Volume 44 1935 > Volume 44, No. 174 > Material representations of Tongan and Samoan gods, by Te Rangi Hiroa, p 85-96
                                                                                             Previous | Next   

- 85
Continued from Vol. 44, No. 1, Page 53.

Throughout Polynesia, inanimate objects were selected or made to represent the gods. These objects were kept in the religious structures or in the personal charge of the priests. They were always on hand whereas the animate representatives were not always available. Simple natural objects such as stones and shells were utilized. In the Tahitian region, red feathers impregnated with the spiritual essence of the god formed the basis for further elaboration in form. More elaborate forms consisted of feather-work, sennit-work, and carved wooden objects. A further development consisted of carving images in wood, stone, or bone. The images developed conventional patterns that on the whole were characteristic in different regions, but yet considerable variation occurred even in the same region.

Considerable diversity occurred with regard to the attitude in which inanimate representatives were regarded. In some regions they seem merely to have formed part of a religious paraphernalia. In other regions they were charged with the spiritual essence of the god and on occasions might even be entered by the god on the solicitation of the priest. The association of the immaterial god with the inanimate object entitles these inanimate representatives to come under the general term of fetish.

The inanimate fetishes reported from Tonga and Samoa range from natural objects such as stones, shells, and teeth, to manufactured articles such as mats, bowls, and weapons. Many of the gods were without such fetishes whilst others had as many as three. The smaller objects were usually wrapped up in bark-cloth smeared with turmeric, placed in baskets, and kept in the house sacred to the god. When worshippers gathered before the temple, the priest uncovered the fetishes and displayed them on a mat. Offerings were made to these material symbols of the deity in order to secure a favourable response to the petition of the devotees. The range of material objects selected as inanimate representatives of various gods in Tonga and Samoa is shown in Table 2.

- 86
Object Tonga Samoa
Whales' teeth 6 1
Sharks' teeth - 1
Trumpet shell - 2
Cypraea shell - 1
Unnamed shell 1 1
Stones 2 3
Fine mats 2 -
Coconut leaf basket - 1
White tapa cloth - 1
Turmeric ball 1 -
Wooden bowl - 3
Throwing club 2 -
Striking club 1 -
Coconut wood spear - 1
Shield (pa) 1 -
Pitching discs (lafo) 1 -
  17 15

Amongst the objects listed in Table 2 there is no representative in human form nor any carved object except perhaps the Tongan club. Yet Wilkes 1 reported that in Samoa carved blocks of wood and of stone were set up in memory of chiefs and worshipped. Collocott 2 also, in speaking of the furniture of the temple of the Tongan god, Mafuta-ae-tau, states that it consisted of “fine mats, weapons of war, and carefully wrought pieces of wood painted with turmeric.” This temple with its contents was burned after the acceptance of Christianity. Though the carved objects mentioned by Wilkes and Collocott may not have been in human form, the statements are interesting in view of the fact that carved images from both island groups are in existence. It is remarkable that neither Collocott nor Gifford were given information by the Tongans regarding the presence of images in the Haapai islands. Turner also made no mention of images in Samoa though such an object is now in the British Museum.

- 87

The existence in Haapai of inanimate representatives of gods, shaped in human form, is established by the documentary evidence of the missionary, John Williams, who visited the Tongan group in 1830. At that time, Haapai was subject to the authority of Taufaahau, who afterward became ruler of the whole Tongan group with the title of George Tupou I. Taufaahau, according to Williams, 3 had despised the Tongan system of worship from his youth and, coming in contact with missionaries during a visit to Tongatabu, he resolved to abandon the gods of his ancestors. On his return to Haapai, he “commenced the work of destruction upon the gods and maraes.” Williams had come to Tonga after work carried on in the Society and Cook groups and by “marae” he meant the Tongan house temple. His misuse of the term “marae” must not be regarded as evidence that the marae form of religious structure was present in Tonga. An opposing faction determined to celebrate a religious festival in honour of the gods whom Taufaahau was desecrating. Williams 4 wrote as follows:

“Taufaahau, resolving to anticipate and neutralize this movement, drove a large herd of pigs into the sacred enclosure, converted a most beautiful little temple, which stood in the middle of it, into a sleeping appartment for his female servants, and suspended the gods by the neck to the rafters of the house in which they had been adored. The idolators, ignorant of his proceedings, came, with great ceremony, attended by their priests, to present their offerings, and found, to their astonishment, a number of voracious pigs, ready to devour anything they had to offer; and the gods, disrobed of their apparel, hanging in degradation, like so many condemned criminals.”

The worshippers retired in indignation but evidently did not attempt to cut down the gods from their elevated position. Taufaahau conducted Williams to the building and Williams 5 continues:

“On observing five goddesses hanging by the neck, I requested this intrepid chief to give me one of them, which he immediately cut down and presented to me. I have brought it to England, with the very string around its neck by which it was hung.”

- 88

Fortunately Williams 6 reproduced a drawing of the goddess. A copy, with the position reversed purposely for comparison with other figures to follow, is shown in fig. 1.

FIG. 1.
Tongan image, after Williams.

The figure has characteristic features which readily distinguish it from wooden images from other parts of Polynesia. The head is poorly drawn but it is evident that the brows, nose, mouth, and ears are clearly defined. The outstanding features, however, are the large pointed breasts, the position of the arms which are free from the body with somewhat pointed hands hanging straight down, and the exaggerated back-projection of an enlarged gluteal region. The slightly flexed legs are short in proportion to the thighs and the feet are notched to represent toes. The figure stands on a base cut out of the same block of wood.

- 89

The image evidently did not pass into the museum of the London Missionary Society in London for it is not amongst the collection handed over by the London Missionary Society to the British Museum. Williams stated that he prized it very highly so he probably kept it for some time until at last it went into circulation along channels which unfortunately did not lead to the British Museum. My attention was drawn to the Williams image by Captain A. W. F. Fuller who has a similar figure in his own wonderful collection. Through his courtesy, I am enabled to reproduce it here in fig. 2. These old authentic artifacts are so valuable for comparative studies that I deem it advisable, when possible, to show views of the front, back, and right side of all figures as in the method of picturing stone adzes.

FIG. 2.
Tongan Image in Fuller Collection. Made of ironwood, well polished. Height, 372 mm.; head-width, 65 mm.; shoulder-width, 144 mm.; width between hand extremities, 119 mm. Head rounded, with well-marked brows, excised elliptical eyes, good nose, and excised ellipital mouth; projecting ears with groove along helix and separate knob to represent tragus. Shoulders wide; arms free from body, extended downward with no flexion at elbows; hands thinned down with thumbs to the front. Breasts prominent and showing female sex. Abdomen not protruding, no navel, and gluteal region prominent. Thigh long and flexed; leg short with well developed calf.
- 90

Comparison with the Williams figure shows a similar attitude of the arms and hands with prominence of the breasts and gluteal region. The Williams figure shows a flat face, flat head, and longer breasts, but these details were probably exaggerated by the English artist who did the drawing. The absence of eyes may also be due to the artist.

Two images, similar to the Fuller image, are in the private collection of Mr. Oldman of London. In a discussion amongst Captain Fuller, Mr. Oldman, and myself, we came to the unanimous conclusion that the images in their collections must have come from the same locality as the Williams image. A strong possibility exists that one of the Oldman images is the one that Williams brought back to London.

Four smaller images cut out of whale-ivory are in the British Museum, and the largest of them was so realistically carved that, at first, I thought it was the work of some European sailor. A comparison with the Fuller image, however, showed that the technique was so similar that there could be little doubt that the two came from the same locality. Three of the ivory images are labelled with the museum series of “Tah.” which was applied to accessions from Tahiti, but the museum authorities place no reliance on the attribution to that locality. The largest of the images is shown in fig. 3.

Except for the details of the ear, the description of the ivory figure is practically word for word identical with that of the Fuller image. The other three smaller ivory figures are shown in figs. 4, 5, and 6. All three have a vertical hole drilled through the back of the head evidently for suspension.

Fig. 6 differs from the previous two, not only in size but in technical details. The downward extension of the arms and the hole through the back of the head link it up with them. The different numbering shows that it was not procured at the same time as the others.

The available data for Tonga here presented consist of three larger images in wood, which may include the Williams image, and four smaller figures in whale-ivory. In all of them, a similar position of the arms prevails.

- 91
FIG. 3.
Ivory Image in British Museum (Tah. 133). Whale-ivory, nicely polished. Height, 121 mm.; shoulder-width, 48 mm. Head rounded; brows, eyes, and nose well marked, with excised mouth; ears project and form complete ellipse without helix groove or separate tragus. Shoulders wide and rounded off; arms free of body, hanging down without elbow flexion; hands with thumbs to the front. Breasts prominent and denote female sex. Abdomen flat, no navel, and prominent gluteal region. Thighs flexed and feet joined by connecting-piece through which hole is pierced.
FIG. 4—Ivory image, British Museum (Tah. 135). Height, 35 mm.; shoulder-width, 19 mm. Resembles fig. 3 in prominent breasts, projecting gluteal region, prominent ears in complete ellipse, and position of downward extended arms with thumbs to the front. The side view not good owing to tilting in photograph. Figure also has raised band around arms and legs which are not well shown in drawing.
- 92
FIG. 5—Ivory image, British Museum (Tah. 134). Height, 34mm.; shoulder-width, 17 mm. Shows all characteristics of fig. 4 except for raised bands around limbs. Head somewhat straight in transverse line but rounded off from before back.
FIG. 6—Ivory image, British Museum (No. 2323). Height, 55 mm.; shoulder-width, 20 mm.; outer thigh width, 19 mm. Face somewhat indistinct; no ears, no breasts, no toes, fingers not formed; no projection of gluteal region; raised band around legs.

All, except fig. 6, are female with prominent breasts and gluteal regions. The definite documentary evidence by Williams and the lack of references to human form in other parts of the Tongan group indicate that female figures with projecting breasts and gluteal region and downward extended arms with thumbs to the front, are to be associated with the Haapai group of the Tongan islands.

- 93

A wooden image presented to Queen Victoria by King Malietoa of Samoa in 1841 has found sanctuary in the British Museum. The image was figured by Kraemer 7 and my first impression on seeing Kraemer's drawing was that the image belonged to the post-European period. However, the year 1841 is very early so far as Samoa is concerned for Christianity was not accepted until after the visit of John Williams in 1830. Doubt as to the purely native origin of the image vanishes after comparing it with the Tongan figures. The Samoan image is shown in fig. 7.

FIG. 7., Samoan Image in British Museum.
Hard wood, light coloured, not polished. Height, 675 mm.; head-width, 146 mm.; head-length, 147 mm.; shoulder-width, 212 mm.; chest-thickness, 84 mm. Well-shaped head with brows, elliptical eye-sockets with shell-opercula inserted to represent eyeballs, wide nose with triangular lower surface with two holes for nostrils, elliptical mouth, and projecting ears with knob interruption to represent tragus. Long body with flat chest without prominent breasts, flat abdomen with depressed navel, and gluteal region not prominent. Arms free of body, extended downward without elbow flexion, five-fingered hands with thumb to front. Short legs and feet with five toes.
- 94

As seen from the figure and description, the Samoan image differs from the Tongan images in a number of details. The head is not so rounded, the body is more elongated with a corresponding shortness of the legs, and the marked protuberance of the breasts and gluteal region is absent. The ears are more roughly made, the eyes have the addition of opercula eyeballs, and the nose besides being broader has the addition of two holes for nostrils. One cannot help feeling that the addition of nostril-holes to the fairly large triangular under-surface of the nose has been influenced more by the technique of the perforated triangular lug at the proximal ends of clubs than by anatomical considerations.

In spite of the differences in technical details, the general pose of the figure with the downward-extended arms and thumbs to the front is strikingly similar to that of the Tongan figures. The ears, though not so carefully carved, have a very important element of resemblance in the presence of a distinct projection to the front which corresponds to the tragus in the Fuller image (fig. 2). The separation of the tragus-element must be regarded as a specialization in western Polynesia because the technique is not present in the many images from the other regions that I have so far examined.

The Tongan wooden images are smooth and highly polished whereas the Samoan image retains the edge-marks from the final adzing-process. The Samoan image is also cruder in craftsmanship as shown by the rough carving of the nose and ears. In spite of these differences, the Samoan image appears to belong to the same art-school that was responsible for the Tongan images. It is unlikely that the same characteristic position of the arms and the specialized ear-tragus could originate in two localities independently of each other. The Haapai locality has been proved beyond question by Williams, and the existence of other images of a like technique shows that the form was not sporadic. Only one Samoan image is known up to the present, and it is more feasible to accept diffusion from Haapai than independent creation in Samoa. If the Samoan image had been made in Haapai, better workmanship and a closer adherence to the Haapai details might be expected. I incline to the opinion that the image sent - 95 to England by King Malietoa was made locally in Samoa by an inferior artist who was either Tongan or someone who was acquainted with the general appearance of the Haapai images.


The use of birds, fish, and human figures as motives in the carvings on certain types of weapons has been definitely established for Tonga. While the human motives vary in form, many of the figures with wide shoulders and dependent arms hanging free from the body suggest the characteristic attitude of the Hapaai images already described. As these figures are small and on the flat, the prominent breasts and gluteal region of the images could not be included. Their presence in so many weapons indicates that the characteristic image-attitude had become an established motive in Tonga. It might well be inferred that the human-motives on the clubs originated with the carving-school responsible for the images in Hapaai. Investigations as to the distribution of club-types in the Tongan group itself might shed light on the subject.

The presence of zoomorphs and anthropomorphs in the carvings on Samoan clubs rests on the sole authority of Churchill. 8 The human figures recorded by him were confined to two clubs in the University of Pennsylvania Museum. I 9 have already expressed my opinion that, from the appearance of the figures and the clubs, the clubs were not Samoan but Tongan. Since then I have had the opportunity of personally examining the two clubs and I have now no doubt that both the clubs are Tongan. In fact, the museum catalogue definitely attributed one of them to Tonga, which reference Churchill must have overlooked. The one positive statement that human figures occur in Samoan club-carvings may thus be dismissed. Out of a large number of carved Samoan clubs examined by me in European and American museums, not one had zoomorphs or anthropomorphs as motives. My experience is shared by Messrs. Fuller and Oldman whose practical experience - 96 in the handling of clubs from Tonga, Samoa, and Fiji must add the greatest weight to their opinions. If the human-figure motive had been adopted by Samoan craftsmen, we would expect to find its presence in some, at least, of the numerous authenticated Samoan clubs in various museums and private collections. The absence of the human-motive is evidence that it never entered Samoan carving-technique as applied to weapons. This lack of a human-motive in Samoan club-carvings may be taken as evidence in support of the argument that the Samoan image is due to Tongan influence.


In order to stress the importance of the position of the arms and some technical details in the Tongan images, a few images from central Polynesia are here figured for comparative purposes. By central Polynesia, I mean the Society islands, but I would also include for the purposes of this study, the Cook and Austral groups. The examples selected are the smaller images in wood as such are more comparable with the Tongan material.


In Tahiti, the inanimate representatives of the gods were termed to'o (cf. Maori, toko) and were usually made of a twined sennit covering over a wooden core with outer decorations of red feathers. The human form was rarely used. Wooden images, termed ti'i (Cf. Rarotongan, tiki), were used by sorcerers to represent their familiar spirits. These were of both sexes, and an example of each is here figured as, from size and material, the technique is more readily comparable with the Tongan images. See figures 8 and 9.

The Tahitian figures are not so realistic as the Tongan, and their main differences consist in the treatment of the ears, the square shoulders with the straight edge across the back, the position of the arms, the nipples instead of breasts in the female figure, the projecting abdomen and navel, and the normal treatment of the gluteal region. It is evident in the two Tahitian figures that sex is indicated by the inclusion of the sex-organ, and the female breasts are not any more developed than the male. In the Tongan figures, the sex is indicated by the size of the breasts and the sex-organ is ignored.

(To be continued.)

1   Wilkes, Charles, Narrative of the United States Exploring Expedition, Philadelphia, 1844, vol. 2, p. 139.
2   Collocott, E. V., Notes on Tongan Religion, J.P.S., New Plymouth, 1921, vol. 30, p. 228.
3   Williams, John, Missionary Enterprises, London, 1839, p. 273.
4   Williams, Opt. cit., pp. 273, 274.
5   Williams, Opt. cit., p. 275.
6   Williams, Opt. cit., p. 274.
7   10Kraemer, A., Die Samoa-Inseln, 2 vols., Stuttgart, 1902, 1903; vol. 2, p. 207.
8   Churchill, W., Club Types of Nuclear Polynesia, Carnegie Institute, Washington, 1917.
9   Te Rangi Hiroa, Samoan Material Culture, Bishop Museum Bulletin, 75, Honolulu, 1930, p. 611.