Home Page
Howard-Rensel Papers

5. Land Tenure and Social Change in Rotuma

Alan Howard

[Published in Journal of the Polynesian Society 73:26-52,1964]


The basic residential unit of political significance in pre-contact Rotuma was the ho'aga, which was apparently a kinship community. [1] The best early account of Rotuman society is contained in J. Stanley Gardiner's article, "The Natives of Rotuma," which was published in 1898. Even at the time of Gardiner's visit (in 1896) Rotuman society had changed to a considerable degree, and was then in a state of transition, but he was able to obtain sufficient information to make a fairly comprehensive reconstruction of the traditional organization. His account is consistent with most of the other available data and therefore appears to be quite reliable. Gardiner was also a man of unusual thoroughness and perception, having been trained in the natural sciences, a factor lending further credence to his account. According to Gardiner's reconstruction the name ho'aga ". . . applied to all the houses of a family, which were placed together, forming if the family was a large one, a small village; it is also applied to the family itself." [2] Each ho'aga was named and was governed by a titled chief, the fa 'es ho'aga.

One can only roughly estimate the number of ho'aga that existed prior to discovery. A comprehensive list of ho'aga names, collected by the District Officer in 1950, Dr. H. S. Evans, included 105 such names. many of which were no longer in use. On checking over Dr. Evans' list in 1960 it was found that informants added a few names, but also professed not to have knowledge of some already included. It is probable that ho'aga were born of expansion of some kinship groups and died of extinction of others, and the figure of a hundred active ho'aga units at any given time seems reasonable. If one assumes the island's population to have been between 3.000 and 4.000 in aboriginal times [3] ho'aga would have averaged between thirty and forty members each.

The ho'aga grouped together to form districts (itu'u). Each ho'aga was ranked within its district, and indications are that the highest ranking fa 'es ho'aga within each district acted as chief of the district (gagaj 'es itu'u). The order of ranking had both ceremonial and practical significance. Ceremonially, the order was recognized in the precedence of kava drinking on festive occasions [4] and in the seating arrangement during district meetings. Practically, the order coincided with degree of authority and a division of labour. The second ranking fa 'es ho'aga in each district was called faufisi. He acted as a lieutenant to the district chief and was known as the chief's "right hand". The faufisi was in charge of all ceremonial affairs involving the district as a unit, including the management of the important kava drinking ceremony. He was also the war leader in times of inter-district strife. In addition, the faufisi generally was in charge of one portion of the district, holding direct authority over several lower ranking fa 'es ho'aga. The third ranking fa 'es ho'aga was known as the chief's "left hand". He was usually in charge of the remainder of the district but he had no specific role in district affairs comparable to that of the faufisi. In the larger districts authority was sometimes sub-divided even further, with intermediate ranking chiefs exercising decision-making license over two or more low ranking ones. There was also a strong tendency for the overall division of labour within a district to follow ho'aga lines. One fa 'es ho'aga in each district was generally designated as tautei, his role being that of fishing expedition leader. Another was ordinarily in charge of the district kohea (kitchen), his job being to organize food preparation during district feasts. All of these positions were hereditary, remaining within the same family (i.e., ho'aga) unless a crisis dictated a change. Cowardice on the part of a faufisi, for example, or ineptitude by a tautei, might lead the district chief, with popular support behind him, to award the role to the holder of another title.

According to legend there were originally only five districts (itu'u): Itutiu, Faguta, Oinafa, Noatau and Malhaha. On two occasions, however, divisions took place, and at the time of discovery there were seven. Legend holds that a portion of the largest district, Itutiu, was given as a gift by the gagaj 'es itu'u to a visiting ho'aga chief from Oinafa. In this manner Itumuta was created. [5] A second story describes a war in which the district of Faguta was defeated by Oinafa, resulting in a division of the former district into two: Juju and Pepjei. [6]

Warfare between the districts was apparently frequent, with a somewhat institutionalized rivalry existing between two blocks consisting of Noatau, Oinafa and Malhaha on the one hand, and Faguta (later Juju and Pepjei), Itutiu and Itumuta on the other. Skirmishes were most frequently initiated by insults in the form of ceremonial neglect. At any given time the districts were ranked in priority, the particular order being influenced in part by the size and power of each district and in part by the results of the last war. The rank order was reflected in priority of ceremonial kava drinking, and the surest way to provoke a war was for a chief from a lower ranking district to instruct his faufisi to call him to be served ahead of a chief from a higher ranking district. This is only an example. Some breaches of etiquette were less contrived in nature. The goal of warfare was never, it seems, land aggrandisement. Gardiner's informants led him to the following conclusion:

"There were no advantages to be gained from the war by the winning side. The villages of the vanquished might be sacked, but they were seldom burnt; their plantations might be overrun, but there was little willful destruction. All pigs were, of course, regarded as legitimate spoil. The vanquished would perhaps promise to pay to the conquerors so many baskets of provisions or so many mats and canoes, a promise which was always faithfully and speedily performed, even though they might accompany the last part of the payment with a fresh declaration of war. The victorious side obtained no territorial aggrandisement, as it was to the common interest of all to maintain the integrity of the land, and the victors might on some future occasion be themselves in the position of the vanquished. Nominally first fruits were claimed by the victors from the chief of the vanquished, or perhaps the victors might depose the conquered chiefs, and put nominees of their own in their places." [7]

The Rotumans also had been experimenting with island-wide leadership for some time before discovery. Early European visitors met with a man designated as sau (loosely translated as "king" in the literature) who presumably held sway over the entire island. According to custom, a new man of chiefly rank was supposed to be elected to the position every six months (one Rotuman year), each district putting up a man in rotation. The sau was said to possess special magico-religious powers, and a man's success was judged on the basis of the island's prosperity during his term of office. If the island prospered his term might be extended. During his reign the sau was granted the privilege of demanding food and services from anyone on the island. If, however, his demands were too severe, they were more likely to be met with battle than compliance, and a sau was often called upon to back up his demands with force. In any case, it is apparent that the sau's authority was by no means accepted unconditionally, and that the institution had not really crystallized at the time of discovery. [8]

Each fa 'es ho'aga controlled the lands occupied by his group, being termed the pure (one who decides) over the lands. Concerning the problem of land tenure, Gardiner writes:

"No private property in land formerly existed; it was all vested in the pure for the time being of the hoag; the district generally had no rights over it. It usually consisted of four kinds: bush, swamp, coast, and proprietary water in the boat channel; common to the hoag, too, were wells and graveyards. Every member of the hoag knew its boundaries, which consisted of lines between certain trees or prominent rocks, posts, and even stone walls. In the bush land every hoag possessed property; it lay on the slopes of hills and in valleys between at some slight distance from the coast, from which it was separated by a stone wall, running round the whole island. On it taro, yams, bananas, plantains, and a few cocoanut trees were grown for food, while the paths into it and through it were planted with the Tahitian chestnut, the fava tree, and the sago palm. The Tahitian chestnut and fava were favourite boundary marks owing to their size and longevity. Swamp land is only possessed by Noatau, Oinafa, Matusa, [9] and Itomotu. It is low lying land, on extensive beach sand flats, which exist in these districts. The tide always keeps it wet, percolating through the sand, and in it is grown papoi, or broka, against famine. The possession of a good-sized strip always caused and gave to the hoag a position of importance; its boundaries were stones at the sides. Coast land lay outside the surrounding wall, to which the hoag had a strip from and including the foreshore. On it as near as possible to the coast the house or houses of the hoag were placed, while the rest of the land was planted with cocoanuts for drinking purposes. Hifo trees are stated to have been planted formerly to show the boundaries, but they more often now consist of stones or cocoanut trees, the ownership of which is a constant source of dispute. Districts and even villages were sharply marked off by walls down to the beach. All had the right of turning out their pigs on this land, and each hoag had to keep in proper repair the parts of the wall adjacent to it. Each had, however, usually an enclosure of its own land for its own pigs, when young. The proprietary water ran from the foreshore to the reef, a continuation of the strip on shore. At Noatau and Matusa, where it is very broad, it was to some extent cross-divided. It consists of a sand flat covered by 10-12 feet of water at high tide. On it fish of all sorts are caught by traps and various devices, and Shellfish are gathered. As these form no inconsiderable portion of the daily food, the value of this property was always very considerable. The reef--i.e., the part on the outside exposed at the low tide was the common property of all. It was explained to me that fish, crabs, etc., cannot be cultivated owing to the heavy breaking seas, but are sent up by the atua, or spirits. [10]

With regard to the role of the pure and the management of the land, Gardiner writes:

"The manager of this land for the hoag, its pure, is usually the possessor of the family name or, if he is too young, its oldest living member. His duty is to divide out the bush land year by year to the different households of the hoag for planting purposes, and to settle all disputes between its members. Further he has to take care that a proper number of cocoanuts are planted to take the place of the old trees, and to see that the walls are kept in proper repair. The swamp land is cultivated by the whole hoag, but if one part of the boat channel is especially fed by one member, she gets an especial right there. On occasions, when the whole hoag is interested, such as the repairing of the great wall on the island, the planting of the papoi land, or house-building, the pure has the power to call all its members out. The first fruits of each cultivated patch were brought to him: a basket of taro or yams, or a bunch of bananas. For all marriage or other feasts of any members of the hoag he was the head, and generally nothing could be done without consulting him. Over the land the chief of the district had no rights, except to order necessary repairs to fences or the keeping up of paths . . . Any encroachment on the land was very vigorously resented; it was referred to the district chief to settle, and his decision loyally adhered to." [11]

When a member of a ho'aga died, the land which he had under cultivation ordinarily reverted back to the ho'aga, i.e., the pure once again assumed primary license over its disposition. Under such circumstances there were no genuine problems of succession. A dying man was not, however without some rights of disposition. According to Gardiner:

"If a man's sons and sons-in-law were living and planting with him, on his deathbed he might apportion out the planted land to each, but the land was none the less under the hoag and subject to the payment of the first fruits to its pure. If he planted more cocoanuts than required by the hoag, he had the entire usufruct of these trees during his lifetime, quite independently of the land below them for planting. If in old age a man was neglected by his descendants or hoag, and taken care of by a stranger, he often gave him for his lifetime the usufruct of these trees and the crop of any plantation, he may have before his death cultivated; it only extended to the single crop; subsequent planting was not allowed." [12]

Upon the death of a pure the problem of choosing a successor was paramount. The pure of the land was also the ho'aga chief, and without a head man the group could not function as a social unit. Although Gardiner did not explicitly discuss the principles of succession it is not difficult to reconstruct them on the basis of later documentary evidence. [13] Four formal principles dominated the selection: (1) Initial priority goes to members of the deceased's pure's own generational level. i.e., his classificatory siblings. (2) The pure must be male. (3) The children of older siblings have priority over the children of younger siblings. (4) Elder siblings have priority over younger siblings.

A hypothetical genealogy should help to illustrate the operation of these principles:

Let us presume that A is pure and that he has no younger siblings. When he dies, he should be succeeded by his eldest son, B. When B dies, he is succeeded by his younger brother C, who in turn, is not succeeded by his own eldest son, but by D, the eldest son of his elder brother B. Upon D's death, his younger brother E is next in line, followed by F and G, the elder and younger sons of C. Female siblings and their offspring do not stand in the line of succession.

In addition to these formal principles, the following informal principles were undoubtedly given consideration: (1) The pure should be healthy and industrious--and therefore capable of vigorous leadership. This would be required because he was at once a work leader and the leader of the ho'aga in case of war. It is therefore probable that sickly persons and old men approaching senility were passed over in favour of more vigorous men. (2) The pure should be a "good man"--that is, someone who is kind, generous, and has a good manner. Rotuman custom prescribes that a good chief be kind to his people and take their desires into account. A man who was overly aggressive was likely to arouse defensive resistances and uncooperativeness. If the genealogically senior man was not the proper type of person, he was generally pressured out of succeeding to the family title. (3) Availability must also have been taken into account. If the man next in line had left the island and not yet returned, or if he had lived away from his home ho'aga for an extended period of time, he might well have been passed over for an available man.

Ho'aga membership seems to have been more a matter of residence than descent, in the sense that adopted persons and individuals married into a ho'aga were regarded as members as long as they lived in the community. A person no doubt also retained rights in the ho'aga of his birth and, after residing with his spouse in her community (or vice versa), was entitled to return to his home ho'aga and to sufficient land for subsistence purposes. It may well be that a person was able to claim land (at least residually) in several ho'aga, provided an appropriate genealogical connection to the pure existed, but we have no evidence on this point one way or the other.

There is also very little evidence detailing the dominant pattern or residence choice after marriage. Gardiner asserts that the dominant pattern was for the groom to move to his wife's ho'aga but states that this was not universal. [14] All the evidence, including the current Rotuman attitudes, indicates that the preferred pattern was uxorilocal, but there is no doubt that a number of other cultural principles altered this preferred pattern to a considerable degree. Among them were probably the following: (1) If the groom were of high status, and particularly if he stood to become pure of his own ho'aga, the couple would go to live in the groom's ho'aga. (2) If the bride's ho'aga was already overcrowded, while that of the groom was not, the couple would go to live in the groom's ho'aga. (3) If the marriage were not properly arranged (e.g., if the couple eloped or the girl became pregnant before the marriage had taken place) the couple would decide between themselves where to make their domicile. Since the girl's relatives might be shamed by her actions, in many cases a virilocal choice probably was made.

For the most part it was probably desirable for the ho'aga to recruit males, since its strength depended upon male labour and military capacity. It would therefore be advantageous, provided sufficient land was available, for the ho'aga to encourage their women to bring their husbands into the group, while at the same time it would be desirable for the ho'aga to maintain a cadre of in-born males, whose allegiance would be undivided. The residence patterns which must have emerged from this situation were probably very much affected by the "social economics" of supply and demand (of men and women), political expediency, and like considerations. It is also likely that residence after marriage was not necessarily fixed and that disputes and other vicissitudes of life resulted in temporary or permanent shifts.

If the analysis presented here is correct, the pre-contact Rotuman ho'aga had much in common with the Maori hapu and the Samoan aiga. It was a localized corporate group composed of a cadre of patrilineally related males and a number of cognatically and affinally related kinsmen. Perpetuation of the group lay in the hands of the patrilineal core, the family title being passed within the agnatic line. Since the supernatural spirits upon which the welfare of the group depended were the ancestral ghosts who "belonged" to the ho'aga, those with the most direct descent, the patrilineal core, must have been responsible for ritual propitiation and in charge of sacred activities for the group. Cognatic kin and affines, although participating in many corporate activities, were no doubt marginal to some and possibly excluded from others. [15] The latter probably also had a higher rate of inter-group mobility.

The above summary constitutes a reconstruction of what I believe were the essentials of social structure in aboriginal Rotuma. In the next section of this paper I will attempt to elucidate the processes of social change that resulted from European contact prior to the cession of the island to Great Britain in 1881.


Contact with European culture began with discovery of the island in 1791 and steadily increased throughout the nineteenth century. Among the ships that began to call on the island regularly were labour recruiters, mission ships and trading vessels. The labour recruiters found the Rotumans most eager to leave the haven of their homeland, and scores of young men were shipped to various plantation communities throughout the Pacific. Others willingly signed on as sailors. Mission vessels made stops at various intervals before mid-century, but it was not until 1864 that the first European missionary, a Methodist minister by the name of W. Fletcher, established a permanent mission station on the island. He had been preceded by some Tongan catechists of the London Missionary Society but they had made only little progress in converting the Rotumans. Not long afterwards, in 1868, some French priests established a permanent Catholic mission station on the opposite side of the island. Serious commerce did not begin on Rotuma until about the sixth decade of the nineteenth century, but once the value of coconut products was realized, trading vessels began to call regularly, picking up Rotuman produce for export and bringing in items of European manufacture.

One of the consequences of contact was depopulation. In this respect the Rotumans were affected much in the same way as other Oceanic peoples. The exact degree to which depopulation took place cannot be precisely measured, but that it was considerable was agreed upon by all later observers, and the population trends reflected in the census reports following Cession confirmed that such a process was still in operation. [16]

The decline in population is attributable to two separate but interrelated factors. The first is the extensive degree to which Rotuman men took advantage of the opportunities opened to them for emigration, either as labourers or as crewmen. Large numbers of young men left the island and although most returned after a few years of adventure, the early statistics show a tendency for emigration to exceed immigration. The Colonial Administration recognized the seriousness of the emigration problem and ordered an inquiry even before Cession had actually taken place. Arthur Gordon, who had been designated Deputy Commissioner, made a report on the situation in December, 1879. He found after taking a census of only four of the seven districts that 177 men were absent from the island. The chiefs were unanimous in their agreement that some kind of regulation was necessary to control emigration. [17] As a consequence of the inquiry Gordon and the chiefs passed two provisionary regulations, one prohibiting boys under sixteen to leave the island and the other restricting emigration to unmarried men over sixteen years of age, only with the permission of his district chief and for a period not to exceed one year. These regulations were not enforced, however, and emigration was never effectively controlled.

The second factor to which the decline in population can be attributed involved the combination of an increased death rate resulting from exposure to new diseases such as measles, dysentery and tuberculosis, to which the Rotumans had a low immunity; and a decreased birth rate, due in part to the decrease in the ratio of men to women resulting from heavy male emigration. According to Gordon's report approximately one-third of the 177 men missing from the island were married. [18]

It is apparent that the decline in population, and the decrease in available able-bodied men in particular, seriously affected the ability of some ho'aga to function properly as political and economic units. If one accepts the estimate of approximately thirty to forty persons per unit made in the previous section of this paper, ho'aga manpower was already spread quite thin. It is not at all unlikely, therefore, that in several cases emigration and death depleted manpower to the point where a ho'aga could no longer function as a social unit. The main technique for adapting to these circumstances seems to have been for previously distinct groups to consolidate. When this occurred, the ho'aga involved ceased to be kinship communities, or at least became more diffuse kinship groups. This hypothesis--that ho'aga consolidated--is supported by the fact that the present ho'aga in Rotuma generally include more than one chiefly title (as togi) although only one acts as headman, or fa 'es ho'aga. The case for consolidation is also supported by evidence gathered from elderly informants, many of whom professed to know of such instances having taken place. Some arrangement had to be worked out between the chiefs, particularly regarding the problem of succession. In some cases succession apparently remained, in theory at least, in the line of the senior titleholder; in other cases a system of rotation was worked out.

One consequence of this process is that land was no longer vested in ho'aga units alone. The consolidations that took place were for the purpose of achieving a more effective mode of social organization, rather than for uniting what would then be overabundant land resources. A situation must have been created, therefore, in which some ho'aga contained a headman who was pure over some of the land in his area, but not all, the remainder being in the hands of the headmen of the other amalgamating groups.

Another acculturation process, the growth of the commercial economy, completely altered the significance of land as a resource to the Rotuman people. Previously land had been used solely for subsistence purposes, and individuals had little interest in specific pieces of land. Within the ho'aga a system of shifting agriculture was used, and the role of the pure in reapportioning the land from time to time was probably to determine which fallow areas should be opened up and which areas in use should be allowed to fallow. Coconut trees were for the use of the whole ho'aga, although Gardiner mentions lifetime usufruct rights for those planted by a specific person beyond the needs of the ho'aga. [19] But as coconut products took on commercial value, the land upon which they grew correspondingly increased in value, and since coconut trees are a long term proposition, vested interests developed in specific blocks of land. If communalism had been strictly adhered to, the income from coconuts would have gone to the chief for redistribution, but no such pattern was established. Instead, each individual sold the products of the land on which he worked and kept the cash income to himself. Had the ho'aga been a more strictly defined kinship unit, such as a lineage or clan, the authority of the fa 'es ho'aga might have been sufficiently entrenched to perpetuate a system of communal tenure, but it seems that a pattern of strong authority was not characteristic. As a consequence, ho'aga land holdings tended to fragment under the pressure of individual interests.

The activities of the missionaries also resulted in deep-seated social changes. They were unable to grasp the concept of collective ownership as it applied in Rotuma and proceeded to treat land problems as though lands were individually held. By the time the missionaries established themselves changes in that direction had undoubtedly already taken place, but their selective interpretation of Rotuman custom, based upon the needs of the mission, was an accelerating factor. One of the missionaries' first tasks was to acquire land for church sites, and to do this the faithful were induced to make gifts to the mission. This was often done without the consent of other family members who shared rights in the land, and numerous disputes arose as a result. The giving as gifts of communal land to missions, and the building of churches on such property, was a significant factor in the generation of antagonisms between the Catholics and Methodists--antagonisms which culminated in open conflict in 1878.

Gardiner, in summarizing the changes that had taken place in the system of land tenure, cites some of the effects of missionization:

"Since the introduction of missionaries...much land has been seized by the chiefs, who, as a rule, in each district were its missionaries, as fines for the fornications of individuals. A certain amount of coconut oil was then given by the chiefs to the Wesleyan Mission, apparently in payment for their support. The mission in the name of which it was done, though generally without the knowledge of the white teachers, was so powerful that the hoag had no redress. The mission and chiefs obtained this power as the result of many wars waged against the adherents of the old religion; the confiscation of all the lands of the vanquished was proposed by the mission, but resisted by all the chiefs. Much land left to and bought by the Roman Catholic Mission is similarly situated; the individuals had no right to dispose of it without the consent of the whole hoag." [20]

By the time Cession took place, after which further sales or gifts of land by Rotumans to non-Rotumans was prohibited, the two missions had acquired a combined total of an estimated 132 acres of land. [21]

The combination of these three influences--the decline in population, the growth of the commercial economy and the effects of missionization--resulted in a major disruption of the traditional system of land tenure. The previous land-holding unit, the ho'aga, was completely altered in character. Whereas it had previously been a kinship group, it now became a territorially based socio-political unit. As a result land rights came to be dissociated from residence in a ho'aga, and the land itself was apparently divided up severally among the surviving members, each becoming pure over the area in which he planted and worked. This undoubtedly did not take place as a consciously executed plan, but rather as a gradual process involving a growth of vested interests in specific blocks of land, and a loss of authority on the part of the fa 'es ho'aga. The validity of this argument rests upon the postulate that authority and property rights in pre-contact Rotuma were linked to the system of kinship relations, rather than to an independent legal or political system. The evidence strongly suggests that this was indeed the case. Another basic change concerned the rules by which land rights were transmitted. Under the ho'aga system land had been inalienable, but along with the severance of the positions of fa 'es ho'aga and pure over the land, the right of the latter to dispose of land under his control came to be recognized. Individual pure began to treat land as private property, dividing it up and selling it, making gifts of it, and willing it to whomever they wished. As a result, ho'aga lands were fragmented into small separate blocks.

Despite the shift toward individuated land holdings the relationship between kinship and land rights was by no means eliminated. What happened is that land became a form of negotiable property, and since custom closely prescribed kinsmen's rights over one another's property, the operating principles were simply extended to include rights over land and the products from the land. The transmission of land rights thus came to be dominated by two sets of principles, one based on the pure's right to dispose of his property in accordance with his own wishes, the second based upon the rights of kinsmen over one another's property. It was this situation that prevailed when the British Government took over responsibility for governing Rotuma in 1879.


From the time that the Rotuman chiefs first made known their desire to cede the island to Great Britain, until notification of the Queen's formal acceptance, the administration of Rotuma was in the hands of the British Western Pacific High Commission. Mr. Arthur Gordon was appointed Deputy Commissioner and given the task of formulating a set of regulations to provide for the initial governing of the island. Gordon arrived in Rotuma in November, 1879, and held a series of meetings with the chiefs. On January 2, 1880, he called a meeting for the express purpose of framing some initial laws. As a result, laws relative to murder, assault, theft, quarreling, slander, and the buying and selling of liquor were passed. In order to provide for the enforcement of these laws Gordon, with the chiefs' consent, appointed three native magistrates, or gagaj ne pure (chief decision-makers), comprising two Methodist and one Catholic so as to correspond with population ratio of the two religious groups. He also gave instruction for two prison houses to be erected to punish offenders. [22] These laws were not originally written down because, no doubt, of the difficulties of translation, but Gordon stated his belief that the chiefs had "fully committed them to memory" and explained them to their people. He also had them explain that "whereas formerly each district had its own laws, now there is a general code for the whole island." [23]

In November, 1880, after the chiefs had been notified of the Queen's acceptance but before Cession officially took place, a proclamation appeared in the Royal Gazette providing for a temporary Constitution of Rotuma. Under its terms Rotuma ceased to be governed by the Western Pacific High Commission and became part of the Colony of Fiji. The Constitution included the following provisions:

"1. Existing laws and customs are to be observed and followed with necessary modifications, and the powers of the Magistrate are to be extended to foreigners and natives.

"2. The Laws of Fiji are to be followed as far as is practicable, with modifications necessitated by local circumstances (together with such Native Regulations as may be expressly extended to Rotuma).

"3. Ordinances and Native Regulations as under are to have force at once:

Quarantine Ordinance; Customs Ordinance and Tariff; Licensing Ordinance; Board of Health Ordinance; Native Regulation No. 2 of 1877 Respecting Courts; Native Regulation No. 12 of 1877 regarding Marriage and Divorce; Native Regulation No. 13 of 1877, regarding Births and Deaths.

4. A Council of Chiefs is to be set up, consisting of the Resident Commissioner as Chairman and the Head Chief and one Councilor of each District, but the Resident Commissioner is not bound to act on their advice.

"5. The Native Magistrates already appointed by the Chiefs under the High Commission are to retain their offices and functions subject to the approval of the Resident Commissioner.

"6. The sale and purchase of land except between natives of Rotuma is forbidden and invalid.

"7. The sale of spirits is prohibited.

"8. The recruiting of labourers to serve out of the Colony is prohibited.

"9. The procedure to be followed by the Magistrate and the Native Magistrates is to be as follows: Cases concerning Europeans are to be conducted by the Stipendiary Magistrate according to the procedure of the Stipendiary Magistrates Ordinance; and in native cases by the Gagaj ne pure following the procedure for Provincial Courts in Native Regulation No. 2 of 1877 concerning Courts. Any breach of the provisions of this Proclamation renders the offender liable to a fine of £50." [24]

The Constitution of 1880 remained in force, with the Council of Chiefs, under the chairmanship of the Resident Commissioner, exercising legislative powers, subject to the approval of the Governor, until March 18, 1882, when the Rotuma Ordinance dated January 18, 1882, came into effect. Among other things, the Ordinance provided that:

"1. There shall be a Court, called the Resident Commissioner's Court which shall be entirely independent of the Supreme Court of Fiji, and answerable only to the Governor in Council; in the case of death sentences the Court shall refer the case to the Governor and the Chief Justice conferring. The decision thus made shall be returned to the Resident Commissioner who shall then pass judgment.

"2. There shall be appointed for the island of Rotuma a board to be called "The Rotuma Regulation Board" consisting of the Commissioner and any number of native Rotumans not exceeding ten and not less than five, appointed by the Governor, or provisionally by the Commissioner subject to the Governor's confirmation." [25]

The duties of the Rotuma Regulation Board were defined as follows:

"To consider all such questions relating to the good government and well being of the natives as may be directed by the Governor or may seem to them to require their attention and the Board shall have power to make regulations upon any subject which may have been considered by them. Such regulations may contain provisions for the infliction of fines and punishments."

It was also stipulated that:

"All Regulations of the Board shall be laid before the Legislative Council and, if approved, shall be published in the Gazette and have the force of law." [26]

There was nothing in the Ordinance dissolving the Council of Chiefs, however, and it continued to operate as an advisory body, without legislative powers. In actual fact the influence of the Council of Chiefs superseded that of the Regulation Board, and the Board made into standing regulations what had been passed ultra vires by the Council. [27] The Ordinance also allowed for the continuance of the positions of Native Magistrates, but granted them no specific powers.

The administrative structure described above remained in force until 1927, when "An Ordinance to Make Special Provision for the Government of Rotuma" was passed. The new ordinance did not substantially alter the 1882 legislation, except that the Resident Commissioner's Court was required to submit evidence from cases in which an indictable offence had been committed to the Chief Justice of the Supreme Court of Fiji for verdict and sentencing. [28] This had the effect of giving the Supreme Court jurisdiction over Rotuma by making the Resident Commissioner a Commissioner of the Supreme Court. [29]

A number of further refinements were written into the Rotuma Regulations of 1939, but it was not until 1958 that the administrative structure received a thorough overhaul. The 1958 Ordinance abolished the Rotuma Regulation Board and granted the right to make regulations to a reconstituted Council of Chiefs, consisting of the District Officer, [30] who was to preside over meetings, the chiefs of the seven districts, one elected representative from each district, and the Assistant Medical Officer with the longest record of government service. [31] Questions coming before the Council were to be decided by majority vote, with the District Officer having an original vote plus a casting vote in case of a deadlock. [32]

It is vitally important that the history of this administrative machinery in Rotuma be understood as a background for the changes that have taken place in land tenure and social organization since Cession. In effect, the Colonial Administration assumed the responsibility for policy formation and decision-making in cases of dispute. Of special significance is the role of the government administrator in interpreting custom, framing regulations, and making court decisions. Until the 1958 Ordinance was passed the Resident Commissioner, and later the District Officer, held extraordinary powers--a situation markedly enhanced by the island's physical isolation from the rest of Fiji.


The first systematic inquiry into land tenure was made by Resident Commissioner C. Mitchell at a special meeting of the Council of Chiefs in March, 1882. The minutes of the meeting reflect the conglomeration of principles that were then being applied by the Rotumans to problems concerning land. They read as follows:

"Resident Commissioner: Are your customs regarding the holding of land and leaving it to others the same throughout the land or has each district its own customs?

"Chiefs: After a little discussion among themselves, said: 'The customs throughout the island relating to rights in land are the same.'

"Resident Commissioner: How do the different relations succeed to property in lands where a landholder has died without stating to whom he leaves his land, or has died while absent from the Island so that his relations do not know what his intentions were?

"Chief A: [33] His brother if he has any take charge of the land. They have their share in it as well as the children of the deceased; but the children must obey the voice of their uncles. The brothers come first and then the children. If the children are obedient to the will of their uncles at the death of the uncles the whole of the land goes to the children, but if the children grow up and can not agree with their uncles then the land must be divided and the children only get a small share each, the uncle or uncles retaining larger shares, which they can leave to whomever they please. Should the brothers of the deceased have children then the land must be divided and the brothers' children inherit along with the children of the deceased. Even if the children are grown up they must obey the voice of their uncle, and ask him for what they require off the land or the permission to build and reside thereon. This the uncle must consent to so long as the children claim only what is fair and reasonable. Where lands are left to anyone in particular his relations have no claim over the land simply because they are relations, though during his lifetime he is expected to provide for his wife and children as in other countries. A man may leave his lands to a friend or stranger where his relatives neglect him especially in his last illness.

"Resident Commissioner: What is your custom where a man has long been out of possession of land that he has a right in, whether from absence or otherwise and where and where [sic] another or others have held adverse possession of such lands?

"Chief A: No matter how long a man has been absent or kept out of possession of lands that he has a right to he can always claim them again.

"Resident Commissioner: Then there is no such thing as a man losing his rights to lands by long absence or adverse possession?

"Chief A: No. His right to the land is always good if it was good at first, no matter how long others have been in possession.

"Resident Commissioner: Supposing a man has left... and was absent forty or fifty years and during his absence others greatly improved his land by planting cocoanuts, etc., and when he returned he found many cocoanut trees bearing where formerly there had been none, would he have a right to take over the land and reap all these advantages without any consideration for those who during the long period of his absence had so greatly improved the lands in question?

"Chief A: He would take over all the lands on his return without consulting the interests of those who had improved the land. If he held along with others on undivided lands he still has his share .

"Resident Commissioner: What are your customs with regard to undivided land?

"Chief A: When a man dies and leaves his land to several persons he generally leaves one of them in charge or with the chief authority of the lands. The others all have claims in the lands but they must ask the person in charge what they require. The person in charge generally though not necessarily resides on the land. The others must ask the person in charge for permission to build or reside on the land or for any of the produce (usufruct) that they may desire to obtain. So long as the person in charge allows to the others their rights there is no difficulty and the land cannot be divided without his consent, but if he ignores their rights to a fair share of the produce, or to residence, or in any other matter they can call for and force a division of the lands in question, and then each one becomes sole proprietor of the share that falls to him. Where however lands are of small extent and the joint-owners are numerous it is often impossible to divide the land. The person in charge could not refuse to the others the right of residing on the land; of building a house or even a dancing house provided the land was large enough to allow this without monopolising more than their fair share of the land." [34]

Further testimony was given in a meeting held during the following month:

"Chief B: (in response to a question with regard to methods of forcing a division of family lands prior to Cession)...according to customs formerly if the right of division was refused one would die for it.

"Resident Commissioner: Who are included in the word 'Hoisasigi'?

"Chief A: Brothers and sisters and their children for three generations. Here it ends after which relations can only claim through those who as 'Hoisasigi' have rights on the land.

"Resident Commissioner: Can a man especially omit any of his children from rights in his land after his death?

"Chief A: If his children are obedient and attend to him in his old age, the land must be left to all of them.

"Resident Commissioner: What are the rights of adopted children?

"Chief A: Adopted children only have claims if specially mentioned by the person who may have adopted them.

"Resident Commissioner: At what age do children obtain rights in the land?

"Chief A: Children have no rights in the land until after their parents' death .

"Resident Commissioner: What claims has a wife on her husband's lands after his death?

"Chief A: None, unless he specifically mentions her." [35]

Extrapolating the general principles implicit in the chiefs' testimony reveals two basic premises governing land tenure at this time. The first might be stated in the following way: "An individual who is the pure over a land holding is entitled to use, allocate and dispose of the land, within discretionary limits."

The second principle derives from the kinship system and might be stated as follows: "The rights of an individual in a specific piece of land are determined by the nature of his relationship to the pure or a past pure of the property."

The latter premise depends on the operation of a set of rules governing rights of kinsmen in one another's property. These can be summarized as follows: "Children own rights in their parents' property." "Siblings own rights in one another's property."

The chiefs' testimony indicates that these two principles are rank ordered, with the rights of children being prior under normal circumstances to the rights of siblings. Two negative principles are also mentioned in the testimony: "Spouses do not own rights in one another's property." "Adopted children do not own rights in their foster parents' property."

Indications are, however, that both spouses and adopted children may be granted rights in the land by a willing pure. In other words, whereas the rights of consanguineal relatives (i.e., siblings and descendants) were regarded as institutionalized, those of spouses and adopted children were regarded as subject to individual evaluation. One further principle implied in the hearing is: "The rights of any relative in a pure's land may be revoked if he (or she) does not fulfill the obligations prescribed by custom."

It is evident that the determination of rights in land based on the foregoing principles is not likely to be unduly complicated so long as the recently deceased pure held undivided rights in the land, as would be the case if he had gained possession by purchase, gift, or by forcing a division of communally held land. But if the land remains undivided over a period of several generations matters become increasingly complicated. For three or four generations the majority of descendants can determine precise kinship ties with one another, and the relevant rules can be readily applied. It is this grouping that is signified by the word haisasigi. [36] In subsequent generations specific kinship connections tend to become vague or totally obscured between many of the descendants. Often relationship is presumed only because of a common claim in a specific piece of land. The presumption is that if two individuals both have rights in the same land they must have at least one ancestor in common and hence are related. The designation of such distant relatives is covered by the term kainaga which is more inclusive than haisasigi and includes all consanguineal relatives. The rights and obligations between distant kainaga is vague and unspecific. Under such conditions the application of relevant principles to land matters is difficult at best, and often impossible.

Two basic recurrent administrative problems emerged out of this situation. The first was the problem of determining succession to the office of pure. The ideal model which was described in the first section of this paper had by this time been revised to accommodate women, who, while not being favourable subjects for chieftainship of a political unit (ho'aga), were acceptable as pure over land. The following order of priority thus came to be accepted by the Rotumans: elder brother; younger brother; elder sister; younger sister; elder son of elder brother: younger son of elder brother; elder son of younger brother; younger son of younger brother; elder son of elder sister; younger son of elder sister elder son of younger sister; younger son of younger sister; elder daughter of elder brother. And so on. This obviously becomes hopelessly complicated with succeeding generations. Seniority comes to be a matter of opinion and conjecture--a situation opening the door for a considerable degree of manipulation. The circumstances were made even more complex by the fact that with the increased number of land holdings the number of pure positions proliferated until nearly every household head was pure over some land. This meant that even when the senior kinsman could be determined under the priority rules, he was likely to be established already on his own land and have no wish to move. No clear rules existed as to whether a person could or could not assume the rights of a pure over several blocks of land under such circumstances, but in fact many did.

The second problem involved the necessity for determining the degree of control a pure should be permitted to exercise over the land, and to what extent he was obliged to yield usufruct rights to others with a stake in the land. With close relatives the rules of kinship relations made these obligations clear, but with distant kainaga no such rules prevailed. It became a matter for the pure to decide, and thus a matter of personal discretion.

To the Resident Commissioners, carrying with them a European concept of ownership, these problems came to be regarded as an issue of individual versus communal ownership and the personal views held by the successive administrators greatly influenced the manner in which they dealt with land questions.

Mitchell himself reached the conclusion that individual ownership was the dominant pattern. In a report to the Governor he stated:

"The natives here hold their lands individually and not as communities. When a man is ill his relatives attend to him and he generally leaves his land to them, but should his relations neglect him during illness or old age he is at liberty to leave his land to any stranger or ordinary friend who attends to him, to the exclusion of his children and other relatives who may have neglected him. When a man leaves his land to his brothers and children, if the children are young, one of his brothers takes charge of the land and the younger members of the family have their shares generally in divided shares in the property.

"When any of them wish for any of the produce they go to the one in charge and he allows them to take what they require. Women hold land in their own right and when a man marries a woman who has land of her own he often goes and lives in her house if he prefers the place to his own." [37]

In actual practice, however, it soon became apparent that matters were far more complicated than that. Disputes and arguments were continually brought to the Resident Commissioner's attention, particularly over the issue of land sales. William Gordon, who succeeded Mitchell as Resident Commissioner in 1882, pointed out this problem in a dispatch:

"Since my arrival in Rotumah over two years ago a number of sales of lands and foreshores have taken place between natives of Rotumah, the price being generally paid either in money or pigs. Several land cases have been brought before the magistrates and myself in consequence of such sales having been made without the consent of members of the vendors' family, who were joint-owners of such lands or foreshores, along with himself." [38]

As a means of minimizing such conflicts Gordon proposed that a system of registration of land sales be inaugurated, and that

"...previous to registration, the right of the vendor to sell, or in cases of joint proprietorship, the consent of the other joint-proprietors to such sale, be inquired into; and the Commissioner decline to register any sale, in respect of which such right or such consent has not been established." [39]

As a result of Gordon's request the "Rotumah Land Sales Ordinance of 1885" was passed providing for just such a system, but during that year Gordon was replaced by R. Mackay, who developed a very different view. After spending nearly three years on Rotuma, Mackay wrote:

"...although having had the 'Rotumah land sales Registration Ordinance 1885' under consideration for a considerable time I have not intimated such a law to the people, for various important reasons.

"I have always thought the Bill ill-advised and impracticable. Although it fully provides for every justice being done to individuals, (so much so indeed that it would be difficult, without injustice, to register a single contract under the law) its publication would be an advertisement that would greatly unsettle their minds and only cause vexatious trouble, disagreement, and disappointment.

"There would be difficulty in devising a form of registration. The simple entering the name of the land would be valueless and fruitful of ultimate trouble and confusion." [40]

Mackay tended to emphasize the communal rights of relatives as opposed to the rights of the pure to sell or dispose of the land. He wrote that

"...family lands should on no account be permitted to be alienated from the family, even although none of the family members were absent from the island; and in many cases it is difficult to determine family from individual lands." [41]

To support his argument Mackay pointed to the antagonisms that were aroused by land sales:

"Knowing the fruitful crop of bitterness and ill feeling always created by any land sale I, for the last three years, have so strongly discountenanced land sales of any description, that, instead of its having been quite a usual occurrence, it now has become a casual event. I have often given the people to understand that land sales in which absentee share-holders are concerned could not hold good by law. The right of this the people thoroughly recognize. Land sales in Rotumah create bitterness of feeling and ill-will, which is hardly possible to exaggerate, and which time will never erase." [42]

Mackay's proposed solution was that the sale or gift of land be forbidden by law, and that no bequest be permitted of land to other than family members. [43] He did his best to persuade the chiefs in Council to denounce land sales, but initially at least, met with resistance. The minutes of the Council meeting held in August, 1886, are instructive as to the way the chiefs viewed the problem:

"Resident Commissioner: There is nothing but evil continually arising out of land sales...Considering the number of absentee land owners it would be almost impossible to complete a land sale on the island without injustice to someone and again, look at the number of children who have lost their fathers or mothers, or perhaps both, but still have their rights in their parents' land: how could they be permitted to give their consent to a sale? It would be well if you left your lands alone; and I am going to recommend the Governor to forbid any further sales between the people .

"Several Chiefs: We quite agree, Sir, with you in all the evils you mention. We only think of men who are getting old and have no children, but who have many lands. It would be hard if they may not sell, their lands are of no use to them, whereas other property is...

"Chief A: Why not be permitted to sell? It would look as if we had not full control over our lands. In England every land-owner can sell his land.

"Resident Commissioner: Are there any people on the island that have not enough land for their comfortable support?

"Members of the Council: There are...

"Resident Commissioner: What is the reason?

"Members of the Council: There are many that never get money and go down all the time; others have money and buy land all the time and come up.

"Resident Commissioner: If there were a man that had no land, what would be his position?

"Members of the Council: He would become a nobody. A working man for others.

"Resident Commissioner: Consider well, then, what has been said regarding this question, and do not forget that you have been appointed. not for your own aggrandizement, but for the welfare of your people.

"I now put the question whether you advise that sales of land shall be put a stop to, or not.

"Members of Council: We are all against putting a stop to land sales.

"Resident Commissioner: I am very sorry we are not of the same mind regarding this question. But it is a matter with which we can not positively deal. The Governor shall be informed of our opinion, and the question settled in Fiji." [44]

In this exchange we can see the conflict between the two premises clearly expressed. Commissioner Mackay was obviously concerned with protecting the rights of individuals in the land of their ascendants, while the chiefs were concerned that the decision-making rights of the pure not be eliminated. Mackay was persistent in his attack on land sales, however, and his persistence apparently bore fruit, because in November of the following year, in a dispatch to the Colonial Secretary, he reported that the chiefs spontaneously resurrected the issue as a result of some questionable sales that had taken place and unanimously passed a resolution to restrict land sales. [45] He expressed his satisfaction with the results of this resolution in the following year:

"It is now nearly twelve months since the people have been under the impression (which I have not disabused) that they are not at liberty to trafick [sic] with their lands, and results have more than proved the good I anticipated from prohibition of land trafficking." [46]

On the basis of Mackay's advice, Regulation 11 of 1898 was passed. entitled, "For the Better Conservations of Native Lands". It included the following provisions:

"1. It shall not be lawful for any native of Rotuma to sell or dispose of by gift any land, or to lease the same without the approval of the Resident Commissioner.

"2. It shall not be lawful for any native of Rotuma to devise any lands to other than members of his or her family, unless he or she shall be the sole owner of such lands.

"3. It shall not be lawful for any adopted child to be considered as a member of the family adopting it so far as inheriting by will shares in the family lands is concerned unless the person adopting it shall be the last of the family or sole owner of such land.

"4. It shall not be lawful to divide any lands held jointly or in common by the owners thereof without the unanimous consent of all the parties concerned (of an age to understand) be first obtained to such division, and no such divisions shall be made without the consent of the Resident Commissioner, and the lands so divided shall still remain family lands until there is only one of the family left. The Resident Commissioner shall take full notes of any such division, and shall file them in his office.

"5. Subject to the approval of the Resident Commissioner, it shall be unlawful for the pure or overlord of the family, for and on behalf of the members of such family (absentees excluded), and with the consent of the heads of such family, to lease any portion of the family lands for a period of not more than twenty-one years. All such leases shall have the Resident Commissioner's stamp on them as sufficient seal.

"6. Any person found guilty of a breach of the Regulation shall be liable to a fine of any sum not exceeding twenty pounds, or, in default of payment shall be imprisoned with or without hard labor for any term not exceeding six months." [47]

The emphasis on communal rights over the decision-making rights of the pure is readily apparent in the above Regulation. Of special significance was the provision that even though lands could be divided as a result of disagreement, they remained family property with all descendants continuing to own rights in all segments. The exclusion of adopted children from rights in their foster parents' land, even by will, may also be interpreted as a restriction on the pure's right of allocation.

Traditional techniques of problem-solving are not easily eliminated by law, though, and there is evidence that "underground" transactions continued to take place. Shortly after Dr. Hugh Macdonald was appointed Resident Commissioner in 1902, the whole problem of land sales was resurrected. Macdonald was far more lenient in his attitude toward land sales than Mackay had been, and gave a liberal interpretation to the 1898 Regulation. He was somewhat puzzled by the fact that the people preferred to make surreptitious exchanges while he was on leave rather than follow established procedures. [48] He explained the proper procedures in Council [49] and before long "legitimate" land transacting was revived as healthily as ever.

The 1898 Regulation remained in force until 1917 when Ordinance No. 1 of 1917, entitled "An Ordinance Regulating the Tenure and Disposal of Land in Rotuma" was adopted. The 1917 Ordinance contained essentially the same provisions as those present in the 1898 Regulation, except that it made specific provision for the right of the District Officer to summon any person to give evidence under oath "with regard to the right, title or interest of any person or persons in the land in question or otherwise, and for so doing the District Officer shall have all the powers of a magistrate under the Magistrates' Courts Ordinance." [50]

The 1917 Ordinance also included the following definition of land tenure in Rotuma:

"The basis of a land tenure amongst the natives of Rotuma shall be deemed and is hereby declared to be the holding by the natives of Rotuma of their lands by family communities according to ancient custom, the members of each family holding the land in undivided ownership and the acknowledged head of the family being the pure or over-lord of the land and each family community shall be deemed capable of selling, leasing or in any other way disposing of or dealing with any land or any estate or interest therein in Rotuma so far as ancient custom shall allow:

"Provided that no individual member of any family community shall be deemed to have, and it is hereby declared that he has not, any greater interest in any land the property of such family community than an interest for the term of his life." [51]

An analysis of the legal principles embodied in these early ordinances reveals a considerable lack of specificity. Indeed, each incoming government administrator was required to make his own judgments and was virtually free to apply whatever principles he deemed just. The Rotumans, for their part, found within these circumstances a great deal of room for manoeuvre and expedient manipulation.

The system operated on this basis until 1959, when the Administration, responding to a request by the Rotuma Council of Chiefs for a survey of Rotuman lands, passed "An Ordinance to Provide a Land Commission in Rotuma, the Registration and Dealing with the Transmission of Land and Matters Incidental Thereto." This ordinance represented an attempt to establish a workable code to govern land tenure in Rotuma. but virtually any attempt to arrive at a consistent set of principles would have resulted in restricting the flexibility and room for manipulation that the Rotumans had come to take for granted and when finally confronted with the Bill the people overwhelmingly rejected it. [52] The main point of opposition was the restrictions put upon the unlimited extension of bilateral rights. The Administration was faced, therefore, with the perplexing problem of permitting the system to continue to operate without a codified set of legal principles to insure consistent decision-making, and was forced to depend upon the arbitrary decision-making, powers of the District Officer to limit abuse. The successful solution to this situation is still, to my knowledge, problematic.


To grasp fully the changes that occurred as a result of the Colonial Administration one must also examine the structure of the land court and the manner in which disputes were resolved. The first land cases were dealt with in 1880 by an interim Deputy Commissioner, Hugh Romilly. Following the conclusion of these cases Romilly reported:

"The native disputes and land claims are all settled. I have not taken any active part in their settlement myself, and have left it to the three Nangagia (?) ne Pures as Native Magistrates to do. Of course I told the people that if they were dissatisfied with the decisions of the fairness with which the cases were conducted that they could appeal to me. In no case has such an appeal been made. As almost all the cases rested upon questions of relationship I thought it better to leave them entirely in the hands of the Native Magistrates as native relationships in Rotuma is a subject I have not been able to master." [53]

It was not long, however, before suspicions arose regarding the treatment of land cases by the Native Magistrates. Mitchell, in his annual report for 1881, mentions several appeals made to him against decisions by the magistrates on the grounds that the cases had not been sufficiently enquired into before a decision was reached. Inquiries revealed that this was indeed the case, and Mitchell decided to hear all claims personally, with the assistance of one or more of the Native Magistrates. [54]

Mackay, in his persistent effort to regulate land dealings, reported an innovation of his own in a dispatch to the Governor:

"I have made it a rule that any land cases of which there are not nearly so many as there used to be...are to be heard on Council days in the presence of the chiefs.

"By doing this, I think, there is less chance of a miscarriage of justice for there would be some among the chiefs that would give valuable and disinterested evidence; and besides many people would not set forth unjust claims before all the Chiefs of the Island as, perhaps they would under other circumstances. I have endeavoured to explain to the people--who seemed to think no one had a right to be present at a court--excepting those personally interested, not only that they had full liberty to come to court but that I would be glad of their presence." [55]

Two years later Mackay recommended the abolition of the office of Native Magistrates and suggested that two persons--one Catholic and one Methodist--be retained as assessors in land cases. [56] The recommended change was apparently adopted because H. Leefe, who became Resident Commissioner in 1891, reported that land cases were heard by himself. with two chiefs sitting as assessors. [57] This procedure was followed by most of the subsequent government administrators as a rule of thumb, but no hard-and-fast rule was established, and each administrator was at liberty to deal with land cases as he saw fit. Thus Dr. H. S. Evans, who became District Officer in 1949, sometimes sat with the Native Medical Practitioner as an assessor, but generally sat by himself without assessors. [58] F. Ieli, the Rotuman successor of Dr. Evans, preferred to sit with the gagaj 'es itu'u of the district in which the dispute took place, and the chief of the nearest neighbouring district to the disputed land. [59]

An analysis of the early land cases reveals the complexities that faced the Resident Commissioners in their attempts to settle disputes. There were three significant causes of dispute. (1) The fact that the Catholics had been driven out of certain districts after the conflict of 1878, and the Catholic parties wanted to re-claim their land in those districts. (2) A large number of cases resulted from the custom of gift-giving of land to non-relatives, leading to disputes between relatives of the original pure and those who either supposedly received the gift or were descended from the recipient. The fact that these gifts were made orally, and at times with only one or possibly no witness, complicated the matter. Sales of land under similar conditions brought similar results. (3) A further source of conflict stemmed from the retention of rights in property even though a person was absent from the island for a long period of time. It often happened that a person not consanguineally related to a previous pure (e.g., a brother-in-law) would live on the land for a period of time and his descendants mistakenly or falsely lay claim to it, in opposition to the actual descendants of the former pure. This problem was enhanced if a generation or two elapsed before the descendants of the original pure laid claim.

The first issue was settled relatively quickly by the Administration. Commissioner Mitchell held that the demands from the Catholics were perfectly fair and just according to native custom and he ordered the Methodists to permit them to re-occupy their lands. The Methodists had feared that the Catholics would build churches on their land, and were willing to permit occupancy provided the Catholics converted to Methodism, but Mitchell refused to permit such conditions to be imposed. [60] By April, 1883, William Gordon, who succeeded Mitchell as Commissioner, reported that "...within the past few months the Catholics have re-entered on their lands, in places from which, in the past, they have been excluded by the Wesleyans..." [61] By, the end of 1884 the matter was essentially closed. Gordon wrote: [62]

"Since the re-admission of the Catholics to their lands in the districts of Oinafa and Malhaha and parts of other districts, and the maintenance of the government of their rights to erect churches on such lands, every subject of discontent on the part of the Catholics has been practically removed. They have now no unjustly withheld rights to strive for."

Issues raised by the latter two problems were not so readily resolved. Written records were lacking and evidence consisted almost exclusively of opinion, unsupported claims and hearsay. An analysis of the decisions actually made reveals some interesting facts. For example, in an ownership (or pureship) dispute four decisions were possible: (a) a decision in favour of the plaintiff, (b) a decision in favour of the defendant, (c) a decision to divide the disputed land, and (d) a declaration of joint ownership. Table I shows the decisions made from 1881 to 1950, by decade: [63]

Table 1
Land Court Decisions in Ownership Disputes

Decision for:


Number of Disputes
Joint Rights

















It is especially interesting to note that during the first two decades in which land court was held, in those cases in which a decision was made for neither the plaintiff nor the defendant, the alternative of dividing the land was selected far more often than that of joint ownership. During this period forty-four out of forty-nine, or 90% of such decisions were made in favour of the former alternative. After 1900 the decision in favour of joint rights became more popular and between 1910 and 1950 seventy-nine out of ninety-two, or 86% of such decisions were in the latter direction.

The early trend toward land division had the effect of continuing the process of fragmentation that had begun with pre-Cession contact, but as the population began to increase during the 20th century the ratio of people to available land reached a point where fragmentation was no longer feasible. The reduction in land divisions resulted in a shift toward communal rights in existing blocks of land. With the opportunity of forcing divisions removed, the only ways in which a person could assume individual rights over a block of land was to buy it, receive it as a gift, or end up as the last remaining descendant of a former pure who held sole rights in the land. But land presumably could only be sold or given as a gift with the approval of all claimants, and with fewer and fewer groups dying out, the forces operating to create kainaga land out of individually held land were much stronger than those operating to create individually held land out of kainaga land.

As a result of this trend the degree of decision-making license that could be legitimately exercised by the pure was increasingly brought into question. As more land became kainaga land, and as fewer people could obtain land by forcing a division of family land, the problem of a pure's obligation to yield usufruct rights, especially to distant relatives, became prominent. There were no established solutions and each individual could support his point of view on the basis of previous custom depending on which aspect he chose to emphasize--the decision-making rights of the pure or the rights of descendants in their ancestors' property. The administrators, however, tended to regard the pure as a manager of the land--a logical consequence of the view that land was communally "owned" in the Western sense--and favoured the interpretation that pure were obliged to yield usufruct rights to persons with legitimate claims. The statement by Commissioner W. Carew in the Council of Chiefs in January, 1931, is indicative:

"It should be clearly understood that a Pure of family lands on which many people have rights, is in a position only of caretaker not owner, as some Pures seem to think. Nor have they any power to take for themselves all the profit arising out of these lands...Some of these Pures think they are 'over-lords' of the people as well as the lands, and as such suffer from what we call 'swollen heads'.

"In the future this type of Pure will be deprived of his office, and the family can elect another whom they think will act more justly." [64]

Despite the difficulties experienced by the Administration in dealing with Rotuman land matters, the people themselves managed to solve the majority of problems in their own way. It is only when difficulties arise that the Administration gets involved. Elsewhere I have discussed the way in which the contemporary system of land tenure functions, [65] and I will not reiterate here. However it might be pointed out that the current system works fairly well, as a consequence of the built-in ambiguities rather than in spite of them. These ambiguities provide the Rotumans with multiple opportunities, and in a round-about way, with a land security. Thus persons are reasonably sure that if they leave their island and some day want to return after an extended absence they can find a niche for themselves. What seems to happen is that the more conservative Rotumans remain on the island, or return to it after a short stay away. while the more progressive and energetic ones go to Fiji to try to make a place for themselves in urban areas. [66] Were land rights more tightly prescribed some of the conservatives might be forced to leave the island and would probably end up sponging off their urban kin in Fiji. On the other hand some of the progressives might end up with sole ownership of land in Rotuma and be caught in a dilemma--whether to stay in Fiji and lose their land or give up a good job and retain it. With the explosion of population that is taking place among the Rotumans the flexibility of the system allows a good deal of personal experimentation and therefore a maximum of opportunity for individuals to find an appropriate place for themselves with a minimum of insecurity.


Inevitably, the changes described in the foregoing sections were accompanied by changes in the social organization. Most obvious is the breakdown of the ho'aga as a kinship community. The ho'aga has survived as a social entity at least in name, but its composition and function are entirely different in contemporary Rotuman society. Ho'aga are now composed of neighbouring households, irrespective of kinship relations, which band together to perform large-scale tasks under the direction of the fa 'es ho'aga. The latter has no rights in the land of the constituent households, nor does he have authority in any affairs other than tasks initiated by either the gagaj 'es itu'u or the District Officer. In contrast to some hundred ho'aga units in pre-contact society, there are now approximately 700 to 800 land-holding kainaga [67] and about 400 persons acting as pure. Succession today tends to be a matter of turning the land over to someone who wants to use it, rather than to the senior man or woman. [68] More often than not the transference of pure's rights is informal and unrecorded, a fact which makes the system appear to administrators to be chaotic. Even the succession of family titles and chieftainship positions is now governed by a wide variety of social considerations rather than by a clear set of kinship principles. This is, perhaps, partially a result of the reduced desirability of being a chief, for with neither control of the land nor with authority over a corporate group of kinsmen, the role of a chief in Rotuma has become more of a burden than a privilege. [69] Put into other terms. since chiefly position is no longer highly valued, the rules governing its allocation have been allowed to lapse without causing undue strain in social relations.

Yet it would be erroneous to over-emphasize change and to underestimate persistence. Many of the rules remain the same: for example, the ideal rules for succession to chieftainship are identical to the ones described for the aboriginal society, the ideal residence rule is still the same (uxorilocal), and the ancestral spirits are still thought to insure justice with regard to the disposition of family lands. What has changed most is the complexity of the problems to which these rules are applied. Even in the traditional culture the ideal rules were undoubtedly modified by practical considerations, as has been pointed out, but in contemporary Rotuman society the quantity of these considerations has become so great that the ideal rules have lost their jural efficacy. The decision-making models used today have thus become more complicated and more fluid, but one cannot help suspecting that the basic ideas used in solving problems like succession, usufruct and residence are relatively unchanged, while the significant alteration has been in the conditions to which they are applied.


1 This paper is adapted from my Ph.D. dissertation, Howard 1962. I would like to acknowledge the assistance of the Government of Fiji who, despite unusually difficult circumstances, offered considerable assistance both in the field and in making available documentary information. Mr. Ian Diamond of the Central Archives in Suva was particularly helpful, and it would not be too much to say that this paper would not have been possible without his aid. Irwin Howard both assisted in gathering original data and in organizing manuscript materials. [back to text]

2 Gardiner 1898:429 [back to text]

3 Early population estimates range from 5,000 (Tromelin 1829) to 2,000-3-000 (Lucatt 1851). A review of the current ecological situation and an evaluation of the various estimates suggests a figure between 3,000 and 4,000. [back to text]

4 Within a district the second ranking man, the faufisi was customarily served first, after which strict rank order was adhered to. [back to text]

5 Churchward 1937-8:355-357. [back to text]

6 Histoire de Rotuma. Unpublished manuscript of the Sumi Mission Station, Rotuma. [back to text]

7 Gardiner 1898:470. [back to text]

8 The institution was singled out for attack by the Christian missionaries because of the pagan ritual associated with it. The last sau took office in 1870, after which the missionaries successfully induced the people to discontinue the custom. [back to text]

9 The name of one of the settlements in Itutiu. [back to text]

10 Gardiner 1898:483-484. [back to text]

11 Gardiner 1898:484. [back to text]

12 Gardiner 1898:485-486. [back to text]

13 cf. section IV. [back to text]

14 Gardiner 1898:484-485. [back to text]

15 A situation of this type is recorded by Groves 1962 for the Western Motu. He suggests, and I agree, that the pattern may have been widespread in Oceanic societies. [back to text]

16 The first census report was made in 1881, shortly after Cession took place, and showed a population of 2,491 Rotumans. By 1901 this figure had dropped to 2,061, and following the measles epidemic of 1911 fell below 2,000. [back to text]

17 Dispatch from Deputy Commissioner Arthur Gordon to the Governor of Fiji, December 4, 1879. Outward Letters. [back to text]

18 Ibid. [back to text]

19 Gardiner 1898:485. [back to text]

20 Gardiner 1898. [back to text]

21 Records of Land Titles Commission, Rotuma, 1882-1883: Preliminary Inventory No. 2, Central Archives, Suva. [back to text]

22 Dispatch from Deputy Commissioner Arthur Gordon to Western Pacific High Commissioner, February 4, 1880. Outward Letters. [back to text]

23 Ibid. [back to text]

24 Eason 1951:77-78. [back to text]

25 Eason 1951:78. [back to text]

26 Eason 1951:78-79. [back to text]

27 Eason 1951:82. [back to text]

28 An Ordinance to Make Special Provision for the Government of Rotuma. Government Press, Suva, September 1, 1927. [back to text]

29 Eason 1951:92. [back to text]

30 The government officer in charge of Rotuma was designated "Resident Commissioner" until 1934, after which the title "District Officer" was used. [back to text]

31 Ordinarily there is only one Assistant Medical Officer stationed on Rotuma. [back to text]

32 An Ordinance to Amend the Rotuma Ordinance, No. 4 of 1958. Government Press, Suva, April 11, 1958. [back to text]

33 In the transcript the chiefs are named, but for the sake of simplicity they are here designated as Chief A, Chief B, and so on. The chiefs' testimony was given in Rotuman and translated by an interpreter, while the Resident Commissioner's statements were made in English and translated into Rotuman for the benefit of the chiefs. The entire transcript is in English. [back to text]

34 Minutes of the Rotuma Council of Chiefs, March 23, 1882. [back to text]

35 Minutes of the Rotuma Council of Chiefs, April 6, 1882. [back to text]

36 Spelt Hoisasigi in Council Minutes. [back to text]

37 Dispatch from C. Mitchell to Governor of Fiji, February 16, 1882. Outward Letters. [back to text]

38 Dispatch from William Gordon to Colonial Secretary, June 6, 1884. Outward Letters. [back to text]

39 Ibid. [back to text]

40 Dispatch from R. Mackay to Colonial Secretary, November 7, 1887. Outward Letters. [back to text]

41 Ibid. [back to text]

42 Ibid. [back to text]

43 Ibid. [back to text]

44 Minutes of the Rotuma Council of Chiefs, August 5, 1886. [back to text]

45 Dispatch from R. Mackay to Colonial Secretary, November 11, 1887. Outward Letters. [back to text]

46 Dispatch from R. Mackay to Colonial Secretary, September 14, 1888. Outward Letters. [back to text]

47 Eason 1951:98-99. [back to text]

48 Minutes of the Rotuma Council of Chiefs, June 2, 1904. [back to text]

49 Minutes of the Rotuma Council of Chiefs, September 5, 1907. [back to text]

50 Ordinance No. 1 of 1917, Colony of Fiji. [back to text]

51 Ibid. [back to text]

52 In a round of meetings held in 1960, 85% of the people voted to return to the previous system whereby the District Officer exercised virtually full control of land matters. [back to text]

53 Dispatch from Hugh Romilly to Western Pacific High Commissioner, September 25, 1880. Outward Letters. [back to text]

54 Dispatch from C. Mitchell to Governor of Fiji, February 16, 1882. Outward Letters. [back to text]

55 Dispatch from A.R. Mackay to Colonial Secretary, April 19, 1885. Outward Letters. [back to text]

56 Dispatch from A.R. Mackay to Colonial Secretary, September 20, 1887. Outward Letters. [back to t ext]

57 Dispatch from H.P. Leefe to Colonial Secretary, January 17, 1894. Outward Letters. [back to text]

58 Personal communication with C. Legge, then Commissioner, Eastern, Colony of Fiji. [back to text]

59 Ibid. [back to text]

60 Dispatch from C. Mitchell to Governor of Fiji, April 12, 1882. Outward Letters. [back to text]

61 Dispatch from William Gordon to Colonial Secretary, April 12, 1883. Outward Letters. [back to text]

62 Dispatch from William Gordon to Colonial Secretary, November 24, 1884. Outward Letters. [back to text]

63 The statistics used in Table I have been compiled from a number of papers extracted from land court cases. The sources consulted are as follows: (1) Extracts from Land Case Decisions, 1881, 11/11/31, and Land Sales, Gifts, etc. to 14/6/17; (2) Extracts from Land Case Decisions, 27/1/32-24/6/41; (3) Land Cases, 7/7/41-21/9/53; (4) Land Dealings (in districts), 31/5/17-15/5/59; (5) Land Sales, 25/4/19-12/6/25. Sources 1,2, and 3 are successive and do not, among themselves, duplicate information. Source No. 4 extracts land case decisions by district and includes sales, gifts, leases, transfers, as well as disputes. There is not only much duplication in this of the first three sources, but on occasion duplication from one district to another. It does include, however, many cases not found in the first three sources. The fifth source actually goes up to 1953, not 1925 as stated in the title.

An attempt was made to define decision categories in the following manner: a decision for the plaintiff was registered in cases where a change of hands of pureship resulted; a decision for the defendant was registered which maintained pureship with the previous holder; a decision of land-division was registered when the land was divided from one or more holders to two or more holders; and a decision of joint ownership was registered in cases in which the defendant and plaintiff were both judged to be kainaga with equal rights, although pureship did not change hands. [back to text]

64 Minutes of the Rotuma Council of Chiefs, January 8, 1931. [back to text]

65 Howard 1963b. [back to text]

66 cf. Howard 1961. [back to text]

67 Estimate of the Land Commissioner, personal communication. [back to text]

68 For details of succession in Rotuma today cf. Howard 1963b:421-425. [back to text]

69 Howard 1963a [back to text]


CHURCHWARD, C. M., 1937-8. "Rotuman Legends." Oceania, 8:351-368.

EASON, W. J. E., 1951. A Short History of Rotuma. Suva, Government Press.

GARDINER, J. S., 1898. "Natives of Rotuma." Journal of the Royal Anthropological Institute, 27:396-435, 457-524.

GROVES, Murray, 1962. "Western Motu Descent Groups." Ethnology, 2.

Histoire de Rotuma. Unpublished manuscript of the Sumi Mission Station, Rotuma.

HOWARD, A., 1961. "Rotuma as a Hinterland Community." J.P.S., 70:272-299.

------ 1962. "Change and Stability in the Rotuman System of Land Tenure." Unpublished Doctoral Dissertation, Stanford University.

------ 1963a. "Non-traditional Leadership and Conservatism in Rotuma." J.P.S., 72:65-77

------ 1963b. "Land, Activity Systems, and Decision-Making Models in Rotuma." Ethnology, 2:407-439.

LUCATT, Edward, 1851. Rovings in the Pacific, 1837-49. London, Longman, Green and Longmans. Volume I.

Minutes of the Rotuma Council of Chiefs. Central Archives of Fiji and the Western Pacific High Commission, Suva, Fiji.

Outward Letters. Rotuma District Office, Central Archives of Fiji and W.P.H.C., Suva, Fiji.

Records of Land Titles Commission, Rotuma, 1882-1883: Preliminary Inventory No. 2, Central Archives of Fiji and W.P.H.C., Suva, Fiji.

TROMELIN, L. de, 1829. "Visite de l'Ile Rotumah." Journal des Voyages, 13.

Home Page
Howard-Rensel Papers