Around 3000 BP bearers of Lapita became the first people to move beyond the intervisible islands of western Melanesia and settle Santa Cruz and the Reef Islands in the eastern Solomons and the archipelagos of Vanuatu, New Caledonia and Fiji. By 2800 BP there were Lapita settlements in Tonga and Samoa, 4500 km east of the Bismarcks.1
The Lapita dispersal across the southwest Pacific was astonishing in its speed and scale. Founding populations were established almost simultaneously in several previously uninhabited island groups. This must have involved the organising of many long-distance voyages carrying considerable numbers of people, including the building of ocean-going outrigger canoes, the recruitment of crews and passengers, and the transport of useful plants and breeding stocks of domestic animals.
The question arises as to what forms of leadership underpinned these achievements and whether historical linguistics in combination with the ethnographic record can provide clues, in the form of relevant lexical reconstructions. The following discussion will draw on a number of earlier studies which have addressed these questions, including Green (1994a, 2002), Hage (1999a, 1999b), Hayden (1983), Kirch (1984), Kirch and Green (2001), Lichtenberk (1986) and Pawley (1982a).
The Polynesian material is surveyed first because it yields a number of well supported Proto Polynesian terms to do with rank and leadership, indicating that Proto Polynesian society had hereditary chiefs in a system of ranked descent groups, with rank based on seniority of descent. We then turn to other Oceanic-speaking societies. Although the evidence there is less straightforward I will argue that Proto Oceanic society had a system of rank and leadership essentially similar to that reconstructed for Proto Polynesian. I will not discuss evidence for reconstructing terms for rank and leadership in Austronesian interstages earlier than Proto Oceanic. Blust (1995b:500) observes that relevant reconstructions for Proto Malayo-Polynesian are few and semantically ambiguous. Bellwood (1996) speculates on the role of hierarchy in the Austronesian dispersal.
The archaeologists Kirch and Green (2001) draw on linguistic, archaeological and ethnographic evidence to reconstruct elements of ‘ancestral Polynesian society’. They equate ancestral Polynesian society with the communities that spoke PPn. This equation may seem a straightforward matter but there are problems in determining precisely where and when PPn was spoken. Bearers of the Lapita culture settled Tongatapu, in the south of the Tongan archipelago, at about 2850 BP (Burley et al. 2011). Within a century or so Lapita settlements were established on the islands of western Polynesia to the north of the Tongan group, on Samoa, Niuatoputapu and ’Uvea and probably Futuna.2 The divergence between Tongic and Nuclear Polynesian presumably followed these settlements.
However, it seems that Tongic and Nuclear Pn remained part of a dialect complex and that innovations continued to spread across western Polynesia for many centuries after the initial colonisation. The innovations defining the Polynesian subgroup are so numerous that they indicate a millennium or more of unified development after it diverged from all other Oceanic languages. If so, some elements of the reconstructed PPn language must date to a period before distinct Tongic and Nuclear Pn dialects developed, and others date to a later time. I will refer to the former as ‘early PPn’ and the latter as ‘late PPn’. Late PPn society, being spread over several widely separated island groups in the Tonga-Samoa area, would have shown some regional variation.
The linguistic evidence for attributing hereditary chieftainship to PPn society is compelling. Almost every Polynesian language has a reflex of PPn *qariki, which can be roughly glossed ‘chief, person of chiefly rank’ and in most Polynesian societies chiefly rank is determined by seniority of birth. A number of derived forms based on *qariki are also attributable to PPn.
Initial *q, representing a glottal stop /ʔ/, is reconstructed in PPn *qariki because it is retained in the Tongan, Uvean and Rennellese reflexes. PPn *r is reconstructed rather than *l because *r is regularly lost in the two Tongic languages, while they retain PPn *l as l. PPn *r and *l merged as *l in Proto Nuclear Pn. The fact that reflexes of *qariki in Tongic and Nuclear Pn have undergone the sound changes characteristic of each subgroup, yielding Proto Tongic *ʔeiki and Proto Nuclear Pn *ʔaliki, indicates that the reflexes are inherited from an early stage of PPn, i.e. before Tongic and Nuclear Pn diverged, rather than being later diffusions across the dialect boundary between Tongic and Nuclear Pn.
| PPn | *qariki | ‘chief, person of chiefly rank’ | |
| Tongic | |||
| Pn | Tongan | ʔeiki | ‘hereditary chief, man or woman of chiefly rank’; ‘be of chiefly rank’ |
| Pn | Niuean | iki | ‘chief, any important person’ |
| Nuclear Polynesian languages of western Polynesia plus Pukapukan | |||
| Pn | Samoan | aliʔi | ‘chief, lord, man of noble birth’; ‘high chief as opposed to orator or talking chief’ |
| Pn | Samoan | aliʔi-taʔi | ‘be subordinated, subservient to’ |
| Pn | East Futunan | aliki | ‘hereditary chief, main priest’ (for expected ʔaliki) |
| Pn | East Futunan | aliki sau | ‘paramount chief’ (see §4.2.2.1) |
| Pn | East Futunan | aliki fenua | ‘village chief’ |
| Pn | East Uvean | ʔaliki | ‘chief, noble, lord’ |
| Pn | East Uvean | ʔaliki-ʔaŋa | ‘nobility, dignity’ |
| Pn | East Uvean | aliki sau | ‘paramount chief’ (see §4.2.2.1) |
| Pn | Tokelauan | aliki | ‘sacred leader, descendant of founding lineage’ |
| Pn | Tokelauan | kāinga aliki | ‘chiefly lineage’ |
| Pn | Tuvalu | aliki | ‘chief’ |
| Pn | Tuvalu | ulu-aliki | ‘head chief’ |
| Pn | Tuvalu | aliki-ŋa | ‘chief’s reign’ |
| Outliers | |||
| Pn | Anutan | ariki | ‘hereditary chief’ |
| Pn | Tikopia | ariki | ‘chief, clan head; leader’; ‘become a chief’; ‘chiefly’ |
| Pn | Tikopia | ariki fafine | ‘Female Chief (title of chief’s eldest daughter)’ |
| Pn | Rennellese | ʔagiki | ‘chief, headman, old gentleman’ |
| Pn | Rennellese | ʔagiki ʔeteaki | ‘chief-priest, priest under vigorous taboo during certain rites’ |
| Pn | Takuu | ariki | ‘hereditary chief and traditional religious leader’; ‘captain of a ship or leader of a canoe-based fishing expedition’; ‘function as an ariki’ |
| Pukapukan and Eastern Polynesian | |||
| Pn | Pukapukan | aliki | ‘chief, head of a major paternal descent group, priest by virtue of chiefly rank’ |
| Pn | Pukapukan | aliki wolo | ‘high chief’ |
| Pn | Pukapukan | aliki wui | ‘lesser chief’ |
| Pn | Rapanui | ariki | ‘chief, nobility, royal family’; ‘govern, reign, rule’ |
| Pn | Hawaiian | aliʔi | ‘hereditary chief, king, queen, noble’; ‘rule or act as a chief, govern, reign’ |
| Pn | Māori | ariki | ‘first-born male or female in a family of note; chief of a clan (hapuu); priest; leader’ |
| Pn | Mangarevan | aka-riki | ‘hereditary chief of district’ |
| Pn | Marquesan | haka-ʔiki | ‘hereditary chief, leader of a tribe (ma ʔeinaʔa)’ |
| Pn | Rarotongan | ariki | ‘high chief, ruler over a tribe’ |
| Pn | Rarotongan | ariki tumu | ‘paramount high chief; king’ |
| Pn | Tahitian | ariʔi | ‘head or principal chief of a tribal group’ |
The PPn term *qariki evidently served both as a noun ‘chief’ and as head of a derived stative verb *qariki-tia ‘be occupied by a chief or chiefs, have a chief or chiefs present’. Its derivation is of the type described by Pawley (2001), whereby PPn *-Cia (where *C is one of a number of consonants) forms a stative verb equivalent to an English passive participle: ‘chief-ed’, i.e. ‘be presided over by a chief’ or ‘have a chief present’.
| PPn | *qariki-tia | ‘be occupied by a chief or chiefs, have a chief or chiefs present’ | |
| Pn | Tongan | ʔeiki-sia | ‘(of a meeting, society etc) have a chief in it’ |
| Pn | Samoan | aliʔi-tia | ‘(of a village or meeting) be occupied by chiefs or distinguished guests, have a chief or chiefs present’ |
| Pn | East Uvean | ʔaliki-tia | ‘have a chief’ |
| Pn | Tokelauan | aliki-tia | ‘be occupied by chiefs or distinguished visitors’ |
| Pn | Tikopia | ariki-tia | ‘have s.o. as a chief’ |
PPn *faka- inherited two functions of POc *paka-. One function was to form a manner adverb, and this gave PPn *faka-qariki ‘(act) in the manner of a chief’.
| PPn | *faka-qariki | ‘(act) in the manner of a chief’ | |
| Pn | Tongan | faka-ʔei-ʔeiki | ‘in a chiefly manner, like or pertaining to a chief’ |
| Pn | East Uvean | faka-ʔaliki | ‘in the manner of chiefs, majestic, royal’ |
The other function was to form a causative transitive verb, giving PPn *faka-qariki ‘make into a chief’. Note that the E Uvean term is in both sets.
| PPn | *faka-qariki | ‘make into a chief’ | |
| Pn | Niuean | faka-iki | ‘treat as a chief’ |
| Pn | East Uvean | faka-ʔaliki | ‘make oneself master of, usurp’; ‘in the manner of chiefs, majestic, royal’ |
| Pn | East Futunan | faka-aliki | ‘name as chief’; ‘nomination as chief’ |
| Pn | East Futunan | faka-ali-aliki | ‘ennoble’ |
| Pn | Rapanui | haka-ariki | ‘proclaim as a king’; ‘become king’ |
| Pn | Māori | faka-ariki | ‘submit to orders’ |
Several other terms associated with rank or authority can be attributed to PPn and in some cases, to PCP.
The range of meanings associated with reflexes of PPn *sau suggests that it referred to the authority or power of a high chief. The Fijian evidence is consistent with this inference.
| PCP/PPn | *sau | ‘chiefly authority or rule’; ‘ruler, one who has authority’ (Green 2002; Kirch and Green 2001) | |
| Fij | Bauan | sau | ‘high chief’; ‘commandment or prohibition of a high chief’ |
| Fij | Bauan | sau ni vū-ni-valu | ‘war chief’ (vū-ni-valu ‘title of war chief’) |
| Fij | Bauan | vaka-sau-sau | ‘behave as a chief’ |
| Fij | Wayan | sau | ‘high chief or paramount chief who has been officially installed’; ‘ruler, one who has authority over the people’; ‘authority or command of a chief over people or events, the power of high rank’ |
| Fij | Wayan | sau-takini- | ‘exert one’s authority or power over s.o.’ |
| Fij | Wayan | sau ni vanua | ‘ruler or chief of a particular people or place (vanua)’ |
| Tongic | |||
| Pn | Tongan | hau | ‘secular chief, as opposed to sacred chief’; ‘champion, victor, conquerer’ |
| Nuclear Polynesian in western Polynesia | |||
| Pn | East Futunan | sau | ‘govern, reign, have supreme power’ |
| Pn | East Futunan | aliki sau | ‘carrier of the highest title’ |
| Pn | East Futunan | sau malū | ‘peace, time of peace’ (lit. ‘gentle rule’) |
| Pn | East Uvean | hau | ‘rule, ruler’ |
| Outliers | |||
| Pn | Tikopia | sau | ‘select a man for office by traditional process of grasping him and elevating him as chief’ |
| Eastern Polynesian | |||
| Pn | Tahitian | hau | ‘peace, government’; ‘reign’ |
| Pn | Tongarevan | hau | ‘peace’; ‘be in peace, be settled’; ‘the local government, judges and police’ |
| Fij | Rotuman | sau | ‘king, kind of sacred chief’; ‘be king’; ‘royal, pertaining to a king’ (possibly a Pn loan) |
| Pn | Rennellese | sau | ‘divine gifts, abundance of gifts from gods’ |
A PPn term *laŋi having to do with chiefly status is reconstructable. Its precise meaning is uncertain but the broad definition proposed by Green (1994a) is consistent with the range of the reflexes.
| PPn | *laŋi | ‘one of sufficiently high rank to be honoured or treated as one in authority’ (Green 1994a) | |
| Tongic | |||
| Pn | Tongan | laŋi | ‘royal tomb, raised and terraced burial place of sovereigns’ |
| Nuclear Polynesian in western Polynesia | |||
| Pn | Samoan | faʔa-laŋi | ‘address or refer to s.o. by his ceremonial title’ |
| Eastern Polynesian | |||
| Pn | Hawaiian | lani | ‘noble, royal, exalted’; ‘high-born aristocrat, very high chief’ |
| Pn | Hawaiian | lani aliʔi | ‘royal chief’ |
| Pn | Hawaiian | hoʔo-lani | ‘treat s.o.as a chief’ |
| Pn | Māori | raŋi | ‘chief, generally as a term of respectful address: sir’ |
| Pn | Rarotongan | raŋi | ‘supreme in authority, highest authority, power’ |
Many Malayo-Polynesian languages of the Philippines, Indonesia and Malaysia reflect a PMP form *datu. Its reflexes have widely varying glosses, including ‘chief’, ‘headman’, ‘leader’, ‘lineage priest’, ‘head of family, ‘grandparent’, ‘prince’, ‘king’, ‘title given to a sovereign’, ‘term of respect to any man of standing’, ‘shaman’, ‘ancestor in the female line’ (ACD). It is tempting to see a connection to Samoan lātū ‘person in charge of an undertaking’ (Milner 1966); ‘head builder’ (Pratt 1911) and to Fijian rātū ‘honorific particle and title of rank before names of males’ (Capell 1941). However, the resemblance may be accidental. No possible cognates have been reported from any non-Central Pacific Oceanic language. The Central Pacific forms have two long vowels, which suggests that they originally consisted of two morphemes. Furthermore, Geraghty (pers. comm.) advises that the use of Fijian rātū as an honorific before names of men of rank is an innovation of the colonial era. Its older use was as a prefix to a chiefly title in certain regions of Fiji.
Kirch and Green (2001:232-236) reconstruct PPn *fatu kāiŋa which they tentatively gloss ‘leader of the kāiŋa’ [a land-holding residential group, local kin group]. They acknowledge that reflexes of *fatu have a diverse range of meanings in daughter languages, which makes semantic reconstruction uncertain. The range includes: (1) ‘head of family’, (2) low-ranked chief’, (3) ‘master, owner of an estate or resources’, (4) ‘person of mature age, elder’. Reflexes of the compound *fatu kāiŋa sometimes mean ‘senior person or carer in a family, one who oversees use of resources such as food’, sometimes ‘land owner’ or ‘overseer’ and in one case, ‘be a married person’.
| PPn | *fatu, *fatu kāiŋa | ‘leader of the *kāiŋa’ (Kirch and Green 2001) | |
| Tongic | |||
| Pn | Tongan | fatu taŋata | ‘middle-aged man’ |
| Pn | Tongan | fatu fafine | ‘middle-aged woman’ |
| Pn | Niuean | patu | ‘a lowly ranked chief (not an iki) and head of a family group’ (p for †f) |
| Nuclear Polynesian in western Polynesia | |||
| Pn | East Futunan | fatu taŋata | ‘man who has arrived at a mature age’ |
| Pn | Tokelauan | fatu kāiŋa | ‘the obligation of caring for one’s family (kāiŋa) in the traditional way (of both men and women)’ |
| Pn | Tuvalu | fatu kāiŋa | ‘be a married person’ |
| Outliers | |||
| Pn | Tuvalu | fito-ariki | ‘chiefs (collective)’ (unexpected i and o for †a and †u) |
| Pn | Nukuoro | foto-oliki | ‘chief of island’ |
| Eastern Polynesian | |||
| Pn | Hawaiian | haku | ‘lord, master, overseer, owner’ |
| Pn | Hawaiian | haku ʔāina | ‘land owner or overseer’ |
| Pn | Hawaiian | haku hale | ‘landlord, house owner, host, hostess’ |
| Pn | Mangarevan | ʔatu motua | ‘name of a founding ancestor’ |
| Pn | Marquesan | hatu | ‘master, owner of an animal’ |
| Pn | Tahitian | fatu | ‘lord, master, owner’ |
| Pn | Tongarevan | hatu | ‘lead a work’; ‘lord, owner’ |
| Pn | Tongarevan | hatu henua | ‘land owner’ |
| Pn | Rarotongan | ʔatū | ‘lord, master, owner, landlord’ |
| Pn | Rapanui | hotu matuʔa | ‘primary founding ancestor of Rapanui’ |
It is tempting to propose a reconstruction PCP, PPn *tuqi ‘ceremonial title of the paramount chief of a region’, occurring before the name of the region. This term has reflexes in Tongan, Futunan, Uvean (all tuʔi), Samoan, Niuean and the Fijian languages (all tui). However, it is not reflected in Eastern Polynesian or Outlier languages and it is likely that it spread relatively recently, i.e. in post-PPn times, across western Polynesia and Fiji.
If ancestral Polynesian society had descent groups what were they called?
In several Micronesian languages reflexes of the term *kainaŋa denote a clan or sib, i.e. a unilineal descent group:
| Mic | Chuukese | kaiɾaŋ | ‘matrilineal sib’ |
| Mic | Puluwatese | yayinaŋ | ‘clan’ |
| Mic | Woleaian | xairaŋ | ‘clan, tribe’ |
A cognate term occurs widely in Polynesian with a range of meanings. The most extensive discussion of the semantic history of this term is Marck (2010). In Tikopia and Anuta it refers to a non-exogamous patrilineal descent group consisting of several exogamous lineages, and in Pukapukan it denotes a matrilineal sublineage. This close formal and semantic agreement with the Micronesian cognates supports attributing *kainaŋa ‘unilineal descent group’ both to PPn and to the stage of Oceanic ancestral to PPn and Micronesian.
However, in several Polynesian languages (Tongan, E Uvean, Rennellese, Hawaiian, Marquesan and Manihiki) the reflex of *kainaŋa refers not to a unilineal descent group but rather to the subjects of a chief, the common people. Marck (2010:617) argues that PPn developed over a long period of time across the islands of western Polynesia and that PPn society and language showed regional variation. He attributes the sense ‘clan, unilineal descent group’ to an early stage of PPn. This was retained in a few Nuclear Pn daughter languages, but in some PPn communities unilineal descent groups were replaced by cognatic descent groups, in which membership could be claimed through either a male or female ancestor, and *kainaŋa was used to refer to these groups. In late PPn times *kainaŋa came to have an additional sense, as a collective term for the populace, the common people of a place, headed by a chief.
| PPn | *kainaŋa | (1) ‘clan, unilineal descent group’; (2) ‘cognatic descent group’; (3) ‘populace, commoners’ (Marck 2010) | |
| Tongic | |||
| Pn | Tongan | kainaŋa | ‘populace, people without chiefly rank’ |
| Nuclear Polynesian in western Polynesia and Pukapukan | |||
| Pn | East Uvean | kainaŋa | ‘people not of chiefly rank’ |
| Pn | Pukapukan | keinaŋa | ‘maternal sub-lineage, headed by its oldest member’ |
| Outliers | |||
| Pn | Tikopia | kainaŋa | ‘clan, a non-exogamous descent group consisting of several exogamous lineages’ |
| Pn | Emae | na-kainaŋa | ‘titled person subordinate to a given person’ |
| Pn | Rennellese | kainaŋa | ‘subject of a chief’ |
| Eastern Polynesian | |||
| Pn | Hawaiian | maka-ʔainana | ‘commoner, populace, people in general’ |
| Pn | Manihiki | mata-keinaŋa | ‘common people, tribe’ |
| Pn | Mangarevan | mata-kainaŋa | ‘assembly, congregation’ |
| Pn | Rarotongan | mata-keinaŋa | ‘a tribe, people of a district, headed by an ariki’ |
Most Polynesian languages have a term or terms for a commoner, a person who is not of chiefly lineage. The literal meanings of such terms fall into three main classes: (1) derogatory terms, e.g. ‘worthless person’ or ‘person of low degree’, (2) those that refer to ordinary people, the populace, people of the land, (3) those that refer to a person of a junior lineage. Although on distributional grounds it seems likely that PPn had one or more terms denoting a person of low rank, no well-supported PPn reconstructions can be made from these comparisons. The partial agreement between Tongan kai fonua, E Uvean kai fenua, and Samoan tū-fanua and tau-fanua points to a PPn compound *N + fanua referring to commoners, where N is a noun denoting person or people.
| Pn | Tongan | tuʔa | ‘commoner, person without chiefly rank’; ‘be common, not of chiefly rank’; ‘(of a chief’s relative) be of lower lineage’ |
| Pn | Tongan | kai fonua | ‘commoner’ (lit. ‘person of the land’) |
| Pn | Niuean | taŋata fakateaŋa | ‘commoner’ (lit. ‘careless person’) |
| Pn | East Futunan | seka | ‘commoner, untitled person, of low birth’ |
| Pn | East Uvean | kai-fenua | ‘commoner, peasant’ (lit. ‘person of the land’) |
| Pn | Samoan | tū-fanua | ‘person of lower rank, commoner’ (fanua ‘land’) |
| Pn | Samoan | tau-fanua | ‘commoner’; ‘owner of land’ |
| Pn | Samoan | taŋata lautele | ‘commoner’ (lautele ‘common, ordinary’) |
| Pn | Nukuoro | ŋati taŋata | ‘ordinary person, commoner, one of slight social value’ (ŋati ‘worthless, empty, useless’) |
| Pn | Takuu | tanata vare | ‘commoner’ (vare ‘ordinary, not special’) |
| Pn | Tikopia | faka arofa | ‘commoner, person not of chiefly class’ (lit. ‘calling forth benevolence, sympathy’) |
| Pn | Rennellese | taŋgani peŋgea | ‘commoner’ (lit. ‘useless person’) |
| Pn | Māori | tūtūā | ‘mean, low-born’; ‘person of low degree’ |
| Pn | Māori | marahea | ‘commoner’; ‘be of low degree’ |
The brief glosses given to reflexes of *qariki in dictionaries of Polynesian languages do not provide satisfactory descriptions of the role of chiefs in Polynesian societies. For more detailed information we must look to ethnographic accounts.
Various types of political systems were found in Polynesia at first European contact. There is an extensive anthropological literature on the evolution of these types. Most such works before about 1980 rely on a comparative typological method, based on the geographic distribution and frequency of types and the logic of transformations between types. Important studies relying chiefly on comparative typology include Williamson (1924), Burrows (1938), Sahlins (1957, 1958, 1963), Koskinen (1960) and Goldman (1957, 1970).
Most of the above-mentioned works take little account of the comparative method used in historical linguistics to reconstruct the sound system and lexicon of the common ancestor of a family of related languages and to construct a family tree (or subgrouping) for the family. However, in recent decades a number of studies have drawn on evidence from historical linguistics to support hypotheses about the history of Polynesian political systems. Among these are monographs by Kirch (1984) and Kirch and Green (2001) and papers by Pawley (1982a), Kirch and Green (1987), Hage (1999a,b) and Marck (2010).
Goldman (1970), a 600-page work comparing 18 Polynesian societies, distinguishes three main types of political systems in Polynesia which he calls ‘traditional’, ‘open’ and ‘stratified’.
In Goldman’s sample the traditional type of political system is represented by Tikopia, New Zealand Maori, Futuna, Uvea, Tongareva, Manihiki-Rakahanga and Pukapuka. Notes on Tikopia and New Zealand Maori follow.
Tikopia is a small high island (4.6 km²) with very limited arable land. Its population of around 1250 at first European contact tested to the limit the carrying capacity of the environment. Nevertheless, as described by Firth (1936), it exemplifies perfectly the essential characteristics of conical clan societies. Tikopia’s population was divided into four ramified patrilineal descent groups or clans (kainaŋa), each headed by a senior chief. The four clans themselves were ranked from senior to junior. Clans divide into paito ‘ramages’ (lit. ‘houses’) consisting of related families tracing descent from a common ancestor a few generations back. There are chiefly families (paito ariki) and commoner families (paito fakaarofa). These are ranked according to traditional ritual privileges and obligations associated with the beliefs in the gods. Commoner families divide into families that have a ritual leader as head (paito pure) and families that have no significant ritual privileges. Firth (1936:313) observes that
differences in the ordinary social position of chiefly and commoner families are not great. They all own lands, they mingle freely, exchanges take place between them on a basis of general reciprocity, there is no “chief’s language”, as in Samoa or Java, kinship terms are used between them, and nowadays intermarriage takes place freely between their members.
Deep genealogies were kept by chiefly lineages. As the genealogically senior man in a descent group, a chief had sacred powers deriving from his mana (mystical power to make things happen) and responsibilities to his people. He communed with the gods on behalf of his kin, for example, to ensure economic prosperity. The chief had the ability to control a ritual circulation of goods. He is the initiator of the grand cycle and is responsible for keeping in motion the distribution and redistribution of food. He was the giver of feasts on his own account and the sponsor of all kainaŋa-wide and island-wide rituals, which also involve energetic consumption. All who gave to him were honoured by being literally drawn into the lines of connection that run between chiefs and gods. For the most part, the chief’s bounty was directly in the interests of his community, for ritual, and for public works.
Despite great differences in geographic context, Maori and Tikopia societies resemble each other quite closely in their systems of rank and leadership. Maori tribes
are classic examples of aristocratic organisations in which the basic principles of primogeniture, seniority of descent, graded rank, sanctity of chiefs, and the sanctity of the male line are well established but which show none of the specialised features – achieved chiefly position, sharp social stratification, and political centralisation – of the Open and Stratified societies (Goldman 1970:30).
Chiefly status was hereditary. Descent groups were ranked by seniority of birth of the founding ancestor. The first-born son of the senior clan outranked lesser chiefs, who came from junior lines. Deep genealogies were kept by chiefly lineages. As the genealogically most senior man in a descent group a chief had sacred powers deriving from his mana and responsibilities to his people. However, a lesser chief from a junior line could extend political power beyond his own clan and add to his mana by demonstrating leadership in secular domains, such as warfare, diplomacy and organisation, which gave him prestige beyond his own lineage.
A very few Polynesian societies, most notably those of Hawaii, Tonga and the Society Is., had at the start of the colonial era a feudal system in which a chiefly class ruled over a class of commoners. Land was held by the chiefs, with commoners as tenants. Marriage between nobles and commoners was proscribed. This extreme degree of stratification most likely developed independently in each of these societies and reached its zenith at the beginning of the colonial era. We will consider the Tongan case.
With a land area of only 696 km², the Tongan group is much smaller in geographical extent and poorer in natural resources than the Hawaiian (16,600 km²) and Society (1535 km²) groups. It contains one relatively large island, Tongatapu, with its satellites in the south, and two clusters of smaller islands in the north, the Ha’apai and Vava’u groups. Nevertheless, Tonga at first contact was densely populated (estimated 40,000) and had developed a highly stratified, feudal society in which all the land was owned by an aristocracy ruling a commoner class of landless tenants, who served as the chief’s labour force and provided warriors in times of war.
Tongan aristocracy combined ranking by seniority of birth in the male line, with a pattern of bilateral ranking based on superiority of sisters and sisters’ sons to brothers and brothers’ sons. The first factor established rank, the second shaped deference and respect.
Chiefly lineages had a classical Polynesian conical clan organisation: major lineages branched into lower-order segments, with succession to titles in the higher-order lineages strictly patrilineal. The supreme chief was semi-divine and sometimes referred to as a god. Adherence to primogeniture was strongest in the highest-ranking lineages. Junior chiefs from outside the leading families could compete to succeed to a title.
In theory, all chiefs were related by descent from an original supreme chief, the Tuʔi Tonga. However, Gifford (1929) could not find genealogies outside the royal line going back beyond about AD 1600. The archipelago was politically unstable between 1600 and 1800, with frequent assassinations of paramount chiefs. Oral history records that Tonga was seldom under a single chief. Rather, a number of high chiefs ruled separate domains. The secular authority of the Tuʔi Tonga was limited to his home island, Tongatapu. Outlying islands owed him first fruits but not political tribute.
ʔeiki is a broader category of chiefly rank below tuʔi. It was extended by courtesy to all kin of a chief. A distinction was made between high chiefs (’eiki motu, ʔeiki taupoto or ʔeiki toho) and petty chiefs (ʔeiki siʔi).
Chiefs made large ceremonial displays of food, the burden of supply falling on the tenants who served as the labour force. Under the Tuʔi Tonga were four high chiefs who served as ministers with ritual duties. Matāpule were ceremonial attendants of chiefs, the highest-ranking matāpule being himself a chief with authority over lower-ranking matāpule.
The Tuʔi Tonga was too sacred to fight. A cadre of professional warriors (toa) drawn from matāpule families were part of the royal court. High chiefs were not priests. Priests performed minor duties at shrines and served offerings to the gods. Their standing depended on rank of the lineage and its gods.
As examples of open societies Goldman cites Samoa, Niue, Marquesas, Mangaia and Rapanui. Although he places Futuna in the traditional category it shows some features of his open class and I include it here. Notes on Samoa and Futuna follow.
Samoan society lacked ramified descent groups. The core kinship unit was the ʔāiŋa, an indefinitely extended bilateral kindred with branches representing hierarchically ranked segments. Every extended family was headed by one or more high chiefs (aliʔi paʔia). Each high chief was paired with a junior chief, tulāfale, who served as orator (‘talking chief’) and came from the female line. There was some recognition of primogeniture and descent lines and patrilineal succession but primogeniture played a weak role in the assignment of chiefly titles. The highest ranked titles were transmitted through the male line (tama tāne). The female line (tama fafine) had the power to curse, through the father’s sister.
Administrative powers in a village were held by the fono, village council of chiefs. The transmission of high ranked chiefly titles was under the jurisdiction of the fono. The chiefly title of tāupōu ‘village maiden’ was held by a high-ranking female who had charge of the aualuma ‘association of unmarried women of a village, with various formal duties’. The Great Fono was a conceptual body, which never met, representing all of Samoa.
The title of tui, followed by the name of a district, was given to the paramount chief of that district. In pre-European Samoa ten districts were recognised. Before the Tongan invasion (possibly ca AD 1200) the Tui Manuʔa title held the highest rank in Samoa. The person and belongings of the Tui Manuʔa were sacred and dangerous, but political and economic authority was lodged in local autonomous districts. In the early 19th century a chief of the Malietoa family had conquered the whole of Samoa but his power was limited to rank and title, not extending to economic control.
The wealth of chiefs was measured in terms of valuable goods, especially ʔie toŋa ‘fine mats’, siapo ‘bark-cloth’, and ʔoloa ‘valuable goods’. Samoa, with its three large islands, had extensive areas of fertile land, allowing production of a food surplus. The distribution of food in feasts accented the social position of chiefs.
The patron-craftsman relationship stood outside the traditional system of genealogical rank. Craftsmen held the title of tufuŋa. The highest status crafts were those of canoe builder, carpenter and tattooer. Craft titles were transmitted in descent lines.
Chiefs were priests, taula. Some priests were concerned with family gods, higher order priests with great gods.
The Futuna or Hoorn group consists of two islands, Futuna and Alofi, that in early colonial times supported a population of 3000-4000, most living on the larger island, Futuna. Land shortage was a pressing problem often leading to war.
There are three levels of rank: aliki sau ‘high chiefs’, aliki ‘sub-chiefs’ and seka ‘commoners, including untitled members of chiefly lineages’. By colonial times the main island was divided into two districts each with its own aliki sau. Each clan in turn divides into a number of ramages (kutuŋa), each having its own chief (aliki or launiu), with the paramount chief being from the highest ranked ramage. The landholding groups are kāiŋa, a term that refers to both the group of landowners and to the segment of land they own.
Seniority of birth was important but not the major factor in qualification for chiefly office. Chiefs came from important families but were selected on the basis of their accomplishments. Distinguished warriors (given the title of toa) were potential chiefs and sometimes usurped royalty.
Chiefs were expected to move freely among their people and to work in the gardens and as craftsmen. Commoners owed their chiefs respect but not servility. But chiefs were set apart in various ways. They were spokesmen for the ancestral gods. They were the instigators and organisers of war. Certain foods – turtles and certain birds and fish – were reserved for them. The fasu institution (special rights given to sister’s children) was restricted to chiefs in earlier times. They had highest status in the kava circle. A village council (fono) dealt with mundane village affairs, chiefs with more elevated matters, such as warfare, and the imposing and lifting of tapu. Symbols of rank were important. There is an honorific vocabulary, used in addressing chiefs.
Kirch (1984) and Kirch and Green (1987) conclude that ancestral Polynesian society had chiefdoms corresponding to the ‘traditional’ type, exemplified by Tikopia and New Zealand Maori, and that those Polynesian societies with very high or very low levels of stratification represent later developments. In this respect they are in accord with some previous commentators on the evolution of Polynesian societies, such as Sahlins (1957, 1963), Koskinen (1960) and Goldman (1970).
Following Sahlins, Kirch and Green argue that the organisational basis of ancestral Polynesian society was the conical clan, a group claiming descent from a common ancestor, ranked and segmented along genealogical lines. Descent groups of this kind are typically associated with terms for same-sex siblings distinguished by seniority of birth (PPn *tuqakana ‘older sibling of same sex’, *tahina ‘younger sibling of same sex’; see §3.5.1.5.1–2). Kirch and Green (2001:231) define PPn *qariki as ‘the senior male, titled leader of a social group, probably the *kainaŋa, who typically inherited his position patrilineally within the senior ranked line of this group, and who acted as the group’s secular as well as ritual leader.’
By virtue of their inherited rank *qariki had mana ‘supernatural efficacy, the power to make things happen’ and were tapu ‘sacred, set apart by taboos’. A chief had important sacred duties to perform relating to the well-being of his subjects, such as ensuring success in growing crops and in communal fishing activities.
Kirch and Green (2001:231) propose that their extended definition of the roles of a priest-chief in ancestral Polynesian society allows one to trace, in some detail, transformations that have occurred in particular societies after the breakup of the PPn speech community. In more conservative societies, chiefs retained both their secular and sacred roles and continued to be associated with descent groups. In many Polynesian societies, however, a functional separation between sacred and secular roles developed. This was particularly so in Eastern Polynesia, corresponding to the widespread importance of a priestly class. Yet another kind of transformation accompanied the breakdown of the ancient *kainaŋa-type social groups and their replacement with strictly territorial groupings, in which chiefs became the owners of such land units: Hawaii and Tonga both exemplify this kind of change.
Dye (1987) is sceptical of Kirch and Green’s (1987) definition of the role of PPn *qariki, remarking that their reconstruction “is modeled on the rights, duties, and modes of succession associated with chiefs of contact-era societies in possession of full land situations” (Dye 1987:445-446). He doubts whether these institutions could have existed in the pioneering phase of settlement, when a founding community of fewer than 100 people occupied an island covered with virgin forest.
However, PPn was not the language of a small first-generation, founding population but of communities which had lived in western Polynesia for many centuries, during which time the innovations defining the Polynesian subgroup developed and spread across this region. The initial colonisation of the different islands in the Polynesian Triangle and of the Outliers must have involved many cases where a small founding population settled uninhabited islands, and it is striking that similar beliefs and practices associated with chieftainship survived these foundation events. That indicates that the founding populations carried the ideology with them.
We turn now to a consideration of rank and leadership in Proto Oceanic society. Reconstruction of the lexicon of POc is less straightforward than is the case for PPn. An etymon can be attributed to POc if it is reflected in (a) at least two first order subgroups of Oceanic or (b) in an Oceanic language and a non-Oceanic Austronesian language.
A problem is that the first-order subgrouping of Oceanic is not clear-cut (see Figure 1.2). There are several candidates for first-order subgroups in western Melanesia: the Admiralties, Mussau and Western Oceanic, and several in southeast Melanesia and the central Pacific: Southeast Solomonic, Temotu, Micronesian, Vanuatu, New Caledonia-Loyalties, and Central Pacific. To meet criterion (a) a conservative procedure is to require that an etymon be reflected in at least one subgroup in NW Melanesia and one outside NW Melanesia.
In an influential and controversial paper Sahlins (1963) sought to explain the larger size of polities found in certain Polynesian societies, running into the thousands, compared with the smaller scale polities, ranging from 70 to 300 people, characteristic of many societies in NW Melanesia. Central to his explanation are differences in the nature of rank and leadership.
The Polynesian advance in political scale was supported by an advance over Melanesia in political structure […] The characteristic western Melanesian “tribe”, that is, the ethnic-cultural entity, consists of many autonomous kinship-residential groups. Amounting on the ground to a small village or a local cluster of hamlets, each of these is a copy of the others in political status. The tribal plan is one of politically unintegrated segments – segmental. But the political geometry in Polynesia is pyramidal […] Smaller units are integrated onto large through a system of intergroup ranking, and the network of representative chiefs of the subdivisions amounts to a coordinating political structure. So instead of the Melanesian scheme of small, separate and equal political blocks, the Polynesian polity is an extensive pyramid of groups capped by the family and following of a paramount chief (Sahlins 1963:287).
Commentators have pointed out that a simple Melanesian/Polynesian dichotomy in systems of rank and leadership is not justified (Chowning 1979; Douglas 1979; Hage 1999b; Sand 2002). Forms of hereditary leadership based on rank are found in various Oceanic-speaking communities of Papua New Guinea, e.g. in the Admiralty Islands (Mead 1934; Otto 1994), in Wogeo in East Sepik Province (Hogbin 1970), in Manam in Madang Province (Wedgwood 1934), in Mekeo (Hau’ofa 1981), Motu (Seligman 1910) and Roro (Seligman 1910) in Central Province, in the Trobriand Is. in Milne Bay Province (Malinowski 1922), in Hahon in Bougainville (Blackwood 1935), and parts of the Solomon Is., e.g. Arosi (Fox 1924), Baegu (Ross 1973) and Sa’a (Ivens 1927), in various societies in Vanuatu (Facey 1981), and New Caledonia (Douglas 1979; Guiart 1963) and throughout Fiji, as well as in most societies in Micronesia (Shimizu 1987).
Sahlins (1963:286) acknowledges that “in eastern Melanesia, New Caledonia and Fiji for example, political approximations of the Polynesian condition became common” but he finds that political structures showing an extensive pyramid of genealogically ranked groups are characteristically Polynesian. Extensive regional trading systems linking scores of local communities have developed in various parts of western Melanesia, but these produced “at best, an ephemeral sort of political association” (Harding 1970:111).
There follow notes on a small selection of Oceanic-speaking societies of Melanesia that have a form of hereditary chieftainship.
Douglas (1979:16-18) describes traditional New Caledonian socio-political organisation as follows:
[It] centred on localized, patrilineal and patri-virilocal clans, and on tribes, aggregates of clans paying allegiance to the chief of one of their number […] clan members claimed common descent and owed allegiance to a hereditary chief, in theory the genealogically senior man of the senior lineage; a tribe was regarded as a group of related clans with a common founding ancestor, which paid allegiance to the senior male member of the senior lineage of the original clan. The chief was a clan’s ‘great (first-born) son’ […] Neither the ideology of common descent nor that of chiefly seniority was necessarily borne out in practice, however […] Both clans and tribes commonly absorbed unrelated individuals and groups, and rationalized the process in kinship terms. Ideology was thus satisfied, but common residence was in practice a key factor in group cohesion. A newcomer was sometimes installed as chief by earlier inhabitants who nonetheless continued to exercise covert authority through their control of land […]
A clan or tribal chief […] enjoyed many prerogatives and possessed authority because of his role as intermediary between the group and its ancestral spirits […]
The chief was neither the sole office holder nor the only decision maker. He shared authority with such dignitaries as the war chief, master of the soil, priests and powerful sorcerers, who with respected elders formed an advisory council which assisted the chief to reach the consensus on which group action was usually based […]
The sanctity of New Caledonian chiefs and the deference to which they were entitled rested on their implied genealogical connexions with deified ancestors.
Douglas comments that in degree of political stratification New Caledonian leadership resembles ‘mid-range’ Polynesian societies (corresponding to Goldman’s ‘traditional’ type) such as Maori, though the descent ideology was probably more flexible than in the Maori case.
Seligman (1910) and Hau’ofa (1981) describe Mekeo society in the early colonial period, when tribal warfare was still ongoing. Key institutions were the hereditary offices of the high chief or civilian chief (lopia faʔa) and the war chief (iso lopia, more commonly simply iso), the official sorcerers (uŋauŋa) and the war magicians (faiʔa). A village is made up of agnatically related subclans, ranked in order of seniority of descent from older vs younger brothers.
Each subclan should have both a civilian chief and a war chief, representing a fundamental division of authority between the civilian and military spheres in Mekeo society. High chiefs do not take an active part in warfare but are regarded as men of peace whose authority rests on their control of sorcerers, and so of life and death. War chiefs are dangerous men of anger and violence valued for their military prowess. High chiefs are ceremonially installed. Hau’ofa (1981:185) observes that in a large subclan the civilian chieftainship is divided between the lopia faʔa and a junior chief (lopia eke). The junior chief performs certain ritual duties on behalf of the high chief in the ufuapie ceremonial feasts.
When large clans segment the junior branch is headed by a junior chief. Once a particular chieftainship is established, be it senior or junior, succession to the office becomes hereditary from father to son, with the qualification that only a successful warrior can succeed to a war chieftainship. The father sometimes chooses a junior son or an adopted son to succeed him but only if the eldest son is not suitable.
Chiefs’ houses differ from those of other villagers in that they are built parallel to the central ground and have special roof designs. Chiefs were not wealthy but derived their prestige and authority from their rank, personality, numerical strength of their group and command of certain kinds of magic.
Mekeo believe that powers of sorcery (uŋauŋa)
were given personally by the deity, A’aisa, to certain men who founded some of their present-day sorcery families. They believe that when A’aisa created their political organization by an edict he appointed two kinds of hereditary leaders, the military and the civilian. For the civilian leadership he appointed some men as chiefs and others to enforce the authority of the chiefs. Sorcery lineages which have risen from these appointments are the most prestigious in the land. (Hau’ofa 1981:230)
Hereditary leaders, particularly sorcerers, usually acquire one or more kinds of non-sorcery magic which they incorporate in their arsenal of powers. These enhance their prestige and at the same time allow them additional sources of wealth.
Baluan, a society of the Admiralty Is., is described by Otto (1994) as thoroughly hierarchical.
Ascribed hierarchy was without a doubt a pervasive trait of Baluan society at the end of the nineteenth century. It was expressed in various ways: birth-order name, kinship terminology, a system of dual rank, ranked titles for leaders, and material signs of distinction (Otto 1994:224)
Hierarchy was not based on ascription or achievement alone, but on a specific articulation of these two principles […] The relative ranking of leaders of high rank was determined by their success in warfare and feast-giving. The status of a lapan [chiefly] lineage would be eroded by inadequate performance, although they could continue claiming the title for several generations. The lapan stories show clearly that lapan status could also be appropriated through violence […]” (Otto 1994:232)
Wedgwood (1934) reports that in Manam society descent is patrilineal, residence is patrilocal and leadership is hereditary, based on primogeniture. There are two named social groups, the aristocracy (tanepoa or tanepwa) and the commoners (gadagada). Membership in these groups is based on birth. The village chief, tanepoa labalaba, literally ‘big chief’, is the “senior male descendant in the male line of the original founder of the village” (Wedgwood 1934:383). The chief’s sons and younger brothers are called tanepoa siʔisiʔi, literally ‘little chief’. There is a strong preference for marriage between chiefly families. Lutkehaus (1990) states that tanepoa were believed to inherit from their ancestors special marou, a form of supernatural power or mana, with both benign and malign aspects. “Its benign form was manifested in a tanepwa’s ability to produce abundant taro in his gardens, to protect his village against illness and strife, to promote his village’s reputation through the organization of large and munificent pig-exchange celebrations” (Lutkehaus 1990:300) and through success in amassing valuables from mainland trading partners. The dark side of marou was its association with sorcery, which a chief was able to control.
The Sa’a occupy the island of Small Malaita in the Southeast Solomons. Their society around the end of the 19th century is described by Codrington (1891) and Ivens (1927). They made a sharp distinction between chiefs (alaha) and commoners (apoloa). Descent is patrilineal. Chieftainship is hereditary. The reciting of genealogies is practised. Ivens recorded one genealogy going back 39 generations. Sons can be adopted to ensure succession. Codrington (1991:50) writes that
[T]he chief’s power at Sa’a comes from his birth and personal qualities, not from his intimacy with supernatural beings and his magical knowledge, [although] he may well have these […] He inherits wealth from his father, and adds to it by the fines he imposes and by the gifts of his people, but no wealth or success in war could make a man a chief at Sa’a if not born of the chief’s family.
Chiefs lived in more substantial buildings than commoners. The toohi ‘chief’s lodge’ is the reception house for all visitors of importance.
Chiefs were not allowed to go into battle. Their main social function was the provision of feasts. The malaohu (initiation of boys into bonito fishing) system was also under their control. The head wife of a chief is chosen from among the girls of neighbouring places who themselves are alaha. His other wives may be apoloa. If the head wife has no male heir, a son of one of the other wives succeeds to the chieftainship.
The persons of chiefs were sacred to the extent that anyone who cursed them or the marks of their chiefly position, their gongs or official houses, was marked down for death. The relics of deceased chiefs, their skulls, were held in especial reverence. The common people paid honour to both chief and priest; the one was the glory of the place and the other served ghosts who either had been chiefs themselves or were connected with chiefs through the bonito fishing. It was customary for the commoner to make offerings of garden produce to the chief and of cooked food to the priest; and no return was made or expected. Chief and priest were exempted from the obligation to make a return for gifts received which always held in the case of commoners.
Hereditary chieftainship is found in societies speaking languages belonging to almost every major subgroup of Oceanic. By contrast, as far as I am aware, none of the more than 700 non-Austronesian language communities of Melanesia have hereditary chiefs. This is a strong distributional argument in favour of attributing hereditary chieftainship to POc society. It is highly unlikely that such an institution would have arisen independently in many different Oceanic societies.
Furthermore, as Hage (1999a,b) points out, POc kinship terminology makes it probable that POc society had unilineal descent groups based on seniority of birth. Milke’s (1958) reconstruction of POc kin terms is diagnostic of a unilineal system, distinguishing as it does between terms for father’s brother (*tama) and mother’s brother (probably *matuqa), and between terms for mother’s sister (*tina) and father’s sister (*aia or *aya), a pattern which cross-culturally strongly equates with unilineal descent groups (85% correlation in Murdock’s worldwide survey; Murdock 1947, 1967). There is a widespread association between such a terminology and descent group organisation and exogamy, such that a parent’s opposite-sex siblings will always be in a different descent group from their same-sex siblings (see §3.2.3).
The highly marked status of the seniority distinction in terms for siblings of the same sex in Oceanic societies also implies the existence in POc society of a conical clan system
in which all individuals and all descent lines theoretically have unique ranks. This ranking is generated by a rule of […] primogeniture in the Oceanic case, which distinguishes elder siblings from younger siblings […] and descendants of elder and younger siblings […] Primogeniture is reflected in the seniority distinction in [Proto Oceanic] sibling terminology and more emphatically in the marked status of the term for elder (parallel) sibling. In a number of [Oceanic] terminologies, e.g. Fijian, the term for younger (parallel) sibling can be used generically to mean sibling irrespective of age (or sex) while the term for elder (parallel) sibling can only have this specific meaning, reflecting the special, elevated position of this relation […] Titles are passed down the senior lines, and members of more senior lines form an upper, chiefly class while members of more junior lines form a lower, commoner class, for example, ’eiki […] and tua as in Tonga, iroij and kajur as in the Marshall Islands, and lapan and lau as in the Admiralties.” (Hage 1999a:208)
A further argument, mentioned earlier, for supposing that Lapita society was stratified stems from the very rapid colonisation of all the major island groups of the southwest Pacific after the initial Lapita movements into this region. This achievement required considerable capital investment in skilled labour to build ocean-going sailing canoes.3 It required the recruitment of crews and passengers in sufficient numbers to form viable founding populations, and the transport of useful plants and breeding stocks of domestic animals. It implies the existence of prestigious leaders who could plan and command these undertakings.
This point is well made by Hayden (1983) who argued that boats and trade are the key to understanding the maintenance of social stratification in Polynesia and other regions initially settled by Lapita people. He notes that the construction of a single medium-sized ocean-going canoe took from one to three years, and canoe builders were specialists trained through a rigid system of apprenticeship.
If POc society had hereditary chiefs, what were they called? Are cognates of PPn *qariki present in other branches of Oceanic?
C.E. Fox (1924, 1970) noted that Arosi, a Southeast Solomonic language of Makira, has a pair of terms araha ‘chief’ and ariʔi ‘eldest son of a chief’ (the latter used in songs). The eldest son became chief before his father’s death. Fox asserted that each of these Arosi terms can be analysed into a personal article a, and an adjective, raha ‘big, great’ or riʔi ‘little, small’. That is to say, a-raha is literally ‘Great one’ and a-riʔi is ‘Little one’, i.e. the chief’s heir. Arosi a is preposed to “names native or foreign; used with nouns to personify; with verbs or adjectives to form a descriptive noun or nickname” (Fox 1970:1). A cognate article is found in some other Oceanic languages, including Polynesian, where it precedes personal names and/or, in some languages, place names (Pawley 1972:58).
In Pawley (1982a), I accepted Fox’s claim that Arosi ariʔi is cognate with PPn *qariki and observed that cognates of Arosi araha denoting a chief or person of a chiefly descent group occur in several languages of the Malaita-Makira subgroup in the Southeast Solomons, supporting the reconstruction of Proto Malaita-Makira *alafa ‘chief’:
| Proto Malaita-Makira | *alafa | ‘chief, person of a chiefly lineage’ | |
| SES | Arosi | araha | ‘chief, man of the chiefly clan’ |
| SES | Arosi | haʔa-araha | ‘raise s.o. to chiefly rank, make a child a member of the of the chiefly clan by his father giving a series of feasts’ |
| SES | Bauro | araha | ‘chief, man of the chiefly clan’ |
| SES | Owa | arafa | ‘boss, chief, lord, ruler, important person’ |
| SES | Owa | arafa ni finua | ‘village chief’ (ni ‘of’, finua ‘village’) |
| SES | Sa’a | alaha | ‘hereditary chief’ |
| SES | Sa’a | alaha-ŋa | ‘rule, dominion, chieftainship’ |
| SES | Sa’a | aʔo i alahaŋa | ‘elder branch of a chiefly family’ |
| SES | Sa’a | puri alahaŋa | ‘the cadet branch’ |
| SES | ’Are’are | araha | ‘chief’ |
| SES | ’Are’are | araha ana komu | ‘chief of a bilateral family’ |
| SES | ’Are’are | araha-a | ‘to rule over’ |
| SES | ’Are’are | araha-ha | ‘be a chief, to rule’ |
| SES | ’Are’are | araha-ŋa | ‘chieftainship’ |
| SES | Kwaio | alafa | ‘wealthy influential leader’ |
| SES | Lau | alafa | ‘a chief’’ |
| SES | Lau | alafa-la | ‘kingdom, region of chief’s authority’ |
| SES | Lau | alafa-na | ‘kingdom, dominion’ |
In line with Fox’s analysis of Arosi araha, I treated Proto Malaita-Makira *alafa as consisting of a personal article *a and a nominalised adjective *lafa ‘big, great’.
Lichtenberk (1986:351) proposed a different analysis. He argued that Proto Malaita-Makira *alafa comes from POc *ta-la(m)pat, where *ta- was a reduced form of the well-attested noun *tau ‘man, person’, preceding *la(m)pat ‘big, great’.4 POc *t is regularly lost in all positions in Malaita-Makira languages.
Lichtenberk also gave a phonological reason for reconstructing *ta rather than *a. This has to do with the order in which certain sound changes occurred in Malaita-Makira languages (Lichtenberk 1986, 1988:41). POc *q had already been lost before PMaMa separated from other SES languages. Next, a sporadic but frequent change of word-initial *a to *θa (theta-prothesis) spread across Malaita-Makiran. Initial θ then became s in several Malaitan languages, l in Kwaio but remained θ in To’aba’ita.5 so that, for example, POc *qalipan ‘centipede’ is reflected as To’aba’ita θāfila (metathesis), Lau saruhe, ’Are’are rarihe, Kwaio lalifa, Oroha saruhe, but Sa’a ɛluhe, Ulawa aliha (Lichtenberk 1986:36). Subsequently POc *t was regularly lost in MaMa. If the protoform had been PEOc *qa-lapat, it would have undergone the following sequence of changes: *qalapat > PMaMa *a-lafa > post-PMaMa *θa-lafa. But the forms listed above show no sign of theta-prothesis, pointing instead to the sequence PEOc *ta-lapat > PMaMa *a-lafa
Lichtenberk was uncertain as to whether his POc *tala(m)pat referred to a hereditary chief or to a self-made big-man. He comments:
The literal meaning of *tala(m)pat was ‘big [person], great person’, and it referred to a leader, as do its reflexes in the [Makira-Malaitan] languages […] Given its literal meaning one might be tempted to conclude that the expression referred to a big-man rather than a chief. However, this conclusion is not warranted. The literal meaning under-determines the denotation of *tala(m)pat. It is compatible with the denotation ‘chief’ just as well as […] ‘big-man’. (Lichtenberk 1986:351)
Lichtenberk was apparently not aware that many Oceanic languages of the Admiralties reflect *lapa-ña ‘chief, leader’, analysable into a nominal root *lapa (homophonous with Proto Admiralties *lapa ‘big, great’) and a suffix *-ña, reflected as -n in many Admiralties languages, which marks singular agreement on an adjective.
| PAdm | *lapa-ña | ‘chief, leader’ | |
| Adm | Mussau | lapa-n | ‘important person, chief’ |
| Adm | Baluan | lapa-n | ‘leader of hereditary status’ |
| Adm | Titan | lapa-n | ‘leader, chief, those of noble blood’ |
| Adm | Nyindrou | laba-n | ‘leader, chief’ |
| Adm | Papitalai | laba-n | ‘chief’ |
| Adm | Loniu | lapa-n | ‘leader, God’ |
| Adm | Papitalai | laba-n | ‘chief’ |
| Adm | Seimat | la-lap | ‘chief’; ‘important, large’ |
As noted in §4.5.1.4, in Manam, a Western Oceanic language of north New Guinea, the village chieftainship is a hereditary position occupied by the senior male descendant of the village founder. The chief is called tanepoa labalaba, where tanepoa is ‘chief’ and labalaba is ‘big’. The latter term comes from POc *lapat ‘big, great’. The chief’s sons and younger brothers are called tanepoa siʔisiʔi, literally ‘little chief’.
A number of Micronesian languages also have reflexes of *lapat, referring to chiefs and genealogical seniority.
| Mic | Marshallese | iroij laba-lap | ‘head of royal lineage’ (lit. ‘very great chief’) |
| Mic | Marshallese | iroij elab | ‘lesser chief’ (lit. ‘great chief’) |
| Mic | Lamotrek | mala-lap | ‘senior representative of a matrilineal clan’ |
| Mic | Woleaian | tame-lap | ‘eldest male of a family’ (lit. ‘great father’) |
The agreement between languages from four widely dispersed subgroups (Malaita-Makira, Admiralties, Western Oceanic and Micronesian) suggests that in POc the adjective *lapat ‘big, great’ also had the senses ‘genealogically senior’ and ‘of chiefly rank’, and occurred as the modifier in one or more compound nouns denoting a chief or person of chiefly rank. One such compound noun was probably *ta(u)-lapat, where the head noun was *tau ‘man, person’.
Note also the title of the high chief of Nadrogā (SW Vitilevu) is kʷālevu, lit ‘the big one’, contrasting with kʷāhewa, lit ‘the small one’, his child.
One other term for chief or leader can be reconstructed, with reflexes in two Admiralties and two New Ireland languages.
| POc | *kabi-ña | ‘chief’ | |
| Adm | Drehet | kapi-ŋ | ‘chief’ |
| Adm | Nyindrou | kapi-n | ‘clan leader, chief’ |
| MM | Sursurunga | kabi(sit) | ‘chief, head man, fight leader’ (meaning of sit not known) |
| MM | Patpatar | kabi-n(sit) | ‘supreme chieftain of one or more villages’ (meaning of sit not known) |
It was noted in §4.5.2.1 that PPn *qariki ‘chief’ resembles Arosi ariʔi ‘first-born son of a chief, chief’s heir’, and that this has led some to conclude that these forms are cognate, each historically analysable as consisting of a personal article and a reflex of POc *riki ‘little’. The sound correspondences are regular. However, Arosi ariʔi has cognates in several Malaitan languages, and, as Lichtenberk (1986) notes, these do not mean ‘first-born son of a chief’.
| SES | Kwara’ae | s-ariʔi | ‘unmarried girl’ |
| SES | Kwaio | la-aliki | ‘unmarried girl, daughter’ |
| SES | Lau | s-arii | ‘unmarried girl, maiden, daughter’ |
| SES | To’aba’ita | θa-ariʔi | ‘unmarried girl, maiden’ |
Lichtenberk argues that the Malaitan cognates continue a Proto Malaitan form *a-ariki that consisted of two morphemes: *a ‘personal article’ and *ariki which he glossed as ‘unmarried girl, maiden, daughter’. Contemporary Malaitan languages have accreted a prothetic consonant to the personal article *a and other roots beginning with the vowel *a.
How are the differences of meaning between the Polynesian, Arosi and Malaitan forms to be resolved? Lichtenberk suggests that the most likely meaning in their common ancestral language was ‘oldest child’. Chowning (1991:63) points to Mussau aliki ‘child’ and Bariai (New Britain) galiki ‘firstborn child’ as possible cognates, and suggests ‘child’ (as an age-grade term, not a kinship term) as the meaning in POc. We may note the existence of the well supported POc etymon *natu- ‘child’ (kinship term, requiring a possessor) but it is usual for Oceanic languages to have distinct terms for ‘child’ as a kinship term and as an age grade term.
At some point in the history of Polynesian it may be that the term *qariki ‘child’ was used as a nickname for the chief himself, and that in PPn it became the conventional term for ‘chief’. A parallel usage occurs in Samoan, where tama ‘boy, youth; son of a woman’ can refer to chiefs of high rank and the compound tama-ali’i (lit. ‘chiefly boy’) as a noun denotes a chief and as a verb means ‘be of noble descent or lineage’ (Milner 1966). The holder of any of the four highest titles in Samoa is known as tama-ʔāiŋa (lit. ‘boy of the lineage’). In Tongan tama ‘child, son especially of a woman’ also has the meaning ‘male of chiefly rank’. In Bauan Fijian ŋone tūraŋa, lit. ‘chiefly child’, is an honorific term for a chief (Paul Geraghty pers. comm.).
Lichtenberk (1986:351) allows that if it were possible to reconstruct a term for ‘commoner’ for POc, this would be strong evidence that POc society had chiefs.
Many Oceanic languages have a term or terms for people of low status but no widespread cognates are found. Often a derogatory term meaning ‘useless, unimportant’ is involved.
| Adm | Nyindrou | lau leleyah | ‘commoner’ (lau ‘person’; leleyah ‘nothing’) |
| Adm | Titan | lau | ‘commoners’ |
| Adm | Baluan | lau | ‘collective term for followers of a lapan’ |
| Adm | Baluan | sayo | ‘person of low status, opposite of lapan’ |
| NNG | Manam | gada-gada | ‘commoner’ |
| SES | Gela | bonaɣa | ‘commoner, of no importance’ |
| SES | Arosi | mʷae taʔa | ‘commoner, man of no importance, man of any clan except araha’ (mʷae ‘man’, taʔa ‘bad, poor’) |
| SES | Lau | nēna | ‘commoner’ |
| SES | Owa | ainuni purua | ‘common people’ (ainuni ‘people’, puru ‘ordinary, common’) |
| Mic | Marshallese | kacuṛʷ | ‘commoner, common people’ |
| Mic | Marshallese | aṛʷmec wān | ‘commoner’ (aṛʷmec ‘person’, wān ‘trivial, common, worthless’) |
| Fij | Bauan | tau-vanua | ‘commoner’ (tau ‘person’, vanua ‘land, place’) |
New Caledonian languages usually have a term that means ‘subject of (a chief)’. These include: in Pije, Fwâi, Nemi, Jawe, Nyelâyu and Nêlêmwa yabʷec, Cèmuhî abʷɛ̄, Xârâcùù kʷara and Iaai kei-. Most of these terms can occur with a collective proclitic, and the Iaai term is directly possessed, e.g. la kei-ñ ‘his subjects’, la kei-k ‘my subjects’ (la ‘collective’).
One candidate for ‘commoner’ is POc *mʷala, which Blust (1981c) suggests may have been a noun denoting a person of low social status. However, the evidence favours reconstructing *mʷala as a stative verb and adjective meaning ‘common, worthless’ and as a noun meaning ‘misfortune, lack of prosperity’.
| SES | Gela | mala | ‘comparative marker with deprecatory force’ |
| SES | Arosi | mara | ‘deprecatory prefix’ |
| SES | Sa’a | mʷala | ‘people, commoners (as opposed to chiefs)’ |
| NCV | Mota | mala | ‘bad, poor’ (often used in depreciation) |
| Mic | Marshallese | wān | ‘trivial, common, worthless’ |
| Mic | Marshallese | aṛʷmec wān | ‘commoner’ (lit. ‘worthless person’) |
| Mic | Mokilese | mʷāl | ‘useless, disadvantaged, low (socially)’ |
| Mic | Ponapean | mʷāl | ‘common, useless, of no consequence’ |
| Mic | Ponapean | aramas mʷāl | ‘commoner’ (lit. ‘person of no consequence’) |
The Proto Oceanic language can be associated with the Lapita culture which appeared in the Bismarck Archipelago at about 3300-3200 BP. Around 3000 BP bearers of Lapita became the first people to settle island groups of the southwest Pacific east of the main Solomons group, and within the space of two or three centuries colonised Santa Cruz and the Reef Is., Vanuatu, New Caledonia, Fiji, Tonga and Samoa. This was an achievement that implies strong leadership, requiring as it did the building of large ocean-going outrigger canoes, the recruitment of crews and passengers and the transport of domestic animals and useful plants.
A number of terms relating to chiefly authority and status are attributable to Proto Polynesian, the daughter of POc that developed in western Polynesia, with some regional variation, in the millennium or so following colonisation by Lapita people. There is strong evidence that speakers of PPn had ranked descent groups, probably termed *kainaŋa, headed by hereditary chiefs, called *qariki, and that rank was based on primogeniture.
In Melanesia, hereditary chieftainship based on primogeniture is also found in societies representing several major subgroups of Oceanic but is absent in societies speaking non-Austronesian languages. This is a powerful distributional argument in favour of attributing hereditary chieftainship to POc society. It is unlikely that that such an institution would have arisen independently in many different Oceanic societies. Furthermore, POc kinship terminology is of a type that makes it probable that POc society had unilineal descent groups ranked by seniority of birth.
Agreement between languages from four widely dispersed subgroups (Admiralties, Western Oceanic, Malaita-Makiran and Micronesian) suggests that in POc the adjective *lapat ‘big, great’ also had the meanings ‘genealogically senior’ and ‘of chiefly rank’, and that it occurred as a constituent in a compound noun denoting a chief or person of chiefly rank, possibly *ta(u)-lapat.