Volume 108 1999 > Volume 108, No. 2 > The changing texture of textiles in Tonga, by Phyllis S. Herda, p 149-167
                                                                                             Previous | Next   

- 149
Would that my doing were as a double mat,
Utufeolo that is doubled
With the weaving of utukaunga.
Unworthy is the mat of Tungua,
But yet 'tis the portion o Falotuma.
Bring then weaving of Matuku,
Efu mat and mat of red feathers,
Fatu mat is a mat not plentiful,
The Tongan kie is in abundance.
Weave fihu in Uiha and Lifuka.
And take Vava'u, and gather them
With the cloth of the plover and twin birds.
A kie of Toku for his adorning,
Or a mat of Late for clothing to Mua.
Finished is the wealth of our land.
Ulamoleka (Collocott 1928:81-82)

The production of cloth in the Tongan islands of western Polynesia was, and continues to be, considered the work of women. In the past this included the making of ngatu or tapa (cloth made from the bark of the paper mulberry tree), and fala and kie (mats woven from pandanus leaves). Certain types of barkcloth and mats are categorised by Tongans as koloa, a term translated as ‘wealth’, ‘possessions’ or ‘what one values’. At one level, koloa made by women reflected the high status sisters occupied vis-à-vis their brothers whose labours and crafts, in contrast to koloa, were deemed ngaue (or ‘work’). At another, more salient level, the making, presenting and exchange of koloa at significant life events, including births, weddings, funerals and title installations, acted out status differentiation between the chiefly ('eiki) and the non-chiefly (tu'a). Koloa continues to be used at significant life events in modern Tonga. In fact, the production of both ngatu and kie continues at a great rate in Tonga with no indication of decline in production in the near future.

- 150

Ngatu and kie are also important items for expressing a national identity within gender and kin relations for Tongans living overseas. Significantly, however, the raw materials for producing textile koloa are not generally available to those living outside of the islands. Consequently, Tongan women living abroad invoke kin relations and obligations to obtain the necessary koloa and have also expanded their textile repertoire to include machine sewn patchwork and appliqué bedcovers or “quilts” in order to fulfill cultural expectations. These innovations, while on the one hand representing a departure from past practices, have also served to complement rather than replace traditional Tongan practices involving koloa. Indeed, it is clear that women's critical role in producing textile koloa has remained central to social and economic exchange in Tonga, although in some cases the parameters of these exchanges have shifted.


Cloth made out of the processed inner bark of the paper mulberry tree has been made in Tonga, as it has throughout Polynesia, for as long as anyone can remember (Koojiman 1972). Its value as a form of wealth in Tonga is also long-standing. Descriptions such as the one made on James Cook's second voyage at 'Eua indicated that it carried clear significance as a prestige item in the later 18th century:

We were welcomed a shore by acclamations from an immence [sic] crowd of Men and Women … they crowded so thick round the boats with Cloth, Matting, etc. to exchange for Nails, that it was some time before we could get room to land; they seemed to be more desirous to give than receive, for many who could not get near the Boats threw into them over the others heads, whole bales of Cloth (Beaglehole 1959:245).

Four years later at Tongatapu, little had changed. The British were somewhat overwhelmed by the amount of ngatu which was given to them— precisely the effect the Tongans wished to impart:

In the after noon a party of us accompanied by Feenough [Finau] went to pay him [Maealiuaki] a visit; we found him amongst a great number of peop[l]e of both sex seated round a large piece of Cloth, not less than 40 yards in length, spread out on the ground …. Before I took leave the piece of Cloth was rolled up and with a few Cocoanuts given to me (Beaglehole 1967, I:127-28).

Conspicuous display was and is an important part of the presentation of - 151 koloa and, as such, large quantities are often amassed and presented in a manner which enhances the prestige of the giver and impresses the receiver. In the late 1790s, the LMS missionaries attended funerary services where Tu'i Kanokupolu Tuku'aho was presented with 35 bales of ngatu as well as a quantity of fine mats (Wilson 1799:244). Fifty years later a Wesleyan missionary recounted how at one chiefly Christian marriage, 10 cartloads of ngatu were presented (Lawry 1851:10). At the marriage of Princess Pilolevu in 1976 a piece of ngatu purported to be the length of a rugby field was made and laid out for the newly married couple to walk upon. The size emphasised the deference to and the distinction of the couple. The importance of a ngatu component in significant presentations continues today with its occupying a prominent place at traditional events such as weddings, funerals and title installations as well as at modern functions such as graduations and 21st birthdays.

There is disagreement over the designation of barkcloth in pre-20th century Tonga. Ngatu is the common term used in modern Tongan, although tapa is understood as well. Tamahori (1963:11) asserts that ngatu is the traditional Tongan name and that tapa represents a later innovation borrowed from other Polynesian languages, presumably through early European contact such as the Cook expeditions. Mariner (Martin 1827, II:206), however, records that tapa referred to unpainted barkcloth while ngatu designated the decorated form.

Kaeppler (pers. comm.) refines this definition by pointing out that in modern Tongan tapa, technically, refers to the unpainted border of the barkcloth while ngatu refers to the entire decorated piece. In the past (up until the 1960s), women often produced ngatu designated tapai ngatu that purposely incorporated a significant amount of “white” or unmarked areas as a background to the principal decoration. There is an added significance of a chiefly woman transforming plain cloth—tapa—into textile wealth— ngatu, which I will return to. At this point, suffice it to say that both words—tapa and ngatu—designated barkcloth in pre-Christian, pre-monarchical Tonga. However, the distinction of the two terms (decorated or not) is significant, and the two terms are not interchangeable.

Making Ngatu

The raw material is taken from the young paper mulberry tree from which sections are cut and set aside for several days in order to facilitate the stripping of the bark. The bark is then sun-dried, stored and soaked before it is beaten with a wooden mallet (ike) against a long wooden anvil (tutua). Through the beating process, the bark is flattened, spread thinly and felted together to produce single sheets of barkcloth composed of several layers. Single - 152 sheets are pasted together to form the final sized piece. While the barkcloth layers are being pasted together, a pattern is rubbed onto it from a coconut fibre block (kupesi) that is attached to the pasting board. Initial decoration is painted on at this point, with final painting being done after the cloth has been fully dried in the sun.

Late 18th and early 19th century technological descriptions of ngatu manufacture could apply to modern procedures (Beaglehole 1967, II:905-6, Tamahori 1963:100-3, Martin 1827, II:202-6, Koojiman 1972:300-8). The differences in technique are minor. In the late 18th and early 19th century, instead of being sundried as it is today, the hiapo bark was baked before being soaked and beaten, and the resulting feta'aki (sheet) used to be glued together with a vegetable dye instead of being pasted together with vegetable starch. In addition, the width of a modern finished piece of ngatu is estimated to be almost double that of the late 18th and early 19th century.

In the past ngatu was made of three or sometimes more single layers glued together, instead of the more recent trend of only two sheets (Martin 1827, II:204-5). (Kaeppler [pers. comm.] suggests, however, that “two sheets” may, in fact, refer to the sheets being pasted across one another with any number of feta'aki making up the crossing sheets.) A feta'aki innovation which is not well regarded in Tonga is the substitution of plastic weave or manufactured sewing interfacing for one of the sheets. Known commercially as Pellon, the interfacing does have a barkcloth-like appearance and feel to it. Clearly labour saving in its intent (only one feta'aki of barkcloth needs to be prepared), the inferiority of the end product, as well as the disdain with which Tongan women speak of it, indicates that it is unlikely that the practice will become widespread. I have also heard reports of ngatu designs being stencilled onto plastic sheeting or canvas, and the subsequent attempt to pass off the resulting cloth as a ngatu substitute. While several women commented that the canvas could be useful around the house, most were contemptuous of the attempt and clearly indicated that the cloth was neither ngatu nor koloa. Another labour-saving innovation in recent ngatu manufacturing was the conversion of a wringer washing machine which pressed the soaked bark tutu and, hence, greatly reduced the amount of beating it needed. However, at this stage it exists only in prototype and most women were sceptical about how efficient or practical the invention was.

There has, of course, been considerable innovation in the decoration of ngatu. Pre-19th century designs tended to be geometrical patterns applied between borders. Abstract designs were introduced in the 1800s and became very popular after the turn of the 20th century (Koojiman 1972:330-40, Kaeppler 1998). Several of these, such as the coat of arms and the hala pini - 153 (‘pine road’ referring to the Norfolk pines planted around the Palace), are now regarded as classic designs and are often used as expressions of Tongan national or ethnic identity. Not all of these innovations are well regarded as aesthetically pleasing. For example, the substitution of modern fabric dyes or paints for traditional vegetable dyes has been tried since the 1970s, but has not really caught on, with most women regarding the results as vulgar. One woman indicated to me, however, that if a high quality dye was used, the results were very good. I have also seen bright orange coloured ngatu for sale in the Nuku'alofa market although both Tongans and papālangi (Europeans) regarded it contemptuously.

Communal Workgroups

The pasting of the feta'aki into the desired length of the finished piece and the concomitant rubbing over the kupesi that imprints the desired design is, and has in the past, been performed by groups of women in a work party known as the koka'anga. The women gossip, joke, sing (often bawdy) songs and generally have a good time while the work is accomplished. Barkcloth is made in segments known as langanga, which were traditionally measured by the handspan of the organiser of the group (kautaha) brought together for the making. Erskine's 1840s description of these groups could easily apply to contemporary Tonga:

In our walk through the village… we came to a house, in the front of which, inside of the enclosure, a number of women were busied with the operation of ‘kokanga’ [sic], viz. the cementing of native cloth, and colouring it with the juice of the ‘koka’ bark,—an occupation which, from the good humour of those assembled, seemed to be a kind of industrious merry-making, resembling an American [quilting] ‘bee’ (Erskine 1853:135).

The final decoration, usually done by a much smaller group of women or a woman alone, consists of freehand painting with vegetable dyes (Koojiman 1972:305). Again, Erskine's description is remarkably similar to contemporary techniques:

We have seen this operation going on on the public green (or malai [sic]) and was surprised by the steadiness of hand of the woman performing it, who makes use of a longish stick, resembling a large camel's hair pencil, and completes the pattern with great dexterity and quickness (Erskine 1853:136).

There is considerable debate in the historical and academic literature - 154 about the communal or co-operative nature of ngatu production (Nayacakalou 1959:108-9, Tamahori 1963:137-53, Koojiman 1972:318-30, Afeaki 1987, Small 1987). While the koka'anga part of ngatu manufacture was probably always communal because of the nature of the work, the issues surrounding the constituency and ranked control of the group as well as the control of the finished piece of ngatu are worthy of review.

Small, in her study of women's mutual aid associations or co-operative organisations in Tonga, argues that the kautaha was an institution designed “for the production of traditional wealth for commoners” and that it was born out of social and constitutional change. She contends that it “died” after the Second World War, to be replaced by the tou langanga, a ngatu production group ostensibly organised around the expanding productive base of the archipelago and its full-scale entry into a cash economy (Small 1987:436-46). Much of Small's detailed descriptive work is interesting and revealing, although her overall analysis falls somewhat short in understanding the changing institutions surrounding koloa production, accumulation and presentation.

There is no question that koka'anga work organisations changed over time, and that these changes reflect socio-economic as well as political and spiritual/religious transformation within Tongan society. Indeed, it would be surprising if they did not. Since the late 18th century Tongans have adopted Christianity, gradually introduced a cash economy, and witnessed a major political transformation with the establishment of a constitutional monarchy. Despite these (and other) major transformations to the Tongan way of life, women continue to koka'anga communally, although how that communality is organised has altered.

Orlebar clearly saw women's koka'anga work groups in 1830: “It [the making of barkcloth] becomes nearly the sole occupation of women, who meet in parties and beat a sort of tune with their clubs that may be heard for miles” (Orlebar 1976:23). Significant in this description is indication of how rank imbued the work organisation. Before the Constitution of 1875, chiefly women were in complete control of the labour and the products of common (or tu'a) women. 'Eiki women would organise a tu'a work group to manufacture barkcloth up to the stage of applying decoration. This included all of the labour involved in producing barkcloth, not just the pasting and imprinting stage represented in the koka'anga. Hence, Orlebar's description of common women being gathered together in order to beat the tutu reflects the way in which the workgroups were organised—namely, in direct response to chiefly need. In contemporary Tonga the beating of tutu, a laborious task, is often done in small groups or, more commonly, alone - 155 when a woman has time during the day. In the mid-19th century, female prisoners were forced to beat tutu as part of their punishment (Erskine 1853:136). This is reminiscent of the labour of female prisoners of war being reserved for beating tutu for victorious chiefs (Martin 1827, I:188). In each case, the prisoners would have been tu'a or non-chiefly women. This organisation for the chiefs can also be seen in the early European accounts which describe plantations being specifically farmed for hiapo (paper mulberry) for ngatu manufacture (Labillardiere 1800:338).

The chiefly woman who presided over the group provided and may have made the kupesi or tablet used to imprint the tapa:

The cobéchi [kupesi], or stamp is formed by the dried leaves of the páoongo [paongo] sewed together so as to be of a sufficient size, and afterwards embroidered, according to various devices, with the wiry fibre of the cocoa-nut husk;…. They are tied on to the convex side of half cylinders of wood…. The stamp being thus fixed, with the embroidered side uppermost, a piece of the prepared bark is laid on it, and smeared over with a folded piece of gnatoo [ngatu] dipped in one of the reddish brown liquids before mentioned, so that the whole surface of the prepared bark becomes stained, but particularly those parts raised by the design in the stamp (Martin 1827, II:204).

Kupesi were the prerogative of chiefly women alone, as was the final decoration of skilled painting on the piece. The significance of this prerogative is twofold. Firstly, the finished piece would be imprinted with the personal design of the particular chiefly woman and, hence, identifiable to her. Secondly, and more significantly, the prerogative of decorating the piece of barkcloth meant that the chiefly woman was responsible for its transformation from tapa to ngatu—she created it through these final processes. It did not become koloa until she had acted upon it in a chiefly manner. This encompassment of non-chiefly women's labour was, certainly, in keeping with Tongan notions of chiefliness. It was her mana, not the labour of the workers, which was imbued within the fabric and made it valuable (i.e., koloa). It was her presence, her skill, her mana that made it count.

Small (1987) argues that a changing conception of how ngatu is valued can be seen in the changing organisation of ngatu-producing work groups. However, this masks the transformation occurring. There is no intrinsic value added to ngatu because it is produced by a group of women. It was the chiefliness of the woman who controlled production which made the significant difference in the past; not the association of women. Unlike ngatu manufacturing, the plaiting of mats is not necessarily done in a group. - 156 Certainly there is not now, and seemingly never has been, formal organisation of women akin to the kautaha or tou langanga. Women do get together to plait and often will work on the same mat, but there is not that same sense of formality as is found in the group nature of ngatu production (Leslie 1996:10-11). Yet fine mats are seen as the epitome of Tongan wealth; they “outrank” ngatu on the koloa scale. Rather than imbued wealth of a co-operative nature, the presence of group work has more to do with the practical issues of pasting huge bales of barkcloth—it is simply not feasible to accomplish it in any other way. However, how those koka'anga groups are organised does reflect changing social ideals.

In the early 20th century Small believes that the kautaha emerged as a women's “tapa production unit” (to use her words) in order to allow common women to produce barkcloth and gain prestige (Small 1987). On one level she is correct, named groups called kautaha did emerge in the early 20th century and chiefly women were asked to provide their patronage in a manner somewhat reminiscent of the control of ngatu production in pre-Constitutional Tonga. The kautaha's patron guided the group and was honoured by the women with prestations, although not to the extent of pre-Constitutional chiefly appropriation. She still provided the kupesi tablets and, according to Tamahori, directed the aesthetic direction that the decoration took (1963:135-37). Again, in a manner reminiscent of the transformation of creating ngatu out of tapa, the chiefly woman's rank created the wealth imbued in the cloth, but without full scale appropriation of the end result.

Gifford makes it quite clear that when women did provide their labour in a directly unreciprocated form, they were, at least in the 1920s, compensated in another way (Gifford 1929:147-48). Koloa was (and is) not simply the accumulation of goods; who made them and what they were were important (and still are for that matter) in terms of how that textile was transformed into koloa. As was previously mentioned, not every form of barkcloth was deemed “valuable”; similarly not every mat was koloa. Textile wealth (koloa) did, and still does, follow rank. However, focusing on western models of independence and self-sufficiency, not to mention cash accumulation, masks the processual analysis of “what Tongans value” (i.e., what counts as koloa) and how textiles are important in this sphere. In this manner, they demonstrate how outside influences were “captured” by a Tongan ideology and domesticated as Tongans sought to meet traditional expectations.

This is not to argue that the production of ngatu is not caught up in the shift of Tonga to a cash economy. Because they fetch between 1,000 and 1,200 pa'anga a bale in the Nuku'alofa market, one would be foolhardy not - 157 to see this as significant in monetary terms. In addition, in recent years, Tongan women have taken out development loans in order to purchase raw materials for ngatu production, using, incidentally, traditional textile koloa for collateral. Upon completion of a group's ngatu production, the women raise funds in order to repay the loan, thereby keeping the koloa they produce. In addition, Tongan women through the Langa Fonua, send ngatu to Auckland where it is sold, on their behalf, through a local Tongan church. In keeping with New Zealand market demands, the ngatu is cut up into smaller pieces which are sold as wall hangings. The prices, however, are much higher than that size ngatu would fetch in Tonga. Despite the significance of these examples of ngatu entering the cash economy, it is important to remember the intrinsic value of koloa is not tied up in cash.

A striking contemporary example can be seen in the manufacturing of ngatu 'uli (black ngatu). It is, by far, the most valued of ngatu and is rarely made. This is said, offhandedly, to be because the intense black dye which is made from candlenut soot is supposedly difficult and time-consuming to make. While this is, undoubtedly, an accurate statement, it is not helpful in gaining an understanding as to why ngatu 'uli is rarely made, especially in terms of the argument that ngatu is, in contemporary Tonga, made for economic—i.e., cash—gain. If this were the whole story, then the investment of time and energy in making candlenut soot would be well recompensed— as ngatu 'uli is greatly prized. Instead, it seems the reason ngatu 'uli is not made by commoner women for personal use or for sale is that it would be fie 'eiki (acting uppity/above one's station) to do so. Even in modern, post-Constitutional, cash economy Tonga it is not acceptable for this highest of barkcloths to be produced, except at the instruction of chiefly women for a chiefly purpose. Kaeppler (1995) includes a photograph of the most beautiful ngatu 'uli which was produced exclusively for and presented at the titular appointment of the King's youngest son to the title of 'Ulukalāla. Even in the photograph its fineness is evident. In terms of design, colour intensity and exceptional skill shown in every aspect of its execution, it is most definitely koloa of the highest order. That it would be made for sale is unthinkable; its value is greater than cash. Its value is based on its high chiefly connections. Again, those same traditional themes—an intrinsic feature of koloa, i.e., ‘what one [as a Tongan] values’, is the chiefly input. If the input is not there, neither is the value.

Kautaha were avidly supported and their members inspired by the late Queen Sālote. The rapid expansion of kautaha groups and their push to construct buildings for meetings and koka 'anga appear to have been at the late Queen's direction. It coincided with the establishment of women's village associations in Western Samoa, a modern transformation of the traditional - 158 Samoan aualuma or village women's grouping. Defined as the female sphere, including a building, in Samoan village life, it comprised the sisters and daughters of the village. Traditionally, it served as a educational institution for girls to learn mat and siapo (barkcloth) making, dancing and oratory (Schoeffel, pers. comm.). In the 1920s it was reconstituted to include individual health issues as well as village hygiene. The similarities between these Samoan village associations and the shift in Tongan koka 'anga work groups are significant. It appears that the two organisations were in contact, and that Queen Sālote promoted the kautaha within Tonga (Fairburn Dunlop, pers. comm.).

The shift of kautaha to the tou laganga as the tapa producing unit in Tonga, coincides, I contend, with the establishment, again by the late Queen, of the Langa Fonua 'ae Fefine Tonga in 1953. The name, tou langanga, is not widely used in Tonga today; most women refer to koka 'anga groups as kautaha. In the 1950s Nayacakalou (1959:109) recorded that tou langanga referred to a “tapa-joining party”, while he applied the term kautaha to work groups in general, composed of either sex.

Through links with the Pan Asian and Pacific Women's Association, Queen Sālote was informed about what was happening elsewhere in the world and sought to implement similar changes in her Kingdom. In addition to social and health care, Queen Sālote, through the Langa Fonua, attempted to raise the quality of Tongan koloa production by commoner women, believing that much of the work was becoming sloppy (Tamahori 1963:241). She also encouraged rural women to supplement their meagre incomes by making handicrafts based on traditional textile koloa techniques for sale in the local and international tourist and handicraft market. Both the kautaha and the tou langanga had clear chiefly (and very high chiefly) patronage— this is what launched the ventures. Spin-off effects in terms of increased access to wealth, increased commoditisation and personal autonomy were, in effect, supported by this patronage.


There are many kinds of mats in Tonga and it is important to note that not all mats are considered by Tongans as koloa. Coarse mats have been used as room dividers and floor coverings as long as anyone can remember; they do not represent wealth. This is not to say that they are not of good quality. George Vason who was resident in Tonga at the end of the 18th century commented that:

The mats were very curiously wrought with a tasty intertexture of different coloured threads: some mats were so closely knitted together, that when - 159 laid on the floor, they were a secure defence from insects, as they were impenetrable to them; and the insects crept under them, in the same manner as under a floor-cloth, or a board (Orange 1840:106).

Women make mats from several types of pandanus leaves: kie, paongo, tapahina tofua and ngafingafi, which differ in colour as well as in fineness of texture. Their purpose and quality is partly based on the type of pandanus prepared and plaited as well as the narrowness of the pandanus strip used. Truly fine mats (kie, ngafingafi) represent, in the past and in the present, the epitome of Tongan textile wealth, which is to say they represent the epitome of Tongan wealth. More value can be assigned to one fine mat than to a large bale of ngatu. In contemporary Tonga fine mats are never sold, whereas bales of ngatu generally are available in the marketplace. The value Tongans placed on fine mats was not lost on early European visitors to the archipelago. In 1793 a Spanish expedition recorded that:

The gift which [Vuna] offered of a large pig, much fruit and roots and some mats of the greatest size and delicacy, were able to give a true idea of his generosity and grandeur (Novo y Colson 1885:261).

The finest of Tongan mats are distinguished, and, hence, their fineness enhanced, as named mats (kie hingoa) with their genealogies intact. Many of these kie hingoa originated in Samoa (where they are known as 'ie tōga or 'ie o le mālō) and were brought to Tonga at times of marriages between the very chiefly individuals of both island groups or at significant life events (births, title investitures or funerals) of the descendants of these chiefly individuals.

Many of the well-known characteristics of kie are important for a general understanding about Tongan textile wealth (koloa) and how it has changed over time. Unlike ngatu manufacture, the plaiting of truly fine mats was, in pre-Constitutional Tonga, the exclusive prerogative of chiefly women. Indeed, part of their fineness was derived from the fact that their makers were exclusively of chiefly birth. The task was time consuming and required considerable skill:

These mats are made entirely by hand, and when very fine and large, occupy two years making…. They are so exquisitely manufactured, that one would suppose them to be woven by a loom (Martin 1827, I:142).

It was as if the chiefly woman plaited herself (or perhaps more accurately her mana) into the mat. I might also add that the quality of these mats was - 160 superb in terms of the type of pandanus used (kie), the preparation of it for plaiting (the narrower the final width of the strip, the finer the results) as well as the skill of the maker. While the rank of the women who prepared the pandanus (in this case kie) for plaiting is unclear and, indeed, non-chiefly women's labour may have been used, the actual plaiting of kie was an important aspect of even very, very chiefly women's lives in pre-Christian, pre-Constitutional Tonga. Early European records chronicle the surprise of their papālangi authors that women of such high rank would be so employed “without prejudice to their rank” (Martin 1827, II:97). Coarser, everyday mats were, of course, historically plaited by non-chiefly women. I think its important to consider this hierarchy of mats because it seems to me that this ranking is an important aspect of the ideology of koloa which is often overlooked by western scholars. Not all mats are koloa. The fineness of a fine mat (ngafingafi, kie hingoa) is recognised and, hence, valued as koloa for a number of reasons including the rank of its maker as well as its positioning vis-à-vis lower ranking mats. In other words, part of the reason they are fine is because they are chiefly.

Further to this is the need to consider the regional significance attached to some kie hingoa, and how this plays out in terms of their ranked value. As previously mentioned, many of these mats were plaited in Samoa and were brought to Tonga at significant life events of Tongan/Samoan elite. Inter-archipelago kin, economic and political links are also represented by the inclusion of highly prized red feathers from Fiji (most particularly Lau) which were woven into the fringes of many of the kie hingoa (Kaeppler 1978). These inter-archipelago unions and their commemoration are important markers of chiefliness—a way, if you like, of binding rank, through blood, across what are now national boundaries and what were, pre-19th century, political and kin boundaries. Once again, the elite are defined against the common or non-elite—separate and distinct from them—by being linked with other elite. Political pre-eminence is also displayed in these kie hingoa, as it is only among the pre-eminent (those who are either of very high blood rank or, more usually—as in the case of the late 19th century 'Ulukālala line and the Tui Nayau of Lau—those who are powerful and desire rank) that these marriages take place. The fine mats presented at these marriages and, later, safeguarded and displayed at significant events, therefore, signal to the world (and especially to other Tongan or west Polynesian elite), political and rank pre-eminence. This display and its associated prestige value are important and are often overlooked by scholars seeking to understand Tonga's past. A Samoan name for fine mat, 'ie o le mālō, clearly reflects this political connection and prestige value—mālō being the traditional designation of the victor in a political challenge.

- 161

Given these kinds of political and kinship associations, it is not, perhaps, surprising that much of the language, as well as the action, surrounding kie hingoa are strikingly similar to those used when discussing hohoko (genealogies). Indeed, kie hingoa can be said to have hohoko. While the identity of the maker may or may not be recalled, the names of renowned individuals who possessed the mat, as well as the occasions when it appeared or was presented, are always remembered and commemorated (see Kaeppler this issue). For example, the late Queen Sālote, speaking of one prized kie hingoa, explained that:

Each line of kings had its own ceremonial mats which were carefully preserved from generation to generation. In fact, our history is written, not in our books, but in our mats…. The ta'ovala I wore when I met Queen Elizabeth on Her Majesty's arrival in Tonga was 600 years old. Worshiped in the 13th century as a symbol of the ancient gods, the mat belonged to the chiefly family of Malupo on the island of 'Uiha (Bain 1967:77).

The value of this and other named fine mats is safeguarded by the high-ranking women who look after them, as well as by the women and men who remember and relate the long and intricate histories of each mat. Weiner (1989:44,53), speaking of Samoa, mistakenly assumes that the high value placed on these mats is because of their antiquity. Although the most prized kie hingoa are very old, it is not their age per se which makes them valuable, rather it is their accumulated experience and the illustrious individuals who have come in contact with the fine mat that add to its greatness and value. They are kept and, hence, old because they are valuable: not valuable because they are old.

Like the making of ngatu, little has changed in terms of the technology of mat weaving, or more accurately, plaiting in Tonga. There have been a few minor innovations to decorative plaiting. Bright coloured yarn or wool is now often plaited in as a fringe. This practice seems to have begun around the turn of the 20th century. More recently, imported commercial fabric dyes have been used to colour ta 'ovala (waist) mats. But this is almost all, except for slight decorative plaiting innovations.

However, in terms of who plaits mats—especially fine mats—there has been significant change. As previously mentioned, in pre-Constitutional Tonga the plaiting of mats was exclusively the work of chiefly women. Theoretically, this changed with the emancipation of the common people in the last quarter of the 19th century following the 1875 Constitution. I say theoretically, because although tu 'a women obtained the legal right to plait fine mats in the Constitution as part of the general emancipation of the - 162 common people, it is my suspicion that it was many decades (and several generations) before they began to actually plait them. Although somewhat difficult to uncover, this type of documentation is important in an understanding how the production of koloa has changed over time. Today in Tonga, the skill of the individual woman who plaits is, perhaps, more important than her birth rank in society in terms of how the mat will be regarded. Certainly, it is what is talked about in terms of the quality of a kie. However, this is not to say that the rank of the maker is not still of crucial importance in terms of how the mat is regarded. What is clear is that chiefly women do not appear to be producing mats in the quantity that their ancestors did.


Tongan women have been crocheting and quilting bedcovers for many decades. There is some debate as to when women began this kind of craft production, but it is generally agreed that it was prevalent in the post-Second World War years and, most probably, existed, to a lesser degree, before the War. Since the 1970s, however, women's “quilts” have acquired a significance beyond that of utilitarian bed linen—their previous categorisation. While not holding the deep cultural value of other textile koloa, quilts are produced and presented as a kind of koloa, albeit an inferior one—koloa si'i, if you will—and represent a technological and ideological hybridity within Tongan notions of ‘wealth’ and ‘what one values’. However, as hybrids, the quilts play on the themes of traditional textile koloa mimicking some, but, certainly not all, of their qualities. In addition the quilts, as a form of textile koloa, provide a connection between the home islands and the Tonga diaspora resident elsewhere, providing a means of connecting with a Tongan identity in a customary manner while also creatively claiming a unique migrant identity.

The first area in which quilts appeared in a koloa type format (i.e., not merely as bed linen) was as grave decoration or art placed at the head of the grave (known as tapu) and as funerary gifts in the early 1970s (Kaeppler, Teilhet Fisk, pers. comm.). Initially these tapu quilts were made by Tongan women living overseas, who found themselves unable to fulfill normal mortuary expectations when a close relative died. Traditionally tapu were fine mats or ngatu which were held taut by two poles and were placed at the head of the grave (Teilhet Fisk 1991:52-53). Living away from the islands in an environment where the raw material—be it hiapo (paper mulberry) for ngatu or kie (pandanus) for mats—was simply not available, and often being unable, usually because of financial constraints, to make the trip back to Tonga for the funeral, these migrant women created textile koloa by - 163 working with cloth and creating quilts. In a manner not unlike that used to transform hiapo and kie into traditional textile koloa, the women processed the cloth by cutting, stitching and appliquéing it, thereby transforming it into something new. Like the traditional textile processes, quilt-making embodied the maker's status, mana and, most significantly, their relationship to the deceased in a manner relevant to them-their 'ofa and faka 'apa'apa was clear in the cloth they constructed, even if the cloth itself looked different.

These first grave art quilts as tapu (the decoration at the head of the grave) were not entirely well received in Tonga; women and men I spoke with remembered being dismayed at the lack of sensitivity towards tradition with regards to these grave presentations. They were shocked when the quilts began appearing as tapu on the graves themselves, along with lengths of manufactured cloth. Nowadays, beautifully pieced and appliquéd quilts, sometimes made in Tonga, adorn graves throughout Tonga (although it must be pointed out that they are most prevalent on Tongatapu). The women who were formerly shocked and dismayed, now view the quilts as attractive, durable and, hence, practical grave decorations that are entirely appropriate in the Tongan setting. The quilts are most often made by mothers, sisters or wives of the deceased and, in some cases, were made during the lifetime of the deceased. Their inclusion as tapu at the grave expresses the double meaning of quilt as koloa si'i and utilitarian bed linen in the everlasting sleep of the deceased.

Traditional textile koloa (ngatu—esp. ngatu 'uli and kie) was presented to the highest ranking chief at the installation of a new titleholder (pongipongi hingoa) in Tonga. In post-Constitutional Tonga, this “highest chief” is, of course, the monarch of the Kingdom of Tonga with the titleholder coming from the ranks of named nobility (nopele). Quilts made their way into the pongipongi hingoa in the 1993 installation of the Tu'i Lakepa. When the audience was seated at the Palace grounds and the kava circle formed, a group of women from the Tu'i Lakepa's kāinga (kinship/political group) entered into the grounds, holding up nearly two dozen brightly coloured, machine-made quilts. The red and white banner that led the presentation of koloa was, in fact, a stuffed or puff quilt with the name “Tu'i Lakepa” emblazoned on it. The innovative decision to include the quilts in the koloa presentation resulted from the combined effort of the mother and sister of the new titleholder. The quilt production by a group of Tongan women who live in the United States was organised by the Tu'i Lakepa's sister, who lived in the Los Angeles area. Her choice of quilts was based on her inability to acquire the raw materials to produce traditional textile koloa, as well as her aesthetic appreciation of Hawaiian quilts she had seen in Hawai'i and the mainland USA.

- 164

Once again, the local Tongan reaction was not entirely positive. While most at the pongipongi hingoa were surprised and delighted with the textile innovation; most were clear that it was not koloa—at least not of the order of the ngatu and kie that were also presented that day. There was also uncertainty whether the quilts were listed on the official inventory of koloa produced that day for the Palace. Most thought that they were, whether or not they truly could be considered as koloa.

Koloa was traditionally, and is still, given to those of higher rank in Tonga as a means of honouring that individual and embodying in a material thing the inherent inequality of the relationship between the giver and the receiver. In keeping with this practice, King Taufa'ahau Tupou IV was presented with an array of valuables when he made an official visit to the Laie campus of Brigham Young University on the island of Oahu, Hawai'i, also in 1993. Among his gifts were many quilts made by the local Tongan women. Significantly, the quilts were not presented with the piles of traditional koloa (including, of course, ngatu and kie) during the taumafakava “royal kava ceremony”, but were presented at the katoanga (festivities including dancing) which followed. The meaning to the women was obvious; while the quilts were “wealth” and were “valued”, they took second place to more traditional forms of textile koloa—they were koloa si'i. Their junior or inferior status was clearly demarcated by their exclusion from the kava ceremony.

Quilts are also beginning to appear as prestations at other significant life events. Recently in Ha'apai, for example, a dozen cot quilts were sewn and presented, in addition to traditional forms of textile koloa, to a woman at the birth of her daughter by the mother's mehekitanga (father's sister). Significantly, the new mother's mehekitanga had bitterly argued with her during the pregnancy, and the community believed that the quilts, given in large numbers and accompanied by traditional textile koloa, represented the mehekitanga's public desire for reconciliation. Similarly, quilts have begun to appear among wedding presentations, again along with more traditional forms of textile wealth.


Adrienne Kaeppler (1978) has argued that the “grammar”, or underlying structure, of Tongan art follows a three part conceptual organisation similar to that of Tongan music with “melody or leading part (fasi), drone (laulalo) and decoration (teuteu)”. She develops the argument by demonstrating how the interconnecting parts function in a polyphony to emphasise the relevant features of any Tongan aesthetic or, indeed, social occasion by identifying the essential feature (through fasi), delimiting its spatial limitations (the function of the laulalo or drone), and elaborating its specific features through - 165 decoration (teuteu). Kaeppler demonstrates that this organising principle can be seen in the performance of dance (faiva), vocal music (hiva), and the production and decoration of ngatu, as well as in Tongan kin and gender structures.

When applied to the introduction of appliquéd or pieced quilts into a Tongan notion of wealth as property, Kaeppler's insightful analysis provides a framework of appreciating this introduction as an expansion of Tongan notions of textile koloa, so that quilts are encompassed within that traditional realm rather than perceived as a deviation from Tongan concepts—as a mere foreign influence. Simply stated, the quilts do not seek to replace or undermine the value of ngatu or kie; instead, they can be appreciated as the teuteu or decoration of traditional forms of Tongan textile wealth—embodying many of the inherent features of ngatu and kie that makes them part of the Tongan conception of ‘what one values’.

This process is most clearly observed in the presentations of quilts to the King at the pongipongi hingoa of the Tu'i Lakepa, and at the katoanga following the taumafakava in Hawai'i. In both instances, the quilts were presented in addition to traditional forms of textile koloa. In each case, ngatu and fine mats remained as the essential feature of the prestation; the quilts enhanced the qualities of these items by duplicating and elaborating them in a lesser form—much like the movements of a tulāfale in a performance of the tau'olunga. The tulāfale supports and celebrates the principal dancer's skill by drawing the audience's attention to the fineness of the principal dancer's movements. Both the tulāfale and the koloa si'i act as teuteu (decoration). In neither case is it the intent of the teuteu to displace or replace the essential feature, rather it is to support its pre-eminence and add to its prestige. It does not disrupt the concept of ‘wealth’ but mimics it: a form of hybridisation congruent with traditional ideals. This form of mimicry or hybridisation is nothing new in Tongan forms of textile wealth. Mariner (Martin 1827, I:296) described the funeral of a chief at the turn of the 19th century where the deceased was “wrapped up in fourteen or fifteen yards of fine East India embroidered muslim [sic]”. Once again, the effect was one of creative play with notions of wealth as well as the enhancement of traditional values rather than one of substitution or foreign hegemony.

What is also abundantly clear in this creative adaptation of tradition is the place of the Tongan diaspora in a modern reckoning of Tongan identity. The impetus for creating quilts as koloa si'i came from migrant Tongan women. It allowed them to participate in prestations to kin relations in Tonga in culturally appropriate ways. The cloth, manipulated and, in a sense, reconstructed by the patchwork or appliqué process, retained the intrinsic value of traditional textile koloa, while also allowing for innovation based - 166 on the modern situation. In each case, migrant Tongan women could have solved their prestation dilemma through commoditisation of textile koloa; they could have bought or arranged with relatives in Tonga to buy bales of ngatu or appropriate mats. However, the women chose to decline commodisation in these cases (I am not saying they would always do this) and, instead, to reinvent how Tongan women express their identity as well as their commitments.

The changing texture of Tongan textile koloa, therefore, needs to understood as a polymorphous narrative of Tongan women's social and kin experiences. While the texture of textile koloa has changed significantly over the last 250 years, it, nevertheless, retains a sense in Tonga of ‘what one values’. In addition to providing agency for modern Tongan gender and kin relations, textile koloa, in its varied forms, enacts for Tongan women aspects of Tonga's past and what it means to be Tongan in the present.


Research for this study was supported by grants from the Marsden Fund of The Royal Society of New Zealand and the University of Auckland Research Committee. I would also like to thank Adrienne Kaeppler, Jehanne Teilhet-Fisk and Elizabeth Wood-Ellem.

  • Afeaki, E., 1987. Women and NFE (Non-formal Education) in rural villages in Tonga. Seminar paper, Canberra, Centre for Continuing Education, Australian National University, 9 March.
  • Bain, K., 1967. The Friendly Islanders. London: Hodder and Stoughton.
  • Beaglehole, J.C., 1959. The Journals of Captain Cook on His Voyages of Discovery. The Voyage of the ‘Resolution’ and ‘Adventure’, 1772-1775. Cambridge: Hakluyt Society.
  • —— 1967. The Journals of Captain Cook on His Voyages of Discovery. The Voyage of the ‘Resolution’ and ‘Discovery’, 1776-1780. 2 volumes. Cambridge: Hakluyt Society.
  • Collocott, E.E.V., 1928. Tales and Poems of Tonga. Bernice P. Bishop Museum Bulletin 46. Honolulu.
  • Erskine, J.E., 1853. Journal of a Cruise Among the Islands of the Western Pacific. London: Dawsons.
  • Gifford, E.W., 1929. Tongan Society. Bernice P. Bishop Museum Bulletin 61. Honolulu.
  • Kaeppler, A., 1978. Exchange patterns in goods and spouses: Fiji, Tonga and Samoa. Mankind, 11(3):246-52.
  • 1995. Poetics and politics of Tongan barkcloth. In D.A.M. Smidt, P. ter Keurs and
- 167
  • A. Trouwborst (eds), Pacific Material Culture. Leiden: Rijksmuseum vor Volkenkunde, pp.101-21.
  • —— 1998. Airplanes and saxophones: Post-war images in the visual and performing arts. In D. Scarr, N. Gunson and J. Terrell (eds), Echoes of Pacific War. Canberra: Target Oceania.
  • Koojiman, S., 1972. Tapa in Polynesia. Bernice P. Bishop Museum Bulletin 234. Honolulu.
  • Labillardiere, J., J.H., 1800. The Account of the Voyage in Search of La Perouse, Undertaken by the Order of the Constituent Assembly of France, and performed in the Years 1791, 1792, and 1793, in the 'Recherche, and ‘Esperance’, Ships of War, Under the Command of Rear-admiral Bruni d'Entrecasteaux. London: John Stockdale.
  • Lawry, W., 1851. Diary. MS. Mitchell Library, A1973 (Microfilm CY Pos Reel 123).
  • Leslie, H.Y., 1996. ‘… Like a mat being woven’. Paper presented at Weaving the World: Cloth in Oceania, Association of Social Anthropologists of Oceania Annual Meeting, Kailua-Kona, Hawai'i.
  • Martin, J., 1827.An Account of the Tonga Islands, in the South Pacific Ocean. With an Original Grammar and Vocabulary of Their Language. Compiled and Arranged from the Extensive Communications of Mr. William Mariner, Several Years Resident in Those Islands. 2 volumes. Edinburgh: Constable and Co.
  • Nayacakalou, R.R., 1959. Land tenure and social organisation in Fiji, Tonga and Samoa. Journal of the Polynesian Society, 68(2):93-114.
  • Novo y Colson, P., 1885. Alrededor del Mundo por las Corbetas ‘Descubiertay’ y ‘Atrevida’ al Mando de los Capitanes de Navio D. Alejandro Malaspina y Don Jose Bustamante y Guerra desde 1789 a 1794. Madrid.
  • Orange, Rev. J., 1840. Life of the Late George Vason of Nottingham. Derby: Henry Mozley and Sons.
  • Orlebar, J., 1976.A Midshipman's Journal, Kept on Board the H. M. S. Seringapatam, During the Year, 1830. San Diego: Tofua Press.
  • Small, C., 1987. Women's Organisations and their Pursuit of Wealth in Tonga: A Study in Social Change. Unpublished Ph.D. thesis, Temple University, Philadelphia.
  • Tamahori, M. J., 1963. Cultural Change in Tongan Bark-Cloth Manufacture. Unpublished M.A. thesis, Department of Anthropology, University of Auckland
  • Teilhet-Fisk, J., 1991. To beat or not to beat, that is the question: A study on acculturation and change in an art-making process and its relation to gender structures. Pacific Studies, 14:41-68.
  • Weiner, A.B., 1989. Why cloth? Wealth, gender and power in Oceania. In A.B. Weiner and J. Schneider (eds), Cloth and Human Experience. Washington and London: Smithsonian Institution Press, pp. 33-72.
  • Wilson, J., 1799. A Missionary Voyage to the South Pacific Ocean, Performed in the Years 1796, 1797, 1798, in the Ship ‘Duff’, Commanded by Captain James Wilson. London: T. Chapman.