The Oceanic Subgroup and Proto-Oceanic Culture

The linguistic evidence indicates that there was just a single early movement of Austronesian speakers into northwest Melanesia that left linguistic traces.

The need to recognize what we now call the Oceanic subgroup of Austronesian was first expressed by the Dutch linguist Hendrik Kern (1886) in a discussion of the relationships of Fijian to several Indonesian and Polynesian languages, but the foundations of modern Oceanic comparative work were not truly laid until the 1930s, when the German linguist, Otto Dempwolff, published his major work (1934, 1937, 1938) on the Austronesian family. In the second volume of this work Dempwolff reconstructed a sound system for the immediate common ancestor of the Polynesian languages and a sample of eighty-two “Melanesian” languages and found that all the languages in his sample reflect a set of sound changes to the Proto-Malayo-Polynesian (“Uraustronesisch”) system he had reconstructed earlier. The quantity of these shared phonological innovations led Dempwolff to conclude that the “Melanesian” and Polynesian languages form a subgroup apart from the Austronesian languages to the west, in the Indo-Malaysian archipelago. Dempwolff labelled the interstage language ancestral to this subgroup “Urmelanesisch”. Another German scholar, Wilhelm Milke, coined the more appropriate “Proto-Ozeanisch” in 1961, and since then “Proto-Oceanic” has been used. Even though some of Dempwolff’s putative phonological innovations have been eliminated as a result of subsequent research, the evidence for Oceanic has on the whole been strengthened and the precision of our phonological, grammatical and lexical reconstructions has improved.[6]

The region covered by “Oceania” in this context is the Pacific east of a line drawn from north to south and dividing Chamorro (Mariana Islands) and Belau (formerly Palau) from the rest of Micronesia and crossing the north coast of Irian Jaya at 138°E longitude. This means that the seam between Oceanic and its closest Austronesian relatives is in the west of New Guinea between the Bird’s Head and the Sarmi Coast. The Oceanic subgroup includes the Austronesian languages of all of Melanesia except the extreme west of New Guinea, all of Polynesia and most of Micronesia (see Map 1). Apart from languages brought by colonialism, languages of Oceania which are not Austronesian are found only in New Guinea and nearby archipelagoes. These are the so-called non-Austronesian or “Papuan” languages of many of the peoples of New Guinea and some island peoples as far east as Savo in the Solomon Islands (and a probable outlier area in the Reef Islands).

Reconstructed Oceanic terminologies give insight into the culture of POc speakers and of the speakers of various interstage languages.[7] It is clear that POc speakers preserved a very high proportion of Proto-Malayo-Polynesian and Proto-Eastern Malayo-Polynesian terms for a range of important cultural domains. The inference must be drawn that there was continuity in many components of the way of life of Austronesian speakers from a Proto-Malayo-Polynesian dispersal centre in Island Southeast Asia to the POc dispersal centre in northwest Melanesia, over a period of uncertain duration, perhaps 1000 years. The lexical reconstructions point to what archaeologists (e.g. Bellwood 1989) refer to as initial “founder settlement” of northwest Melanesia by Austronesian speakers.

Elaborate terminologies for seafaring and fishing and for horticulture and pottery indicate that the economy of POc speakers was based on both the sea and the land, and that some people or some local groups were specialist fishermen, farmers or potters (Pawley 1981; Pawley and Green 1984). Table 2, for example, lists reconstructed POc terms for canoe parts, seafaring and the sea (mostly from Pawley and Pawley 1994). Terms can also be reconstructed for various winds (Ross 1994d) and for numerous fish and sea creatures (Walter 1989; Pawley 1993). Table 3 shows the reconstructed forms for the growth stages of the coconut and for parts of the coconut, whilst Table 4 gives terms for food plants other than the coconut as well as some other horticultural terms (terms in both tables are from French-Wright 1983 and Ross 1993). Lichtenberk (1994) has reconstructed terms associated with food preparation (e.g. *tabiRa ‘wooden dish’, *qumun ‘stone oven’) and Ross (1994c) terms for pots and pottery; some of these are listed in Table 5. We can also reconstruct the names of various land animals (*boRok ‘domesticated pig’, *bawe probably ‘wild pig’, *kadroRa ‘cuscus’, *mwaj[oa] ‘bandicoot, marsupial rat’, *kasuari ‘cassowary’). Reconstructed terminologies can also illuminate social organization, exemplified by the kinship terminology of Table 6 (based on Milke 1958; Chowning 1991): it is noteworthy that there is a gap in the system at ‘father’s sister’ and that only two affinal terms have so far been reconstructed.

Map 1: Boundaries of the Oceanic subgroup of Austronesian languages.

Map 1: Boundaries of the Oceanic subgroup of Austronesian languages.

Table 2. Reconstructed Proto-Oceanic terms for canoe parts and seafaring.(a)

Types of vessel:



large sailing canoe; outrigger canoe (generic)


small outrigger canoe (?) or canoe hull

Parts of a canoe:


*baban, *bapan

plank; canoe plank or strake


thwart; collar-beam in house


strake, probably topstrake (washstrake)


platform of any kind, inc. one erected on outrigger framework


outrigger float


outrigger boom


connective sticks attaching float

*kata(q)e, kate(q)a(b)

free side of canoe, opposite the outrigger




mat; matting sail


boom or yard of (triangular) sail


rear, stern; back of any object


front, bow of boat; front of any object


nose; projecting headboard of prow(c)


snout; projecting headboard of prow(c)

Canoe accessories:



(canoe) paddle


to paddle

*lima(s), nima(s)



scoop or ladle out; ladle, bailer




place rollers under a boat

*ujan, *lujan

to load (a boat); cargo, freight


to steer; rudder


be anchored or moored, be stationary

Canoe making:







to adze, carve


(a) In this and the following tables reconstructions are given in the orthography of Ross (1988), with the addition of the phoneme *pw . Conventions in proto-forms are:

(x) it cannot be determined whether x was present;

(x, y), xx/yy either x or y;

[x] the item is reconstructible in two forms, one with and one without x;

[x, y] the item is reconstructible in two forms, one with x and one with y;

xx, yy both x and y.

-xx xx is always suffixed (in most cases by an inalienable possessor suffix).

(b) The parentheses indicate that the *-q- is not reflected in our data, but is required by the rules governing the canonic form of POc words.

(c) Often with ornately carved figurehead.

Table 3. Reconstructed Proto-Oceanic forms associated with coconut culture.

The coconut and its stages of growth:


coconut (generic); growth stage: ripe, brown but not fallen yet


growth stage: fruit bud


growth stage: young, green


growth stage: green, drinkable


growth stage: ripe, brown but not fallen yet

*maRaŋo, *goRu

growth stage: dry and ready to fall


growth stage: sprouted

Parts of the coconut palm and of the coconut:


coconut water

*bwilo, lasa

coconut shell used as liquid container or cup


coconut embryo


spongy centre of a sprouting coconut

*punut, *penut

coconut husk, fibres on coconut husk


to husk a coconut


frond of a palm


palm branch


dry coconut leaf torch

*Runut, *nuRut

sheath around base of coconut frond, used as a strainer

Table 4. Reconstructed Proto-Oceanic terms for horticulture and food plants (other than coconuts)

Tubers and their culture:


taro (possibly all Araceae)


taro, Colocasia esculenta


giant taro, elephant ear taro, Alocasia macrorrhiza


swamp taro, Cyrtosperma chamissonis


k.o. wild taro (?)


greater yam, Dioscorea alata; yam (generic)


potato yam, aerial yam, Dioscorea bulbifera


banana or taro sucker, slip, cutting, shoot (i.e. propagation material)


taro stem (used for planting)


new leaves or shoots, or taro tops for planting


taro seedling


to plant (tubers)


to cut off taro tops




banana, Musa cultivars


banana, Australimusa group


k.o. cooking banana: long with white flesh (presumably Eumusa group)

Other food plants:


sugar cane, Saccharum officinarum


a kind of edible wild cane or a reed, Saccharum spontaneum


cucurbit (generic); cucumber, Cucumis sativus


ginger, Zingiber officinale


turmeric, Curcuma longa


breadfruit, Artocarpus altilis


breadfruit fruit (?)


pandanus (generic); coastal pandanus, Pandanus tectorius


coastal pandanus, Pandanus tectorius


Pandanus dubius


k.o. pandanus with useful leaves


sago, Metroxylon spp., mainly Metroxylon sagu


sago starch


sago fronds, thatch


Java almond, Indian almond, Terminalia catappa


Tahitian chestnut, Pacific chestnut, Inocarpus fagifer


canarium almond, Canarium spp.


citrus fruit or citrus-like fruit


mango, probably Mangifera indica

*wai, *waiwai

mango (generic)


Malay apple and rose apple, Eugenia spp.


Morinda citrifolia


Pometia pinnata


edible greens, Abelmoschus manihot


wild cinnamon, Cinnamomum spp.


Polynesian plum, hog plum, Tahitian apple, Spondias cytherea


k.o. tree with avocado-like fruit and hard wood, Burckella obovata


New Guinea walnut, Dracontomelon dao


areca palm, Areca catechu

Gardening practices:




soil, earth


to clear ground for planting


to dig a hole


to plant

Some prehistorians evidently find the methods of historical linguistics so arcane or the idea of such detailed lexical reconstructions so incredible, that they prefer to ignore or discount the reconstructions as irrelevant to prehistory. This attitude is no more excusable than that of a linguist who would ignore C14 dates for artefact assemblages because he does not understand how such dates are arrived at or who would discount the relative dating of assemblages in any archaeological site on the suspicion that worms, humans or earthquakes have disturbed the layers.

Table 5. Reconstructed Proto-Oceanic terms associated with pots and pottery.

Types of vessel:



cooking pot, pot (generic)


frying pan


k.o. large pot

Parts of vessel:


*tupa((n, ŋ))

lid, cover

*joŋa(n, ŋ)

plug, bung, stopper


herringbone pattern

Pottery manufacture:





mould (clay etc.)


paddle for beating clay into shape


coil around, encircle; add strip of clay around top of pot

Pottery use:



roast in the fire, fire (pot)


cook by boiling


steam, boil



Table 6. Reconstructed Proto-Oceanic kinship terms.

Consanguineal terms:



grandparent, grandchild




father, father’s brother


mother, mother’s sister


mother’s brother


mother’s brother, sister’s child


child, same-sex sibling’s child


older same-sex sibling


younger same-sex sibling


brother (woman speaking)


sister (man speaking)

Affinal terms:





spouse’s opposite-sex sibling

Note: The problem of reconstructing two terms for ‘mother’s brother’ is discussed by Chowning (1991:65).