Volume 101 1992 > Volume 101, No. 2 > Why were pots imported to Ngulu Atoll: a consideration of subsistence strategy, by Michiko Intoh, p 159-168
                                                                                             Previous | Next   

- 159

Ngulu is an atoll situated between Yap and Belau in the Western Caroline Islands, 110 km south-south-west of Yap and 300 km east of Kayangel Atoll of Belau (Fig. 1). The only inhabited islet is about 500 m long, 300 m wide and 2 - 3 m high. This islet is too small to support enough fresh water in its Ghyben-Herzberg lens to be useful for human consumption. Wetland taro, yams and bananas do not grow on the islet. The only useful vegetable foods which will grow are coconut, a few seeded breadfruit trees, and a form of dryland taro, Alocasia.

Despite the poor plant resources available, Ngulu has been inhabited for the last 1800 years (Intoh 1981) and, judging from pottery sherds which occur throughout the cultural sequence, the people who lived on the islet maintained cultural contact with the inhabitants of high islands where suitable clay was available for pottery manufacture. The use of pottery continued into the historic era, unlike other Central Caroline islands, where earth-oven cooking replaced cooking in earthenware pots.

This paper examines why pots were imported to Ngulu. A functional interpretation is advanced for the pottery arising from a consideration of the subsistence strategy on Ngulu atoll. 1


The people of Ngulu are known to have kept close cultural contact with the inhabitants of Gror village in southern Yap during historic times (Lessa 1950:29). This contact was part of an exchange system known as sawei and permitted the atoll islanders to obtain various natural resources which were not available on their atoll. For example, the Ngulu people obtained from Yap a number of important plant foods such as banana, taro (Cyrtosperma), sugar-cane, Polynesian chestnut and yams. In addition, nonfood resources were obtained, such as turmeric, red soil, bamboo and, above all, pottery. In exchange, the Ngulu people sent to Yap handicrafts, coconut oil and coconut syrup, these things being produced on Ngulu itself.

The origin of the sawei exchange system is not known, and the details of the earlier participation of Ngulu in it are equally obscure. However, it is suggested that some kind of cultural contact with high islands probably existed since the time of initial settlement of the atoll, because the natural resources must have always been extremely

- 160
Figure 1: Map of Ngulu Atoll in Micronesia.
- 161

limited. It is notable that most of the materials transported to Ngulu through the sawei system were items which were necessary to maintain the Ngulu way of life — foods essential for continued existence, turmeric for ceremonial use, red clay for painting canoes, and bamboo for various domestic and constructional uses. It is suggested that items which were not essential on Ngulu, but which were present on Yap, were probably not transported. For example, the Ngulu people do not appear to have traded goods for Yapese stone money. The suggested reasons are that its large size made transport difficult but, more particularly, it had no functional use on the atoll.

It must be remembered that space in canoes is strictly limited, and that people returning to Ngulu with traded material on board are likely to have given preference to food and utilitarian items, rather than nonessentials. The question arises: “Was pottery essential to life on Ngulu?” Judging from the continuous importation of pottery throughout the habitation history, the answer appears to be “yes”. It is interesting to note that potsherds excavated from archaeological sites on Ngulu have been traced to both Yap and Belau (Dickinson 1982). The fact that the Ngulu people approached both high islands nearby for pottery adds weight to the idea that it was functionally important on the atoll.


What then was the function of pottery imported to Ngulu? There were not many plant foods available on the atoll which required cooking. Moreover, it is not necessary to have pottery vessels in order to cook food. It is notable that cooking in pottery was replaced by earth-oven cooking in a number of places in Oceania, particularly in environments where pots were not easily manufactured or obtained by trade (Irwin 1981; Leach 1982). In fact, using the earth oven was a major cooking technique in the Central Caroline Islands where pottery was once used (Le Bar 1964; Shutler et al. 1984; King and Parker 1984). The use of pottery on Ngulu throughout the known period of human habitation may indicate that the pots which were imported had a function which could not be satisfied by using the earth oven. 2 This is further indicated by the fact that the earth oven was present on Ngulu and, in particular, was used for cooking turtles.

Ethnographic records of the use of pottery might help us to understand the apparent necessity for pots on Ngulu in earlier times. In 1980, two Yapese pots were still kept on Ngulu (Fig. 2). These had been obtained from Yap through the sawei exchange network and were kept inside a house as precious goods. Most households possessed more than two pots of similar shape, generally one for making coconut syrup and one for cooking food.

An outline of coconut syrup making is as follows:

  • 1) A coconut inflorescence is shaped into a long cylinder by binding it with plaited coconut fibres.
  • 2) The end is cut off twice or three times a day to allow the sap to drip from the cut surface. In former times a shell scraper was used for this purpose, made from a shell of Cassis sp. (Eilers 1936:226). An example is illustrated in Figure 3.
- 162
Figure 2: Historic period Yapese pottery.
Figure 3: Cassis sp. shell scraper used for cutting coconut inflorescence to assist extraction of sweet syrup.
- 163
  • 3) The drops of sweet sap are collected in a coconut shell container suspended from the extremity of the inflorescence. Barrau notes that the yield is approximately 1.2 litres per day (Barrau 1961:38).
  • 4) The sweet sap which is gathered is either drunk fresh or poured into a pot and then concentrated over a low flame until it forms a thick syrup.

The final simmering process requires some kind of vessel which can be placed over a flame and, in the prehistoric period on Ngulu, an earthenware pot is the obvious candidate for this. The similar simple vessel form which prevailed in both the prehistoric and ethnographic periods suggests similar functions. The slow cooking produces a thick brown syrup which is like molasses. The presence of pottery vessels would have permitted this syrup to have been made, because it would be well nigh impossible in an earth oven.

There is an important advantage which concentrated syrup has over fresh coconut sap. The fresh sap starts fermenting after 12 - 24 hours in the tropical environment. According to Barrau, fresh sap contains 15-16% sugar and, when fermented, contains about 6% alcohol after 15 hours (Barrau 1961:39). However, in the form of concentrated syrup, the sweet liquid can be stored much longer without fermentation setting in. The reason for this requires some explanation.

Successful preservation of food is greatly facilitated by lowering its water content. Riddervold, in discussing water activity and conditions suitable for microbes, comments that water activity is diminished by “adding salt and/or sugar which binds the water molecules and has the same conserving effect as drying” (Riddervold 1988:213). Muller and Tobin (1980:264-5) provide useful quantitative information on the importance of moisture on inhibiting the activity of micro- organisms in foods.

Food Type Moisture Content Water Activity
Fruit 80-90% 0.97
Eggs 75% 0.97
Meat 70% 0.97
Cheese 40-50% 0.96
Jams 30% 0.94-0.82
Honey 15% 0.75
Sugar 0.5% 0.10

They note that there are minimum threshold water activity values below which various micro-organisms cannot survive. Some of these are: salmonella bacteria 0.92, staphylococcus bacteria 0.83, saccharomyces yeasts 0.90, zygosaccharomyces yeasts 0.62, penicillium moulds 0.85, and xeromyces mould 0.60. These figures clearly show why sugar- rich substances preserve so well. Another factor is identified by McGee, who notes that “Because of their generally low water content and because high external sugar concentrations draw water out of microbes by osmosis and dehydrate them, candies are seldom spoiled by the growth of bacteria or molds” (McGee 1984:422).

- 164

It would, thus, appear that there are two factors at work which inhibit fermentation and decay when high sugar content is involved — dryness and osmotic pressure. Coconut syrup, therefore, possesses a considerable advantage over coconut sap: it can be kept for long periods without degeneration of quality, and consumed whenever desired.

Coconut syrup is an important source of energy (Child 1964:189), and is still used in many ways by Ngulu islanders.

  • 1) Fresh sap or diluted syrup is given to children. It is said that a baby who lost his/her mother could still survive if coconut sap were available.
  • 2) Concentrated syrup was used as a gift in the sawei exchange system, and was clearly valued by the Yapese people.
  • 3) Coconut syrup was used in producing coconut candy. This was a high-energy food used during ocean voyages. It was made by cooking the syrup with grated coconut.
  • 4) Coconut syrup was also used in preparing various cooked dishes on Ngulu.

Finally, there was another unique use of coconut syrup on Ngulu; this will be described in relation to the atoll's subsistence system.


The diet on Ngulu consists mainly of coconut and fish. This is supplemented by turtle, breadfruit, Alocasia and tree-fruits such as Crataeva speciosa (Muller 1917:70). Plants of banana and Cyrtosperma are brought back from Yap and planted in a dugout pit near the central part of the islet. These plants are kept until consumed, but do not actually grow in the pit because of the brackish ground water. Few breadfruit trees grow on the atoll. A small amount of fermented breadfruit was observed in a small storage pit in 1980.

Alocasia is a dry-land taro distributed widely in Oceania (Bown 1988:199-200, 229). In many islands, Alocasia is not consumed regularly and is categorised as a famine food (Thaman 1984:115; Parkinson 1984: 216; Sproat 1968:31) because Alocasia is very acrid and is considered to be toxic. Calcium oxalate crystals are present in the plant and, if eaten without suitable preparation, will cause a burning sensation, salivation and swelling in the mouth (Tang and Sakai 1983; Bradbury and Holloway 1988:110-18). The calcium oxalate crystals in such plants can, however, be broken down by sustained cooking. For example, the acridity of Colocasia esculenta can be destroyed by cooking or fermentation (Tang and Sakai 1983:150). However, the concentration of calcium oxalate is far greater in Alocasia and very long cooking is needed to remove the active ingredient. In Malaysia, thorough boiling is considered essential and betel-nut chips or small quantities of slaked lime are sometimes added in order to assist the breakdown of calcium oxalate (Ghani 1982:32); the chemistry of this process is unknown. Alocasia is normally eaten only in cases of emergency. There are a number of important exceptions to the extreme toxicity of Alocasia, for example, - 165 in Tonga and Samoa. In these islands, Alocasia is an important food crop. The tubers of Alocasia are almost free of oxalate crystals, and can be cooked by the normal methods used for other types of taro (Barrau 1961:42). Whether this dramatic change in plant biochemistry has come about by selective breeding in these islands is unknown.

Alocasia is the only starchy food harvested all year round on Ngulu. The Alocasia plants growing there have been described as extremely acrid compared with those growing on nearby Yap, which could possibly be due to the more severe weather which prevails on Ngulu. Despite the high concentrations of calcium oxalate, the plant is still eaten, but only after very careful cooking and, on other islands, toxicity is reduced or removed by sustained or repeated boiling. However, in the case of Ngulu, residents claim that such treatment is insufficient for complete removal of toxicity. Their own method of cooking is unique, and involves repeated cooking with coconut syrup.

The mechanism of how the addition of coconut syrup might work to destroy the toxic effects of calcium oxalate is not at all clear. It is unlikely that the syrup actually reacts chemically to remove the toxic crystals. An additional factor might be that the high concentration of sugar in the syrup serves to ameliorate the irritation in the mouth which would otherwise be caused by the crystals. Another possibility concerns the fact that the temperature of tubers cooked in such syrup would be considerably higher than in plain water. Calcium oxalate loses all its water of crystalisation at 200° C, but requires a much higher temperature to break down into CaCo3 or CaO. McGee (1984:411) provides information showing how boiling point increases with sugar content. An extract is given below.

Sugar wt % °C
0 100
85 113
90 131
100 160

Pure sucrose liquefies at 100° C and begins to break down (caramelise) at 168° C. It is reasonable to infer that detoxification of Alocasia would be a much faster process, when cooked in syrup, but this needs to be confirmed by experimental research.

An interesting parallel to this treatment of the plant in Ngulu has been reported in the Cook Islands. Here, Alocasia was cooked with another plant also rich in sugars, Cordyline terminalis. The tubers of Alocasia were placed in the bottom of a large earth oven lined with hot stones and then covered with a layer of the tap roots of Cordyline terminalis. The sweet juices from the Cordyline ran down on to the taro during cooking (Barrau 1961:42). This seems to be a similar example to Ngulu, where sugar is used to reduce the acrid effects of ingesting calcium oxalate.

Coconut syrup making was also known in other outer islands of Yap, such as Ulithi and Fais; in these places, imported pots from Yap were also used to concentrate the sweet sap, and the syrup was used in cooking as well as for exchanging goods between - 166 islands. It is not known, however, whether the syrup was specifically used for cooking Alocasia on these islands. In the ethnographic period, coconut syrup made on Ulithi was exchanged for tobacco produced on Fais (personal field notes, 1991).


It is suggested that, on Ngulu atoll, the main function of pottery was for simmering coconut sap, and, associated with this, the cooking of extremely toxic Alocasia taro by a specialised process designed to reduce toxicity.

If Ngulu had been rich in plant resources, earth-oven cooking could have been the main cooking method, and pottery would not have been necessary for survival. However, the severe natural environment of this atoll forced the Ngulu people to rely on a highly toxic variety of Alocasia which could be grown locally. The addition of coconut syrup during sustained cooking made it possible for Ngulu residents to eat this carbohydrate-rich plant without adverse effects. In addition, surplus coconut syrup was produced in order to trade with nearby high islands to obtain other food resources and natural materials important for life on the atoll. Among the imported items, of greatest importance was the continued supply of earthenware vessels for producing coconut syrup in the first place.

Coconut syrup making is widely distributed in Ceylon (Browning and Symons 1916), Indonesia (Fox 1977), and both west and east Micronesia (Barrau 1961;Catala 1957:44-6; Luomala 1953:85-6). It was also present in Belau at the time of first European contact. As far as Ngulu is concerned, it is not easy to determine how far back in time this custom prevailed. The distinctive shell scraper made from Cassis sp. shell, which was used in ethnographic times to cut the coconut inflorescence for gathering the coconut sap, has been found in archaeological sites dating back as far as 1000 B.P. on Ngulu. It is possible that, before this period, a bamboo knife was used, and such a knife is still used in Indonesia. In this present context, however, the antiquity of the custom on Ngulu is not important. The main purpose of this paper is to suggest that there is a functional relationship between the existence of pottery on Ngulu and the production of coconut syrup on the island, and that this relationship has been formed because of the harsh natural environment of the atoll.

- 167
  • Barrau, J., 1961. Subsistence Agriculture in Polynesia and Micronesia. Bulletin 223. Honolulu: B. P. Bishop Museum.
  • Bown, D., 1988. Aroids: Plants of the Arum Family. London: Century.
  • Bradbury, J. H. and W. D. Holloway, 1988. Chemistry of Tropical Root Crops: Significance for Nutrition and Agriculture in the Pacific. Canberra: Australian Centre, for International Agricultural Research.
  • Browning, K. C. and C. T. Symons, 1916. Coconut Toddy in Ceylon. Journal of the Society of Chemical Industry, 35:1138-42.
  • Catala, R. L., 1957. Report on the Gilbert Islands: Some Aspects of Human Ecology. Atoll Research Bulletin 59. Washington, D.C.
  • Child, R., 1964. Coconuts. Tropical Agriculture Series. London: Longmans.
  • Dickinson, W. R., 1982. Temper Sands from Prehistoric Sherds Excavated at the Pemrang Site on Yap and from nearby Ngulu Atoll. Indo-Pacific Prehistory Association Bulletin, 3:115-128.
  • Eilers, A., 1936. Westkarolinen: Tobi und Ngulu. Ergebnisse der Südsee Expedition. 1908-1910. Hamburg: Friderichsen, de Gruyter, II, B.9, ii.
  • Fox, J. J., 1977. Harvest of the Palm. Cambridge: Harvard University Press.
  • Ghani, F. D., 1982. The Potential Aroids of Malaysia. IFS Provisional Report No. 11. Regional Meeting on Edible Aroids, Suva, Fiji. Stockholm: Inernational Foundation for Science, pp.29-39.
  • Intoh, M., 1981, Reconnaissance Archaeological Research on Ngulu Atoll in the Western Caroline Islands. Asian Perspectives, 24: 69-80.
  • Irwin, G., 1981. How Lapita Lost its Pots: the Question of Continuity in the Colonisation of Polynesia. Journal of the Polynesian Society, 90:481-94.
  • King, T. F. and P. L. Parker, 1984. Archaeology in the Tonaachaw Historic District, Moen Island. Micronesian Archaeological Survey Report No. 18. Center for Archaeological Investigations, Occasional Paper No. 3. Carbondale: Southern Illinois University.
  • Leach, H. M., 1982. Cooking without Pots: Aspects of Prehistoric and Traditional Polynesian Cooking. New Zealand Journal of Archaeology, 4:149-56.
  • Le Bar, F. M., 1964. Material Culture of Truk. Yale University Publications in Anthropology No. 68. New Haven: Yale University Press.
  • Lessa, W., 1950. Ulithi and the Outer Native World. American Anthropologist, 52:27-52.
  • Luomala, K., 1953. Ethnobotany of the Gilbert Islands. Bulletin 213. Honolulu: Bernice P. Bishop Museum,
  • McGee, H., 1984. On Food and Cooking: The Science and Lore of the Kitchen. London: Unwin Hyman.
  • Muller, H. G. and G. Tobin, 1980. Nutrition and Food Processing. London: Croom Helm.
  • Müller, W., 1917. Yap. Ergebnisse der Südsee Expedition. 1908-1910. Hamburg: Friederichsen, de Gruyter, II.B.ii.
  • Parkinson, S., 1984. The Contribution of Aroids in the Nutrition of People in the South Pacific, in S. Chandra, (ed.), Edible Aroids. Oxford: Clarendon Press, pp.215-24.
- 168
  • Riddervold, A., 1988. On the Documentation of Food Conservation, in A. Riddervold and A. Ropeid (eds), Food Conservation. London: Prospect Books, pp.210-18.
  • Shutler, R. Jr., Y.H. Sinoto, and J. Takayama, 1984. Preliminary Excavations of Fefan Island Sites, Truk Islands, in Y. H. Sinoto (ed.), Caroline Islands Archaeology: Investigations on Fefan, Faraulep, Woleai and Lamotrek. Pacific Anthropological Records No.35. Honolulu: Bernice P.Bishop Museum Press, pp.1-64.
  • Sproat, M. N., 1968. A Guide to Subsistence Agriculture in Micronesia. Agricultural Extension Bulletin 9. Saipan: Division of Agriculture, Department of Resources and Development, Trust Territory of the Pacific Islands.
  • Tang, C. and W. S. Sakai, 1983. Acridity of Taro, in J. Wang, (ed.), Taro — A Review of Colocasia Esculenta and its Potentials. Honolulu: University of Hawaii Press, pp.148-63.
  • Thaman, R. R., 1984. Intensification of Edible Aroid Cultivation in the Pacific Islands, in S. Chandra (ed.), Edible Aroids. Oxford: Clarendon Press, pp.102-22.
1   An earlier version of this paper was presented at a seminar, organised by Tomoya Akimichi, National Museum of Ethnology, Japan, and at the XVII Pacific Science Congress in Honolulu (in a session organised by Barry Rolett and Yosihiko Sinoto). I wish to express my thanks to Shuji Yoshida, Peter Mathews, Janet Davidson, Nancy Pollock and Katharine Luomala for their valuable comments made at these meetings.
2   It has been pointed out to me that coconut syrup could have been made on Ngulu using a stone boiling technique, and that this has been recorded for Kiribati (Nancy Pollock, personal communication, 1991). To my knowledge, this technique has never been used on Ngulu atoll. In any event, cooking the sap in pots would be a far more efficient process than using heated stones in wooden bowls.