Volume 108 1999 > Volume 108, No. 4 > Shorter communication: Tongan headrests: Notes on terminology and function, by Jeffrey Dhyne, p 411-416
                                                                                             Previous | Next   

- 411

Incorrect terminology has been used consistently in the literature when describing Tongan headrests. There has also been a corresponding misinterpretation of their function.

Until 1982, those few sources published in English that dealt specifically with Tongan headrests relied on English terminology. Guiart (1963:351) used the word headrest, as did Force and Force (1971: 150). The word neckrest (or neck rest) was used by Kaeppler (1978:228-30), Gathercole, Kaeppler and Newton (1979:93,175), and Mack (1982:250-52). But Mack also used non-English terminology. He used the word sauthoko for a type of headrest with a pair of bowed legs at each end of a rectangular headpiece and the phrase kali fafapo for a type with a single wide leg at each end (Mack 1982:250-52). Edler (1990:40,42) also used the words sauthoko and fafapo for the same two types, and Wardwell (1994:178) used the word sauthoko.

Although Mack did not reference his use of these terms, it is evident from his bibliography that the source was Thompson (1940). In discussing the material culture of the Lau Islands of Fiji, Thompson (1940:190) described an old type of headrest called kali thinggi. It was made of three pieces—a headpiece and two pairs of legs—tied together with coir. According to Thompson's informant, headrests of similar shape were formerly made in one piece and called sauthoko. Thompson (1940:191) also wrote, “The hahapo is a Tongan type of headrest (fafapo in Tongan) made of a single piece of mbau wood.”

Sauthoko is not a Tongan word. It is Fijian and as such is inappropriate to use in reference to Tongan headrests. And Thompson's identification of the word fafapo as the Tongan equivalent of hahapo is incorrect.

Churchward's Tongan Dictionary (1959:248) defines kali as a “native wooden pillow or headrest (carved, with two short legs)”. He lists five types: “k. kofe (made of bamboo), k. loa (long enough to be used by several people at the same time), k. toloni, k. hahapo”, and “k. lei”. Lei means whale's tooth. He then lists three related words: 'olunga, “headrest” or “to use as a headrest, to rest one's head on” (p. 565), tekihi, “block of wood prepared and used as a pillow or head-rest”(p. 477), and lalango, “pillow or headrest for a chief (or a corpse)” (p. 280). He gives a second definition of kali as “bow-legged”.

St Cartmail (1997:52-57) made use of Churchward's definitions but also incorporated information from a letter in the archives of the Auckland Museum by Dr T. Childs dated 23 September 1944. Childs wrote that there were four types of kali: kali laloni, kali toloni, kali loa, and kali hahapo. He did not offer verbal - 412 descriptions of three of the four types but included a simple sketch of each. According to his sketch, kali laloni consisted of three pieces: a headpiece and two pairs of of bowed legs. Kali toloni also had two pairs of bowed legs but appear in his sketch to be made of one piece of wood. Kali hahapo had one wide leg at each end. He described kali loa as consisting of a long bamboo pole mounted on legs.

Figure 1: Kali toloni, Auckland Museum 31812. (Photograph courtesy of Auckland Museum)
Figure 2: Kali hahapo, Auckland Museum 8638. (Photograph courtesy of Auckland Museum)
- 413
Figure 3: Kali loa, private collection. (Photograph by Jeffrey Dhyne)

With very few exceptions, Tongan informants in 1996 and 1997 knew only the words kali and kali loa. Although most Tongans have not seen a kali loa because of the now infrequent use of headrests in general, it is remembered in the proverb fielau he na'e 'olunga he kali loa (because he/she rested his/her head on the kali loa), said of a good child, one who has learned well from his/her mother's “pillow talk”.

One informant, Sitiveni Feau Fehoko, a professional carver from Vaipopua, Tongatapu, told me there are three types of kali. These are kali hahapo, with one wide leg at each end, kali toloni, with a pair of bowed legs at each end, and kali loa, which are large enough to be used by more than one person. Kali loa usually consist of a bamboo headpiece with a pair of wooden legs attached to each end. He also assured me the word fafapo is not Tongan. It does not appear in Churchward's Tongan Dictionary.

Another informant elaborated on the cultural significance of kali. Born in 1908, Uluaki Tangata resides in the village of Niutoua on western Tongatapu and is the current holder of the chiefly title Tamale. Holders of this title have been the historians of the Tu'i Tonga since at least A.D.1200. According to Tamale, who spoke in the present tense, a kali toloni is a symbol of a young woman's virginity. When a woman is to be married, the family makes the best kali it can, a kali toloni, since, as Tamale stated, it requires more work. On the day of the marriage, the woman is taken to the man's house with a basket (kato alu) containing a fine mat, scented coconut oil and a kali toloni. Tamale also stated that kali toloni, which are more delicate, are used by women and kali hahapo are used by men. The differential use of headrests based on gender was known to only one other informant. Setaleki Iloa, a professional carver from Vaini, Tongatapu, stated to me that he had been told this many years ago by an old man in the Ha'apai Group. This is in contradiction to a statement by Childs (1944) that kali laloni and kali toloni were used by chiefs, and kali hahapo were used by commoners.

- 414

There are two other types of Tongan headrests which are quite rare. According to Kaeppler (1978:228-30), the few known examples of these types may have been collected during Cook's voyages. One type has two legs extending out from under each end of a rectangular headpiece. A variation of this type has two legs at one end and only one leg at the other. A second type has five legs, two at one end and three at the other, extending directly from the ends of a comparatively narrow headpiece.

Figure 4: Headrests possible collected during Cook's voyages: (top) from a drawing by Cleveley in the British Library, possibly of a headrest in Florence, (middle) headrest collected by William Griffin of the Resolution during Cook's third voyage, in the National Maritime Museum (Greenwich), and (bottom) headrest in the Museum für Völkerkunde (Vienna), drawn by Kelly Shows.

References to or drawings of headrests by early Western visitors to Tonga are rare. Other headrests identified by Kaeppler (1978:228-30) as possibly being collected during Cook's voyages are what contemporary Tongans would classify as kali toloni. Cleveley, the artist on Cook's second voyage, drew such a headrest on Tongatapu in 1774 (Joppien and Smith 1985:190). Johann Forster, also with Cook on his second voyage, drew a kali toloni on Nomuka, Ha'apai, in 1774 (Joppien and Smith 1985:219).

Just over three decades later, William Mariner (Martin 1827:97), who resided in Tonga from 1806 to 1810, stated, “A pillow to sleep on, in these islands, consists merely of a rod of wood about an inch in diameter, and a foot and a half long, and raised about half a foot by two diverging pieces at each end.” Drawings from the first voyage of d'Urville, who visited Tonga in 1827, depict two types of headrests. One drawing - 415 (St Cartmail 1997:48) has a cylindrical headpiece with a pair of bowed legs at each end and appears to be the type described by Mariner (Martin 1827:97). An example of this type with whale tooth inlay is in the collection of the Auckland Museum (Registration number 31811) and has been identified by Fanua (1987) as a kali toloni lei. The second drawing from d'Urville (St Cartmail 1997:42) is of a kali hahapo. It differs from kali hahapo found in Western collections and in contemporary Tonga in that the legs curve down from the headpiece instead of forming a sharp angle.

Since, according to Mack (1982:252), the kali hahapo (which he inappropriately referred to as the “M-form neckrest”) is known only from 19th and 20th century collection dates and does not appear to be evolved from other Tongan forms, he suggested it was a post-European introduction from the Society Islands. Two examples of a similar type from the Society Islands are illustrated in Force and Force (1971:120). “During the first half of the 19th century, many Polynesians, including high-born, traveled widely in Oceania on whaling ships and other Western vessels. These Polynesians frequently collected objects from other island groups, and it would not be surprising if this M-form neckrest found its way to Tonga in this manner, where its novel shape gained wide popularity” (Mack 1982:252). It is just as probable, if not more so, that the diffusion was from Tonga to the Society Islands, but until comprehensive studies of Fijian and Society Island headrests are undertaken, nothing definitive can be said concerning inter-island relationships or chronological sequences of headrest types except that the number of types in Tonga has diminished since European contact.

It is evident that, at least since European contact, Tongans have used three criteria—form, function and material—to classify headrests. In contemporary Tongan society, the primary consideration appears to be form, the fundamental differentiation being between kali toloni with four legs, and kali hahapo with two legs. Kali loa are defined by function—being large enough to be used by more than one person—although in form, they may resemble either kali toloni or kali hahapo. Since kali loa were frequently made with a bamboo headpiece, those that were may also be referred to as kali kofe. And any type of headrest with whale tooth inlay may be referred to as kali lei. None of my informants differentiated between headrests with four legs made of a single piece of wood and headrests with four legs made of three pieces as did Childs (1944). None were familiar with his term kali laloni. Churchward (1959:280) defines laloni as the “bottom row of flowers on an ornamental girdle”.

I have observed Tongans using headrests on several islands in the Ha'apai Group. It is the head (the base of the skull) that rests on the headrest, not the neck. Consequently, the preferred English word should be headrest, not neckrest.

There is also a problem with the stated function of Tongan headrests. In his discussion of Tongan headrests, Mack (1982:250) stated, “Wooden neckrests were used in the Society, Samoan, Tongan and Fijian Islands. The purpose of these pillows was to keep the men's elaborate coiffures from flattening during sleep.” Edler (1990:40), writing about Tongan headrests, stated, “Headrests of various kinds were used throughout Western Polynesia. The immediate use for headrests was practical—to protect the elaborate coiffures worn by men from being damaged or flattened during sleep.” And Wardwell (1994:178), also writing about Tongan headrests, stated, “The basic function of such neckrests was to protect the elaborate coiffures of their owners.”

- 416

Mack did not reference this statement either, but, from an examination of his bibliography, the source appears to be Wilkes (1845:345). Wilkes, writing specifically of Fiji, stated: “In sleeping they use a pillow made of a piece of bamboo or other species of wood, about two inches in diameter, with four legs; this is placed immediately under the neck, and is sufficiently high to protect their large head of hair from being disarranged.”

The observation that headrests were high enough to protect the hair of the user is not sufficient evidence to allow one to say this was their primary function. But the subject is Tongan headrests, and observations made in Fiji, no matter how they may be interpreted, are not necessarily applicable to Tonga. Tongans did not have the huge, frizzy hair styles for which the Fijians were well-known. In the absence of any historical records or anthropological accounts to the contrary, I suggest we accept the view of contemporary Tongans that headrests function, and have always functioned, as “Tongan pillows”, to be used while sleeping or resting.

  • Childs, T., MS. 1944. Letter to Auckland Museum. Ethnology Department files, Auckland Museum.
  • Churchward, C. Maxwell, 1959. Tongan Dictionary. Nuku'alofa: The Government of Tonga.
  • Edler, John Charles, 1990. Art of Polynesia. Honolulu: Hemmeter Publishing Corporation.
  • Fanua, Tupou Posese, MS. 1987. Notes on Tongan collection in the Auckland Museum. Ethnology Department files, Auckland Museum.
  • Force, Roland W. and Maryanne Force, 1971. The Fuller Collection of Pacific Arts. New York: Praeger.
  • Gathercole, Peter, Adrienne L. Kaeppler and Douglas Newton, 1979. The Arts of the Pacific Islands. Washington: National Gallery of Art.
  • Guiart, Jean, 1963. The Arts of the South Pacific. New York: Golden Press.
  • Joppien, Rudiger and Bernard Smith, 1985. The Art of Captain Cook's Voyages, Volume Two: The Voyage of the Resolution and Adventure 1772-1775. New Haven and London: Yale University Press.
  • Kaeppler, Adrienne L., 1978. “Artificial Curiositiesbeing An Exposition of Native Manufactures Collected on the Three Pacific Voyages of Captain James Cook, R.N. Honolulu: Bishop Museum Press.
  • Mack, Charles W., 1982. Polynesian Art at Auction 1965-1980. Northboro: Mack-Nasser Publishing.
  • Martin, J., 1827. An Account of the Natives of the Tongan Islands in the South Pacific Ocean. Compiled from William Mariner. 3rd edition. Edinburgh: Constable.
  • St Cartmail, Keith, 1997. The Art of Tonga. Honolulu: University of Hawai'i Press.
  • Thompson, Laura, 1940. Southern Lau, Fiji: An Ethnography. Bernice P. Bishop Museum. Bulletin 162. Honolulu.
  • Wardwell, Allen, 1994. Island Ancestors: Oceanic Art From the Masco Collection. Detroit: University of Washington Press, in association with the Detroit Institute of Arts.
  • Wilkes, Charles, 1845. Narrative of the United States Exploring Expedition During the Years of 1838, 1839, 1840, 1841, 1842. Philadelphia: Lea & Blanchard.