- 364

I wish to acknowledge the assistance received from Prof. A. R. Radcliffe-Brown in enabling me to criticise this ceremony scientifically. For many of the suggestions in this monograph I wish also to acknowledge my indebtedness to Dr. M. N. Srinivas of Exeter College, Oxford. I use the word “ceremony” in this monograph in the loose sense of a body of beliefs and rites centreing around the drinking of kava, not because it has any easily defined scientific meaning but because Collocott used it in the article from which much of the description is taken. 1 The kava ceremony takes place on almost any occasion but the drinking of it was formerly confined almost entirely to chiefs and to chiefs' attendants but is now widespread as a social act among all sections of the population. However, in former times, and to a certain extent now, there seems to be more to the drinking of kava than a purely social act and it is this relation between the material facts of the drinking of the kava root and the more intangible sentiments associated with it that I wish to illustrate in this monograph.

In this monograph I have adopted the following scheme. Commencing with a description of Tonga and Tongan social

- i

- ii Page is blank

- 365

structure insofar as it is relevant to this thesis, I have attempted to show how the material tools used in this ceremony are determined by and related to certain other aspects of Tongan social structure and are incapable of being comprehended by themselves. I have then tried to show that it is more probable that the root was introduced via Micronesia and the eastern islands rather than via the Solomons and Fiji to Tonga. If this were so (and it can never be more than an hypothesis) one of the reasons for its being carried by the early voyagers was because it was the centre of a ritual complex even at that early stage in Polynesian history. Evidence for this thesis seems to be forthcoming from Hornell's examination of the canoes of Oceania. Since the most essential of the reasons for the kava ceremony being brought to Tonga depends on its ritual and symbolic character, I have attempted to show how in a modern formal kava, ceremony, that of the Tu'i Kano-kupolu, such a character does exist. The ritual value of the ceremony is shown by certain empirical tests of the degree of sacredness. In addition the various parts of the ceremony have a certain symbolic meaning, which can be determined partly by the obvious explanation given to the ceremony by the participants, but also by certain other ritual techniques which even the participants do not fully comprehend, but which the anthropologist can discover. The place of the kava ceremony in the most formal body of rites of the Tu'i Kanokupolu is reproduced in kava ceremonies with other people as president but where, because the ceremony is less formal and less sacred, many of the effects seen in the ceremony of the Tu'i Kanokupolu are less easily analysed. However, it can be. seen that in the contrast between the kava ceremony of the Tu'i Tonga (the religious head of Tonga) and the Tu'i Kanokupolu (the secular head) certain rites are divectly opposed to one another and that these are symbols of the greater sacredness of the one chief over the other.

I then go on to give various explanations of why this ceremony should have been held at war and at marriage but show that these other explanations are really secondary and not primary explanations because they only explain the results of what actually does happen but cannot explain why the kava ceremony takes the specific form that it does.

- 366

Since this paper has gone to press I have read Melanges d'histoire des Religions by Ms. H. Hubert and M. Mauss (Paris, Felix Alcan, 1909), part I, Essai sur la nature et la fonction du sacrilége, in which many of the theories I have demonstrated here have been worked out as they applied to Vedic and Old Testament ritual. I had no previous knowledge of this book.

The islands of Tonga lie between 15° and 23° 30″ S. and 173° and 177° W., about 400 miles to the south-west of the Samoan group and about 200 miles from the nearest islands of the Fijian group. It has three main groups of islands, Vava'u, Ha'apai and Tongatapu and numerous smaller ones ranging from active volcanoes like Tofua to disappearing islets like Falcon island. As a rule the islets are low flat coral just above the surface of the sea but there are some mountainous groups like the island of E'ua with a maximum height of 640 feet. Although the variation in minor details of custom is quite considerable over such a large territory, all are united by the fact that they owe a common suzerainty to the sovereign at the capital, Nuku'alofa at Tongatapu. At the same time Tongans have been intrepid voyagers and there are definite traditions and evidence that canoes in historic times have visited Tikopia, Samoa, Ellice islands, Fiji and elsewhere. It is to be expected, therefore, that Tongan custom has been considerably affected by contact with other islands and that there is a possibility that the original settlers of Tonga might have remained for some considerable time in a neighbouring group.

Like so much of the material of anthropology, with the spread of European ideas and influence, native customs, which are so much better understood now than ever before, are rapidly dying out or their former variations and significance forgotten. To this tendency, Tonga is no exception. As an attempt to comprehend Tongan customs and ideas within the last hundred years, I hope in this monograph to examine one institution, the kava ceremony, which is widespread in the Polynesian area, and show by reference to its probable origin, diffusion, and contribution to Tongan social structure, in what way the study of such institutions may be best approached by the social anthropologist. In order to do - 367 this it will be necessary to give a brief description of certain ranks and institutions in Tongan society derived from published sources. But it is here that one of the main difficulties arises because many significant facts have not been recorded and all that one can do is to hope that the hypothesis here put forward can be verified from other areas by those with the facts at their disposal and the theoretical background necessary. I have accordingly made use of three main authorities 2 with subsidiary facts from other authors. After preparing the main body of this monograph, I found that Mariner's description of the ceremony has been ably summarized by R. Piddington 3 and, in order not to duplicate accounts, have used Collocott's description of the ceremony.

Tongan society was, at the time of Cook and Mariner, a feudal society, in many ways similar to the English Middle Ages. At the top of the hierarchical structure were two principal chiefs, the Tu'i Tonga and the Tu'i Kanokupolu. The former was concerned principally with religious duties and to him an annual first-fruits ceremony was dedicated (inachi). The latter was concerned with administration. However, at the time of Mariner, the religious head was rapidly losing power and authority and during the nineteenth century he lost his right to the first-fruits. Today his line is included with that of the Tu'i Kanokupolu. The Tu'i Kanokupolu kept his position by periodically waging war against those lords which rebelled against his authority but during the nineteenth century, a brilliant general, Taufaahau, succeeded by a series of campaigns in uniting all of Tonga under his authority. Each chief held his position from those higher in rank than himself but this right was very much strengthened by his own rank and traditional authority. Land was, strictly speaking, the property of the Tu'i Tonga and later of the Tu'i Kanokupolu only, but because it was of no use unless worked, all land had a - 368 number of serfs or commoners on it, who were given the right to work it in exchange for certain feudal duties in the form of produce, marks of sovereignty, religious obligations and so on to their superiors. However, unlike the English feudal system, serfs were not tied to the land on which they were born but could move from one chief to another. However, if they returned to their former chief they were likely to be killed. Commoners were not thought of as going to an after-world as chiefs were. Although the rights of chiefs were theoretically infinite over their serfs, in actual fact the right of revolt by the subjects often proved a limiting factor to a tyrant's cruelties, although revolt against a superior was not only a political but also a religious matter.

Rank in Tonga was relative rather than absolute. The key to the understanding of it lies in the Tongan family, which has many peculiarities not found in other Polynesian societies, except Samoa. Within every family of husband, wife and children, there existed a system of ranking in which every member was of slightly different status from everyone else. Within every generation of the same parents, the elder brother is chief (eiki) to the younger. The eldest sister has a higher status than the eldest brother. This is a ritual inequality. At the same time the authoritarian control of the family lies in the hands of the father and every woman in Tonga is under the control of some male. This is a political inequality. At the same time descent is traced through the paternal line and residence is patrilocal and, as far as can be ascertained, has always been so. Individuals who are brother and sister (in the descriptive and to a lesser extent in the classifacatory sense) avoid each other after about the age of six until old age and do not live in the same house together. It follows from this that brothers and sisters are usually brought up by other than their true parents after the age of six if they have any brothers or sisters of the opposite sex. These family relationships are reflected in the structure of Tongan society in which the most politically important lineages are those who can trace their descent through direct patriliny from the first Tu'i Tonga through the eldest son and round whom are grouped inferior relatives tracing descent through younger sons. On the other hand the most ritually important members of the family are always a man's sister and those descended from - 369 his sister. The contrast between the ritual and political status of the members of a family is realised in the relationship known as the fahu relationship, that is to say the relationship between the mother's brother and the sister's children (and their children to a lesser extent). The sister's children can appropriate the goods of the mother's brother without permission and part of the expenses of their education is paid by him in modern times, and in ancient times. Prof. A. R. Radcliffe-Brown has shown how the fahu ate the sacred food presented by the uncle to the gods. 4 There seems reason to believe, however, that, while in modern times, the main emphasis is an economic one, in ancient times it was mainly a ritual one. It would be difficult to conceive that, if in a family the political and ritual power was thus split up between the male and female members of the family, that it could continue to exist but for this right of fahu. Prof. Radcliffe-Brown has demonstrated that where the Tu'i Tonga delegated his political power to the Tu'i Kanokupolu, he at the same time strengthened his own position by marrying the daughter of the Tu'i Kanokupolu so that the children should have the right of fahu over their own mother's brother (that is to say the Tu'i Kanokupolu). 5 It appears that similar dynastic marriages occurred among many other members of Tongan society, where they had sufficient rank to make it worth while. 6

Consequently no person is theoretically ever a real commoner but is related to a chiefly line either through a distant of less distant series of younger son relationships.

- 370

There were in ancient times various terms to indicate rank which I will not mention here. The only words necesary to mention here are eiki (chief), meaning a relatively higher ritual relationship than the person speaking though often used of political superiority also, tu'i (king) meaning a position usually derived from control over a certain territory now or in the past and given to high ranking chiefs, and matapule. The duties of matapule are similar to those of the Samoan talking chief (tulafale) but inclusive of a wider range of duties. It is important to note that a very great number of them are traditionally descended from foreigners (usually Fijian or Samoan immigrants) and hence are ritually outside the Tongan social structure. They are at the same time a mark of the importance of the chief to whom they are attached. Mariner states that the sons of the most distinguished matapule belonged to the commoner class. 7 The authority they wield is derived from the chief they represent, and hence should they eat the sacred food, it is as though the chief ate it himself.

Regarding the history of Tonga relevant to this monograph, it is believed that in the 11th century there was only one spiritual and secular head of Tonga known as the Tu'i Tonga. Legend places the first kava ceremony to be held in the reign of the tenth Tu'i Tonga, Momo, about 1150. About 1770, after the assasination of his father, the 23rd Tu'i Tonga bestowed the title of Tu'i Ha'a Takalaua on his younger son, who assumed the secular reins of office, leaving only the ritual duties to the elder son. Similarly about 1610, the Tu'i Kanokupolu title was bestowed on a younger son of the Takalaua for his victory in a war. His lineage is known as the Ha'a Ngata. At the present day the only remaining principle lineage is that of the Tu'i Kanokupolu, who also holds the title of Tu'i Tonga. The Tu'i Ha'a Taka-laua's title died out except as a junior lineage in 1799. During the nineteenth century, Tonga was split by civil wars, accentuated by religious differences as either faction was converted by Wesleyan or Roman Catholic missionaries Finally a brilliant general. Taufaahu, the Tu'i Kanokupolu, later known as George I, united the islands again under his sovereignty by a series of campaigns. In 1900 the Tongan - 371 government signed a treaty with the British government in which they surrendered the right to control their foreign affairs for the privileges of British protected citizenship. This treaty established the legal equality of all Tongan citizens, a revolutionary step. It can also be seen as a culmination of the increasing participation by commoners in ritual occasions while the drinking of kava is more common now than formerly. This is due perhaps either to the ceremony becoming more profane and less essential to the maintenance of the Tongan social structure, or to the progressive elevation of the Tongan commoner in the social hierarchy. The Church declares commoners have souls which are eternal, whereas formerly only chiefs possessed eternal life. As a corollary commoners must clearly be chiefs in part and share in a chief's privileges, one of the most important of which was the right to drink kava.

It can be seen that if one is to make a functional study of the kava ceremony that its place in the Tongan social structure is very different from what it was 150 years ago. This difficulty is an unavoidable one and I have accordingly tried to confine my use of material to the period between 1900 and 1925 with reference to earlier ceremonies where it seems relevant. I have avoided the use of comparative material except when it seemed directly relevant to elucidate an hypothsis and accordingly conclusions cannot be applied in toto to Samoa, Fiji or other islands in Oceania.

The material required for the performance of a majority of the kava ceremonies is ceremonial food (fono), the fau fibres used in straining the kava root, water (vai) and the bowl itself. All the materials used in this ceremony have some ritual value. The meaning of the food and kava is dealt with subsequently (page 385). All that needs to be said here about the kava root is that its botanical name is piper methisticum, that it undoubtedly originated outside Oceania and that its presence in so many Oceanic islands is most probably due to its being transferred by canoe rather than by ocean currents. Its drug-like qualities are very much less than are popularly supposed.

The fau fibres are denied from the hibiscus tree. The word, fau, is apparently either a synonym for or entomologically connected with the word, hau, a ruler or conqueror. 8 - 372 It is interesting to note that in modern times for ordinary communal village drinking, coconut fibre is used for straining 9 and this shows to what extent kava drinking has become merely a social and not a ritual occasion. This use of coconut fibre seems to be a recent innovation as there is only mention of hibiscus fibre in earlier authorities. Unfortunately there is no evidence to show what the exact relationship of the conqueror to the hibiscus tree was in ancient times but there appears to be a definite connection as it is used also in Samoa on the occasion of an important visitor being present. 10

Most of the Tongan islands are completely flat and hence water is collected, either from the rain water from the rooves of thatched huts, or from puddles in the ground. Before the construction of churches with their large roofs, there were many occasions when drought occurred and water had to be carefully rationed. It is therefore interesting to note that a MS. published by Gifford 11 mentions a tradition of a country in which there was a fountain called Vaiola (vai = water, ola = prosperity, used sometimes of rain falling after a drought). Whoever plunged into it would become healed of any infirmity but no one knew where the well was situated. It is possible that the bringing of water in small containers to the kava mixer suggests the value and rarity of that which was brought. Moreover, in ancient times only the chief or a high matapule washed his hands in a special bowl of water peculiar to him alone. A commoner usually washed his hands by rubbing them with the pith of the stock of the fusi banana. 12

The kava bowl itself as the centre of the ceremony was a very sacred object and treated with extreme care. It was of varying sizes. The bowl in the Pitt-Rivers Museum, Oxford, is from Nei'afu in the Vava'u group presented by J. H. Worthington in 1922. It has a diameter of 16 inches and is 7½ inches from the ground to its upper edge. It has a single projection on the underneath of the bowl on one side - 373 with a hole bored through it, used for hanging the bowl from a wall by a cord. This hole seems also to have ritual significance perhaps emphasising the difference between the front and the back of the bowl. The bowl being almost circular would otherwise have no easily recognizable front and back. It rests on fourteen legs so that the only part of the bowl to touch the ground is the bottom of the legs. In modern times it seems as though the bowls are interchangeable with those in Fiji and many modern kava bowls have been made in Fiji and exported to Tonga. (Kramer states that the bowl usually had six legs in Samoa, 13 but one Samoan bowl in the Oxford Museum collected by Vice-Admiral Gurner has only four.)

In the past there might quite conceivably have been constructional differences between bowls made in different parts of the Pacific but there seems no way of finding whether this is so today from the Fijian-Samoan-Tongan area. From the bowls in the Pitt-Rivers Museum it seems possible that the Fijian bowls were of a somewhat heavier construction with thicker rims. Of those in the Museum there are nine with four legs (of which one was Tongan, one Samoan and seven Fijian), one Fijian with ten legs and one Tongan with fourteen legs. It thus appears that no generalisation can be made to cover all the bowls. Most of the photographs in Kramer show the bowls with four legs. From a personal opinion by a Fijian chief in Oxford, I am informed that the majority of Fijian bowls had six legs. In Tonga in the ceremony to be described subsequently there are three legs on the bowl but most of the bowls I saw myself in Tonga had four or six legs. Undoubtedly a considerable measure of exchange takes place and has taken place between these islands.

F. W. Christian 14 points out that it is far easier to make a kava bowl with a slightly flattened bottom from a purely technological point of view than to affix legs to it. But if the bowl contains kava, valuable not only for its possible drug-like qualities but also for its semi-divine nature, it would seem only natural that the container should not touch the ground. The most obvious way to meet this difficulty seems to be by means of legs. Perhaps the fear that the - 374 drops of kava may touch the ground when the fau fibre is being wrung is also the explanation why there is one Fijian kava bowl in the Pitt-Rivers Museum which is cigar-shaped rather than round. The reason for the number of legs seems to be a more difficult problem but it may perhaps be connected with the fact that some numbers are more sacred than others in Tonga. Unfortunately evidence on which such a statement could be based is lacking in any detail not only because there is practically no mention of any such sacred numbers in writers other than Gifford, but also because one would logically expect a specially sacred number to be used only on the most formal ceremonies. But the collectors of kava bowls mention only the area from which the bowls come in the most general way but not the ownership of the bowl nor the nature of the ceremony in which they are used. However, my Fijian informant is emphatic that in Fiji a bowl which would be used in the granting of a special title or in an inaugural ceremony would never be used in a less important ceremony but would be specially stored in a special part of the chief's house, where only certain people could touch it. This is most probably true also of Tonga. Gifford 15 states that upon entering a certain cave a number of shadows of the visitor are seen. If only one or three shadows appear, it is a sign that death is near the one shadowed; if two or four shadows appear good health is assured. The six legs on a bowl are arranged in two sets of three in all the ones examined. He points out that ten also may be a lucky number and that after ten sharks have been caught, the fishermen may dispense with further ritual precautions. In view of the paucity of the material on the subject, however, no generalisation can be made.

In order adequately to understand the importance of the kava ceremony in Tonga it will be necessary not only to describe the place of the ceremony in Tongan social structure but also to attempt to show its origin if possible and also its possible connection with other islands in Oceania.

The botanical origin of the plant appears to have been in the New Guinea-Indonesia area or further north and accordingly it probably arrived in Tonga (a) by being trans- - 375 ported by man in canoe; (b) by Tonga's being attached to a main land mass on which kava was developed or (c) by drifting or otherwise being transported from island to island. Tonga being a comparatively recent geological land mass in its present area is precluded from developing this variety itself. Most of the islands consist of coral lime-stones or volcanic cones. The possibility of the kava root drifting to Tonga seems slight owing to the distance. It seems doubtful as to whether it could exist in a bird's crop for a sufficient while to sprout on many of the more distant islets. Recent research seems to show that even the coconut can drift in the sea for only a short while before its propagating powers are lost. 16 Of the Caroline islands it is said: “Though coconuts are of frequent occurrence in the drift, it is interesting to note that on the eastern or weather side of the island where they are washed up, there is not a single coconut grove near the water's edge, while on the western or lee side where groves have been planted they grow so near the sea that their roots are often bared by the waves.” 17

Accordingly without excluding the latter two possibilities I will asume that the kava. root was brought by canoe by early Polynesian invaders in the same way as we know they brought other plants.

The kava plant occurs in nearly all the Pacific except the central part of the Solomons and New Zealand. Its existence in Easter island is uncertain and its previous existence in the Gilbert 18 and Marshall islands 19 is uncertain or unknown. On the accompanying map is shown the places where the kava. ceremony is definitely known to exist but it also occurs in other islands which are not specifically - 376 marked but which are not important for the purposes of this argument. It occurs in Tonga, Samoa, Tokelau islands, 20 Ellice islands, 21 Cook islands, 22 Society islands, 23 Marquesas, 24 Caroline islands, 25 Manus, 26 New Guinea, 27 and in Melanesia at Fiji, Santa Cruz islands, 28 New Hebrides, 29 and elsewhere south of these islands. It also occurs in some of the “Polynesian” outliers though information here is scanty. Tikopia 30 is the only well documented account although there is no mention of kava in Ian Hogbin's account of Ontong Java, from which its absence there may fairly be deduced. 31 Its absence also seems probable therefore in Sikiana, the Mort-locks and Nuguria as these groups were all reputed to be settled from the same spot and there seems reason to believe this account. 32(The original island was known as Ngiua.) - 377 Accordingly the main Central Pacific area where kava is not found is in the Solomons islands area between Manus and Duff island. Fox, referring to San Cristoval in the Solomons states: “Beyond the mere fact that the drinking (of kava) once took place, which has been confirmed by a number of independent witnesses, I can learn nothing of the ceremony. 33

The obvious route for the kava plant to reach Tonga from New Guinea is through the Solomon islands chain whether it was transferred by human or non-human means. Its absence seems to be additional evidence that if it came through the Solomons chain it must have been by man as many other plants are found equally throughout the whole group. But if we accept the hypothesis that it was associated with an invading people passing through the Solomons, we have to explain how it is the centre of an important ritual in Fiji, of no importance in Manus and yet has been so completely forgotten in between that even the plant seems to be non-existent. Rivers 34 states that kava drinking probably arose through the needs and conditions of the immigrants who found the kava root more satisfactory as a drug than the betel nut but this suggestion is difficult to accept, (a) because both the betel nut and the kava root grow in the same area in New Guinea and Samoa and yet in the first the root is of no importance whereas in Samoa the betel nut is of no importance, (b) because the kava root is not in fact more potent as a drug than the betel nut (page 381).

In Polynesia however, the kava plant is always associated with a value out of proportion to its drug-like effect (which is almost negligible after the root has been dried). In the later part of the monograph I will attempt to show why this is so but suggest here that if the root had a necessary part to play in the Polynesian society as it was then, that would be a sufficient explanation of why it was transported by the early voyagers. Its great importance in Fiji can be explained from borrowing from Tonga or Samoa and it is interesting to note that the Fijian tradition as to the origin of kava states that it sprang up originally from the grave of a Tongan leper. 35 This hypothesis - 378 would also provide an explanation of its absence from the centre of the Solomons and its differing importance in Fijian and Manus religion. I would accordingly suggest that kava was introduced into Tonga from the east rather than the west. This hypothesis would have to overcome two further difficulties, (a) to explain the variety in the ceremonies between Fiji and other southern Melanesian colonies and (b) to explain how the kava ceremony reached the central cultural area of the Polynesians.

With respect to (a) Rivers states that in the Bank's islands kava is usually drunk between pairs of men and bears no resemblance to the large ceremonial distribution of the Central Pacific. However, it nevertheless has associated with it a certain degree of sacredness, but plays a very unimportant part in social life. However, it appears to me to occupy a far more important place there than in the northern Solomons (Manus). I suggest accordingly that it seems more probable that immigrants from Fiji and Tonga in recent times brought the plant with them and that its partial adoption seems more probable in this way than through any hypothetical early invasion. It is obvious that at first at least it would not be accepted in toto by a different people and that the public recognition of rank, which takes place at the kava ceremony in Polynesia and Fiji, is already adequately catered for in Bank's island by the system of secret societies. (These secret societies are absent in Fiji.) However, that suggestion would have to be more carefully examined by a field-worker in the area.

With respect to (b), one can almost say that the most distinguishing feature of the Polynesian social structure is its kava ceremony. Tikopia seems almost certainly to have acquired the ceremony from Tonga, from where there are canoe connections in recent times. Its absence in Ontong Java is useful subsidiary evidence in my opinion to show that the population probably originally came from Micronesia (where kava is unevenly distributed and plays a varying part in social life) rather than Polynesia. If we reject the possibility that the kava root was brought through Melanesia from whence it must have been distributed as far away as Hawai'i we have three other alternative routes: (a) via the Carolines, Gilbert islands, Samoa, Tonga; (b) via the Carolines, Marshalls, Hawai'i, Marquesas, Society - 379 islands, Samoa, Tonga; (c) via the Carolines islands, Marshall islands, Hawai'i, Penrhyn islands, Samoa, Tonga. Which of these alternative routes was used cannot be solved by a kava distribution map alone. But before mentioning the history of Polynesian migration in this area, I wish to bring forward a possibly insuperable objection of Peter Buck's to any of these routes and that is the distance between Kusai'e in the Carolines to Ra'iatea in the Society islands or to Samoa. 36 He states that the only introduced plants that will grow on atoll islands are the coconut and a coarse variety of taro and hence that all other plants must have reached Polynesia via Melanesia. But this objection is not as serious as it sounds because the distance is by no means too great for one voyage as we know that Tongans sailed to Tikopia carrying kava, an almost equivalent distance and that Samoans have reached the Gilberts and returned. Moreover, kava is found in the Ellice islands, a group en route. Again there seems no real reason why kava should have had to be planted on the intervening islands as the soil required to keep the root fresh and growing could have been carried in the canoe. 37

If we agree with Rivers in assuming a close connection between an original Polynesian immigrant people (whatever they may be called) and the kava ceremony but reject his theory that kava was brought via Melanesia, we are not assuming that other Melanesian immigrants did not bring other customs which were transferred to Polynesia via Samoa and Tonga. It seems possible for example that the right of fahu was influenced or affected by a similar right of vasu in Fiji. Hornell 38 traces a large number of invasions into parts of Polynesia by means of various canoe connectives. He has three main types which he calls Proto-Poly- - 380 nesian, Melanesian-Polynesian and Tangaroan. 39 He associates the Proto-Polynesian craft as most nearly similar to the present Hawaiian and Tuamotuan craft. Of three alternative explanations, (a) that the Proto-Polynesian arrived in and colonised all the Polynesian groups via Hawai'i, (b) that Hawai'i itself was reached by the direct route, the rest of the migration passing by a different and more lengthy one, and (c) that all of the newcomers arrived after coasting along the shores of the Melanesian islands stretching from New Guinea to Fiji, and that Hawai'i was colonised by way of the Society islands, he supports the first on comparative grounds. He ascribes the Melanesian-Polynesian type of canoe connectives to Melanesian influence on the Proto-Polynesian canoe in the west Central Pacific area and then passed back after a possible Polynesian conquest of Melanesian settlers in Tonga, Samoa or eastern Fiji. However, from the point of view of this monograph, this argument tends to show that if the original Polynesian canoe was introduced via Micronesia and is distinctively different from canoes introduced via Melanesia, it is as likely, if not more likely, that the distinctively Polynesian custom of kava drinking was also introduced by that route.

Peter Buck 40 supports this general conclusion. He believes the Polynesians in several waves emigrated to the Society islands, whether through Hawai'i or not, through Micronesia, from where they penetrated as far east as South America (to obtain the sweet potato) and as far west as the eastern islets of the Solomons and to New Zealand. However, he believes the great majority of the plants travelled clown the Solomons chain to Fiji from where they were distributed by voyagers from Samoa or Tonga.

To summarize the argument thus far:—

The kava ceremony exists as an important ritual complex in all of Polynesia (excluding New Zealand and perhaps Easter island), Fiji and in the southern Solomons. In those islets in the Solomons settled by Polynesians (Tikopia) it is of very great importance but in those possibly settled by Micronesians (Ontong Java) it appears to be absent. In New Guinea and the northern Solomons (Manus), kava exists without being the centre of any ritual - 381 complex where its place seems to be taken by the betel nut. The kava plant is found in the Micronesian chain of islands, close enough for it to be transferred from one to the other by canoe, where it seems to have an importance intermediate between Polynesia and New Guinea. 41

If it were associated with immigrant peoples passing through the Solomons, it is necessary to explain (a) how it became the centre of a ritual complex as otherwise it would almost certainly not be transferred by immigrants by canoe as it is not sufficiently strong for a drug nor useful as a food otherwise, (b) Why it is found to be most important in the area furtherest away from its presumed place of origin. 42 (c) Why it left no trace of its presumed existence in the island chain itself. These three questions seem to be soluble far more easily if we state that kava was introduced via Polynesia.

If it were associated with immigrant peoples passing through Micronesia to Hawai'i or the Society islands it is necessary to explain (a) the long journey via atolls on which it might not have been able to grow, (b) why it became associated with the Polynesian social structure everywhere in Polynesia except New Zealand. I have attempted to show that as far as (a) was concerned, the distance between atolls covered with soil is no greater than in respect to a similar route between places in the Solomons in which the kava root is found. The answer to (b) I hope to illustrate later. Its non-appearance in New Zealand is due, 1 suggest, to the peculiar social structure of the New Zealand Maori in which the whole relationship of chief and the family with the social structure differs very greatly from anywhere else in Polynesia. The name kava, however, remains in the name of another native plant.” 43

- 382

In support for the Polynesian route for the introduction of kava to Tonga, I can claim support from the independent evidence of Hornell's canoe examination and also the facts of known canoe connections between different parts of Oceania such as between Tonga and the Gilbert islands. 44

However, it must be stressed that if the bearers of the “Proto-Polynesian” canoe actually did bring the kava root to the Society islands, it must at that time have had a certain sacred character equal to the social value of stored food, which had to be carried within limited space. It might perhaps have had as an auxiliary use, the appeasing of the gods en route in which voyagers threw kava root overboard to the shark god, Sekatoa, swimming alongside 45 Unless the root had some value, there would otherwise have been no point in transferring it from one island to another.

Accordingly it seems that even if there were an early invasion into Polynesia via the Solomons, it is likely that the kava root as a ritual complex was introduced via Micronesia.

The centre of all the Tongan kava ceremonies is the drinking of kava either from a folded banana leaf in formal ceremonies or from a half-coconut shell on less formal occasions. It is as though through the actual drinking of the kava that the rank of the person concerned is confirmed or the god consulted. Again kava is always the liquid drunk at such ceremonies, with the one exception that in olden times a god, who was ritually connected with a certain liquid drank that liquid through the intermediaryship of a priest. The priest of Kautae, for example, drank sea water while president, but the members of the circle still drank kava. Kava is always presented to a person if one wishes to show respect to that person, usually with a little food, and both food and kava have a special meaning, which makes them more socially valuable than they would otherwise be.

The following description of the kava circle of the Tu'i Kanokupolu is largely from Collocott. 46 Some of it is more formal and capable of being voluntarily changed than other - 383 parts of it. I have divided it up into definite stages although it is probable there is no clear division in the minds of the participants in actual fact. The ceremony cannot be understood without examining the position of the participants in the accompanying diagram. It will be noticed that there are three main parts of the circle of varying sacredness, alofi, fasi, and toua.

(1.) Before the constituted alofi, food and kava are brought and placed before the king and chiefs. Motu'apu-aka, the matapule to the right of the Tu'i Kanokupolu directs they be counted. The matapule sat fairly close to the chief except in the case of the Tu'i Tonga, when they were at least six feet away. Even the usual freedom of the matapule in ritual matters was apparently insufficient to offset the Tu'i Tonga's sacredness in his ceremony. 47 The pigs and kava are distinguished aloud in weight and size.

- 384

Pigs and kava in Tonga ave not regarded as ordinary food. They acquire a certain ritual value, which is shown by their use only for special purposes or at special times. Wheeler 48 says poor people never thought of killing a pig for themselves, these being mostly raised for the use of chiefs. Gifford 49 says:

In this earlier period also, kava drinking was almost a prerogative of chiefly rank and met a demand only from such people. Today pigs are used mainly on special

occasions such as weddings, deaths, visits of important people or sometimes for Sunday dinners but rarely for meat alone. Their value, moreover, depends on their size, which was carefully distinguished in counting. Kava similarly depended for its value on size and weight 50 It would be regarded as presumptuous for a commoner who grew the larger type of kava root to keep it for himself, so he would - 385 present it to his chief. In addition a pig, when it was ceremonially divided, was regarded as more sacred in some of its parts than others. The back of the pig was always presented to the highest ranking dignitary present. In the marriage kava went usually to the bride or the bride's fahu.

The counting of the kava and the pigs therefore was one means by which the formality and gravity of the ceremony was emphasised, the greater the number of pigs and kava, the more important the ceremony. But more important than this is the meaning of the kava root. Kava was poured out on the ground in heathen times as an offering to the gods. Before engaging in an athletic contest each contestant would throw a piece of kava (mesini) over the fence into the enclosure of his patron deity. Kava was customarily presented to a chief by one of his retainers on being sent for. It thus has a meaning the acknowledgement of a ritual or politicial inequality and hence often the meaning of sovereignty also. The erroneous presentation of kava by one commoner to a chief who was not his overlord once precipitated a civil war. 51 Visitors of low rank brought a gift of kava, which was handed to the chief by his matapule. 52 Finau, at the time of Mariner, when Tu'i Kanokupolu, met an insurgent messenger from a hostile island but on satisfying him of his chiefly rank, the messenger pulled off his turban, gave Finau a piece of root and kissed his feet. 53

(2.) Motu'apuaka, the presiding matapule, directs the toua to be put in order. The toua sits behind the kava bowl and consists of (a) undistinguished visitors of low rank, (b) commoners, (c) chiefs who are descended from minor lineages of noble houses, who are not included in the Tu'i Kanokupolu's fasi or alofi, (d) younger brothers or sons to those entitled to sit in the ring. Tungi, the prince consort of Charlotte, the Tu'i Kanokupolu in 1918, and the successor to the throne sat in the toua and although separate from the other chiefs. Of those in the toua, he was the first to receive the kava cup. The relatives of the president usually sit in the toua. The toua, however, is not just a crowd of spectators, like those watching a football match, but an - 386 integral part of the whole ceremony. Its presence and absence marks the formal inauguration of the ceremony at the beginning and its dispersal at the end.

(3.) Motu'apuaka directs kava to be brought, which is placed before the mixer with the root towards him and the branches away. A man then loosens and divides the roots into big sections with a sharpened stake. When this is finished, the kava is turned about so that the kava. root faces the king and the presiding matapule. By comparing the use made of the kava root in this and other ceremonies, it may be possible to understand it. In the Tu'i Tonga's kava the roots of the kava are placed away from the President and the branches towards him. Now in the Tu'i Tonga's kava ceremony also no person could strike a wrongdoer, but at that of the Tu'i Kanokupolu and the Tu'i Ha'a Takalaua, it was the duty of those responsible for keeping order to punish wrongdoers, with the wedge for splitting kava or with the stick for carrying burdens (on which the kava plant would be carried to the ceremony) or with a branch of the kava plant itself. That is to say the Tu'i Tonga had the right of sanctuary, which the other chiefs had not associated with a special arrangement of the kava root. Robertson Smith 54 points out that a sanctuary does not become sacred because miracles will be performed there but that a place or person only became a sanctuary or holy, when it or he gave unmistakeable signs of being inhabited by a god. He says:

When the Tu'i Tonga had power in Tonga, a yearly miracle was performed by which the whole of nature was renewed at the inachi or first fruits ceremony. Without describing the whole ceremony, in one part of it, the priests carry a special variety of yams (known as “sick” yams) before the Tu'i Tonga, succeeded by real yams of a sort eaten by chiefs, followed by the families of the priests shouting, “We have come from work,” after which cere- - 387 mony the Tu'i Tonga immediately departs. This shows clearly the specially sacred nature of the Tu'i Tonga as the human “sanctuary” of the gods, by which “sick” yams are renewed for the coming year. The right of sanctuary, therefore, is a result of the sacredness of the Tu'i Tonga's kava ceremony, which does not exist in the Tu'i Kanoku-polu's. Evil doers are usually a source of danger to a community and must be removed. However, at sanctuaries, the place or person is so holy that evil doers cannot do harm. But at secular ceremonies the evil doer is normally removed from the ceremony and killed or otherwise punished. However, they have endangered not only those present by their possible impulses but, more important, have lowered the ritual status of the participants by their impurity. Thus at the Tu'i Kanokupolu's and Tu'i Ha'a Takalaua's ceremonies, to lessen the danger as far as possible they are removed by means of the sacred kava plant. Thus, because of the lower ritual status of the Tu'i Kanokupolu to the Tu'i Tonga, the kava ceremony can only be held if all forces likely to hinder the successful performance of the ceremony are removed. Thus the right of arrest, although originating from the Tu'i Kanokupolu, is performed through the inter-mediaryship of the kava root.

Now the notion of danger resulting from touching a chief superior in rank to yourself can be removed by the ceremony of moe-moe, when one kisses the soles of the feet of a ranking superior to the one whom one has inadvertently touched. Of the leading chiefs, therefore, a list can be made of those requiring to moe-moe to the one above which is a classification of rank. For the upper ranks it is:

• Tamaha
• Tu'i Tonga fefine
• Tu'i Tonga
• Tu'i Ha'a Takalaua
• Tu'i Kanokupolu
• Tu'i Tonga's daughter
• Tu'i Tonga's sister

When the Tu'i Tonga was the ruler of Tonga, he became responsible for the religious rites and the Tu'i Ha'a Takalaua for the ordering and administration of Tonga. As the Takalaua lost his importance his duties were taken over by the Kanokupolu. But all three have Tu'i Tonga blood in them but the latter only through the younger son.

- 388

Thus associated with the Tu'i Tonga we have superior rank, the kava roots placed away from the Tu'i Tonga and the right of sanctuary at the kava ceremony. Associated with the Tu'i Ha'a Takalaua and the Tu'i Kanokupolu we have inferior rank, the kava roots placed towards the president and the punishment of wrongdoers by means of the kava root or something that has touched it. But in any case Gifford says quite clearly that it is implied that the Tu'i Tonga's alofi was tapu in a way in which the Tu'i Kanoku-polu's was not. 55 But the idea of rank is a comparative rather than an absolute one. In a family, the higher ranking sister may be termed eiki by her brothers although a commoner (tua) in the social hierarchy. The Tu'i Tonga is thus a sacred figure because of his continual descent from the higher ranking side of the family and is a contrast to the commoner Tu'i Kanokupolu and Tu'i Ha'a Takalaua. The Tu'i Tonga is not and never was a god in the same way as the Roman emperors were deified. All authorities are unanimous in saying that what power the Tu'i Tonga receives is from the gods not because of his godlike nature but because he is ritually the most suitable person to receive it. It is thus possible that the difference in the method of splitting the kava roots as well as the different position of the kava bowl is an acknowledgement of this difference in relative rank. The right of sanctuary in the religious heads' kava circle and not in the secular heads' no doubt parallels the difference in duties between the three chiefs, but it also exhibits the comparative sacredness of each to those below him.

(4.) Motu'apuaka orders the root to be broken up further when men sitting by the kava, all relatives of the family of Ata, break off conveniently sized pieces, which they throw into the toua where they are scraped clean and beaten up with small stones. Formerly they were chewed to pulp but under missionary influence, this was abandoned. It is interesting to note that in the commoner village of Panga'i in Vava'u in modern times, the stones used to pulp the kava are returned to the owner of the bowl immediately after use, and are never used for any other purpose. 56

- 389

(5.) After the kava has been broken up, either Ma'ata-fahi or Tuakalau, both sons of Kapukava, orders the bowl, water, strainer and cups to be brought. The bowl is rinsed. Usually the well is not very near and many people rise from the toua and rush off to procure it. If they are not very quick, another lot of people rush after the first lot to procure some more. In former times they carried it in small coconut shells or other small containers, contriving to spill half the contents on the return journey. The explanation of this given was that it showed the power of the chief to have all these retainers under his control.

(6) The bowl is placed before the man who is to mix and strain the kava and the root placed in it. The bowl is three legged and the small hole used to hang it up by is placed towards the mixer in the Tu'i Kanokupolu's kava circle and away from him in the Tu'i Tonga's. The mixer grasps two legs of the bowl and tilts it towards the matapule so that he may see the contents. He then receives orders to commence to mix it. His two assistants pour in the water until they receive orders to stop. The man then strains out the solid part of the kava with the fibre of the fau tree. In straining the kava there are various elaborate movements. The elaborateness of the movements depend on the size of the kava and the formality of the occasion. Brierly states that the raising of the fau to let the liquid filter through into the bowl was considered the most critical part of the whole proceedings as a man's reputation as a kava maker would be lost for ever if he let a drop of the liquid fall outside the bowl. 57 This may be perhaps because of the danger in letting the sacred drops fall on the secular ground.

(7.) While the kava is being strained the food presented to the President for use in the ceremony (fono) is being distributed. Out of a basket a pig is selected which is cut up ceremonially before the king. It is distributed as follows:—

• Back of the pig to the President.
• Head and one hind leg and one foreleg are for Lauaki

- 390

• and the chiefs and matapules on the left of the President in the main kava ring.
• Another hind leg, foreleg and hindquarters are given to Motu'apuaka and the chiefs and matapules on the right of the President in the main kava ring.
• A yam and the chest of the pig are put in a basket and taken to Ve'ehala. (This last portion is termed the vae-tolu, the third portion.)

Both sides of the circle thus receive an equal sort of portion. In Samoa the hindquarters of a pig are traditionally assigned to the orators, the equivalent approximately to the Tongan matapule. Although Ve'ehala receives the third portion, it would be incorrect to presume that it is therefore an inferior position in the fasi alofi to the alofi itself, and that the various parts of the pig can be graded downwards in that order. For example Ma'afu is the highest ranking member of the Ha'a Havea whereas the Tu'i Kanokupolu and not Ahio is in the Ha'a Ngata. Yet Ahio may receive a different part of the pig. The implications of this part of the ceremony seem rather to emphasise the different duties of the members of the varying Ha'a. Again, strictly speaking, matapule have no rank in themselves but what is acquired from their chiefs. It is reputed in early times that many matapule were foreigners because of their exemption from Tongan rules of precedence, moe-moe and so on. The only person permitted to eat with a chief was his matapule (except the Tu'i Tonga's matapule.

However, while this pig is being distributed as above, the rest of the fono food is being subdivided and distributed to individual chiefs and matapule. For example, the baskets with the food in them assigned to Ve'ehala's section (the vae-tolu), goes to Ma'ahu (and his matapule), Ata, Vaha'i and finally out of the ring to the kav ngaohi fono (lit. the attendants to make the sacred food.) Each person retains a portion of the food and then passes it on further, a certain amount is not taken to all of them separately. This food that is retained by each member of the vae-tolu is not eaten by them but by certain relatives after the ceremony is over. These relatives are, in the first generation, those who are fahu to them (children of the sister's son) but in the second and subsequent generations, those entitled to eat the food may not be strictly fahu but “higher ranking grand- - 391 children.” The idea of fahu was not narrowly limited to one generation only as it is in modern Tonga. 58 Its importance in this ceremony lies not in its economic but in its ritual importance, the best example being that mentioned by Prof. A. R. Radcliffe-Brown. 59 But in the instance of this kava ceremony, all grandchildren on the female side, whether male or female and possibly including commoners would have access to this fono food. 60 Collocott suggests the final selection was made on grounds other than strict descent. But whether this was so or not, the actual jural relationships (as distinguished from blood) must have been acknowledged by custom and confirmed in part by the ceremony. Gifford suggests that in ancient times the fono food was actually eaten by those to whom it was presented. 61 If this were so it would be an interesting conjecture, unable to be verified, as to whether the nature of the fahu relationship was undergoing an increased emphasis in its economic aspect at the expense of that of authority and seniority 62 If there are no fono eaters present, the chief may invite anyone to partake of the fono after the ceremony. Foreigners may eat the fono without invitation. The foreigner is regarded as exempt from most tapus binding on the Tongan, even if they are only descended in the second or third generation from foreigners, provided that they have kept themselves together as a special “foreign” group.

During the distribution of the fono food, various speeches are made of a formal type, preceded usually by an invocation to certain ritually important chiefs (fakataputapu). The matapule of the Tu'i Kanokupolu must speak sitting, in order not be be above the Tu'i Kanokupolu. 63 The - 392 speeches themselves vary with the reason for the kava. At this inaugural ceremony of the Tu'i Kanokupolu, they took the form of an acknowledgement of the Tu'i Kanokupolu's right to rule Tonga. There were three main speeches by Tovi of the village of Kanokupolu (an important place with a minor chief on Tongatapu) on collecting shellfish, certain types of which must be sent to the Tui Kanokupolu, by Faoa of Kanokupolu on reef-fishing and bonito catching on the same theme and Ata on land cultivation. All emphasised the fact that the upkeep of the land was a religious duty in order that the ruler could have plenty of food and hence continue to rule over them.

(8.) During all this time, the food and kava distributions (as distinct from the ritual apportionment of the fono food) was taking place, each chief's name not being called but that of his matapule.

(9.) Meanwhile the straining of the kava is finished and the serving of the kava commences. Mariner 64 and Thomson 65 state that the kava dregs retained in the fau fibre after the straining is complete, are afterwards taken away by the cooks (commoners) and chewed over again to make a fresh infusion for themselves. This is clearly because commoners would acquire a higher than normal ritual status by using it thus. 66 The strainer is very cleverly twisted over each cup as it is raised to be filled so that it is taut and every drop of liquid is expelled from the taut fibres. It sometimes happens that more than one fibre must be used but in this ceremony described by Collocott, apparently one fibre was sufficient to fill everyone's cups. 67

- 393

(10.) Of the two men seated at the side of the bowl one calls as each cup is filled, the other removes little pieces of kava that adhere about the rim with a small coconut leaf and using only the hand on the side nearest the President. The caller cries, “The kava is lifted,” and some people from the toua approach and have kava wrung into the bowls they are holding by the server. Motu'apuaka replies, “Take it for Lauaki.” He claps his hands and gets it taken to him. The second cup is for Niukapu and the third for the Tu'i Kanokupolu when the command is varied by “Bring it here for Tupou.” (Taliai Tupou is the patron god of the Tu'i Kanokupolu.) The highest chief's kava is always ordered to be brought, not taken.

The fourth, fifth, sixth and seventh cups are taken in order to chiefs and matapule, alternating from one side to the other. Cup four to Nuku, cup five to the matapule next to Niukapu, cup six to the matapule next to Nuku, cup seven to Ahio, cup eight to the fasi alofi to Ve'ehala. The ninth cup returns to normal distribution with three more on each side; then to the chief in the fasi alofi opposite Ve'ehala, namely Ma'afu; three more on each side then Ata; three more on each side then Vaha'i; three more on each side then Tungi; three more on each side then to Tolo in the toua, who was specially brought from Vava'u to Tongatapu to represent the Ha'a Ngata Tupu. Then the matapule ordering the distribution receives a drink himself. He does not at once drink it but sits with it between his hands, resting on the ground. The drinking of the three on each side followed by a cup which breaks the regular distribution continues. When the kava is nearly finished, the caller drinks and after a few more cups his vis-a-vis on the other side of the bowl. Then the cups will go to the toua. Finally Ata says,” Fae kava,” and the caller cries,” There is the remainder of the kava.” The matapule replies, “Leave it for the kava server.” The matapule, who has been holding his cup since it was given to him now drinks. The server wipes the bowl with the fibre strainer, another man takes away the bowl and the server drinks his own cup. The two men at the side of the bowl turn and face the King. A man goes to bring away the matapule's cup and after he has sat, Vaha'i, without rising slides forward on his rear and tells the matapule that that is the end of the kava. The matapule - 394 responds, “Remove the toua.” Those in the toua then go. No one in the alofi may leave until the toua has gone. That completes the formal kava ceremony of the Tu'i Kanokupolu. The food remaining is then eaten or taken away. This method of distribution of kava between alternate groups was used similarly of some early white navigators to the Tongan islands in which Tongans and white men were served alternately.

According to Gifford 69 in the Tu'i Tonga's ceremony, he received the second cup of kava and one of his matapule, the first, the cup was carried to him in silence, his title not being called. As the bearers of the filled cup to those seated in the alofi were not permitted to cross the ring from one side to the other, there were two bearers, one going to the right of the bowl, the other to the left. 70 The bearers announced the filling of each cup in the Tu'i Tonga's ring, whereas in the Tu'i Kanokupolu's it is announced by a man who sits continuously at the side. The order of being served depends upon the composition of the kava circle, that is, the precedence of a chief in the Tu'i Kanokupolu's kava circle might differ from his precedence in the kava circle of another chief.

In analysing this last section of the kava ceremonial, it is quite clear that if the statement by various observers that - 395 the ranking or honour given to the kava drinkers was the same as that order in which the kava was served, then they are incorrect except in the vague sense that if you were President, for example, you would drink one of the first four cups or that if you were of high rank or a white visitor, you would be likely to get a cup earlier rather than later in the ceremony. The presence of the Takalaua in this kava ceremony is that of a junior branch of the Tu'i Ha'a Takalaua, as he would normally be excluded from participation in the Tu'i Kanokupolu's circle if his kava ceremony were still being held. The early presentation of kava to Niukapu, Nuku and their matupule is probably due to their common parallel descent along with the Tu'i Kanokupolu from the Tu'i Ha'a Takalaua. They are all three known as Ulotolu (three-headed). If Nuku and Niukapu are present but no higher chief, both preside. Note that no person fahu to anyone in the circle can sit in the circle. Tungi, for example, related to the Tu'i Kanokupolu through the sister of Taufaahau (George I) and a descendant of a woman is thus fahu and hence of higher rank than the Tu'i Kanokupolu but yet he sits outside the circle. Many of those outside the alofi have special duties like the lineage of the Ha'a Ngata Tupu, whose job it is to protect the Tu'i Kanokupolu and to punish any offenders against the law of the land.

At the installation of Queen Salote (Tu'i Kanokupolu) she was preceded by a man who ran ahead, brandishing a spear, crouching and looking around. During the kava ceremony, he was free of all restrictions. He smoked, lounged and walked close before and behind the Queen's person and when the pig's liver was placed before her he impaled it on his spear. This liver was part of the main fono food after the kava ceremony and after its being offered to the Queen, the food was distributed among the high chiefs to be eaten. This man, whom I will call a herald although his actions differ from those we usually ascribe to a herald, comes from one of the traditional Fijian foreign clans. It is said that a man named Soakai, also a Fijian, had similar rights in the Tu'i Tonga's kava circle. This custom can be compared with one in Samoa 71 in which certain high chiefs had personal attendants who performed - 396 such tasks as shaving, burying them when dead and so on. The interesting fact is that many of these Samoan heralds also behaved outrageously at the kava ceremony also. One Salelesi, a descendant of a Tongan foreigner used, when his village was preparing kava to perform all sorts of services for the visiting King, ate the portion not eaten by the King of the food set before him and had the privilege, when food was brought in, of selecting and taking away a large pig and quantities of taro from anyone's allotted portion.

In order to understand this ceremonial it will be necessary to define several of the terms that I use. These are taken from Prof. A. R. Radcliffe-Brown: 72

The attitude of one person to an object outside himself is determined in part by the relative status of himself and the outside object. The idea of status may be defined as the totality of all the individual's rights and duties towards the object, as recognised by the social usages (laws and customs) of the society to which he belongs. If this conception of status between subject and object is infringed it may be punished by certain material fixed penalties visibly imposed by society (law) or by a feeling that if the rules governing the status are infringed, a supernatural punishment will result. This punishment may be regarded very lightly as Hogbin has shown with respect to kipua in Ontong Java 73 or else it may have an immediate terrible destructive force, as it was once believed would happen in many parts of Polynesia, were the temple gods destroyed. However, the peoples and objects concerned with the preservation of this varying ritual status are not always equally sacred through- - 397 out their lives. The kava ceremony clearly falls into the category of the ascription and destruction of the ritual status of those objects and participants in it.

The concept of ritual status may belong to certain objects naturally because of their relationship to other objects. In Tonga a sister of a chief is of greater ritual status than her brother or anyone lower in rank than she. Should he touch her by error he would immediately find someone in rank higher than she to perfom the moe-moe feet-touching ceremony to. We are told that this applies even to the Tu'i Tonga, who, in the accidental event of touching the Tamaha could moe-moe to her feeding bowl. The reason why a low-ranking Tongan could not keep a pig or kava root of the largest size for himself is probably because of the ritual danger of doing so to himself. In Tikopia, Firth mentions that no Tikopian could keep more than one or two bonito-hooks for himself and that even when he did so, he pretended they belonged to the chief and that he was only borrowing them. 74 The chief, being of higher ritual status than the borrower could handle bonito-hooks without danger to himself whereas a commoner had to observe the convention for his own safety. This duality between those objects which have ritual value and those things which do not seem to bifurcate Tongan society. I am not concerned to explain why this is so, if, indeed, it is ever possible to do so, but to show first, how in this ceremony and second, in many other kava ceremonies, there is an attempt to bring these two aspects together.

The ceremony commences with certain persons of ritual value present in a certain formal order. That these people are in a state of high sacredness can be seen by two facts among others. Mariner 75 says:

The people in the ceremony normally associate together in various other gatherings but they clearly have this sense - 398 of something “different” just as the modern churchgoer puts on his best to go to Church although he sees only those people whom he sees in everyday life there. The second fact is that at the Tu'i Tonga's kava ceremony it is so important not to interrupt it that a special bowl is put outside belonging to the Tu'i Tonga, so that anyone who suffers from ritual danger by touching a chief lower in rank than the Tu'i Tonga may moe-moe to it, without disturbing the ceremony. Both these facts are inferential to this ceremony because of lack of documentation.

Meanwhile food is to be distributed and eaten by the people in the alofi formerly and now by the fahu relatives. Now a Tongan chief normally ate alone (or with his mata-pule, traditionally a foreigner), because immediately he ate the food, it acquired a similar ritual status by contact with himself and was hence a source of danger to anyone eating it subsequently. Since all the food is presented to the highest ranking chief for re-distribution it became of the same ritual value as himself instead of the former profane value. In this ceremony, this transformation is performed by the act of being counted by the chief's matapule.

The toua is now put in order when the formality of the ceremony is emphasised by contrasting the “sacred” chiefs with the “profane” toua, although certain people in the toua must, in actual fact, be of higher rank than some in the alofi. (However, the status of the individual chiefs in the toua as contrasted with the generic in the toua is recognised later with the individual distribution of the cups.) The area before the bowl is also more sacred than other parts of the ground as witness in the most sacred ceremony recorded of the Tu'i Tonga, that the commoner kava bearers cannot cross this area. In the informal ceremony given to the Queen, the Tu'i Kanokupolu, the food was placed on this area after counting, then the kava selected from this food was beaten and mixed, and then all the food was moved to one side to be dealt with in a manner to be described later.

It is as though the ceremony is designed to build up ritual value and then to destroy it. (See diagram 34.) The sign of the building up of the ritual status of the ceremony is shown by such indications as counting the food by the President's matapule, to whom the food is nominally given, the ritual eating of a portion of the pig which is not normal - 399 food, the use of specially sacred kava roots and gestures, the participation in the ceremony of the fono eaters, ritually superior to those in the sacred alofi and so on. When the actual kava ceremony is performed the ritual status of the participants is at its maximum but as it is given to those of lower and lower rank, the sacred kava gradually becomes lit to be drunk by those of lower rank approximately in order. This hypothesis could be verified if it could be shown that if the kava were not distributed approximately in order of rank, an undefined misfortune would befall the participants. However, the fact that if the kava was insufficient to supply the assembly, its redistribution commenced again with the President and not from where the first serving finished gives credence to this fact. But lacking any such direct statement, I can only bring forward Gifford's evidence that in the illness of the Tu'i Tonga a prayer is said for his recovery after the dilution of the kava before distribution, which can only be done by a chief or a matapule. 76 The words used are:

• Server: “The kava is diluted.”
• Director of Ceremonies: “Diluted with a blessing.”
• Server: “The kava is diluted.”

This addition to the ceremony is only intelligible if the maximum sacredness occurs at this point. Gifford also uses another phrase cited from a MS. of the Tu'i Tonga's mata-pule in which he says, “And the kava ring faces him (the Tuitonga) and the sacredness of persons commences with him and continues to the end.” 77 Lacking the original Tongan text, it seems to me “end” means here the end of the alofi as the MS. refers only to the kava circle proper.

That the kava circle actually imparts to the participants a greater than normal ritual status is shown by the kava mixer. Normally the kava mixer, being a commoner, would not be permitted to drink the kava in this important ceremony, but by actually preparing and distributing the kava, by contact they acquire a degree of sacredness sufficient to drink the kava after the high chiefs but equivalent to certain of the low chiefly ranks. That is to say that whereas the normal order of drinking the kava might be high chief, chief, low chief, commoner if the latter were in the toua, - 400 now because he is in the aloft and because he is mixing the kava the order is high chief, chief, commoner mixers, low chief, commoners in the taua. 78 An interesting comparison with this alteration of sacredness during the kava ceremony is cited by Mariner with reference to funeral observances. 79

That is to say that the period of taboo for dead chiefs is ten months, which is the maximum, but in view of the comparative ritual status of the toucher of the dead man and the dead man himself (here acquired by birth rather than through a ceremony) only a three, four or five months' taboo is expected. However, in the case of the Tu'i Tonga, he is so much more sacred than anyone else in Tonga that it is as though the comparative ritual status of everyone in Tonga is as nothing compared with the sacredness of the Tu'i Tonga. Hence the maximum period of taboo is enjoined.

In using the term “sacred” I am referring only to the ritual status of the participants in the particular ceremony referred to. This idea is quite clearly understood by the Tongans themselves. In the MS. referred to earlier, the Tu'i Tonga's matapule writes:

- 401 “Ei” is a method of showing respect. By Tu'i Tonga here is meant the lineal descendants of the Tu'i Tonga's junior line. The Tu'i Faleua is the Tu'i Palehake, usually the King's younger brother. We have here an example of how the actual pattern of the kava ceremony is transferred to other Presidents of lower sacredness in which its comparative nature is clearly illustrated.

The more important the ceremony the greater the contrast between those who have and who have not ritual status in the ceremony. This has already been illustrated by the different customs of the Tu'i Tonga's and the Tu'i Kanokupolu's kava ceremony such as the placing of the bowl with the hanging knob away from or towards the President. But one interesting problem that arises from the ritual value of the Tu'i Kanokupolu's kava circle is why the Queen should have an herald (page 395).

Now the usual emotional state accompanying ceremonies of great importance to the participants is the extreme gravity of the occasion, exhibited by the slow stately step and the conformity to pattern. The chiefs taking part in the ceremony would never dream of allowing themselves to smile. And yet the herald is smiling, grinning, shouting, insufficiently clad and carrying a spear, expressly forbidden to others. His behaviour is quite impossible for a Tongan, who would expect immediate death for such misdeeds. But this herald is a foreigner, a member of a Fijian clan, and thus exempt from Tongan supernatural penalties. Foreigners are never classed as Tongans ritually as long as they keep their separateness as a group. Accordingly this Fijian herald, in behaving in such a profane way is so profane that he is incapable of being identified with the high ritual status of the other participants in the ceremony, and yet is in no danger from the Tongan gods.

Now referring back to the main ceremony, we have left a problem to be solved. Although we have desacralized the kava by easy stages through decreasing rank, the food has been left in a higher ritual status than normal by being placed in the centre of the ring. In the process of redistribution this food must be touched and eaten by profane hands and those responsible for redistributing and eating this food must be in no danger from touching it. The herald solves this problem. He has carefully refrained from touching - 402 any of the food during the ceremony but now it is completed he sticks his spear into the food and thus makes it once again palatable.

A similar functionary seems to exist in Samoa. Unfortunately there is no mention of his part in the final act set of the kava ceremony but Kramer gives the following facts about him. 80 Salelesi is the personal servant of the Tuiatua and is the only one allowed to eat from the King's food and what he leaves. He is supposed to be descended from a Tongan foreigner, Lesi. When Malietoa Laupepa died, he was buried in a coral chamber in the reef but the only attendants admitted to the vault were Tuamasina and Salelesi. At the nomination of the Tuiaana, at the kava ceremony, Salelesi sat near Mata'afa on his right. After the kava was ready

One major distinction between Salelesi and the Fijian herald in Tonga is that the former was entitled to a special position at important kava ceremonies, but this can be explained by the fact that he appeared also to have had some of the duties of a matapule, which is to say that his rank was acquired from the chief whose servant he was.

The hypothesis here put forward is that in the most formal ceremony there is in Tonga, the ritual status of the participants is shown by the nature of the rites which accompany the ceremony, such that the necessity of their continuation lies in the contribution each symbolic part can give to the comprehension of the whole ceremony. Certain of the symbols in the ceremony seem to have no meaning while others are so clear that no exposition seems necessary. The undoubted change in certain of the symbols over the course of time is probably clue to an alteration in the structure of Tongan society. This assumption is, in the case of - 403 Tonga, unverifiable, for lack of enough material facts on which such an hypothesis could be based.

In the last section, I attempted to show in what way the kava ceremony is representative of a body of rites whose purpose is to impart and remove a certain ritual value from the participants and other objects in the ceremony. Although I did not explain why it. was necessary for the kava ceremony to have this value, in this section an attempt will be made to put forward other interpretations of this ceremony which, although failing to explain why the ceremony should take this particular form, help to reinforce and strengthen the ceremony as a whole.

The kava ceremony is held on a variety of occasions. If the ceremony as a whole occurs on a large variety of social occasions, there must be some underlying common factor on all those occasions which require its use. But the kava ceremony just described is almost the most formal used and on occasions when the ceremony is not performed with the same care, it is not to be expected that there is the same necessity for it full contribution to the integration of society. Its specific contribution to the functioning of society will accordingly be more difficult to define on less formal (or less sacred) occasions. The occasions mentioned in the table below are mentioned in the footnotes but limitations of space preclude any detailed reference to the occasions.

A.—SOCIAL OCCASIONS.
• 1. Morning kava party of chiefs from one area together with their matapule and attendants. 82
• 2. At visits of chiefs from elsewhere. 83
• 3. Before war.
• 4. Pigeon snaring. 84

B.—CEREMONIAL OCCASIONS.
• 1. At conclusion of fono (?) 85
• 2. Inauguration of new chiefs.

- 404

• 3. At meetings with a god or gods for divination. 86
• 4. Inachi and tow-tow ceremony. 87
• 5. Pact of revenge. 88
• 6. Peace offering. 89

C.—TIMES WHEN ONLY COMMONERS ARE PRESENT OR RITES DE PASSAGE.
• 1. Marriage.
• 2. Death.
• 3. Modern naming at one year of age.
• 4. After seeing the shark god, Sekatoa, in the water. 90
• 5. Fuakava. 91
• 6. Evening or morning drink for sociability in modern times. 92
• 7. To remove tapu on shark house after a shark cutter has been used for the first time for shark fishing in modern times. 93
• 8. At birth to celebrate the conclusion of the mother's confinement at which no paternal relatives are present in modern times. 94

The above are only a few occasions on which kava has been prepared, selected from both modern and ancient times. There are probably many other occasions which I have omitted. The seriousness and importance of the occasion is shown by the formality of the occasion rather than by the high rank of the participants. Collocott explains how in an informal ceremony at which the Queen was present, she received the first instead of the usual third cup. 95

Now before comparing the various reasons for the ceremonies mentioned above, I will briefly cite two descriptions of the place of the kava ritual complex in warfare and in marriage.

- 405

Warfare: In the attempt to Christianise Tonga in the last century, Taufaahau (later George I) attempted to conquer the non-Christian opposition by force of arms. He first of all conquered the people of Mu'a and Nukua'alofa but was unable to conquer the fortress of Hihifo, through lack of adequate troops, lie returned to Nomuka, a small island in the Ha'apai group and enlisted the aid of Havili, one of the greatest fighting men of the time and other Ha'apai troops. He made Havili his general at the head of the combined reinforcements and the whole fleet of canoes sailed for Atata island, near the Hihifo fortress. The war leaders were not all chiefs but consisted of some commoners among whom were Taifonoifua of Fonoifua, Falase of Nomuka, Suitus of Lofonga, Tafolu of Lifuka and Tuimalohi of Vava'u. Albert Taufu described the scene on Atata to Gifford as follows:

Taufaahau was entitled to the first rather than the third cup of kava on Ataa because here he was the leader of a war party and the ceremony was more informal than the Tu'i Kanokupolu's. No one could have accepted Taufaahau's challenge to drink his cup of kava of course. It was rather a histrionic way of appointing Havili to his command. The reef outside Kolovai dries up at low tide and is very painful for the feet to walk on. It stretches for as much as one to two miles offshore. The tide appears to have been going out and it shows their large enthusiasm for coming to grips with the enemy, when, rather than wait for the tide to come in again, they travelled over the reef bare foot. Taufaahau's war party was extremely heterogenous even for Tonga. Not only were there fighters from Ha'apai and Vava'u in the same party but there were commoner war leaders, undoubtedly having under their command many of higher rank. Even within the fighting units, there must have been a strong group for returning back immediately from fear of a counter-attack from the recently subjugated Mu'a and Nukua'alofa areas. In fact when the Nomukans returned home, they found that some warriors from Mu'a had “slaughtered the people, women and children and left the same day.” Taufaahau also, on returning to Lifuka found a revolution in progress against his authority. After the battle we find the Hihifo troops overcome after a furious resistance in which they almost succeeded in beating the invaders off. The turning point in the battle was the gibe,” Why don't you assist the troops instead of mending sails?” to which Havili replied after the battle, “Have you forgotten the kava that I drank on Atata?” with the obvious implication that if he had drunk the kava under such circumstances, he could not lose the fight. The key to the victory is the kava.

- 407

At first one has a number of disparate and separate forces, Ha'apai troops, Vava'u troops, chiefs, commoners and so on, loosely united by the personality of Taufaahau and their recently acquired new lotu Wesleyan Christianity. The first thing to be done was the preparing of kava, which was done immediately before the battle, where the participants were sacralized, and where the head chief surrendered his right to lead his men into battle by symbolically offering the first cup of honour to his war leader, Havili. 97 As soon as they had finished their kava they sailed straight to Hihifo, the place to be attacked under one leader and as one unit (tangatatau). One could represent it diagramatically in - 408 some such form as I have tried to do in the diagram opposite where the closeness of the lines represent the amount of social interaction between persons. The order of events is clearly (1) disparate units loosely joined; (2) ceremony; (3) tangatatau led by Havili, an acknowledged leader; (4) victory.

Marriage Kava: The information used in this section is recorded by Tukuaho from his father, Tungi. 98 It refers to marriage between a chief's son and a chief's daughter. Now in any society, a person has different social relations with different persons by which he regulates his behaviour to those persons. One of the most important ways in which he decides on his relationship to others is by means of kinship determined by birth, marriage or adoption. But at marriage, not only is a person entering upon a new and important relationship but all those related to one of the marrying parties becomes related to the other by the ceremony. Marriage thus becomes a formal confirmation of new relationships among all those concerned. Many of these new relationships in Tongan society take a fixed pattern. Among the more important of these is that of fahu, by which the future children of the bride have a claim to appropriate much of the goods of the bride's brother. It is little wonder that the relatives of those marrying have a decisive voice in forbidding the marriage, but, even after marriage, it is in the interests of the bridegroom's parents and of the bridegroom, through whom property will descend as well as titles, that the children will also be given the other half of the inheritance due to them by the maternal relatives. Marriage is thus the confirmation of new relationships between two groups of people and has three main parts: (a) separation before the marriage; (b) transition in which all the relatives get to know each other without finally committing themselves and (c) reintegration in which the new relationships are confirmed. In Tonga the final part of the marriage was not confirmed until the first-born child

- 409

was sufficiently strong to be moved from one village to another should it be desired. Marriage can be thus defined as the creation of new relationships by the combined sanction of those connected with the bride and bridegroom by kinship. Since the ultimate necessity of these new relationships is to continue them in the new generation, the birth of a healthy child is the final stage in validation.

Now it is clear that one of the means by which groups of persons are linked together is by means of the kava ceremony so it is interesting to note in what part of the total ceremony it occurs.

- 410

• 1. Kava is drunk at the bridegroom's house, the bride being the president of the ceremony, on the first official meeting of the bride and bridegroom's friends. The bridegroom prepares food but it is not eaten but taken to the bride's home.
• 2. Every day there are presentations of food and kava until the marriage is held.
• 3. On the night before the day on which the marriage ceremony actually takes place, the bridegroom's sisters come with kava and with some of the bride's people to drink kava on the bridegroom's village green. (Marriage in Tonga is patrilocal.) It is immaterial whether the bride is present or not.
• 4. On the marriage day after the bride and bridegroom have changed twice into clothing, the first time the clothing being provided by the bride's people, the second time by the bridegroom's people, the bride goes to the green where the bridegroom's sister utters the invocation, “May you bear children, yea, even twins.” Then the bridal kava is served with the bridegroom as president. As fono food for the kava, food is provided both by the bride and the bridegroom, which all cat. Each type of food presented is symbolic of either the bride's or bridegroom's side of the family. The back of the pig is given to the bride's father's sister's child.
• 5. The second day after marriage the bride presides at her kava ring with relatives of the bridegroom present.

I have not analysed in any detail the other aspects of the marriage ceremony to show in what respect the relatives of the bride and bridegroom confirm the marriage nor have I described the distribution of the wealth of the two families, food and manufactured articles (ngaue and koloa). The total series of marriage rites have the effect of coalescing by the supernatural kava ceremony, the newly created social relationships. In the diagram opposite, on each side of the family are mutual rights and responsibilities surrendered to or gained by the new generation and the triangular group of sanctions at the centre of the marriage is the mechanism of validation confirming the mutual contract between the relatives concerned. The relationships here are more fixed than in the previous case of the war party but the effect of the kava rites on the people concerned must have a common element of meaning which makes it appropriate to the three ceremonies described in detail.

In all the ceremonies described and in all those mentioned on page 403 certain other symbols are used in addition to the kava symbol. 99 It will be worth while - 411 separating the essential from the unessential part of the system of rites until we have the highest common factor so to speak. Is food exchange an essential part of all the ceremonies? In the instances mentioned on page it occurs in A1, A2, B2, B3, B4, B6, C1, C2, C3, C7, C8, but not in A3, A4, B1, B5, C4, C5, and C6. Its absence can be understood in A3 as war parties might not carry food with them, expecting to find some present, in A4 and B1 we are not informed whether food was served, in B5, C4, C5 and C6, food is not part of the ceremony. I think from the evidence we can state that food was usually served but not always. The occasions on which food was not served seems to be only when one or two people were present or when the ceremony was informally composed of a random selection of people in modern times. In B5 only one person was present, in C4, fifteen Tongans in the accidental presence of a God never contemplated giving the God food nor eating any themselves, in C5 only one person was present and in C6 food is not essential. From these examples it seems clear that food is only an essential part of the rites associated with the kava ceremony when a number of people are deliberately present or to stress the formality of the occasion. We can also reject the necessary presence of any special food such as pig, although having a ritual value it would naturally be associated with other ritually valuable objects or occasions.

Is the order in which the kava served of fundamental significance in all the ceremonies studied? In A1, A2, A3, it is so. In B2, B3, C1, C2 it seems that certain people had to be offered kava at a certain moment in the proceedings but that otherwise wide latitude was allowed to the officiating matapule. In A4, B1, B4, B5, B6, C3, C5, C7 and C8 either one person only is participating or we have insufficient information. In C4 and C6 the order of serving the kava seems unimportant. Here the evidence seems inconclusive because the ceremonies at which an exact order is to be performed may only be a further stage of the occasion on which individual validation of titles occurs. The two contrary occasions may be dismissed as they are modern performances of the ceremony by commoners, since at one time commoners could not participate in kava ceremonies. It seems clear, however, from studying the instances, that the - 412 more formal the ceremony, the greater the significance of certain persons receiving the bowl at a certain moment. For example, at all ceremonial occasions the President receives one of the first three cups depending on who the presenter of the original kava root was to the assembly.

In all ceremonies mentioned kava is prepared in a special bowl by someone other than the active participants. Accordingly I will define a kava rite as the preparation of the kava root by a certain method of straining it through fibre by someone other than the active participants, the degree of formality being shown by the importance attached to the distribution of the drink to certain members of the circle. We have already noted that the presentation of the kava root by itself without being accompanied by any rite has as a meaning the acknowledgement of ritual inequality between donor and receiver. We have also noted that a very large number of persons and objects in the Tongan social structure possess ritual value per se. Also certain occasions in the Tongan world require certain bodies of rites to be performed in order to fix certain relationships in a new way with a supernatural as well as a normal agreement between the parties concerned. This is done by means of person and objects of ritual value having their ritual value accentuated by the kava rite and then returned to normal ritual value. The rite uses as its central symbol the kava root, which already has a meaning of ritual inequality. This seems to me to be a sufficient answer as to why the kava rite as the centre of a larger ceremony continues to be performed in Tonga. I will now mention a few other explanations of this ceremony which have been put forward by other anthropologists and writers.

Some early missionaries considered the beverage intoxicating and that kava, was accordingly drunk for the sensation of doing so. Turner, 100 for example, calls kava an “intoxicating draught from an infusion of the chewn root of the 'ava plant.” However, he apparently does not regard the plant at all intoxicating in comparison with “foreign liquors” 101 because he says, “Long may the natives be preserved from the curse of drunkeness.” If they were drunk with kava previously, they could hardly be saved from the - 413 curse of it after European spirits had been introduced. It seems that, apart from a certain soporific effect and, among habitual drinkers, a paralysis of the lower limbs, it has almost no effects. Moreover, it is served immediately on being mixed and is under no circumstances allowed to ferment.

It would thus appear that, were the main purpose of the kava ceremony principally the mere enjoyment of kava, then its soporific effect and floury taste contrasts ill with the variety of food otherwise available. From personal enquiry it seems few modern Tongans thoroughly enjoy the taste of kava on first sipping it, and this apparent contradiction is paralleled by the attitude towards roast pig, which, although always served at ceremonial feasts, is seldom eaten until the more toothsome dainties are consumed. And yet after a feast the question a visitor is asked, “What sort of pig was served?” It thus appears that although the drinking of kava may be one reason for the ceremony, it is not a principal or adequate reason. Moreover, such a theory would fail to explain why there was no appropriate ceremony for coconut drinking for example, which is far easier to obtain, more frequently and apparently pleasanter to drink.

Because the kava drinking has a certain ceremonial significance the simplest example of kava drinking (like C6) is more than a merely technical action. Hence any explanation of its continuance must take into consideration the symbolic or expressive element in the ritual. Having considered and rejected the material reason for the continuation of kava drinking, I will now attempt to suggest that the key to the understanding of this ceremony's continuance lies in its ceremonial side.

In Tonga there is no tradition of the kava plant having been brought to earth by the gods. 102 It was first used as a drink by the tenth Tu'i Tonga but there is no record as to whether there was a previous ceremony of this sort, or when the ceremony was founded. In this instance, therefore, its supernatural origin can be no explanation of its continual - 414 celebration. 103 There is, moreover, no historical tradition stating how or whether the kava ceremony was brought from another island. There can be thus no known historical explanation of the origin of the ceremony.

A second explanation which is put forward by the Tongans themselves is that the ceremony continues to be held because it is faka-Toga, in conformity with Tongan custom. Although this explanation is adequate to explain why a particular ceremony is thought to be performed as it was before, the anthropologist can hardly be satisfied with this explanation in view of the wide variations in custom over the last few years and the distinction between the different ceremonies at the same time. Although it is possible that the reason that the Tu'i Tonga's ceremony is now abandoned is because of an “accident of history,” there may also be inductive laws of change in society, which it is his task to attempt to formulate and verify.

A third explanation of why this ceremony is performed is that put forward by the London “functionalist” school of social anthropology. They find a sufficient explanation of the continued existence of a ceremony such as this in the inter-relationships of the various factors concerned. In the kava ceremony of the Tu'i Kanokupolu, for example, it is probably correct to say that the performance of the ceremony reinforces the “functioning” of the ranking system, that it acts as a stimulus to economic organisation both in its demand by the three chiefs to produce more and by the organisation required to bring the necessary fono food forward at the right time, that the religious integration of secular and sacred is reinforced and so on. But this statement, when analysed, is seen to really assert that every aspect of society tends to be interconnected and reinforced by another aspect. But this is not a very important point unless there is a method of determining to what extent - 415 any recurrent activity, such as the kava ceremony contributes to the maintenance of structural continuity. It is probable, for example, that if, among all known societies, there exists political organisations of one sort or another, that for any society to exist there must be a form of political organisation. But this political organisation must be studied separately in each society in relation to other specific sections of that society and its exact relationship determined. In this ceremony the question that needs solution is, “Why should this ceremony have as its centre the drinking of kava rather than coconuts?” rather than a general statement like “the political organisation of the state as exemplified in the seating at the kava ceremony cannot be studied apart from the economic incentives required to assemble the food at a certain spot at a certain time.”

Another explanation of this ceremony is that put forward by the ethnologist, that a certain people brought the kava drinking complex with them from another area and that this complex remained in spite of possible subsequent invasions by a later people. The fact of diffusion seems quite clear but the amazing diversity of the ceremony in different parts of the Polynesian and Melanesian area puts forward the equally puzzling question of the mechanical process by which this diffusion occurs. It may well be that the solution to this problem lies in the realm of symbolism.

In human societies, a symbol is said to have a meaning when it presupposes knowledge of another object, idea or process recalled by the symbolic object, but which is not inherent in the symbol itself. In anthropology this connection between symbol and meaning is a social representation. For example Haddon states that, “Among the North American Indians the pipe is generally the symbol of peace . . . and the hatchet the common symbol for war.” 104 That is to say that the meaning of the pipe among the North American Indians is peace. But that there is no inherent quality of being a pipe which has the meaning of peace. It is the society of the North American Indians that associate the state of being at peace with a pipe. But symbols are not exclusively connected with inanimate objects but may be - 416 represented also by other facts in culture, gestures and so on. We have seen how in mixing the kuva in the bowl, the size of the root and hence the formality of the occasion are exemplified by the gestures of the kava mixer. Sometimes however, the meaning of a symbol is not so easily found out and may not be present in the minds of the participants themselves. It is very possible that the reason for presenting the kava root to certain peoples on certain occasions is not understood by the Tongans themselves. When this happens, one method of finding out the meaning of the symbol is by a comparison of the diverse occasions in which the same symbolic actions occur. 105

When a rite or series of rites are studied as in the case of this kava ceremony, we are faced with a number of different actions or symbols, the meaning of which has to be discovered. Having done this for some of the symbols we see that certain groups of symbols become associated together and it is only at this stage that the true function of the ceremony can be understood. Since the members of a group must adapt themselves to varying conditions in their social structure, the studying of function is really the study of the mechanism by which the members of society adapt themselves to changes in the geographical and human environment. If the environment changes so rapidly that the function of each varying part of society can no longer perform its job of keeping a certain minimum social integration, the society disappears. One method by which the anthropologist can discover that the function of an institution or custom is changing is by means of an objective study of the symbolism of a society. But in changing society, the varying meaning of symbols can only be discovered by studying the same society over a period of time and there have been few occasions when this has been done in the past. Consequently the first and most important task is to show to what extent each institution, custom or belief - 417 contributes to the maintenance of structural continuity. I have attempted to suggest a method by which this may be done in respect to this particular ceremony.

The fact that the kava people came to Tonga might have been largely accidental and the distribution of the kava ritual in the Pacific can accordingly only give the anthropologist a historical connection of invading peoples but cannot explain the equally important problems surrounding the symbolic nature of the kava root and the reason why its meaning is an essential aspect of Tongan social integration. Certain symbols have a tendency to congregate around sentiments and it seems as though societies have to express certain meanings although the outward symbols which express those meanings change. Were there people in Tonga before the introduction of kava, for example, the sentiment of dependence might have needed to be expressed in such a way that the kava root filled the symbolic requirements of the meaning of ritual inequality, or by its greater efficacy in this manner displaced any former symbol. It may well be that the conditions for cultural diffusion lie in this realm of symbolic social requirements. 106

Finally I wish to stress that this monograph is only an hypothesis and needs verification in a society where the theorist can actually ascertain many of the points here only derived from documents. The unfortunate part is that the number of peoples who are yet a sufficiently small unit to be studied comprehensively is ever growing smaller as they become more and more absorbed into the larger scale European social structure. Although it is not impossible to study this European social structure, it is very much more difficult to relate rites to this structure and it is this necessity that makes field work now so essential.