This chapter asks what linguistic evidence can tell us about Proto Oceanic speakers’ settlement patterns and relation to territory, both land and sea.1 Where were settlements located? What sort of residential communities existed? Did people live in sizeable villages or in dispersed hamlets? What sort of buildings were constructed, in terms of function and architecture? What sort of territorial units were recognised?
A conjunction of archaeological and linguistic evidence places the primary dispersal centre of Proto Oceanic in the Bismarck Archipelago and associates this language with the archaeological tradition known as (Early Western) Lapita, whose bearers spread very swiftly across the southwest Pacific around 3000 BP, reaching Tonga by 2850 BP. Archaeology tells us a good deal about the preferred habitation sites of early Lapita settlements in the Bismarck Archipelago (Green 2003; Kirch 1997, 2000; Spriggs 1997a; Summerhayes 2010). All the sites are on the coast, close to beaches and fringing reefs that, respectively, would have provided landing places for canoes and for obtaining seafood. However, archaeology has yielded little evidence about the internal organisation of Lapita dwelling sites, as very few have been excavated extensively enough to reveal the arrangement of houses and other structures and areas of use.
Ethnographic evidence on settlement patterns and use of territory in Oceanic speaking communities is rich (comparative studies include Forge 1972, Hogbin and Wedgwood 1953, Oliver 1989). Patterns that recur in the ethnographic accounts include the following:
However, as ethnographic comparisons refer only to recent times they can do no more than suggest hypotheses about the situation in POc times, some three millennia ago. We turn to lexical reconstructions as a source of evidence.
POc speakers were a maritime people, for whom the sea was an important economic resource. They had an extensive vocabulary for fishing methods and technology (vol.1, ch.8) and for outrigger canoe parts and sea travel (vol.1, ch.7). More than 140 POc names for marine fish taxa and more than 40 names for marine invertebrates have been reconstructed (vol.4, chs 2 and 4). Relevant terms for the seascape and landscape (vol. 2, ch.4) include *tasik ‘sea’, *masawa(n,ŋ) ‘open sea’, *laman ‘deep sea beyond the reef’, *ŋalun ‘mounting wave, ocean wave’, *bayau ‘ocean swell’, *loka (N) ‘high sea, heavy breakers’, (V) ‘be rough, of sea’, *maqati ‘low tide, dry reef,’ *Ruap ‘high tide’, *sakaRu ‘coral reef’, *motu(s) ‘detached reef’, *laje ‘coral’, *mʷaloq ‘submerged rock or coral reef’, *namo ‘lagoon inside reef, deep pool in reef’, *mata (qi, ni) sawa(n,ŋ) ‘passage through reef’, *(b,bʷ)iker ‘beach, esp. sandy beach’, *nusa ‘island’, *tobʷa ‘bay’, *(i,u)cuŋ ‘cape’.
POc speakers were also farmers. They cultivated a range of ground and tree crops (vol.1, ch.5; vol. 3, chs 9–13) and kept pigs (vol.4:237–240) and chickens (vol.4:283–287). Terms for clearing garden land (*quma, *poki), for a plantation (*topa) and fallow land (*talun) contrasted with one for bushland, inland country (*qutan). There were terms for planting (*tanum), planting in holes in the ground (*asok), weeding (*papo), scattering seeds (*kabu(R)), garden fence (*kaRi), and boundary marker in a garden (*bayat) (vol.1, ch.5).
That POc speakers occupied permanent dwellings is indicated by a host of terms to do with house construction. Terms for several kinds of buildings can be reconstructed. Blust (1987) compared four different cognate sets, widely represented in Austronesian languages, that refer to kinds of domestic buildings. He reconstructs Proto Malayo-Polynesian (PMP) *Rumaq ‘dwelling house’ *balay ‘public building’, *kamaliR ‘men’s house’ and *lepaw ‘granary’.2 The first three terms were continued in POc.
| PAn | *Rumaq | ‘dwelling house’ (Blust 1987) | |
| POc | *Rumaq | ‘dwelling house’3 | |
| Adm | Lou | um | ‘house’ |
| Adm | Loniu | umʷe | ‘house’ |
| NNG | Arawe | a-rumuk | ‘village’ |
| NNG | Vehes | ɣumak | ‘house’ |
| PT | Motu | ruma | ‘house’ |
| MM | Bali | rumaka | ‘house’ |
| MM | Petats | luma | ‘house’ |
| SES | Lau | luma | ‘family house’ |
| SES | Arosi | ruma | ‘house’ |
| SES | To’aba’ita | luma | (1) ‘traditionally, house where a woman and her children lived; today, family house’; (2) ‘building’ (̄in compounds) |
| NCV | Mota | imʷa | ‘house’ |
| NCV | Raga | imʷa | ‘house’ |
| NCV | Nokuku | ima | ‘house’ |
| NCV | Kiai | ima | ‘house’ |
| SV | Lenakel | n-imʷa | ‘house’ |
| NCal | Dehu | uma | ‘house’ |
| Mic | Kiribati | uma | ‘any kind of building, anything with roof’ |
| Mic | Marshallese | yimʷe- | ‘house’ |
| Mic | Ponapean | iimʷ | ‘house’ |
A host of reconstructed POc terms for house parts and construction, including the following, indicate that at least some dwelling houses were substantial buildings, with heavy superstructure, gabled, with floor raised on piles (for supporting cognates see vol.1, ch.3):
Some Oceanic-speaking societies distinguish by name more than one design of dwelling house. The Arosi speech community of Makira distinguishes at least the following kinds of ruma ‘house’ (Fox 1978):
| ruma huri | a round house, taller than the well known Santa Cruz round house |
| ruma gaura | a double, long house |
| ruma ora | a house with five pairs of posts or more |
| ruma pʷarapʷara | a house roughly built, with roof of fronds instead of thatching |
| ruma raŋi | a house with two roofs, the second roof only above the upper part of the first, made by extending the rafters upwards |
| ruma okera | like ruma raŋi but with the two roofs close together |
| ruma sinakuhi | house with rounded end |
| ruma waiho | simplest form of oblong house with three pairs of posts |
Although no compound term consisting of *Rumaq plus modifier has so far been reconstructed for POc, the latter retained two PMP terms for buildings other than main dwelling houses.
POc *pale, continuing PMP *balay, has in some daughter languages replaced *Rumaq as the usual term for a dwelling house. However, in languages where both terms are retained reflexes of *balay and *pale generally refer to a simpler kind of building, such as an open-sided boat shed, shed for storing yams or garden shelter. The replacement was perhaps due to speakers habit of humorously referring to their dwelling house as a humbler kind of building.
| PMP | *balay | ‘public building’ (Blust 1987) ; ‘unwalled building’ (Waterson 1993) | |
| POc | *pale | ‘building for storage or public use, open-sided building, shed’ | |
| Adm | Lou | pal | ‘canoe hut’ |
| Adm | Mussau | ale | ‘house’ |
| NNG | Bebeli | bele | ‘house’ |
| NNG | Yabem | ale | ‘house’ |
| NNG | Lukep | para | ‘yam house’ |
| MM | Tolai | pal | ‘house, room’ |
| MM | Tolai | pia na pal | ‘place where there are houses, village’ (pia in compounds, ‘a place’) |
| MM | Tangga | pal | ‘small house or shed, storehouse for temporary storage of food’ |
| MM | Mono-Alu | hale-hale | ‘public building’ |
| SES | Arosi | hare | ‘shed for yams’ |
| SES | Arosi (West) | hare | ‘house with side of roof only, made in garden’ |
| SES | Bugotu | vaðe | ‘house, building’ |
| SES | Bauro | hare | ‘canoe house, men’s house’ |
| SES | Sa’a | hale | ‘yam shed outside a garden’ |
| SES | Kwaio | fale | ‘hut for childbirth’ |
| SES | Gela | hale | ‘house’ |
| NCV | Raga | vale | ‘house, hut, garden house’ |
| NCV | Nokuku | vale | ‘shelter’ |
| NCV | Nokuku | val-val | ‘garden shelter’ |
| NCV | Kiai | vale | ‘shed, shack’ |
| PMic | *fale | ‘meeting house’ (Bender et al. 2003a) | |
| Mic | Puluwatese | fǣl | ‘meeting house’ |
| Mic | Woleaian | fal, fale- | ‘men’s house, club house’ |
| Fij | Bauan | vale | ‘house’ |
| Pn | Samoan | fale | ‘house’ |
| Pn | Hawaiian | hale | ‘house’ |
POc *kamali(R) ‘men’s meeting house’ has reflexes in languages of the Admiralties and North and Central Vanuatu. Among the Ponam, a Titan speaking community of Manus, kamal is the name both of a patrilineal descent group (a land-owning group) and of the men’s house of such a group (Carrier & Carrier 1989). In Vanuatu the reflex of *kamali(R) refers to a building used by men of a village as a recreational centre, e.g. for drinking kava, and as a sleeping and living area for unmarried men and boys and male visitors.
| PMP | *kamaliR | ‘men’s house’ (Blust 1987) ; ‘granary’ (Tryon 1995) | |
| POc | *kamali(R) | ‘men’s meeting house’ (Blust 1987) | |
| Adm | Titan | kamal | ‘patrilineal property-owning group; men’s house of such a group’ |
| Adm | Nyindrou | kamen | ‘men’s house’ |
| PNCV | *kamali | ‘men’s house’ (Clark 2009) | |
| NCV | Mota | gamal | ‘club house of supʷe (graded society) or of a single high rank’ |
| NCV | Raga | gamali | ‘men’s house’ |
| NCV | Southeast Ambrym | n-emel | ‘men’s house’ |
| NCV | Lewo | kumali | ‘village, men’s meeting house’ |
| Pn | Emae | kamali | ‘men’s house’ |
| NCV | Namakir | na-kamal | ‘men’s club house, dancing ground, village meeting place’ |
A term for ‘canoe house’, *v(a,o)lau, is reconstructed for PCP. This term is formally cognate with a verb meaning make a sea voyage, which has antecedents in POc *palau(r) ‘make a sea voyage’ and PMP *pa-laSud ‘go down to the sea or coast.’
| PCP | *v(a,o)lau | ‘canoe house’ | |
| Fij | Bauan | volau | (1) ‘canoe shed, with ridged roof but no ends or sides’; (2) ‘carpenter’s workshop’ |
| Fij | Wayan | volau | (1) ‘boat shed’; (2) ‘workshop, tool shed’ |
| Pn | Tongan | ala-folau | ‘canoe shed’ (l in ala unexpected) |
| Pn | East Futunan | a-folau | ‘boat house’ |
| Pn | Samoan | ā-folau | ‘long house, used, e.g., for receiving guests’ |
| Pn | Tikopia | a-forau | ‘canoe shed’ |
| Pn | Māori | farau | (1) ‘temporary shed of tree branches’; (2) ‘canoe shed’ |
| Pn | Rarotongan | ʔōrau | ‘canoe shed’ |
In Sa’a and Ulawa, Malaita, SE Solomons, chiefs had a canoe house, taoha, built at the landing place of canoes. This served the double purpose of a house for the chief’s decorated canoe and a house where men congregated and slept (Ivens 1927:7, 34).
A term for ‘house site’ or ‘house foundation, consisting of a mound or platform of earth or coral rubble’, is attributable to PEOc. It has reflexes in Southeast Solomonic, Fijian, and Polynesian. It is cognate with a verb meaning ‘to pile up, heap, make a mound or wall’. House mounds are associated with houses without raised floors, which are characteristic of eastern Melanesia, Polynesia and Micronesia.
| PEOc | *apu | ‘house foundation’ | |
| SES | Gela | avu | ‘house site’ |
| SES | Arosi | ahu | ‘mound of earth, heap of things’ |
| Fij | Bauan | yavu | ‘house site, house foundation’ |
| Fij | Wayan | avu | ‘house site, house foundation’ |
| Pn | Tahitian | ahu | (1) ‘platform of stones, often stepped, with religious functions’; (2) ‘pile up stones, put up wall of a marae’ |
| Pn | Rapanui | ahu | ‘large platform of stones, with religious function’ |
| Pn | Māori | ahu | ‘platform of stones’ |
Ivens (1927:375) writes of Sa’a communities that
[i]n certain hamlets … there is more or less of an appearance of house mounds, but speaking generally there is nothing to correspond to the Fijian yavu … This is owing to the frequent changes of location in the past, and to the absence of anything like settled towns.
It is noteworthy that there is no very well supported POc reconstruction whose primary sense was ‘village’. There are several candidates, chiefly *pera, *koro, *malaqai and *panua, but objections can be raised against each.
| POc | *pera | ‘? settlement, open space associated with a house or settlement’ | |
| NNG | Manam | pera | ‘house, room’ |
| SES | Bugotu | vera | ‘courtyard, open space in a village’ |
| SES | Tolo | vera | ‘village, home, country, place where one lives’ |
| SES | Ghari | vera | ‘village’ |
| SES | ’Are’are | he-hera | ‘open space in front of the houses for walking’ |
| SES | ’Are’are | herā | ‘agglomeration of houses, village’ |
| SES | Arosi | hera | (1) ‘open space for dancing, usually to the east of burial ground for chiefs’; (2) ‘any burial space surrounded by stone walls’ |
| SES | Baegu | fera | ‘hamlet, a named locality’ |
| SES | Kwaio | fela | ‘skull house’ |
| SES | Lau | fera | ‘land; village; habitation, home, artificial island (for habitation)’ |
| SES | Lau | mā-fera, mae-fera | ‘hamlet, 2 or 3 houses’ (mā, mae ’classifier for round objects) |
| SES | Lau | fera fū | ‘mainland, solid land’ |
| SES | Lau | fera daudau | ‘artificial island’ |
| SES | To’aba’ita | fera | ‘traditionally, house where a woman and her children lived; today, family house’ |
| PNCV | *vareqa | ‘outside, public space’(Clark 2009) 4 | |
| NCV | Mota | varea | ‘village, place of a village settlement’ |
| NCV | Nokuku | varea | ‘home, village’ |
| NCV | Namakir | vareʔ | ‘outside’ |
| NCV | Uripiv | varea | ‘outside’ |
| NCV | Nguna | varea | ‘chiefs’ meeting house’ |
It can be seen that *pera is widely reflected in Southeast Solomonic languages, with meanings that range from ‘village’, ‘hamlet’, ‘habitation’, ‘house’, ‘named locality’ to ‘open space for dancing, courtyard’, ‘burial space’, ‘skull house’ and ‘land.’ Elsewhere in Oceanic only one secure reflex has been noted, in Manam, a North New Guinea language, where it is the usual term for ‘house’.
Some of these disparate meanings are better associated with other POc terms, e.g. ‘house’ with *Rumaq and *pale (§5.3), ‘open space in a village’ with *malaqai (see below), ‘land’, ‘inhabited place’ and perhaps ‘settlement’, with *panua (§5.5.1).
The reflexes of POc *koro in Vanuatu and Polynesian indicate that it referred primarily to any enclosure that is fenced or protected by barriers and that it has later come to mean ‘village’ in languages of the Admiralty Islands, Fiji and Tonga.
| POc | *koro | (1) ‘fenced-in area’; (2) ‘? settlement fortified by barrier’ | |
| Adm | Leipon | kor | ‘village’ |
| Adm | Titan | kor | ‘home, village, settled land, farm, the earth’ |
| NCV | Nguna | kooro | ‘enclosure, pen, blowhole’ |
| NCV | Nguna | na-koro | ‘yard’ |
| NCV | Nguna | na-ko-koro | ‘fence, hedge, windbreak’ |
| NCV | Nokuku | kokoo | ‘garden’ |
| NCV | Namakir | kor | ‘fence, rail’ |
| Fij | Bauan | koro | ‘village, an eminence’ |
| Fij | Wayan | koro | ‘settlement, village, hamlet, town’ |
| Fij | Wayan | loma ni koro | ‘centre of a village’ |
| Pn | Tongan | kolo | ‘village, town; fortress; temporary fence around open grave’ |
| Pn | Niuean | kolo | ‘fort, tower, lookout point’ |
| Pn | Samoan | ʔolo | ‘fort, shelter, tower’ |
| Pn | Tikopia | koro | ‘fort, barrier against sea’ |
| Pn | Rarotongan | koro | ‘fenced or walled-off area, enclosure, yard, fence, palisade, surrounding wall’ |
Another candidate, *malaqai, is attested throughout Polynesian, where it denotes an open space in the middle of a village or in front of a house used for public activities, and in a North New Guinea language and a Papuan Tip language, where it refers to a village. Possible cognates appear in Tangga, a Meso-Melanesian language, but the data are problematic.5
| POc | *malaqai | ‘? public space in a village, village plaza’ | |
| NNG | Yabem | malaʔ | ‘village, place of residence, dwelling place’ |
| NNG | Yabem | malaʔ-gɛdɔ | ‘part of a village, a group of village houses’ (gɛdɔ ‘part’) |
| NNG | Yabem | malaʔ-luŋ | ‘village square, meeting place’ (luŋ ‘middle’) |
| PT | Wedau | melagai | ‘village’ |
| Pn | Tongan | malaʔe | ‘village green, playground, open market place’ |
| Pn | Samoan | malae | ‘open space in the middle of a village, meeting ground’ |
| Pn | East Futunan | malaʔe | ‘public area, open space in front of houses’ |
| Pn | Tikopia | marae | ‘open space for public assembly, including dance festivals’ |
| Pn | Māori | marae | ‘village common, courtyard, enclosed space in front of a house’ |
| MM | Tangga | male | ‘village’ |
| MM | Tangga | male-lil | ‘village square’ |
| MM | Tangga | māli | ‘dancing square within the confines of every village settlement’ |
| Mic | Kiribati | marae | ‘open space, public place (Pn loan)’ |
Another term whose reflexes in certain contemporary languages refer to a village plaza is POc *mʷalala. However, in POc this probably had a more general meaning such as ‘land cleared of vegetation (but not planted or built on)’, as well as serving as a stative verb meaning ‘be cleared of vegetation’.
| POc | *mʷalala | ‘cleared land, clearing’; [V] ‘be cleared of vegetation, vacant’ | |
| Adm | Baluan | malala | ‘clearness; free from weeds’ |
| NNG | Manam | malala | ‘market place, assembly place’ |
| MM | Nakanai | malala | [N] ‘area within a village used as a dance ground; garden area cleared but not yet planted’; ‘be cleared of vegetation’ |
| SES | Arosi | mʷarara | ‘space (between things), hole, opening’ |
| Mic | Chuukese | mannaan | ‘grassland (open, treeless)’ |
| Mic | Ponapean | mall | ‘natural clearing in a forest’ |
| Mic | Satawalese | melal | ‘cleared ground’ |
| Fij | Bauan | ŋalala | ‘be spacious, empty, free, at liberty, exempt’ |
| Fij | Wayan | ŋʷalala | ‘be vacant, unoccupied, empty, free, exempt’ |
The semantics of POc *panua are discussed in §5.5.
POc *panua, which continues PMP *banua, is very widely reflected in Oceanic languages, indeed more widespread than any of the terms reconstructed above. The persistence of *panua suggests that it denoted something of very fundamental importance in the early Oceanic worldview. However, reconstructing its semantic history is a considerable challenge because its reflexes in present-day languages show an even more puzzlingly disparate range of meanings than *pera, *koro, and *malaqai. Some idea of how reflexes of *panua vary in meaning within and across different subgroups of Oceanic is given by the following sample.
| Adm | Mussau | anua | ‘land’ |
| Adm | Penchal | panu | ‘village’ |
| NNG | Bariai | panua | ‘village’ |
| NNG | Bam | anu | ‘earth’ |
| NNG | Gedaged | panu | ‘village, settlement, hamlet, place, (modern) town’ |
| NNG | Gedaged | panu-panu | ‘everybody, the whole world’ |
| NNG | Kove | pana | ‘people’ |
| NNG | Tami | panu | ‘house’ |
| PT | Muyuw | ven | ‘land, earth, country, village, hamlet, town, home, place, locality, district, weather’ |
| PT | Dobu | anua | ‘house’ |
| PT | Kilivila | valu | ‘any place, land, village, uninhabited land’ |
| PT | Molima | vanua | ‘house’ |
| PT | Molima | vanua-pou | ‘residents of a village’ |
| PT | Motu | hanua | ‘village’; ‘world subject to diurnal cycle’ (in compounds) |
| PT | Mekeo | panua | ‘social division in a village’ |
| MM | Nakanai | la-valua | ‘the men’ |
| MM | Nakanai | valua-gu | ‘members of my club’ (-gu ’my) |
| MM | Bali | vanua | ‘island’ |
| MM | Vitu | vanua | ‘garden’ |
| MM | Tabar | vanua | ‘house’ |
| MM | Taiof | fan | ‘village’ |
| MM | Tangga | fān | ‘many people, everybody’ |
| MM | Teop | van | ‘land’ |
| MM | Marovo | vanua | ‘house’ |
| MM | Vangunu | vanua | ‘house’ |
| SES | ’Are’are | hanua | ‘land (not sea); district, place, country; island; the territory where a person lives and where his possessions are, including houses, food, trees, water, graves’ |
| SES | Lengo | vanua | ‘village’ |
| SES | Arosi | hanua | ‘island, village’ |
| SES | Lau | fanua | ‘land, the earth, world, weather’ |
| SES | Kwaio | fanua | ‘place, village, shrine-territory’ |
| SES | Kwaio | mā-ʔe-fanua | ‘segment of a territory, sub-clearing of a settlement’ |
| SES | Sa’a | henua | ‘land, country, village, site of village, place’ |
| SES | Sa’a | i hanua | ‘on land, shore (not sea)’ |
| SES | Sa’a | taʔa ni hanua | ‘people living inland’ |
| NCV | Mota | vanua | ‘land, island, village, place’ |
| NCV | Nokuku | venua-na | ‘house, home, village’ |
| NCV | Nguna | vanua | ‘bounded plot of land for gardening’ |
| NCV | Nguna | na-vanua | ‘land, country, island’ |
| NCV | Neve’ei | ne-vanu | ‘place from which one originates’ |
| NCV | Paamese | hanuo | ‘person, human being’ |
| NCV | Lewo | vanua | ‘outside’ |
| SV | Anejom̃ | n-henou | ‘taro swamp’ |
| SV | Kwamera | ru-kʷanu | ‘home, residence, house, village, hamlet’ |
| SV | Lenakel | na-uanu | ‘village’ |
| Mic | Woleaian | faẓüw | ‘land, island’ |
| Mic | Puluwatese | fanɨ | ‘land, island, islet, country’ |
| Fij | Bauan | vanua | ‘land, country, community, place, confederation of clans’; ‘world subject to diurnal cycle’ (in compounds) |
| Fij | Rotuman | hanua | ‘land, country, place, native land, home, people’; ‘world subject to diurnal cycle’ (in compounds) |
| Fij | Rotuman | hanua noho | ‘dwelling place, village’ (noho ‘dwell’) |
| Pn | Tongan | fonua | ‘land, country, territory, place, people (of the land), grave’ |
| Pn | Samoan | fanua | ‘land, field’ |
| Pn | East Futunan | fenua | ‘people, a people, nation, territory, land’ |
| Pn | Nukuoro | henua | ‘land mass, island, country or any other geopolitical unit’ |
| Pn | Tikopia | fenua | ‘land; island; country; inland, as opposed to shore; people of a land, folk; general physical environment; abroad’ |
| Pn | Rennellese | henua | ‘land (not water), island, unknown land (poetic), people of the land’ |
| Pn | Tokelauan | fenua | ‘land (owned by someone, or a land mass), country (geo-political unit), people of the village’ |
| Pn | Hawaiian | honua | ‘land, earth’ |
This sample exhibits about 20 more or less distinct senses associated with reflexes of *panua. The distribution of these senses across major subgroups of Oceanic is as in Table 5.1.
In his discussion of PMP *banua Blust (1987) compared cognate sets from three putative subgroups of Malayo-Polynesian: Western Malayo-Polynesian (WMP), Central Malayo-Polynesian (CMP), and South Halmahera-West New Guinea (SHWNG) as well as from Oceanic. Languages from the first three subgroups exhibit roughly the same range of glosses as can be found in Oceanic, with a few additions. For example:
| WMP | Belau | beluu | ‘country, district, place’ |
| WMP | Kapampangan | banwa | ‘year, sky, heaven’ |
| WMP | Bikol | banwaʔ | ‘town, country’ |
| WMP | Cebuano | banwa | ‘fatherland, town, village’ |
| WMP | Sangir | banua | ‘land, district; people; state; sea; weather’ |
| WMP | Tondano | wanua | ‘village’ |
| WMP | Malay | benua | ‘large expanse of land, empire, continent; mainland in contrast to island’ |
| WMP | Old Javanese | wanwa | ‘inhabited place or area, settlement’ |
| WMP | Nias | banua | ‘sky, heaven; thunder; village; homeland; fellow villager; serf’ |
| CMP | Selaru | hnu(a) | ‘village’ |
| SHWNG | Numfor | menu | ‘village’ |
| ‘land (not sea or sky)’ | Adm, PT, MM, SES, NCV, Mic, Fij, Pn |
| ‘earth, ground (soil)’ | NNG, Pn |
| ‘world (subject to weather, day and night)’ | NNG, PT, SES, Fij, Pn |
| ‘weather’ | PT, SES |
| ‘island’ | MM, SES, Mic, NCV, Pn |
| ‘country, territory’ | PT, Mic, Fij, Pn |
| ‘uninhabited land’ | PT (one language) |
| ‘place, area, district, region’ | NNG, PT, SES, NCV, Fij, Pn |
| ‘village, settlement, hamlet’ | Adm, NNG, PT, SES, NCV, SV |
| ‘social division in a village’ | PT (one language) |
| ‘house’ | MM, NNG, PT, NCV, SV |
| ‘residents of a village’ | PT (one language) |
| ‘community of people’ | NNG, MM, Fij, Pn |
| ‘men’ | MM (one language) |
| ‘everybody’ | NNG, MM (one language in each) |
| ‘person, human being’ | NCV (one language) |
| ‘confederation of clans’ | Fij |
| ‘garden, plot of garden land’ | MM, NCV, SV (but rare in each) |
| ‘taro swamp’ | SV (one language) |
| ‘field’ | Pn (one language) |
| ‘shrine-territory’ | SES (one language) |
| ‘grave’ | Pn (one language) |
If POc *panua had a single meaning, what was it? If it was polysemous, as its reflexes in many daughter languages are, which senses did it have? What paths led to the diverse senses found in contemporary languages?
In tackling these questions one meets a number of methodological problems, starting with the descriptive data. It must be said that the primary data, the definitions of *panua reflexes given in bilingual dictionaries and wordlists of contemporary Oceanic languages, are often deficient. Most sources do not offer systematic definitions. Instead, the ‘definitions’ are typically single word glosses – serving as rough translations equivalents in the European target language. (This problem is not particular to dictionaries of Oceanic languages. It is systemic in that bilingual dictionaries are typically designed to be translation aids rather than to provide analytic definitions.) And such single word glosses as ‘land’, ‘earth’, ‘country’ and ‘people’ are themselves imprecise, because each has more than one sense. For example, the entry for land in The Shorter Oxford English Dictionary gives many senses, of which the following are relevant to our concerns:
Cruse (1986:80-81) writes that
Linguists who have worked in lexical semantics can be broadly divided into two categories: on the one hand there are those who believe that a word form is associated with a number (perhaps finite, perhaps not) of discrete senses; and, on the other, there are those who believe that the discreteness is illusory.
The second group prefer to think of variant readings of a single lexical form as forming a spectrum of senses, a continuum without clear boundaries, much like the colour spectrum or a dialect continuum. However, it is possible to have a foot in both camps. There are domains where the evidence favours analysis into discrete senses and domains where it favours analysis in terms of sense spectra. There are standard diagnostics for polysemy. These include the existence of synonyms, antonyms and contrasting forms restricted to particular senses, the construction of sentences with qualifying elements that are sensitive to sense differences, and the restriction of particular senses to occurrence in a small set of collocations or minor constructions.
Blust was impressed by the correspondence between the glosses for reflexes of *banua in Iban, a language of Borneo, and ’Are’are, of Malaita in the SE Solomons. There is an ethnographically rich dictionary of Iban (Richards 1981) which gives the following definition:
menoa/menua area of land held and used by distinct community, esp. longhouse (rumah), including house, farms, gardens, fruit groves, cemetery, water and all forest within half a days journey. Use of the menoa is only gained and maintained with much effort and danger, and by proper rites to secure and preserve a ritual harmony of all within it and the unseen forces involved; home, abode, place, district, country, region; menoa laŋit ‘the heavens, abode of Petara and other deities.’
The definition of ’Are’are hanua given by Geerts (1970) mentions a very similar list of elements. Blust observes that in both Iban and ’Are’are the reflex of *banua
refers to an inhabited territory that includes not only the human population and dwellings, but also plant and animal forms that contribute to the maintenance of the human community. Drinking water is mentioned in both glosses, as well as the burial sites of the deceased….
He concludes that PMP *banua (and by implication POc *panua) had a single but complex meaning.
PMP *banua, then, probably referred to an inhabited territory which included the village and its population together with everything that contributed to the life-support system of that community (Blust 1987:100).
| Sense | Usage notes | |
|---|---|---|
| 1 | ‘Land, ground, as opposed to sea’ (waitaci), ‘water’ (wai, ruwai) or ‘sky’ (lomālagi). near syn. qwele. | Contrasts with locative phrases i vōvō, ‘ashore, on land’ (in contrast to ‘at sea’) and i ata, ‘inland ‘(in contrast to ‘on the coast’). |
| 2 | Land in the sense of land mass, large tract of land, territory, country. | This sense may take modifiers indicating attributes, such as being fertile, stony, uninhabited, mountainous. |
| 3 | Homeland, someones home region or country. | Requires a possessive pronoun. |
| 4 | A particular delimited place, spot, area, district, region, zone. | Near synonym tiki. |
| 5 | A community, the people belonging to a place, a land-owning kin-group. Refers to a collective ‘people’, never to a single person. | Ara sā sevutia na ledra tovatova i na vanua. ‘They presented the first fruits of their gardens to the community.’ |
| 6 | The chief of a community, the representative of a community. Only metaphorically, in compounds such as bilo ni vanua (lit. ‘cup of the community’), ‘high chief, one who has been formally installed’ and aqona ni vanua (lit. ‘kava of the community’), ‘the first cup of kava in a ceremony, drunk by the chief’; ‘kava ceremony to welcome a visiting chief or for installing a new chief’. | Ei na rugutia vinā me sā somia na aqona ni vanua. ‘He’s well suited to be made chief.’ (lit. ‘He’s well suited to drink the kava of the community’). |
| 7 | A political confederation of clans under a chief. | (Possibly borrowed from Bauan Fijian.) |
| 8 | As the subject in certain verbal constructions concerning atmospheric, climatic and living conditions: the world, atmosphere, that which is subject to the diurnal cycle, weather and climate, e.g. bogi na vanua, ‘be dark, night’, qwataqwata na vanua, ‘be dawn’, siga na vanua, ‘be daylight, sunny, clear’. | qwele, ‘ground, land’; tiki ’place, region’, cannot be substituted in these constructions. |
This complex but unitary meaning, he suggests, is ‘fragmented’ into various more specific senses. Our view is that a stronger case can be made for treating POc *panua as a highly polysemous term, whose senses are distinct lexical units, differing from each other in features of grammar and in their semantic relations (synonymy, antonymy, etc.) to other lexical units. While the standard tests for polysemy cannot be applied directly to POc we can in principle apply them to contemporary languages for which the data allow systematic treatment of sense discriminations.
In Wayan, a dialect of Western Fijian, some eight senses of vanua can be distinguished (Pawley and Sayaba 2022). Most, if not all of these can be shown to contrast by one or another diagnostic criterion, as indicated in Table 5.2.
We can go a fair way towards making sense of the great diversity of glosses in the daughter languages by assuming that POc *panua had a range of senses corresponding roughly to those shared by the Fijian languages and certain languages of various other subgroups, including Muyuw of Papuan Tip, ’Are’are of S.E. Solomonic, Tongan and Tikopia of Polynesian, and Rotuman.7 These are shown in Table 5.3.
The senses attributed to *panua do not include ‘settlement’ or ‘village’. Closest to these is 3(a) ‘territory belonging to a community, inhabited place’. It seems likely that in POc this broader sense encompassed habitation sites and their residents but that in certain daughter languages it was narrowed to refer specifically to the cluster of buildings and associated features that make up a village. This development in turn provided a platform for a further narrowing to ‘house’ in a number of languages.
To sum up, the comparative lexical evidence assembled in sections 5.4 and 5.5 does not tell us whether POc speakers occupied sizeable villages or dispersed hamlets.
| Sense | Distribution | |
|---|---|---|
| 1 | Land in the sense of land mass, large tract of land, territory, country. This sense may take modifiers indicating attributes, such as being fertile, stony, uninhabited, mountainous. | Adm, PT, MM, SES, NCV, Mic, Fij, Pn |
| 2 | A land mass or defined territory and whatever features are an integral part of it (forests, lakes, rivers, settlements, etc.). | PT, Mic, Fij, Pn |
| 3a | Territory belonging to a community, inhabited place. | PT, SES, Fij, Pn |
| 3b | One’s homeland, home place. | PT, Fij, Pn |
| 4 | Community associated with a territory, people of a community. | NNG, MM, Fij, Pn |
| 5 | Place, area, district, region. | NNG, PT, SES, NCV, Fij, Pn |
| 6 | (in certain multi-word expressions) The world: that which is subject to the day-night cycle, weather and climate. | NNG, PT, SES, Fij, Pn |
| Language | Phrasal expression + gloss | Notes | |
|---|---|---|---|
| NNG: | Manam | anua izama ‘morning, daybreak’ | anua now means ‘village’ |
| PT: | Motu | hanua-boi ‘night’ (N) | |
| hanua idaradara ‘evening glow’ | -dara ‘ascend’ | ||
| SES: | Sa’a | sato e kʷaʔalie henue ‘the sun has risen’ (lit. ‘sun rises on (the) world’; kʷaʔali- of heavenly bodies, ‘to rise on s.t.’) | from POc *panua ‘land‘ + *boŋi ‘night ‘ (hanua now means ‘village’ |
| SES: | Lau | fanua sato ‘sunny weather’ | sato ‘sun’. Lau fanua no longer refers to land or settlements but occurs in several compounds specifying weather conditions |
| Fij: | Bauan | sā boŋi na vanua ‘it is night time, nightfall’ | [is night the world] |
| sā karobo mai na vanua ‘it is twilight’ | [is dusk the world] | ||
| sā siŋa na vanua ‘it is daylight, it is sunny’ | [is day/sun the world] | ||
| ā siŋa-levu na vanua ‘it is midday/the sun is high’ | [is big-sun the world] | ||
| Fij: | Rotuman | hanua ræn ‘daylight, dawn’ | [world day] |
| hanua ke pöŋ ‘(until) nightfall’ | pöŋ ‘night’ | ||
| Pn: | Rennellese | henua pō ‘night time’ | henua ‘land, people of the land’ |
| Pn: | Tikopia | ku pō te fenua ‘darkness has come’ | [has become dark the world] |
We note in passing that sense 6 attributed to POc *panua, the world of the diurnal cycle and weather, is present in various Oceanic languages in certain multiword expressions containing reflexes of *panua. A sample is shown in Table 5.4.
From this material we can reconstruct a family of POc verbal constructions of the type of *qaco na panua ‘be(come) daylight, be sunrise’ and *boŋi na vanua ‘be(come) dark, nightfall’ (where *qaco and *boŋi, otherwise ‘sun’ and ‘night’, are verbs), and a parallel set of complex nominal constructions of the type of *panua qaco ‘sunrise, sunny conditions, daytime’, *panua boŋi ‘nightfall, night time’ (vol.2:40-41, 295).
Ann Chowning has written as follows of the range of variation in historically attested systems of land tenure in Melanesia.
Title to land is usually vested in a corporate group, membership of which is likely to be based on descent, residence or some combination of the two….
The details of systems of land tenure [in Melanesia] differ greatly from society to society. Some permit permanent alienation and individual ownership; others do not. It is not uncommon to find distinctions between gardening land, village land or house sites, and bush land, with different systems of rights applied to each, not to mention the rights that apply to sacred places, grave sites, paths, water supplies, sago swamps, and fishing areas. (Chowning 1977:39)
The strongest candidate for an early Oceanic term for a landholding corporate group is *kainaŋa (§4.1.2.6), which has reflexes in Micronesian and Polynesian. Goodenough (1955) observed that in Micronesian languages this term typically refers to a land-owning matrilineal descent group. Bender et al. (2003a) offer the following, less precise semantic reconstruction for PMic:
| PMic | *kayinaŋa | ‘clan, folk, tribe, stock’ (Bender et al. 2003a; Goodenough 1955) | |
| Mic | Chuukese | kainaŋ | ‘matrilineal descent group, clan’ |
| Mic | Puluwatese | yayiŋan | ‘clan’ |
| Mic | Lamotrek | kailaŋ | ‘a named exogamous matriclan’ |
| Mic | Woleaian | gairaŋe | ‘clan, tribe, tribal division’ |
| Mic | Ponapean | kɛynɛk | ‘clan, lineage, extended family’ (final -k unexpected) |
For PPn we cite the reconstructions proposed by Marck (2010) and Kirch and Green (2001), which slightly modify those proposed by Koskinen (1960), Marck (2008) and Pawley (1982a, 1985).
| PPn | *kainaŋa | (1) ‘descent group, headed by an *qariki “chief”’; (2) ‘the subjects of a chief, the common people’ (§4.1.2.6; Marck 2010); ‘a land-holding exogamous descent group tracing descent from a common ancestor and headed by an *qariki’ (Kirch and Green 2001) 8 | |
| Pn | Tongan | kainaŋa | ‘populace, people without chiefly rank’ |
| Pn | East Uvean | kainaŋa | ‘people not of chiefly rank’ |
| Pn | Anutan | kainaŋa | ‘clan, membership based mainly on patrilineal descent’ |
| Pn | Tikopia | kainaŋa | ‘clan, a non-exogamous descent group consisting of exogamous lineages’ |
| Pn | Rennellese | kainaŋa | ‘subject of a chief’ |
| Pn | Pukapukan | keinaŋa | ‘maternal sublineage, headed by its oldest member’ |
| Pn | Hawaiian | maka ʔainana | ‘populace, common people (in contrast to those of noble birth)’ |
| Pn | Marquesan | mata ʔeinaŋa | ‘people, the people, subjects’ |
| Pn | Rarotongan | mata kainaŋa | ‘a settlement, the inhabitants of a district’ |
| Pn | Tahitian | ʔeinaʔa | ‘a body of followers, servants, people united by the same service’ |
PPn developed in the various islands of the Tonga-Samoa area after these were settled late in the first millennium BC, probably as a dialect network that remained quite cohesive for many centuries. Marck argues that PPn social organisation probably showed regional differences, with some smaller, less stratified island communities retaining the original use of *kainaŋa to refer to unilineal descent groups (as in Anuta, Tikopia, Pukapuka) and larger, more stratified societies developing cognatic descent groups, leading to a shift in the meaning of *kainaŋa from ‘descent group headed by a chief’ to ‘subjects of a chief, populace, commoners’. His arguments resemble those of Burrows (1939) who inferred that early Polynesian social grouping “consisted of descent groups which occupied and controlled territories and that this system was transformed repeatedly and in various ways” in different parts of Polynesia (Kirch and Green 2001:208).
The term *tau ‘person, human being’ is well attested in both PMP and POc. A number of compound nominals containing *tau have been reconstructed, including POc *tau mate ‘dead person, corpse, ghost’, *tau paqoRu ‘young adult of marriageable age’ and PEOc *tau tasi ‘fisherman, expert mariner’ (Pawley 1985; vol.5:39–42). In some compounds *tau has the sense of ‘owner’ or ‘person intimately associated with an entity’, e.g. POc *tau (ni) waga ‘owner of a canoe’ and PPn *tau fale ‘owner or occupant of a house’. This use of *tau is often associated with phrasal constructions reflecting the form *tau ni N, where the linker ni marks an associative relation, e.g. Dobuan, a Papuan Tip language, has such terms as to-ni-ʔasa ‘owner of a village’, to-ni-butu ‘owner of a feast’, i.e. ‘master of ceremonies’, to-ni-to-ni-bwaʔa ‘sprites’, lit. ‘little owners of the land’ (Grant 1953) and Molima, another Papuan Tip language, has to-ni-bwaʔo ‘owner of a garden’ and to-ni-waga ‘canoe owner’ (Chowning 1958).
The compound *tau panua ‘native of a place, land owner’ is reconstructable for PEOc. Although no reflexes have so far been noted beyond EOc, it is likely that it was present in POc. Both the constituents and the construction type are attributable to POc.
| PEOc | *tau panua | ‘person belonging to a place, land owner’ | |
| SES | Arosi | au henua | ‘man born in and belonging to the place’ |
| SES | Sa’a | eu-henue | [N] ‘householder, neighbour’; [V] ‘be a native of a place, be a resident’ |
| NCV | Mota | ta-ɣ-vanua | ‘joint owner of a village’ (i.e. ‘one of the land-owning locals’; -ɣ- reflects an earlier construct linker *ki) |
| Pn | Samoan | tau-fanua | (1) ‘commoner (as opposed to chief)’; (2) ‘owner of land, landlord’ (in contrast to tau-fale ‘householder, prospective owner of house under construction’) |
| Pn | Tikopia | tau-fenua | [N] ‘wealthy man’; [ADJ] ‘wealthy’ |
Some Polynesian languages reflect a structurally parallel, functionally equivalent compound, *taŋata (qi) fanua, in which *tau ‘person, owner’ is replaced by *taŋata ‘person’.
| Pn | Tongan | taŋata ʔi fonua | ‘native, person who really belongs to the country’ |
| Pn | Tikopia | taŋata fenua | ‘man of the land, man of status’ |
| Pn | Māori | taŋata fenua | ‘land owner, native of a place’ |
From *panua, sense 3, and the compound *tau panua we can draw the unsurprising inference that POc speakers were divided into communities with recognised territories over which members had rights.
Linguistic evidence cannot definitively answer all the questions asked at the outset of this essay but it can tell us some things about Proto Oceanic speakers settlement patterns and relation to territory.
A strong prima facie case can be made that the POc speakers preferred to live close to the sea. Fishing and reef foraging was central to the economy and there was an extensive vocabulary relating to canoes and sea travel. Three kinds of buildings can be identified by name: *Rumaq, main dwelling house, *kamaliR, men’s meeting house and *pale, a less substantial building, such as a shed for storage or other non-residential purposes. *malaqai may have referred to a village plaza or public space in a settlement, but the semantic reconstruction is not secure. A term for canoe shed, *(a-)v(a,o)lau, is reconstructable for PCP but not for POc. There is no term attributable to POc whose primary sense was ‘village’ or ‘settlement’ and it is unclear whether POc speakers lived in scattered hamlets or substantial villages; *panua appears to have been used to refer to any inhabited place, as well as to the whole territory belonging to a community, including land cleared for gardens, and to the people of a place. The compound *tau panua ‘person belonging to a place, land owner’ is reconstructable for PEOc and on logical grounds it can be inferred that it was present in POc but lost in non-EOc languages.
A term for a land-owning descent group, *kainaŋa, is attributable to PROc, being reflected in both Micronesian and Polynesian languages but not in Western Oceanic, SE Solomonic or the Admiralties. As such, this term can be associated with the bearers of Lapita culture who moved into Remote Oceania but not with the Early Western Lapita tradition found in the Bismarck Archipelago.