Volume 46 1937 > Volume 46, No. 182 > Additional wooden images from Tonga, by Te Rangi Hiroa, p 74-82
                                                                                             Previous | Next   

- 74

IN a previous paper, 1 describing wooden and whale-ivory images from Tonga, reference was made to two wooden images in the private collection of Mr. W. O. Oldman of London. After the paper appeared, Mr. Oldman sent me photographs and details of the two images mentioned and of a third unique image that had come into his possession after I saw the other two in 1933. Just as this paper was being written, Mr. Oldman sent me a further photograph, of a whale-ivory image that is also in his collection. Mr. Oldman has kindly given me permission to reproduce the four additional images and so make the series of Tongan images up to date for the information of students of Polynesian culture.

Two of the wooden images and the ivory one are in the erect position with pendent arms, but the third wooden image shows a unique departure in being carved in the sitting position. The taller of the erect images is shown in fig. 1.

The resemblance of fig. 1 to the figure in Williams' work 2 is so great that they seem to represent the same image. The presence of the cord around the neck in both figures gives further support, but all doubt as to identity is dispelled by the written inscription on the back which reads as follows: “Goddess of Lifuga hung by Tafahau on embracing Christianity. Hapai, July 1830.”

Lifuga (Lifuka) is the principal island in the Haapai group where in 1830 John Williams met Tafaahau who later became George Tubou I of Tonga. It was on that occasion that Tafaahau 3 cut down one of the five goddesses suspended by their necks from the roof of a house and

- 75
FIG. 1.
TONGAN WOODEN IMAGE Oldman collection, number 530. Height, 14.5ins.; shoulder width, 4.75ins. Head-vertex rounded; face markedly flat and meeting the sloping planes above forehead and on either side of face in distinct edges which meet below in pointed chin. Eyes poorly represented by horizontal slits; nose long, clearly defined, and projecting out from flat face surface; mouth formed by horizontal slit similar to eyes; ears formed by projecting chevron flange open to the front. Practically no neck; shoulders slope downward and outward to meet outer line of upper-arm in obtuse point; smooth slope from neck to back. Arms straight and pendent, ending in pointed hands without defined fingers but evidently in natural position of thumbs to front., Breasts large, long, pointed, and projecting forward and downward. Body slim with back and sides meeting in well-marked edge on posterior axillary line but sides and front rounded off. Abdomen not protruding, no navel, and horizontal edge defining upper line of a raised public triangle, base upward, which may represent female sex-organ., Gluteal region projects backward to an exaggerated degree with upper and lower planes meeting in a marked mesial horizontal edge. Thighs long, legs short with marked backward projection of calves similar to gluteal region. Feet rotted in front; heels project backward., The feet and base were originally in the same block of wood but when acquired by Mr. Oldman both feet and base were so worm-eaten that a new base had to be made. A length of sennit braid is tied around the neck and a paper inscription in ink is pasted on the back.
- 76

gave it to Williams. It is probable that the inscription on the back of the image figured was written by John Williams himself.

The second erect figure is shorter than the one described but is much wider across the shoulders. The top of the head is somewhat flattened from side to side, but in other details it corresponds closely to fig. 1. See fig. 2.

FIG. 2.
TONGAN WOODEN IMAGE Oldman collection, number 531. Height, 13ins.; shoulder width, 6.25ins. As compared with fig. 1, the flat vertex of the head, the wider and squarer shoulders, and shorter thighs are noticeable variations. The flat face, pointed chin, eyes, mouth, nose, and ears are similar and so are the short neck, pointed shoulders, pendent arms, prominent pointed breasts, slim body with posterior axillary edge, flat abdomen, and projecting gluteal region and calves with horizontal mesial edges. Some slight pitting occurs in the wood in the umbilical region; it does not appear to be a purposive attempt to define the navel-depression., The raised pubic triangle present in fig. 1 is absent. The lower outer edge of the hand-terminal is notched with four incisions to represent five fingers with the thumb to the front. The feet are represented by low flat pieces projecting at the outer sides as well as the back; the right foot is notched in four places on the front which extend the depth of the shallow front surface while the left has four slight notches on the upper edge of the front surface; each foot is thus represented with five toes. The figure and pedestal are cut out of one piece of wood.

The Fuller image and the ivory figures in the British Museum previously described were much more realistic than the two Oldman images, and but for Mr. Oldman's kindness in making further information available an erroneous standard for Tongan art would have been created. It will - 77 be seen in the carving treatment of figs. 1and 2, that though certain features of the more realistic images are present, the treatment of the eyes, mouth, ears, shoulders, and hands have been highly conventionalized. The well-formed ears of the Fuller image and the largest ivory image in the British Museum are here represented by chevron projections carved with exactly the same technique as the suspensory lugs in Tongan and Samoan kava-bowls.

Mr. Oldman writes that the remarkable small figure in a sitting position came from a very old collection of Mr. O. Belsham which was acquired about 1846. It is carved from a dense, hard, dark-brown wood, and the features and limbs show deliberate mutilation, the face showing distinct cuts made with a knife or an axe. See fig. 3.

FIG. 3.
TONGAN WOODEN IMAGE Oldman collection, number 532. Height, 7.5ins. a, front: head and face with pointed chin of general shape of figs. 1 and 2 but probably face not so flat; both arms cut off but probably both were pendent as there is no trace of hands being connected with the body; left lower limb well flexed, with knee rotated inward and curve of thigh and leg show artistic execution; right leg cut off through thigh and stump indicates right lower leg was flexed below left leg with foot to left. b, back: head outline and short neck show similarity to figs. 1 and 2; slim body, graceful waist, and well-rounded buttocks. c, right side: head in profile with similar lines to fig. 2; nose apparently cut off; ear as raised flange with curve instead of chevron angle; breasts apparently cut off; curve of buttocks plainly shows sitting-posture with left thigh higher than right and section of right thigh and legs indicate that right leg was probably beneath left as suggested by front view (a).
- 78

The original label that went with the sitting image reads as follows: “Household Goddess of the Emperor of Tonga and part of the dress worn by him when he worshipped the Devil.” The label confirms the locality of the image, and, as Tafaahau became the ruler of Tonga, the image evidently formed one of the group that Tafaahau abandoned in 1830. It is the only image that I know of that has been carved in a sitting position and is hence unique. In spite of the mutilation, the lines of the body and the thighs show good workmanship.

The ivory image is thus described by Mr. Oldman in his letter: “Very old and much worn with handling, stained a deep orange-yellow of beautiful tone and with symmetrical deeper coloured markings on front of body and lower legs and on posterior. Arm has been fractured. I consider it one of the finest little gems I possess.” A line drawing of the print is shown in fig. 4.

The Oldman ivory image, while typically Tongan, is not quite so realistic in treatment as the largest ivory image in the British Museum or the Fuller wooden image. From the treatment of the eyes and the comparatively wide width of the shoulders, it stands between these more realistic images and the more conventionalized types in the Oldman collection shown in figs. 1 and 2.

The dress worn by Tafaahau when he took part in the religious service of his own culture, rather snobbishly described as “worshipping the devil,” consisted of bark cloth, and from its narrow width it was evidently a sash or a perineal band (malo). As specimens of cloth of authentic early manufacture should be regarded as type specimens for future study, a drawing made from Mr. Oldman's photograph is shown in fig. 5, a.

Mr. Oldman included in his pictures, two clubs with projecting bosses on the expanded heads of the clubs. These bosses are shaped in a form similar to the flat faces with pointed chins of the Tongan images, and they have elongated triangular projections which correspond in position and shape to the noses of the images. Though eyes and mouth are not represented, I agree with Mr. Oldman in his opinion that this conventionalized motive has been derived from the faces of the Tongan images. It is probable, therefore,

- 79
FIG. 4.
TONGAN IVORY IMAGE Oldman collection, number 533. Height, 5ins. (217mm.); shoulder width, 2¼ins. (57mm.). Head-vertex rounded; face rounding off to sides; eyes have brow and lower lid curves but eyes themselves represented by straight incised lines slanting slightly downward and outward; nose long, wide, and rather low in forward projection, flawed on left side of point; mouth straight, horizontal incision wider in middle; chin rounded; ears formed by raised circular flange with front notched to represent tragus. Neck extremely short. Shoulders wide with straight horizontal upper line but outer points rounded off. Arms straight and pendent with hands wider than wrist, pointed, with incised lines to define fingers, thumbs to front., Breasts large and rounded; abdomen natural curve with circular depressed navel. Gluteal region exaggerated. Thighs slightly flexed; lower leg normal proportion without exaggerated calves; feet notched to represent toes. Low pedestal in one piece with image, projecting on right but not at back.
- 80

that these clubs illustrate types that were used in Lifuka, the locality with which the images are definitely associated. See fig. 5, b and c.

FIG. 5.
TONGAN CLOTH AND CLUBS a—Tongan bark cloth: period of 1830 before foreign influence affected native patterns; oblique parallel lines forming long chevrons. b—Tongan club with expanded head showing bosses in form of face and nose, similar to those of wooden images in figs. 1and 2. c—Tongan club similar to b but with raised zig-zag vertical lines between the bosses.

Anthropology in the branch of material culture has certain advantages and disadvantages. The advantages exist in the fact that artifacts now no longer obtainable in their country of origin have been preserved in museums and private collections. The disadvantages are that many valuable specimens, after passing through various hands, have become inaccurately labelled as to locality and, furthermore, that owing to their not having been described in anthropological literature, they are inaccessible to the student who cannot make a tour of the world's museums and private collections.

In order to promote accurate study by students and correct identification by museum authorities, it is necessary in the cause of science that really old artifacts, of which the history and locality have been accurately determined, should be recorded in the literature as type-specimens. Once - 81 type-specimens have been definitely established by authentic history or the careful analysis of the technique by an expert, such type-specimens can be used to determine the locality of other specimens of a similar appearance but concerning which the history and attributed locality are uncertain. There is as much scientific need for specimens in ethnological museums to be accurately labelled as to locality as there is in museums of natural history.

So far I have described nine specimens of Tongan images, four in wood and five in whale-ivory. The four whale-ivory images already described 4 are in the British Museum (Tah. 133, Tah. 134, Tah. 135, and 2323). Their history beyond their actual acquisition by the Museum is unknown. From the catalogue numbers commencing with Tah., it is evident that three of them were attributed to Tahiti, but owing to lack of information the British Museum authorities would not vouch for this locality. The attribution of the images to Tonga rests on the detailed resemblance of the largest of them (Tah. 133) to a wooden image in the A. W. F. Fuller collection (no. 791), which also has been described. 5 The diagnosis was first made by Captain Fuller himself, and my examination supports his diagnosis. The attribution of Captain Fuller's specimen (no. 791) was based by him on its resemblance to the specimen in Mr. Oldman's collection (no. 530) and the opinions of both Mr. Oldman and Captain Fuller that the Oldman specimen was identical with the image figured by Williams. 6 The piece of braid around the neck and the inscription pasted on the back of the image, establish their identity without any doubt to my mind. Here we have a specimen whose history and locality are sufficiently proved to establish it as the type-specimen. From the type-specimen, number 530 in the Oldman collection, the locality of Tonga given to the other erect image in the Oldman collection (no. 531), the wooden image in the Fuller collection (no. 791), and the Oldman ivory image (no. 533), can be definitely accepted. The Fuller image varies in certain details as to eyes, ears, and the rounding of the shoulders, being more realistic - 82 than the Oldman type. The Fuller image may be regarded as a variation and by it the largest whale-ivory image in the British Museum (Tah. 133) may be definitely assigned to Tonga and not to Tahiti as its number would infer. From this specimen, the other three variations of whale-ivory images in the British Museum may also be attributed to Tonga. The localities of seven images with indefinite histories have thus been determined, primarily from the Oldman type specimen (530) of which the history and locality may be accepted as proved.

The history of the sitting image, while not so connected as that of the standing type, is supported by the specimen of bark cloth, the description written in missionary phraseology, and certain structural details. No doubt can be entertained as to the locality being Tonga and the specimen may be regarded not only as a type but also as a unique.

As a record, the nine Tongan images so far known are given in the following table with their height and shoulder width in inches:

Collection Material Height inches Shoulder width inches
W. O. Oldman, 530 wood 14.5 4.75
W. O. Oldman, 531 wood 13.0 6.25
A. W. F. Fuller, 791 wood 14.6 5.55
W. O. Oldman, 532 wood (sitting) 7.5 -
W. O. Oldman, 533 ivory 5.0 2.37
British Museum, Tah. 133 ivory 4.75 1.85
British Museum, Tah. 135 ivory 1.4 0.75
British Museum, Tah. 134 ivory 1.35 0.7
British Museum, 2323 ivory 2.2 0.8

If the four wooden images so far described formed part of the collection of five seen by John Williams, 7 it is intriguing to speculate as to whether or not the one remaining specimen also found its way into the safe haven of a collection. If anyone recognizes it from the illustrations in this paper, I would be obliged if the owner or discoverer would inform me.

1   Te Rangi Hiroa, “Material Representation of Tongan and Samoan Gods”; Journal of the Polynesian Society, vol. 44, p. 12, 1935.
2   Williams, John, Missionary Enterprises, London, 1839, p. 274 (also Journal of the Polynesian Society, vol. 44, p. 10).
3   Williams, John, Missionary Enterprises, London, 1839, p. 275.
4   Te Rangi Hiroa, op. cit., pp. 13, 14.
5   Te Rangi Hiroa, op. cit., p. 11.
6   John Williams, op. cit., p. 274.
7   John Williams, op. cit., p. 275.