were also in general use in the Society and Austral Islands, it would appear that the change in fashion first took place in one of three adjoining groups of islands and spread to the other two. The Society Islands was the most likely focus from which diffusion took place. It is interesting to note that Te 'Erui, the second ancestor to arrive in Aitutaki from the Society Islands, brought with him an adz named 'Aumapu about which there is a chant (pe'e) and a dance. When the dance was demonstrated to me by the descendants of Te 'Erui, they boasted that their ancestor brought the first adz to Aitutaki. When I asked how the first ancestor, Ru, made his canoe, they replied "Tei 'ea te pe'e" (Where is the chant?). They admitted that Ru must have had adzes but suggested they must have been a poor kind not worthy of being recorded by a chant. There is a possibility that Ru had the old quadrangular form and that Te 'Erui came in later with the triangular form whose newly acquired status was celebrated in song.
Though the toe haft is ancient, it underwent in the second period a certain amount of variation in the treatment of the heel for aesthetic reasons. Some hafts have a prominent heel and were probably left so by the craftsman because he did not think it necessary to remove the upper projection on the original block or because he thought that a heel gave better balance or symmetry to the haft. In Niue, the toe hafts retained a prominent heel with a small upper knob for decorative purposes. The purely decorative heel reached a peak in the New Zealand carved hafts of ceremonial jade adzes (toki pou tangata) in which the heel was made large enough to be carved into a human figure. However, the toe of these hafts was shortened for better balance and the haft entered the medium class whereas the working adzes retained the toe type of haft.page 445
The heel haft (fig. 101, e) associated with the inverted triangular adz head is present naturally enough in a secondary central area comprising the Cook, Society, and Austral Islands (fig. 269). Ample evidence that the heel haft is pre-European is afforded by the Bishop Museum photograph catalog in which there are records of over 60 old Tahitian adzes in various museums and collections that all have the heel haft. Several old Mangaian working adzes show the heel haft (fig. 111), and all the Mangaian ceremonial adz hafts (fig. 107) have a foot without a toe. An apparently old Rarotongan hafted adz (fig. 106, a) and another with a different lashing pattern (fig. 113) are without the toe. The modern hafted adzes from the Austral Islands (1, pp. 155, 156) show the same technique as regards the haft. The origin of this form of haft is intimately associated with the change in the form of adz head. Owing to the projecting median edge on the back of the adz head, a deep groove had to be cut on the foot of the haft to form a bed for the butt, a technique that was not necessitated by adz heads with a flat back. It is obvious that if a triangular adz was fitted entirely to the toe, the deep groove would have seriously weakened the immediate wooden support. To obviate this weakness, the adz butt was shifted up onto the body of the foot where the wood was much thicker because it was continuous with the shaft junction (fig. 101, b, 2; e). The toe part ceased to be of use and was dispensed with thus forming the heel type of haft.
The lashing pattern used with the heel haft in Rarotonga and Aitutaki was the triple triangle (fig. 271, c, d and pl. 9, A), and it may be assumed that the same pattern was used in Atiu, Mauke, and Mitiaro. In Mangaia, a local variation of the triple triangle was evolved (fig. 271, e, f and pl. 9, B). The triple triangle was also the general pattern for the Society Islands as proved on the large number of old hafted adzes in museums. The material from the Austral Islands is not so convincing. A heel haft in the British Museum (13, p. 13, fig. 5, d) carved in the technique of Raivavae, has the adz head lashed on with what looks like a badly executed open triple triangle pattern but the haft appears to be of post-European manufacture. A number of hafted adzes in Bishop Museum collected in Tubuai in 1922 are lashed with the double loop pattern (fig. 271, a, b). A Samoan hafted adz in the British Museum (E.1112) has the toe haft with the adz butt lashed to the toe with transverse turns. Above the transverse turns, an overlapping triple triangle pattern has been worked around the heel and the upper part of the toe, evidently for decoration for it takes no part in lashing the adz butt to the haft. In Polynesia, except for the unique Samoan specimen, the triple triangle pattern has been recorded only on the heel hafts of the Cook and Society Islands adzes.
The use of the two forms of the multiple lozenge pattern in Mangaia (fig. 112 and pl. 9, C) and in Rarotonga (fig. 113) may be regarded as examples page 446of the striving by expert craftsmen to apply a more complicated known technique for artistic effect and the display of their skill.
Hoop iron was a common medium used by early European voyagers for exchange or for presents. The Polynesians were quick to recognize the superiority of metal over stone; and hoop iron, next to hatchets and axes, was sought after to fashion into adzes. Later, plane blades were bought for the same purpose and with increasing trade in metal goods, the stone adzes ceased to be used. Many of the hafted adzes used in the second period were acquired by traders and collectors and found their way into museums. Hafted adzes that had been used in the native culture speedily became exhausted, but Polynesians have from time to time lashed original adz heads to new hafts in order to meet the demand created by later foreign collectors. Many of these later adzes have found their way into museums, where they create problems as to whether the shape of the haft and the form of lashing conform to an old pattern. When old specimens are available for study, the technique used with the later ones may be checked but where authentic old adzes were not collected at an early date in a particular area, the later hafted adzes from such an area do not convey proof as to the original hafting. In modern hafted adzes, the adz heads are usually authentic old specimens that have been picked up; but in Mangaia and Rarotonga, new adz heads have been turned out with a grindstone. It is because of the lack of old material that I have tentatively placed the medium type of haft in the third period.
The medium type of haft (fig. 101, d) was developed from the toe haft in a secondary western area comprising Tonga, Samoa, Futuna, and Puka-puka (fig. 269), all of which have cultural affinities with each other and traditional accounts of frequent contact. The neighboring islands of Fiji, Rotuma, and Niue adhered to the toe haft. The Samoan hafted adzes available for study are not old, and among those in Bishop Museum the toe haft occurs as an alternative form, although some of the lashing turns pass around a well-formed heel (73, fig. 206). The Tongan medium hafted adzes in Bishop Museum and the British Museum were evidently hafted in post-European times as were those from Futuna and Pukapuka. The question of the age of the medium haft thus required evidence from hafted adzes that could be definitely dated as having been collected by early voyagers. Through the courtesy of Mr. Helge Larsen of the State Museum, Copenhagen, I received photographs of two old hafted adzes. One (fig. 270, a) is apparently a Polynesian adz with a toe haft. It is said to have been in Denmark before 1690 A.D. If the date is correct, the adz could have been collected only by the Dutch voyagers Schouten and Lemaire (1615-17) or Abel Tasman (1643). Schouten page 447and Lemaire traded with natives in Tonga (Haapai group) and Wallis Island. Tasman traded with the natives in Tongatabu. From the photograph, the adz head could be Tongan. The other adz (fig. 270, b) is smaller but the haft is also of the toe type. This adz is stated to have been in Denmark before 1700 A.D. A third adz (fig. 270, c) in the National Museum, Rome (no. 1087), also with a toe haft, is definitely stated to have been collected in Tongatabu by the D'Entrecasteaux Expedition in 1793. Two other adzes attributed to Tonga in the Rome National Museum (nos. 2887, 1088) were collected much later and both have the toe haft. The toe haft having been established for Tonga at the end of the eighteenth century and later, it seems likely that the change to the medium haft was effected after European contact when hoop iron and plane blades took the place of the stone adz heads. Samoan carpenters of today use plane blades attached to medium hafts by a sennit lashing with diagonal crossing turns around the toe and heel of a short handled haft. I believe that the change from the toe haft to the medium haft occurred early in post-European times as a more suitable support for plane blades and that the medium hafts with stone adz heads that occur in museums were made as curios by craftsmen, who simply copied the late development that was then in use with the metal tools they were using.
Figure 270.—Tongan (?) adzes with toe hafts: a, b, National Museum, Copenhagen; c, National Museum, Rome (no. 1087). a, said to have been in Denmark before the year 1690; long toe with transverse lashing turns, adz blade of black stone "with a polished edge and upper side"; 570 mm. long; slight knob forming heel, b, said to have been in Denmark before 1770; very long toe with transverse lashing turns, and heel rounded off; blade also of black stone with polished edge and upper side (front); haft, 430 mm. long, c, collected in Tongatabu by D'Entrecasteaux Expedition; also long toe with transverse lashing; heel rounded like that of b.
The modern hafts made in the Cook and Austral Islands to mount adz heads for foreign collectors are of the heel type. In Tubuai, a gross exaggeration of the heel (1, p. 146) has taken place, evidently to make them more spectacular.
In the western area, the shortening of the toe of the medium haft made it impossible to secure adz heads to it by transverse turns alone. The adz butt moved up onto the foot proper, and additional lashing turns had to move up accordingly. Figure-of-eight turns were made around the shortened toe and the heel with the oblique crossings over the front of the butt. By arranging the turns in a proper sequence alternately above and below the first crossing, a lozenge pattern was developed (73, p. 360); and if the first crossing was made low on the butt with the subsequent crossings above it, a pattern of ascending chevrons was produced (73, p. 359). These patterns offered no problem for they were in use in the house lashings of wall plates to wall posts. In the adz haft, the shaft corresponded to the wall post and the foot to the wall plate, and the adjustment of the house lashing technique to include an adz head on the foot of a haft was made readily.
In the Cook Islands, the adzes hafted for trade adhered to the local form of heel haft and the triple triangle form of lashing was perpetuated in Aitu-taki.
The carved hafts of Mangaia interested foreign collectors who stimulated their manufacture. Steel tools made elaboration easy, and the native craftsmen quickly perceived that the more elaborate the article the greater the foreign demand. Thus the small enlargements on the lower ends of the older hafts of class 2 expanded into the enlarged pedestals of class 3, and the shafts underwent a corresponding increase in diameter. Though older elements such as narrow panels and the K-motif were retained in part, such innovations as the octagonal form of shaft, the square holes in the pedestals, and the multiple lozenge and rows of triangles in carving were introduced. It is possible that the holes were introduced by old craftsmen who remembered the perforations in the slab gods and that the change from the lozenge shape to square was necessary to preserve balance in the areas to be carved. The commercial value of a religious association was realized and maintained and the "peace ax" myth had selling value. In the course of time the lashing technique deteriorated and the local form of triple triangle was forgotten. The manufacture of stone adz heads ceased with the introduction of steel tools, but wood was easy to work and so the making of adz hafts continued for over a century. When the local supply of stone adz heads was exhausted, foreign importations took their place and the last stage of degeneration was reached when pieces of stone were roughly shaped on a trade grindstone.
Development of Lashing Patterns
Polynesian craftsmen who used the toe hafts evidently felt no artistic urge to depart from a practical technique. The New Zealand craftsmen were second to none in the manufacture and use of stone adzes, and they were evidently content with the efficiency of the seizing technique they employed with quadrangular adzes. Their attitude was shared by the Hawaiians, Marquesans, and others who retained the rectangular adz and the toe haft with its simple transverse lashing.
In the western area, the lashing technique with medium hafts had to undergo change and the craftsmen undoubtedly made use of the technical knowledge that they had acquired with house building.
Figure 271.—Evolution of the Mangaian triple triangle. a, b, double loop technique, Austral Islands; c, d, triple triangle, Cook and Society Islands; e, f, Mangaian triple triangle, Mangaia. a, front: lower (1) and upper (2) loops. b, side: lower and upper loops crossing on side to form one large triangle. c, front: lower loop (1) continued throughout; upper loop (2) continued for a few turns and then split (3) as shown by crossings in middle line. d, side: split loops in their vertical turns (4) around shaft work inward and by alternately crossing with upper and lower oblique turns create an upper and lower triangle in addition to front triangle. e, front: lower loops (1) persist but upper loops (3) split immediately so that full upper loops of previous two lashings (2) are eliminated. f, vertical turns around shaft that are parallel in d are diverged into two bands (4, 5) and cease to cross with lower oblique series (6) which eliminates lower triangle seen in d; diverging two series (4, 5) cross alternately and create a new triangle (7) below upper original triangle of d.
In the central area, the use of heel hafts also necessitated some change in the lashing technique. The simplest form of lashing used with heel hafts is the Austral Islands method that I have termed the double loop (fig. 271, a, b). Here again, the craftsmen did not have to invent something new, for the double loop technique was used extensively in lashing together parts of houses and the cross booms of canoes. That the double loop was the precursor of the triple triangle is supported not only by its simpler technique but also by the fact that after the fixation of the commencement end (fig. 102), the commencing turns of the triple triangle were made with the double loop method (fig. 103). The change was then made by dividing the upper loop into two stages (fig. 104), which by creating a different sequence of courses resulted in the triple triangle lashing (figs. 105 and 271, c, d). It is probable page 450that the double loop lashing was used first with the heel hafts in the Society Islands and accompanied the heel haft and triangular adz head to the Austral Islands. Lacking old hafted adzes, we cannot say that the triple triangle lashing also reached the Australs. The simpler form of the double loop lashing known to the present inhabitants may be a direct inheritance from the past. The three-fold complex of triangular adz head, heel haft, and triple triangle lashing certainly reached the Cook Islands.
In Mangaia, perhaps owing to the great interest and skill developed in sennit work, the orthodox triple triangle pattern was modified by splitting the upper loop at once and by dividing another series of turns alternately into two diverging sets. This second splitting altered the sequence of courses and resulted in the Mangaian form of triple triangle (fig. 271, e, f).
In Mangaia was also developed the ceremonial adz, and it is probable that the local variation in the form of lashing was developed in connection with this form of adz. Evidently the lashing pattern became so popular that it was applied to working adzes and thus supplanted the older form of triple triangle. Associated also with the ceremonial adzes was the development of the pattern of oblique crossings (fig. 108), which were purely decorative and took no part in securing the adz head to the haft. Evidently, the decorative lashing pattern became popular and was applied to working adzes, as evidenced by the specimens preserved in collections. It is possible, of course, that those given to collectors were show pieces decorated for the occasion.
In the true ceremonial adzes (fig. 245), the shaft was plotted into narrow panels and carved with some form of the K-motif. If it is true that these adzes had a religious significance, it is reasonable to infer that the technique of narrow panels with the K-motif was derived from the carved slabs that represented tribal gods. The technique as applied to adz hafts having become established, it was used with some adzes of class 1 (fig. 244) purely for ornamentation.