Previous Next

Degei’s Descendants: Spirits, Place and People in Pre-Cession Fiji

4. Understanding Traditional Fijian Society

This monograph is particularly concerned with variation between polities, as following a general geographical pattern. This pattern represents a broad continuum of polities of degrees of complexity, with the simplest in the western areas of Fiji, and the most complex in the east. Previous studies (Schütz 1962, Pawley and Sayaba 1971, Geraghty 1983a, 1983b) have contrasted the east and the west in terms of history, linguistics and mutual intelligibility of communalects. They have indicated a broad dichotomy between an eastern group of communalects and a western group, whilst pointing out that within each group there is a chain of communalects of differing degrees of mutual intelligibility. Pawley and Sayaba (1971) proposed a geographical and demographic explanation which explains a considerable amount, and Pawley (1981) added some social features. This study puts forward a proposal for a pan-Fiji continuum of differing degrees of complexity and stability of polities, taking into account the various internal and external factors already referred to as well as the patterning of communalects.

These factors affected different polities to different degrees at different times depending on local circumstances which were liable to vary. Such differences can sometimes be explained by recourse to the current oral accounts; whereas other differences, especially those involving outside pressures such as those from Tonga, could be explained by recourse to sources other than the current accounts and by taking into account the proximity of the polity to the source of the pressure.

Current understanding

The first aim of my investigations is to determine how Fijians currently understand and explain the origins, development and interaction of their various polities in pre-Colonial times.

To this end, current oral accounts were recorded in the course of discussions with representatives of all the polities in the three field areas comprising Rakiraki in North-eastern Viti Levu; Nadi/Nawaka/Vuda in Western Viti Levu; and the western archipelago of the Yasawa Group. For comparative purposes, the project took into account, first, polities in areas other than the study areas; and, secondly, inter alia, the itukutuku raraba, being the narratives of origins and movements of polities as recorded by the Native Lands Commission (NLC) at the turn of the 19th/20th centuries. This will be discussed further in this chapter. In the course of the research, a record was made of the current oral accounts of Fijian myths, traditions, histories and symbols associated with eighty-seven yavusa in the Nadi/Nawaka/Vuda and Yasawa areas, and with thirty-six yavusa in the Rakiraki area. These were recorded by myself first when District Officer in the areas in the early 1950s and later as a researcher at ANU in the 1990s. In compiling my own accounts, I was constantly referring to the itukutuku raraba (traditional accounts) of the NLC (n.d).

Common features and themes for polities

The term ‘polity’ refers to both yavusa or descent groups, and socio-political constructs or (con)federations, known as vanua (in the sense of vanua 2) and matanitū. An analysis of the oral accounts recorded indicates how those Fijians with whom the matters were discussed, currently understand and explain:

  • • the origins, structure and dynamics of polities in pre-Colonial times;
  • • the basis of pre-Colonial socio-political and military leadership;
  • • pre-Colonial intra- and inter-yavusa relationships; and
  • • the internal and external influences that brought about and affected such relationships.

From these accounts and from accounts of polities in other parts of Fiji, a number of features and themes emerged as common to the pre-Colonial polities studied. They flow throughout the monograph, and form the nerves and muscles of the discussions and arguments.

Common features

Several common features of polities emerged from the accounts of pre-Colonial polities both in the study areas and in other parts of Fiji. These accounts indicate that generally:

  • • each yavusa had myths of origin and ancestral spirits;
  • • each yavusa exhibited certain symbols of group unity and identification;
  • • each yavusa had a pattern of social hierarchy, and a recognised basis for leadership;
  • • a number of yavusa often joined together wholly or partly, and formed a socio-political federation referred to as a vanua;
  • • a number of vanua sometimes joined together and formed a socio-political confederation referred to as a matanitū;
  • • there were patterns of linkage and bonding between polities at all levels;
  • • there were recognised channels of communication between certain polities at yavusa, vanua or matanitū level; and
  • • the spirit world of ancestral and other spirits, such as war spirits and defending spirits, was important in:
  • • achieving and maintaining unity and ‘proper behaviour’;
  • • ensuring prosperity and continuity;
  • • validating group activities and appointments to chiefly office of leadership; and
  • • maintaining some degree of stability in a society faced with outside influences such as Tonga, Christianity, and European visitors and settlers.

Common themes

Two common themes emerge from the accounts of pre-Colonial polities in the study areas. They are the following contrasting sets of ideologies and realities:


(i) the ideology of social unity and integrity; and

(ii) the realities of social fusion and fission;


(i) the ideology of allocation of power based on the concept of inherited sau or mana (spiritual power); and

(ii) the reality of achieved kaukauwa or qwāqwā (secular power) based on the war club and the spear, leadership disputes and external leadership.

A key focus of the research has been to assess the linking of, and synthesis between, these two contrasting sets of ideologies and realities.

Social unity, fusion and fission

The current oral accounts may reflect an ideology of social unity and integrity at yavusa level, but they also indicate how a yavusa was seen to develop not only on the basis of natural increase but also through the realities of social fusion and fission. The latter came about not only internally through factors of ambition and jealousy, but also on the development of good and bad relationships with other polities. Current accounts of the past are also important to a yavusa as symbols of its unity and identification. They recount its supposed victories and successes which are a matter of pride, and its supposed defeats. The victories and successes are recounted in order to explain how a yavusa became the paramount of a federation of polities consisting of a number of yavusa. This could have been either by defeating its neighbours or by associating with weaker yavusa which came to it for protection or by forming military alliances with other yavusa. The defeats could have been included as an explanation for the yavusa's formal association with the strong, respected polity which defeated it and brought it under its authority. Currently the yavusa may well take pride in this association, however it may have developed in reality. In this way, the accounts can turn successes and defeats to the advantage of a yavusa in its position in current Fijian society.

Spiritual and secular power

The current accounts also indicate an underlying ideology of allocation of power based on the concept of maximum mana inherited through the most direct male line of descent from the original ancestral spirit. They also indicate how leadership may have been based not only on inherited mana and descent but also on the realities of internal leadership changes or disputes, as well as the acceptance (forced or voluntary) of external leaders. Such themes are relevant to the understanding of Fijian society past and present, and also, to a limited extent, to an exploration of Fijian society in the wider context of neighbouring Tonga, Rotuma and even Samoa as forming a quadrilateral interaction zone in late prehistoric and proto-historic times. There is archaeological evidence of such interaction before the period covered by this monograph. The forms of structure and leadership of Fijian polities are considered in the general perspective of neighbouring Oceanic polities. The project is also placed in the context of relevant literature and comparable accounts such as those recorded by the NLC, relating to Fijian polities, places and spirits.

Factors affecting variation in polities

By the period immediately preceding Cession in 1874, the polities in the Yasawa Group, the west and northeast of Viti Levu, and eastern Fiji, especially Bau, Rewa and Cakaudrove, had developed to different degrees of socio-political complexity. Of these, the simplest were generally in the west and the most complex were in the east. It is easier to understand any particular polity in relation to a continuum rather than a dichotomy of complexity.

The simplest form of polity was an independent yavusa, a group who claimed descent from a single original ancestor in spirit form (kalou vu) or in human form (simply vu); and who maintained its internal unity and its independence from any external authority. The most complex form of major polity was a matanitū or confederation of several vanua, or federations of yavusa, with a recognised leading yavusa and an accepted paramount chief. Such a confederation included a number of bati or military allies, and of qali or tributaries, being minor federations of yavusa (or vanua), or single yavusa. Such allies or tributaries might have been conquered or have sought protection from hostile neighbours or have simply recognised the socio-political advantages of a formal association with a strong and respected leading group and a powerful and generous paramount chief.

In between these extremes, investigation shows that there were forms of polity which manifested varying quantitative degrees of complexity of federation of yavusa or parts of yavusa or of groups of yavusa, and which experienced varying qualitative degrees of stability.

An analysis of available evidence based on current accounts has indicated that during the earliest period to which these accounts relate, the simplest form of polity, the independent yavusa, or descent group, was, generally, the earliest form of polity recognised. More complex forms, such as vanua based on a federation of yavusa, were usually a later development. However polities developed in the pre-Colonial period, they retained certain common features and themes as already described. In spite of these common themes and features, it is highly unlikely that there had ever been a golden age of homogeneity. Indeed, an analysis of current Fijian understandings of the structure and dynamics of pre-Colonial polities indicates a considerable variation in the degree of complexity of these polities. Variation was manifested in such factors as socio-political unity, and structure and dynamics including leadership especially as socio-political federations developed, at any rate towards the end of the pre-Colonial period. An analysis of the factors affecting the variation between polities can be undertaken from the point of view of whether they were internal or external to the system of polities being studied. Internal factors are those which can be considered within the parameters of the three elements of the ideological concept of vanua (in the sense of vanua 1); that is, spirits, places and people. These elements of vanua 1 permeated all forms of polity, including the yavusa or descent group and such socio-political constructs or federations referred to as vanua (in the sense of vanua 2) or as matanitū.

Internal factors: spirits, places and people

Variations in pre-Colonial socio-political unity and structure, in fusion and fission, and in federation and confederation, may have been due in part to factors internal to the polities. Such factors might have been ease of geographical access between polities, availability of planting land in the areas involved, insults and quarrels, and the expansive careers of ambitious and able leaders with military and naval powers, such as Cakobau, Vunivalu of Bau.

Another significant factor affecting the variations in the unity, structure and dynamics of polities is the extent to which they were connected with important centres of the spirit world, often based on dramatic natural features. For instance, the settlements of the various yavusa of the polities in the Rakiraki area lie just below the glowering Nakauvadra range and one of the most respected of Fijian spirit centres at the peak of Uluda. Many of the yavusa of the polities in the Vuda area were adjacent to the spirit path along the dramatic Tualeita Range and a respected spirit centre based on caves at Edronu at the west end of the Range. Uluda is associated with the spirit Degei and the other spirits who settled there with him and later spread throughout Fiji as progenitors of many yavusa. Edronu is associated with the first mythical arrivals in Fiji, many of whom went on and settled on the Nakauvadra range. The Rakiraki and Vuda yavusa have close traditional connections with these places. They gained considerable spiritual and political prestige from their close association with these spirit centres and from their spiritual connections with other yavusa whose progenitors had spread from these centres to other parts of Fiji.

Oft-recurring internal factors in the emergence of differences between polities as they developed are:

  • • local ambitions and quarrels within a yavusa leading to fission and fusion with other yavusa or to the establishment of a separate yavusa; and quarrels and rivalries;
  • • recognition by a yavusa that another yavusa was particularly strong and worthy of respect and subservience; and
  • • regional pressures from other yavusa with ambitious leaders wishing to expand their sphere of influence through the development of socio-political relationships or through warfare.

External factors

Variation between polities may also have been due once again to factors external to the polities being studied, such as:

  • • the influence of Tonga;
  • • the introduction of Christianity;
  • • the demands of European settlers and traders for land and security; and
  • • the need for the Cakobau Government to assert its authority outside Cakobau's traditional areas of authority especially in the independent west.

Oft recurring external factors in the emergence of differences between polities as they developed are:

  • • external pressures from expansionist Tongans; Christianity and overseas missionaries; Cakobau, first as Vunivalu of Bau and then in 1867 as European-crowned King of Bau, and later in June 1871 as proclaimed King of Fiji; and the Cakobau Government of Fiji established at the same time with the backing of some European settlers;
  • • the ideology of Tongan paramount authority and eventual monarchy, of which ambitious Eastern chiefs, especially of Bau, had had first hand experience during visits to Tonga;
  • • proximity to spiritual central places such as Uluda and the Nakauvadra Mountains or the cave complex of Edronu at the west end of the Tualeita Range, and the interplay between the spiritual unifying force derived from a common place of origin of ancestral spirits, and the secular sense of security derived from association with a strong and protective paramount;
  • • the degree of availability, strength and loyalty of allied military and naval forces which could enable an ambitious leader to expand his traditional sphere of influence and to maintain his position of paramountcy in the face of internal dissidence and external hostility; and
  • • the realities of kaukauwa or secular power as symbolised by the war club and the spear, and the ideology of sau or mana or spiritual power as a legitimising force.

As far as the east is concerned one of the most significant factors in the dynamics of polities was the highly ambitious and remarkably able chiefs and the military and naval forces at their command. A second was a series of external factors, namely the external influences of Tongan ideology of paramountcy; Tongan military power and expansion ambition; and the acceptance of Christianity, perhaps often largely in return for Tongan military assistance.

As far as the central regions are concerned the most important factors were the proximity to the main spiritual centre of the Nakauvadra as a unifying force; warfare to the south of Nakauvadra resulting in people being forced over the range to take refuge with the polities on the north side; and quarrels between ambitious leaders who gathered people together to assist them in their warfare.

As far as the western area and the islands were concerned the polities were noteworthy for the spirit of independence and local pride which prevailed, especially as regards the east. This spirit was symbolised by the myths of origin of several groups which relate not to the Nakauvadra but to the central spirit place at Edronu near the traditional first landing, where those who did not go on up to the Nakauvadra remained. Edronu was regarded in Vuda as more important than the Nakauvadra because it was an earlier and therefore more respected site. The external factors listed had a minimum effect on the west until the time of Cakobau and the sometimes-forced introduction of Christianity which was regarded as a political device of Cakobau and of the eastern polities. These outside influences were regarded with grave suspicion in the west as being likely to affect their independence, except perhaps in those areas in Vuda. Traditions of a very early arrival of Christians in Vuda and the apparently supporting evidence of ancient buttons carefully preserved in Viseisei Church cause the Vuda people to believe that they had become Christianised before the east.

Assessment of current accounts

Sources of data

The second aim of this research is to assess the reasonableness and accuracy of the recorded oral accounts of current Fijian understanding of the origins, and structure and dynamics, including leadership, of late pre-Colonial yavusa and the more complex forms of polity.

These accounts may be assessed by taking into account:

  • • evidence for internal and external consistency and inconsistency;
  • • early written records;
  • • limited archaeological evidence resulting from surface (and somewhat superficial) surveys; and
  • • linguistic information obtained from speakers of local communalects.

This is the basis of the evaluation in this research of the current oral accounts of the yavusa in the three geographical areas in the northeast and west of Viti Levu and in the Yasawa group. This evaluation focuses on questions such as whether the accounts have some absolute chronological and historical basis; or whether they are merely myths of origin and tales of half-remembered, half-fictional military and political successes aimed at validating current attempts at self-aggrandisement. In the assessment of the reasonableness and accuracy of current Fijian understanding, account is given of the consistency or inconsistency between accounts by different yavusa when referring to the same event. The following sources of data are also used:

  • • accounts recorded by the Lands Claims Commission (LCC) between 1875 and 1882; by the NLC at the turn of the 19th/20th centuries; and by written accounts by early visitors, missionaries and settlers as well as occasional personal accounts written by members of some yavusa and preserved in villages;
  • • information about the locations and features of about 200 archaeological and natural sites associated with these yavusa and their associated spirits as recorded in the course of my explorations; and
  • • basic linguistic information I recorded in the 1950s and the 1990s, in respect of eleven communalects recognised by the polities in the two mainland study areas, as well as data collected by me from various parts of the Yasawa group. The main sources of information about Yasawa communalects are from the work of Raven Hart on the island of Yasawa and of Andrew Pawley on the island of Waya.

Particular attention is paid to inconsistencies as well as consistencies arising from this use of data other than current oral accounts. Attempts are made to explain such inconsistencies, especially in regard to the extent to which current accounts either deliberately follow the NLC records or deliberately contradict them. Generally, even the earliest places mentioned in the oral accounts have been found by exploration to exist in the archaeological landscape. A comparison between features of such sites, such as the size and nature of spirit mounds, can be used to show a correlation between the importance of such a site according to the oral account and the characteristics of such features on the ground. People currently regard their communalect as a symbol of unity and identification. Investigations support the reasonableness of a communalect as such a symbol, and patterns revealed by the distribution of the communalects seem to show a parallel with the nature and patterning, either geographical or social, of Fijian society as revealed in the oral accounts.

An analysis of these sets of data reveals links between the spirit world, archaeological and natural sites, people and communalects. These sets of data can be integrated in ways relevant to a favourable assessment of the reasonableness and accuracy of current accounts of how Fijian society understands and explains the origins and the structure and dynamics, including leadership, of late prehistoric and proto-historic Fijian society.

Integrating the data

The multi-disciplinary and integrative methodologies used in this monograph are able to:

  • • use oral traditions to help to locate archaeological and natural sites;
  • • use the archaeological evidence of located sites to complement, and help to assess and evaluate the reasonableness of such oral traditions;
  • • compare the patterns, both social and geographical, of communalects and socio-political groups; and
  • • elucidate what are currently claimed to be the myths of origin of specific yavusa, integrating these myths with the evidence of oral history, archaeology and such symbols of identification and unification as ancestral and other spirits and their associated sites and manifestations, as well as particular totem-like categories of tree, living creature or food, and the jumping-off places of spirits of the dead.

The value of such an analysis of integrated data from several study areas is that the project is not dependent on simply one set of data, but on a series of integrated sets of data involving various disciplines. Such an interdisciplinary approach has been helpful in the final analysis of the information, not only from the point of view of those immediately concerned (that is, the living members of the groups subject to the investigation), but also from the point of view of the investigator as a sympathetic but critical outsider concerned with analysing the data from the various sources in order to assess the reasonableness and accuracy of current oral accounts.

When sources appeared to differ in explanation or in a material degree, such differences were first referred to informants, and the reasonableness of the varying accounts and the circumstances in which they were obtained were discussed. Where the accounts differed and the matter could not be resolved satisfactorily, both accounts were recorded with some argument for the preferred version, taking into account all the evidence available, even if such an account was to the disadvantage of the informants. For instance, Cakaudrove chiefs told me that they were descended from Tongans. The NLC records, Sayes’ (1982) doctoral research and my own investigations in Ra and southern Vanua Levu suggest otherwise. It is almost certain that the Tui Cakau and other leading Cakaudrove chiefs were in fact descended from some Rakiraki people who came to Cakaudrove and were singularly successful in helping those already there in their local quarrels. The locals, out of gratitude and recognising the strength of these Rakiraki people, made them paramount. Such an origin from a place of scant repute as far as the east is concerned would quite understandably be unacceptable, indeed insulting, to the proud Kings of the Reef.

Polity dynamics: general comments

My research concentrates on exploring those factors, both local and external, which generally affected the origins, development, structure and interaction of pre-Colonial forms of polities such as the yavusa or descent groups, and vanua 2 or socio-political federations, as well as the dynamics of their restructuring by fusion and fission. It investigates how pre-Colonial Fijian society, the geographical landscape and the Fijian spirit world were intermeshed. It identifies internal factors which created unity at the yavusa level and led to relationships of varying degrees of complexity between yavusa. The concepts of unity, identification and the sense of belonging; hierarchy, leadership based on descent, and reciprocal obligations; and spiritual support and sanctions are seen to be golden threads running through Fijian society. These threads were strengthened, tarnished, weakened or disentangled because of a variety of circumstances arising from time to time and from area to area. These circumstances often led to situations in which the protagonists involved needed to take measures for self-preservation and mutual protection. Such situations might have arisen out of a variety of circumstances. Investigations have focused on, first, breaches of propriety and failure to meet obligations; secondly, personal ambition and arrogance; thirdly, leadership based on factors other than descent, such as personal suitability; fourthly, military or political expansion based on secular force; and, fifthly, on the influences of external economic, military, political, spiritual and religious ideologies and forces.

The degree of interplay of these concepts in differing circumstances and situations shows variations in the complexity of socio-political structure. These variations were manifestations of a flexibility characterising, in reality, the structure and dynamics of pre-Colonial Fijian society. Whatever varieties might, however, have occurred in particular polities in Fijian society, there were also common features which formed the basis of such a society. These features are the spiritual, social and physical dimensions of the ideological concept of vanua 1.

The results of explorations of the oral accounts of the origins and development of yavusa recorded in the course of this project in the three main study areas indicate that there were general ideal principles and concepts underlying Fijian society before the Lands Commissioners made their investigations and pronouncements. Further, the basis of pre-Colonial Fijian leadership, at any rate during the two centuries or so before Cession was as much dependent on circumstances and achievement as it was on descent. In practice, the interpretation of, and degree of adherence to, such principles was varied and pragmatic. It emerges that the dynamics of late pre-Colonial Fijian society were, not unexpectedly, much more fluid and flexible than those of its Colonial period successor, the structure of the polities of which had been regularised by the NLC and the Colonial Government, and regulated by Colonial legislation.

Previous Next