It is claimed that one of the distinguishing features of the Lapita culture is the movement out into the Pacific of a wide range of domesticated plants, and that Lapita represents, in the Bismarcks and Solomons, the first clear evidence for an agricultural base to the economy, a base apparently little changed until this century (see Pawley & Green 1973, Shutler & Marck 1975, Spriggs 1997a).1
Early this century, ethnographers such as Malinowski (1935a, 1935b) for the Trobriands, Blackwood (1935) for north Bougainville and Ivens (1927 (reissued 1972)) for the Southeast Solomons, described communities in which the people were, above all, gardeners. Their gardens provided a basis for their settlement size and determined the nature and division of labour. Much of the ritual and magic practised related to the stages of cultivation and growth and harvest. Their produce excessive to daily needs fuelled their festivals, it was their standby in times of food scarcity, and their trading currency. Reciprocal obligations to kin were expressed in foodstuffs exchanged. Years were calculated in terms of annual crop cycles—an event might have ocurred two yam crops ago or could be expected after the next galip nut harvest.
Over one hundred POc food plant terms have now been reconstructed (see Food Plants in vol 3, an earlier version of which has been published as Ross 1996d). Here, however, we are concerned, not with the terms for the plants themselves (these are listed where relevant, but without supporting data) but rather, with terms for gardens and for activities associated with gardening.
A considerable debt is owed to the unpublished 1983 thesis of Renwick French-Wright, Proto Oceanic Horticultural Practices, a work which travelled much the same road as this chapter does, and many of whose cognate sets have been included here. Because more stringent subgrouping requirements apply today before a reconstruction can be identified as POc, some revision of his reconstructions has been necessary—usually involving downgrading from POc to PEOc.
Almost all of Melanesia lies within the humid tropics and experiences high rainfall. Apart from areas round Port Moresby and the west coast of New Caledonia, which receive less than 1,000 mm per annum, most of Melanesia averages above 2,500 mm, while south-west New Britain receives over 6,000 mm. Seasonal variation of rainfall is due primarily to the changing trade wind cycle, and can best be described as variation between ‘fairly wet’ and ‘very wet’, although major droughts can occur from time to time. As Ann Chowning (pers.comm.) points out, one of the primary components of garden magic is weather magic. There is little seasonal variation in temperature. Most Austronesian settlement is close to the sea, where maximum temperatures are around 30-32 °C, with minima around 23 °C. Associated with these factors are high humidity and cloudiness (McAlpine, Keig & Falls 1983: 3, 61-66). The natural vegetation cover is tropical forest. Even in the drier areas, savanna woodland is the natural plant cover.
There is broad general agreement that the Austronesian-speaking people who first ventured into Melanesia practised a swidden style of gardening, centred on the cultivation of tubers, particularly taro (Colocasia esculenta) in the wetter areas and various varieties of yam (Dioscorea) in the drier regions (Brookfield & Hart 1971, Yen 1973). Some of their practices, however, may have been adopted from Papuan speakers who had practised agriculture in New Guinea for millennia before Austronesian speakers arrived. Brookfield and Hart (1971) made an exhaustive survey of forty-four places, fourteen of which are Austronesian settlements2, in order to identify, as best they could, the environment and land use in ‘Old Melanesia’, the region as it was before external forces began to radically modify the man-environment systems. These provide a good representative sample of traditional methods across the region, and incidentally cover the major linguistic subgroups from New Guinea to Fiji, together with the Samoic outlier of Tikopia. The factors they considered were location (distance from the sea), elevation, topography, rainfall regime, biotic environment (soil and vegetation type), wild food sources, traditional crops, cultivation methods, cultivation frequency and crop segregation. With the exception of the locations in Fiji, New Caledonia and Tikopia, which have developed more intensive systems, all these places employed what was ‘a set of variations on the classic swidden or shifting cultivation system.’ (p. 108)
All gardens are mixed, although usually dominated by a single staple, and crops are planted by making only small holes in the soil, sometimes with limited spot-tillage by levering the soil with a stick before inserting the seed tuber or cutting, or occasionally, as at Roviana, making small mounds. Gardens are generally used only once, though sometimes subsidiary crops are planted after the main crop, and remain longer in the ground. Useful trees continue to provide fruit for some time after the garden has been relinquished to wild growth. Crops are usually planted among the fallen trees and ash … [although] ash is sometimes thrown out of the garden before planting, or else piled around stumps where it is of no direct use to the plant. Occasionally ash is piled around particular plants which are thought to benefit from such treatment, but not around others.
There is sometimes a little crop segregation, but it is quite subsidiary … Most gardens are untidy, littered with the fallen trunks of trees and quickly invaded by weeds. Because of wild and domesticated pigs, most gardens are fenced. Weeding is sometimes done in the early months but is never continued beyond the lifting of the main crop. (Brookfield & Hart 1971: 107-108)
Rarer techniques within what was still essentially a swidden system included tillage in some grassland areas to eradicate grassroots and aerate the soil; the planting of yams in specially prepared pits which were composted with a mixture of leaves, grass and soil (e.g. in Mapos Buang (NNG)); and employing rudimentary irrigation (e.g. in Seniang (NCV)). Locations in Fiji, New Caledonia and the Solomons were distinguished by more advanced irrigation systems, permitting some continuous cultivation of the same plots.
For swidden agriculturalists, land could be classified in one of three ways—as land under cultivation or ready for cultivation, as old garden land now lying fallow, and as uncleared ‘bush’ or forest. The following reconstructions illustrate.
| PAn | *qumah | ‘swidden; work a swidden’ (ACD) | |
| POc | *quma | ‘garden; to clear land for a garden’ (Grace 1969) | |
| NNG | Gedaged | um | ‘garden, cultivated land’ |
| NNG | Adzera | gum | ‘garden’ |
| NNG | Mengen | kume | ‘prepare a garden’ |
| NNG | Sengseng | kum | [v] ‘work (in general)’ |
| PT | Molima | ʔuma | ‘planted garden’ |
| PT | Sudest | uma | ‘garden’ |
| PT | Motu | uma | ‘garden, enclosed, cultivated plot’ |
| MM | Bulu | ɣ⟨in⟩uma | ‘garden’ (< POc *q⟨in⟩uma ⟨NOM⟩garden) |
| MM | Nakanai | (ma)huma | ‘garden’ |
| MM | Tolai | uma | ‘garden’ |
| MM | Kia | (n-un)uma | ‘garden’ (< POc *na ART + *q⟨in⟩uma ⟨NOM⟩garden) |
| MM | Roviana | uma | ‘make a garden’ |
| MM | Roviana | (in)uma | ‘a garden’ (< POc *q⟨in⟩uma ⟨NOM⟩garden) |
| SES | Gela | uma | ‘clear away the bushes in making a garden’ |
| SES | Arosi | umʷa | ‘weed a garden’ |
| NCV | Mota | umʷa | ‘clear away growth from a garden, first stage of preparation’ |
| NCV | Nguna | uma | ‘cut bush, clear land’ |
| Fij | Wayan | uma(ni) | ‘turn the soil over’ |
Three MM languages (Bulu, Kia and Roviana) reflect POc *q⟨in⟩uma ⟨NOM⟩ + ‘garden’ (on nominalisers, see Ch. 2, §3.2.1), suggesting perhaps that at least in the early MM linkage, *quma was a verb meaning ‘make a garden’, *q⟨in⟩uma the product of that activity. The following item betrays a similar pattern.
| POc | *(b,bʷ)aku(r,R) | ‘work a garden’ (Dempwolff 1938: POc *(p,b)agur (V) ‘hoe’) | |
| PT | Gumawana | bagula | ‘work in a garden’ |
| PT | Gumawana | baguli | ‘plant (s.t.)’ (perhaps < POc **(b,bʷ)aku(r,R)-i) |
| PT | Kalokalo | bagula | ‘garden’ |
| PT | Bwaidoga | bakula | ‘garden’ |
| PT | Sudest | bakubaku | ‘ground (cleared near house)’ |
| PT | Kilivila | bagula | ‘work in a garden, make a garden’ |
| SES | Arosi | bʷaʔu | ‘tableland above the shore where the gardens are’ |
Ross (1994a) has reconstructed Proto Central Papuan *v⟨in⟩ayula ‘do work’ (Hula inayulu, Lala vinaula, Kuni bilaula, East Mekeo pinauna). Central Papuan languages reflect POc final consonants only in borrowings from other PT languages, so *v⟨in⟩ayula points to PPT *b⟨in⟩aqura, POc *(b,bʷ)dmaku(r,R). The presence of the nominaliser *⟨in⟩ suggests that *(b,bʷ)aku(r,R) was a verb.
| PMP | *talun | ‘fallow land’ (Blust 1972b; Dempwolff 1938) | |
| POc | *talu(n) | ‘fallow land, land returning to secondary growth’ | |
| SES | Gela | talu | ‘forest land which has been previously cultivated’ |
| SES | Kwaio | alu | ‘garden of second or third crop’ |
| SES | Kwaio | alu (sīsī) | ‘an old garden plot returning to secondary growth, beginning to be overgrown’ |
| SES | Sa’a | alu | ‘last year’s yam garden’ |
| SES | Arosi | aru | ‘an overgrown garden; land formerly used for a garden; a dug garden’ |
| Pn | Niuean | talu-talu | ‘land out of cultivation’ |
| Pn | Rennellese | tagu-tagu | ‘begin to be brush-covered, of a fallow garden’ |
| Pn | Māori | taru-taru | ‘weeds, herbs’ |
| PAn | *quCaN | ‘fallow land’ (Blust 1989: 174) | |
| POc | *qutan | ‘bushland, hinterland’ | |
| Adm | Mussau | utana | ‘garden’ |
| NNG | Manam | (a)uta | ‘inland’ |
| PT | Motu | uda | ‘bush, forest’ |
| PT | Bwaidoga | ɣudana | ‘forest’ |
| PT | Misima | ulan | ‘forest’ |
| MM | Nakanai | huta-huta | ‘general term for small plants and leaves; trash’ |
| SES | Tolo | uta | ‘garden’ |
| NCV | Mota | uta | ‘bush, forest, unoccupied land; the inland country’ |
| NCV | Nguna | uta | ‘inland’ |
| NCV | Southeast Ambrym | ut | ‘place, area, land, shore, island, homeland, weather’ |
| Mic | Kosraean | wət | ‘area inland or towards the mountains’ |
| Fij | Rotuman | ufa | ‘land (from the sea); interior (from the coast)’ |
| Pn | Tongan | ʔuta | ‘inland (from shore); shore, land (from sea)’ |
| Pn | Hawaiian | uka | ‘inland (from shore); shore, land (from sea)’ |
The Mussau and Tolo reflexes mean ‘garden’: this change of meaning is probably due to the fact that, in Melanesia, gardens are often remote from the village and surrounded by bushland, so that to go to the garden is to go into the bush.
| PEOc | *wao | ‘forest’ (French-Wright 1983) | |
| SES | Gela | ao | ‘forest, land never brought under cultivation’ |
| Fij | Rotuman | vao | ‘forest, large number of trees or big plants growing together’ |
| Pn | Tongan | vao | ‘forest, bushland, scrub, land in its natural uncultivated state’ |
| Pn | Tahitian | vao | ‘wilds, wilderness’ |
| Pn | Māori | wao | ‘forest’ |
The meaning of earth as soil would seem to be the central meaning of POc *tanoq (cf. POc *panua whose extended meanings included (1) inhabited area or territory; (2) community together with its land and things on it; (3) land, not sea; see Ch. 3, §3.7).
| PMP | *taneq | ‘earth, land’ (Dempwolff 1938) | |
| POc | *tanoq | ‘earth, soil’ | |
| Adm | Loniu | (ko)tan | ‘earth’ |
| Adm | Lou | tan | ‘loose soil’ |
| NNG | Gedaged | tan | ‘soil, ground, land, garden, earth, world’ |
| NNG | Takia | tan | ‘ground, earth, land’ |
| NNG | Kove | tano | ‘earth, sand’ |
| PT | Motu | tano | ‘earth, soil, country, land’ |
| PT | Minaveha | tano(pi) | ‘earth, ground, world’ |
| MM | Nakanai | talo | ‘down’ |
| MM | Meramera | tano | ‘down’ |
| SES | Bugotu | tano | ‘earth, ground’ |
| SES | Sa’a | ana | ‘ground, garden ground’ |
| SES | Arosi | ana | ‘ground, earth, soil, the land’ |
| NCV | Raga | tano | ‘earth’ |
| NCV | Lewo | tano | ‘earth, land’ |
| Mic | Kiribati | tano | ‘earth, ground, soil’ |
| Mic | Woleaian | tal | ‘earth, ground, soil’ |
The form POc *tano(q) given in vol.1,119 has now been revised to *tanoq. Evidence supporting final *-q lies in the retention of a final vowel in Kwamera (John Lynch, pers. comm.) and Iaai (Françoise Ozanne-Rivierre, pers. comm.).
| POc | *pʷaya | ‘(cultivable) soil’ | |
| NNG | Mengen | pae | ‘soil used to blacken teeth’ |
| PT | Kilivila | pʷe-pʷaya | ‘real soil’ |
| PT | Muyuw | pʷe-pʷay | ‘ground, land, earth, soil, dirt’ |
| PT | Molima | pʷaya-pʷaya | ‘dust’ |
| MM | Tolai | pia | ‘earth, land, dirt’ |
| MM | Ramoaaina | pia | ‘plantation’ |
| SES | Sa’a | pʷai(nā) | ‘the garden ground just above the beach’ |
In some Central Pacific languages, reflexes of *gʷele carry a similar range of meanings.
| PCP | *gʷele | ‘earth, soil’ | |
| Fij | Wayan | gʷele | ‘earth’ |
| Fij | Bauan | gele | ‘earth, soil’ |
| Pn | Tongan | kele-kele | ‘land, dirt, soil, earth, ground’ |
| Pn | Rennellese | kege | ‘earth, ground, dirt, land, soil, world’ |
| Pn | Māori | kere- | ‘earth (in compounds only)’ |
New gardens are traditionally prepared along the following lines. First, the undergrowth and vines and small trees are cleared. Before the introduction of metal implements, this would have been done with a large digging stick used as a crowbar (§5.5) and perhaps a stone axe (POc *kiRam, Ch. 4, §4.1.1), together with smaller cutting implements made of bamboo or shell. Large trees may be left standing, although branches may be lopped off to admit maximum sunlight. Alternatively, they may be slowly killed off by ringbarking or by burning rubbish heaped round their base. Debris is usually left to dry and then burnt, with the ashes fertilizing the soil. The ground can then be raked or picked over, and remaining rubbish removed, together with any weeds which have survived the burning.
It seems from its reflexes that POc *quma meant ‘clear land for a garden’ (§3.1). In addition, the following terms are reconstructable.
| PAn | *tebaS | ‘cut, clear vegetation’ (Dempwolff (1938) reconstructed *teba ‘land appropriated by clearing of forest’) (ACD) | |
| POc | *topa | ‘land cleared for a garden; land formerly planted as a garden’ | |
| Fij | Wayan | tova-tova | ‘garden plantation’ |
| Fij | Bauan | tova | ‘flat piece of land, formerly planted with yams’ |
| POc | *salu- | ‘slash, hoe; a hoe, an adze’ | |
| NNG | Gedaged | salu | ‘hoe made out of shell’ |
| NNG | Yabem | salu(na) | ‘stone adze with small blade’ |
| SES | Gela | halo | ‘a small k.o. adze’ |
| NCV | Mota | sal | ‘cut with a slashing cut’ |
| Fij | Rotuman | saru | ‘dig (as a garden), till, break up finely’ |
| Fij | Bauan | saru | ‘strike grass and reeds with a heavy stick’ |
| Pn | Tongan | halu | ‘scarify soil’ |
| Pn | Samoan | salu | ‘brush up, as rubbish; scrape out, as coconut’ |
| POc | *poki | ‘clear the ground for a garden site’ (French-Wright 1983) | |
| PT | Lala | voi | ‘gardening (general); (clear garden in grassland after burning)’ |
| MM | Tangga | pok | ‘cut down, cut off’ (pok palaŋ ‘clear clumps of bamboo and other low scrub from a garden site’; palaŋ ‘garden’) |
| MM | Simbo | poki | ‘clear ground; garden, clearing’ |
| SES | Sa’a | hui | ‘clear a taro garden’ |
| POc | *sani | ‘complete clearing of garden; stripping, of leaves’ | |
| Adm | Loniu | cani | [VT] ‘clear, cut down bush or sugarcane’ |
| Adm | Loniu | (pono)sani | ‘clear out, sweep, straighten up a garden after heavy clearing is completed; trim trunk of tree before chopping up’ |
| PT | Motu | dani | ‘cover, as weeds a garden’ |
| Fij | Bauan | sani | ‘remove leaves’ |
| Pn | Tongan | hani | ‘remove all the leaves’ |
| POc | *sara | ‘clear (vegetation, rubbish) from a garden’ | |
| MM | Nakanai | sala | ‘scrape away garden debris with the hand’ |
| NCV | Mota | sara | ‘pass, draw along, sweep, be swept away’ |
| NCV | Nguna | sara | ‘sweep’ |
| Fij | Wayan | ðoara | ‘clear a garden (burn off dried vegetation) for planting’ |
| Fij | Bauan | ðara | ‘clear a walk of rubbish’ |
| Fij | Bauan | ðara-ðara | ‘cleared, clean of weeds’ |
| PWOc | *lamo | ‘clear a garden site’ (Ann Chowning, pers.comm.) | |
| NNG | Sengseng | lom | ‘clear a garden site’ |
| NNG | Mangseng | lom | ‘clear a garden site’ |
| NNG | Amara | lomo | ‘clear a garden site’ |
| MM | Nakanai | lamo-lamo | ‘clear a garden site’ |
| MM | Meramera | lamo-lamo | ‘clear a garden site’ |
| Adm | Baluan | lolome(k) | ‘the communal activity of planting tubers’(McEldowney 1995: 472) |
| PEOc | *kokoda | ‘gather weeds or rubbish, prepare a garden’ (French-Wright 1983) | |
| SES | Arosi | ʔoʔora | ‘prepare a garden; unplanted but prepared, of a garden’ |
| NCV | Mota | gogor(ag) | ‘gather together (weeds +)’ (-ag < POc *-aki APPL) |
| Fij | Bauan | gogot(a) | ‘gather up in handfuls; clear up rubbish’ |
| Pn | Māori | koko | ‘scoop up’ |
The following reconstruction may refer particularly to the action of burning grass, undergrowth and debris. There are other burning terms more general in their application.
| PMP | *zeket | ‘burn (fields +)’ (ACD) | |
| POc | *soko(t), *sokot-i- | ‘burn (grass, rubbish +)’ | |
| Adm | Lou | sakot | ‘burn on’ |
| MM | Bulu | roɣo | ‘(fire) burn’ |
| MM | Lihir | so | ‘(fire) burn’ |
| MM | Lihir | sakt | ‘burn (grass)’ |
| MM | Barok | soŋot | ‘burn (grass); bake (on fire)’ |
| MM | Tangga | sok | ‘(fire) burn’ |
| MM | Nehan | suk | [ADJ] ‘burnt’; [v] ‘singe’ |
| SES | Arosi | togo | ‘make up a fire, set more wood on’ |
Reflexes of POc *tunu ‘roast on embers or in fire; burn (grass +); make decorative cicatrices by burning the skin’ (Ch. 6, §3.2) are commonly used in NNG and MM languages to refer to burning off land for clearing.
| PMP | *babaw | ‘weed (a garden +)’ (ACD) | |
| POc | *papo | ‘weed (a garden +)’ (ACD) | |
| NNG | Yabem | wao(ŋ) | ‘weed’ |
| PT | Gapapaiwa | vao | [VT] ‘grow, plant’; [N] ‘garden’ |
| SES | Gela | vavo | [VT] ‘weed’ |
| SES | Lau | fofo | ‘weed with a knife’ |
| SES | Sa’a | hoho | ‘cut undergrowth’ (hohola ‘a garden cleared for yams’) |
| SES | Arosi | haho | [V] ‘weed’ |
| NCV | Mota | wowo(r) | [V] ‘weed’ |
| Fij | Bauan | vovo | ‘dig the ground between yam mounds’ |
| POc | *rabut, *rubat | ‘uproot’ (French-Wright 1983 has *(R,d)amput ‘uproot’) | |
| MM | Tolai | rubat | ‘root up, pull up by the roots, raise taro +’ |
| MM | Roviana | rabutu | ‘pull out by the roots’ |
| SES | ’Are’are | rahu-i | ‘pull out, down, of grass, shrubs, trees, plants’ |
| SES | Tolo | rubo(a) | [v] ‘weed’ |
See also POc *puti- ‘pick, pluck (feathers), pull out (weeds +)’ (Ch. 9, §6.1).
Both *kali and *keli are reconstructable for POc, and both apparently refer to digging as a general activity, including planting and digging up (tubers)3.
| PAn | *kali | ‘dig’ (Blust 1983–84a: 15) | |
| POc | *kali | ‘dig’ | |
| POc | *kali-aki- | ‘plant (garden) with (taro +)’ | |
| MM | Minigir | kali | ‘dig’ |
| MM | Tolai | kal | ‘dig (out)’ |
| MM | Tolai | kalie | ‘plant, dig in’ |
| SES | Arosi | ʔariʔae | ‘plant taro in a garden’ |
| Pn | Tikopia | kari | ‘dig’ |
| Pn | Māori | kari | ‘dig for, dig up’ |
| Pn | Hawaiian | ʔali | ‘dig’ |
| POc | *keli | ‘dig, harvest (tubers)’ | |
| NNG | Lukep | keli | ‘dig, unearth (used for the action of harvesting yams and sweet potato)’ |
| NNG | Wogeo | (i)kel-kel | ‘dig, harvest (tubers)’ |
| NNG | Manam | ʔeli | ‘dig postholes, yamholes +’ |
| NNG | Sengseng | kel | ‘dig, dig up, as yams’ |
| PT | Kilivila | keli- | ‘dig, harvest (tubers)’ |
| PT | Motu | ɣei | ‘dig, dig up (yams +)’ |
| MM | Bali | ɣeli | ‘dig, harvest (tubers)’ |
| MM | Lihir | kel | ‘dig, harvest (tubers)’ |
| MM | Roviana | ɣeli | ‘dig (yams +)’ |
| SES | Gela | ɣeli- | ‘dig, harvest a crop of yams +’ |
| SES | Lau | ʔeli | ‘dig (yams, postholes +)’ |
| SES | ’Are’are | eri | ‘dig, harvest’ |
| NCV | Raga | geli | ‘dig, harvest (tubers)’ |
| SV | Kwamera | eri | ‘dig, harvest (tubers)’ |
| Fij | Wayan | keli | ‘dig up, dig out soil’ |
| Fij | Bauan | keli | ‘dig (a hole)’ |
| Pn | Tongan | keli | ‘plant (yams)’ |
| Pn | Rennellese | kegi | ‘dig, esp. tubers’ |
| Pn | Samoan | ʔeli | ‘dig, dig up, harvest (a root crop)’ |
| Adm | Mussau | kai | ‘dig, harvest (tubers)’ |
| Adm | Mussau | (ka)kaia | ‘bury’ |
| MM | Patpatar | kil | ‘dig (with a stick)’ |
| MM | Ramoaaina | kil | ‘dig, harvest (tubers)’ |
| MM | Label | kil | ‘dig, harvest (tubers)’ |
| SES | Lau | ʔili | ‘dig, harvest (tubers)’ |
| SES | Arosi | giri | ‘dig with a stick’ |
| NCV | Lonwolwol | kil | ‘dig, dig for’ |
| SV | Lenakel | il | ‘dig, harvest (tubers)’ |
Although pigs usually have an unwanted, destructive effect on gardens, in some areas pigs are allowed to root in ground that has been harvested of its tubers. This contributes to turning over the soil in preparation for a second crop.
| POc | *suar, *suar-i- | ‘root up the ground, as pigs do’ | |
| NNG | Manam | suari | ‘dig, grub up, turn up the ground’ |
| NNG | Gitua | zuar | ‘(pig) root about’ |
| MM | Nakanai | sua | ‘push up the hot stones of an oven with a stick’ |
| MM | Nalik | suar | ‘dig’ |
| MM | Tabar | cuor | ‘plant (sweet potato +)’ |
| SES | Lau | sua-sua | ‘heap of earth thrown up by crabs’ |
| SES | ’Are’are | sua(hi) | ‘burrow, root the earth (of pigs)’ |
| SES | Arosi | sua | ‘root up earth, as a pig’ |
| Pn | Tongan | hua | ‘root in earth (of pigs)’ |
| Pn | Samoan | sua | ‘dig over the ground, to free it from roots’ |
| MM | Vaghua | sula | ‘dig’ |
Two kinds of digging stick may be physically differentiated—the longer, heavier version used to break up the surface of the soil, and the smaller dibble used to poke holes in the soil for planting. These are seldom, if ever, differentiated in word lists and it may be that most languages use the same term for both.
| POc | *waso | ‘digging stick’ | |
| MM | Nakanai | uaro | ‘digging stick’ |
| MM | Tangga | wās | ‘digging stick’ |
| SES | Lau | kwato | ‘digging stick in planting yams’ |
| SES | Sa’a | wāto | ‘digging stick (for making holes when yam planting or for husking coconuts)’ |
| SES | ’Are’are | wato | ‘stick used for husking coconuts and digging holes for planting of food’ |
| SES | Arosi | wato | ‘a digging stick, a stake for husking coconuts’ |
| NCV | Uripiv | nu-was | ‘digging stick’ |
| NCV | Mota | as | ‘pierce, stab, prick’ |
Note also Sengseng (NNG) e-suk ‘digging stick’ and East Fijian oukit-a (V) ‘root, dig up or loosen the ground with a stick’, both reflexes of POc *(su)suk-i ‘pierce, prick, sew’ (Ch. 4, §3.2.1); also the Polynesian reflexes of POc *kojom ‘husking stick’ which refer to ‘digging stick’ (Ch. 6, §5.5).
Great damage can be done to gardens wherever there are bush pigs or wallabies. Although wallabies are not found in the Solomons or places east, wild pigs are found throughout virtually all of Melanesia. In these places, preparation of a garden usually includes construction of a strong fence. In addition to POc *baRa ‘fence, wall, enclosure’ (see Ch. 3, §3.6), a term which evidently encompasses all kinds of fences and walls, there are several terms which may apply more specifically to garden enclosures made of logs or stakes.
| PAn | *qa(l,R)ad | ‘fence, palisade’ (Ross 1994a: 459) 4 | |
| POc | *qaRa(r) | ‘fence’ | |
| NNG | Numbami | ala | ‘fence (e.g. garden fence)’ |
| NNG | Kove | ala | |
| PT | Bwaidoga | ala | |
| PT | Motu | ara | ‘fence of upright stakes’ |
| NCV | Raga | ara | |
| NCV | Nguna | na-ara | ‘wall of bamboo or cane’ |
| Pn | Tongan | ʔā | ‘fence, wall, enclosure’ |
| Pn | East Uvean | ʔā | ‘palisade’ |
Ross Clark (pers.comm.) has pointed out that PPn *loto-qa is reconstructable (where *-qa reflects POc *qaRa(r)); this is evidently a fossilised compound, literally ‘inside the fence’, which in a number of Polynesian languages (Emae, Samoan, Tongan, Tikopia) means ‘a garden’.
| POc | *kaRi | ‘garden fence or partition’ | |
| NNG | Tami | kali | ‘fence’ |
| NNG | Kairiru | kar | ‘fence’ |
| PT | Molima | ali | ‘fence, esp. of garden’ |
| PT | Kilivila | kali | ‘garden fence’ |
| SES | Lau | (sā)kali | ‘fence round a garden’ |
| SES | ’Are’are | ari-ari | ‘a separation, partition in the garden’ |
| POc | *bayat | ‘fence, boundary marker’ | |
| POc | *bayat-i- | ‘make a garden boundary’ | |
| PT | Gumawana | bayata | ‘garden boundary made by terracing rocks’ |
| PT | Gumawana | bayasi | ‘make a garden boundary’ |
| PT | Iduna | bai | ‘stick used as garden boundary; roof baton’ |
| MM | Tolai | bait | ‘enclose with a fence’ |
| MM | Tolai | ba-bait | ‘fence’ |
| SES | Arosi | bai-bai | ‘large logs put round a finished garden’ |
| NCV | Tirax | pae | ‘fence, wall’ |
| Fij | Bauan | bai | ‘fence round garden or town’ |
| Pn | Tuvalu | pae | ‘stones round an earth oven’ |
| Pn | Tikopia | pae | ‘wall-like accumulation of stones’ |
| Pn | West Futunan | bae | ‘a stone fence’ |
The two staples, taro and yams, would have primarily determined the nature of gardening activities, although the degree to which each is cultivated and the varieties favoured may vary from place to place. The two follow a different growth cycle. Malinowski (1935b: 296) describes the situation in the Trobriands:
Taro differs considerably from both yams and taytu [small yams] in the time it takes to mature, the conditions in which it thrives and in its capacity for being stored. Its period of growth is much shorter… There can… be three taro crops in a year, as against one for both yams and taytu.
Ivens (1927 (reissued 1972):355) describes the practice at Sa’a and Ulawa, in the Southeast Solomons, where the staple crops are yams, hana (a prickly yam regarded as distinct from a yam) and taro, the latter being grown in a separate plot.
The bush (inland) people grow taro and hana… My lists contain names of sixty varieties of yams, twenty of hanas and a hundred and twenty of taro, the large number of taro showing that the people were originally dwellers of the high lands. All of these varieties are cultivated and the differences between them recognised. In a taro garden the tubers are pulled as required, the tops are replanted, the suckers are planted out and clearing goes on all the year through, but yams and hanas have a regular season and only one crop. New ground is cleared for the yams and hanas every year, and the same ground is never planted twice, fertilizers being unknown. The old yam gardens serve for bananas and pawpaws, and for the planting of betel pepper, and the edible reed awalosi. Different words are used to denote the clearing for a yam garden, and the clearing for taro.
In a description of traditional gardens on Manus Island in the Admiralties, McEldowney (1995:98) describes the yams Dioscorea esculenta (suwe) and Dioscorea alata (meyen) as completely dominating, with only limited amounts of taro (Colocasia esculenta) sometimes found along the boundary walls of the garden.
Food plants other than tubers would require little labour. The most significant are breadfruit (also a staple in some places), some banana varieties, sugarcane and greens. Sago is eaten in the absence of other staples. The coconut might also be regarded as a staple, the plant’s use, as with sago, extending beyond that of a foodstuff. Together with other fruit and nut trees, all of these are found in the wild, sometimes ‘managed’ as semi-wild, but also planted in garden plots, where they remain while the plots undergo their cycle of fallow.
William Clarke (1994: 13) has commented on the importance of arboriculture to Oceanic peoples:
A striking characteristic of what can now be interpreted as an independent origin of agriculture in western Melanesia is its emphasis on arboriculture—the culture of trees—orchards. Trees were transported Pacific-wide, carried by the itinerant colonist-cultivators, incorporated into all production systems and begot the typical village environs of Polynesia and Micronesia (Yen 1990). Some nut and fruit trees and sago dropped out before reaching the farther islands of the eastern and northern Pacific. Archaeological evidence for a well-developed arboriculture at least 3,500 years ago comes from the Mussau Islands, north of New Ireland (Kirch 1989). Arboreal species existing at this considerable time depth included: coconut, two or three species of Pandanus, Inocarpus fagifer (the ‘Tahitian chestnut’, which remains one of the most important Oceanic arboricultural species), Canarium indicum (a nutritionally substantial ‘almond’-producing tree in Melanesia), Spondias dulcis (the vi-apple, now of very wide distribution in the tropical Pacific), Pometia, Pangium, Terminalia, Burckella, Calophyllum, and others, including hardwood trees widely prized for woodworking.
We have reconstructions going back at least as far as PMP for two kinds of taro (Colocasia esculenta and Alocasia macrorrhiza), the greater yam, bananas, breadfruit and sago. Other food plants, such as sugar cane, may have been indigenous to New Guinea. The sweet potato is of South American origin, and was probably introduced into New Guinea-north-west Melanesia after AD 1550, though it was introduced into Polynesia in pre-contact times (Yen 1973).
Ross (1996d) includes the following POc reconstructions for items he classifies as “staples and related food plants”:
| taro | ||
|---|---|---|
| POc | *talo(s) | ‘taro, Colocasia esculenta’ |
| POc | *mʷapo(q) | ‘taro’ |
| POc | *piRaq | ‘giant taro, elephant ear taro, Alocasia macrorrhiza’ |
| POc | *bulaka | ‘swamp taro, Cyrtosperma chamissonis’ |
| POc | *kamʷa | ‘k.o. wild taro (?)’ |
| POc | *(b,p)oso | ‘k.o. taro’ |
| yams | ||
|---|---|---|
| POc | *qupi | ‘greater yam, Dioscorea alata ; also used as yam generic’ |
| POc | *pʷatik | ‘potato yam, aerial yam, Dioscorea bulbifera’ |
| POc | (?) *mamis | ‘k.o. yam’ |
| POc | *mʷaruqen | ‘k.o. yam: wild yam (??)’ |
| POc | *udu(r,R) | ‘k.o. greater yam’ |
| POc | *pʷasepe | ‘greater yam’ |
| PWOc | *gobu | ‘potato yam, Dioscorea bulbifera (?)’ |
| PWOc | *(q,k)amisa | ‘lesser yam, Dioscorea esculenta’ |
| PEOc | *damu | ‘k.o. yam’ |
| coconut (Cocos nucifera) | ||
|---|---|---|
| POc | *niuR | ‘ripe coconut; coconut (generic)’ |
| breadfruit (Artocarpus and Parartocarpus spp.) | ||
|---|---|---|
| POc | *kuluR | ‘breadfruit, Artocarpus altilis’ |
| POc | *baReqo | ‘breadfruit fruit (?)’ |
| POc | *beta | ‘k.o. breadfruit’ |
| PEOc | *ma(R)i | ‘breadfruit’ |
| bananas | ||
|---|---|---|
| POc | *pudi | ‘banana, Musa cultivars’ |
| POc | *joRaga | ‘banana, A ustralimusa group’ |
| POc | *sakup | ‘k.o. cooking banana’ |
| POc | *bʷera | ‘banana type’ |
| POc | *baqapun | ‘k.o. banana’ |
| POc | *tawai | ‘k.o. banana’ |
| PWOc | *bʷatiq | ‘k.o. banana’ |
| sugarcane | ||
|---|---|---|
| POc | *topu | ‘sugarcane, Saccharum officinarum’ |
| POc | *pijo | ‘k.o. edible wild cane or reed, possibly Saccharum spontaneum’ |
| sago | ||
|---|---|---|
| POc | *Rabia | ‘sago, Metroxylon spp.’ |
| greens | ||
|---|---|---|
| POc | *bele | ‘shrub species, Abelmoschus manihot (syn. Hibiscus manihot)’ |
No doubt a variety of green leaves were eaten but other reconstructions (e.g. *kusa(q)) do not permit a definition more precise than ‘k.o. edible greens’.
| other fruit trees | ||
|---|---|---|
| POc | *wai, *waiwai | ‘mango (generic)’ |
| POc | *pau(q) | ‘mango, probably Mangifera indica’6 |
| POc | *koRa | ‘wild mango, Mangifera minor’ |
| POc | *quRis | ‘Polynesian plum, hog plum, Tahitian apple, golden apple, Spondias cytherea (syn. Spondias dulcis)’ |
| POc | *molis | ‘citrus fruit or citrus-like fruit’ |
| POc | *kapika | ‘Malay apple and rose apple, Eugenia spp.’ |
| POc | *tawan | ‘Pometia pinnata’ |
| POc | *natu(q) | ‘k.o. tree with avocado-like fruit and hard wood, Burckella obovata’ |
| nut trees | ||
|---|---|---|
| POc | *raqu(p) | ‘New Guinea walnut, Dracontomelon dao’ (syn. Dracontomelon mangiferum, Dracontomelon edule) |
| POc | *(w,v)ele | ‘cut nut, Barringtonia sp.’ |
| POc | *[ka]ŋaRi | ‘canarium almond, Canarium spp.’ |
| POc | *qalip | ‘canarium almond, Canarium sp.’ |
| POc | *talise | ‘Java almond, Indian almond, Terminalia catappa’ |
| POc | *qipi | ‘Tahitian chestnut, Pacific chestnut, Inocarpus fagifer’ (syn. Inocarpus edulis).’ |
No garden is complete or likely to succeed without an abundance of colourful and aromatic herbs and shrubs. In addition to their ornamental qualities, these herbs and shrubs play a major role in various forms of garden magic and ritual. Reconstructions for two such are *laqia ‘ginger, Zingiber officinale’, whose role is magical rather than culinary, and *yaŋo ‘turmeric, Curcuma longa’, which has both an edible root and a rich orange pigment, of ritual significance in many Oceanic societies.
Taro is propagated either by replanting the cut-off top of a tuber or by transplanting the suckers that grow around the ‘mother’ taro. Alternatively, these are simply left in the ground after the mature taro has been harvested. Reconstructed terms for this propagation material may refer to other plants in addition to taro. They are POc *(s,j)uli(q) ‘banana or taro sucker, slip, cutting, shoot (i.e. propagation material)’; POc *wasi ‘taro stem, used for planting’; POc *bʷaŋo ‘taro tops, top shoots of areca nut, coconut’ and POc *up(e,a) ‘taro seedling’.
| PAn | *suliq | ‘tendril, sucker’ (Blust 1972b) | |
| POc | *(s,j)uli(q) | (1) ‘banana or taro sucker, shoot (i.e. propagation material)’; (2) ‘slip, cutting’ (Ross 1988) | |
| Adm | Lou | sili-n | ‘sprout: sprout of banana or pineapple’ |
| Adm | Loniu | cili | ‘sprout, esp. banana shoot’ |
| NNG | Tami | jili | ‘taro sucker’ |
| NNG | Lukep | suli- | ‘a banana shoot’ |
| NNG | Manam | suli | ‘banana slip, cutting’ |
| PT | Dobu | suli | ‘taro’ |
| PT | Tawala | huni | ‘taro’ |
| PT | Motu | dui | ‘banana plant’ |
| MM | Nehan | hon | ‘taro’ |
| SES | Gela | duli | ‘banana sucker’ |
| NCV | Mota | suli(u) | ‘sucker from roots of a plant’ |
| SV | Anejom̃ | ni-sje-n | ‘taro shoot with leaves’ |
| Fij | Wayan | ðuli | ‘banana or taro sucker, shoot (i.e. propagation material)’ |
| Fij | Bauan | suli | ‘banana or taro sucker’ |
| Pn | Tongan | huli(ʔi talo) | ‘taro sucker’ (ʔi < POc *qi ‘generic possessive preposition’; talo ‘taro’) |
| POc | *wasi | ‘taro stem’ (Ross 1996d) | |
| Adm | Mussau | asi | ‘taro’ |
| Adm | Loniu | wos | ‘taro stem, used for planting’ |
| NNG | Rauto | i-sin | ‘taro’ |
| PT | Ubir | wasi | ‘taro’ |
| PT | Nimoa | wusi | ‘taro’ |
| SV | Anejom̃ | n-ase(n-tal) | (-n CONST, tal ‘taro’) |
| POc | *bʷaŋo | ‘new leaves or shoots, (or taro tops for planting?)’ (Ross 1996d) | |
| NNG | Manam | baŋ | ‘taro’ |
| PT | Tawala | pam | ‘edible green leaves (e.g. taro leaves)’ |
| SES | Lau | gʷaŋo | ‘taro tops (for planting)’ |
| SES | Arosi | bʷaŋo-bʷaŋo | ‘the top shoots of betel nut/coconut, taro for planting’ |
| POc | *up(e,a) | ‘taro seedling’ | |
| NNG | Mutu | (do)uwe | ‘seed’ |
| NNG | Tami | uwe | ‘taro seedling’ |
| NNG | Yabem | uwe | ‘seedling’ |
| NNG | Kove | uwe | ‘top of taro or pineapple, coconut sprout, all intended for replanting’ |
| PT | Are | ube | ‘taro tops for planting’ |
| PT | Gapapaiwa | uve | ‘taro tops for planting’ |
| PT | Tawala | uwe | ‘taro seed shoots’ |
| PT | Motu | uhe | ‘the end of yam, kept for planting, any seed for planting’ |
| PT | Molima | uveya | ‘taro top for replanting’ |
| MM | Nakanai | uve | ‘taro top for replanting’ |
| SES | Arosi | uha | ‘taro sp.’ |
| NCal | Pwapwâ | upe | ‘taro seedling’ |
| NCal | Pwapwâ | uva | ‘taro’ |
| NCal | Pije | uwe(qʷaco) | ‘taro seedling’ |
| NCal | Fwâi | uve(qio) | ‘taro seedling’ |
| NCal | Fwâi | uva | ‘taro’ |
These four items refer to parts of a plant which are used in its propagation. Except perhaps for POc *wasi ‘taro stem, used for planting’, they refer to the relevant parts of other plants, as well as taro. However, as Ross points out (1996d: 180), in each case generic reference is to taro. In several Central Papuan languages (Motu and its relatives), reflexes of POc *(s,j)uli(q) have become the generic term for banana.
One ‘cutting’ term has been reconstructed which may refer to the severing of anything from its parent body, such as hair or leaves, but whose reflexes in North New Guinea, the Southeast Solomons and Micronesia have also come to apply specifically to the action of cutting off tuber tops.
| POc | *koti | ‘cut off (hair, taro tops +)’ | |
| NNG | Manam | ʔoti(ŋ) | ‘cut off taro tops for planting’ |
| NNG | Sengseng | kot | ‘cut leaves for roof material or bark from a tree’ |
| MM | Nakanai | koti | ‘cut hair’ |
| SES | Gela | goti | ‘cut off, as taro head in planting’ |
| SES | Arosi | ʔoi | ‘cut off taro tops for planting; to scrape off or peel with a shell’ |
| Mic | Woleaian | xos | ‘cut a tuber top’ |
| Fij | Bauan | koti | ‘clip, shear, cut off small things’ |
| Pn | Niuean | koti | ‘pinch, snip’ |
| Pn | Hawaiian | ʔoki | ‘cut, clip’ |
Propagation of yams is generally by seed yams, small healthy yams selected from the previous year’s crop, or by slices of larger yams. Ross has reconstructed PWOc *kapul ‘seed yam’.
| PWOc | *kapul | ‘seed yam’7 | |
| PT | Dobu | awona | ‘seed yam’ |
| PT | Kakabai | ko-koya | ‘yam’ |
| PT | Misima | ka-kaəun | ‘seed yam’ |
| PT | Kilivila | kaula | ‘yam’ |
| MM | Tiang | ko | ‘lesser yam’ |
| MM | East Kara | ko-kau | ‘lesser yam, Dioscorea esculenta’ |
| MM | Patpatar | kau-kau | ‘lesser yam, Dioscorea esculenta’ |
| MM | Nehan | ko-ko | ‘yam’ |
Ross suggests that this can probably be combined with the cognate set below.
| PNGOc | *(k,kʷ)apii | ‘k.o. yam’ | |
| NNG | Tami | kʷapil | ‘k.o. yam, Dioscorea vulg.’ |
| NNG | Numbami | kowila | ‘greater yam’ |
| NNG | Yabem | kili | ‘greater yam’ |
| NNG | Mapos Buang | ker | ‘yam’ |
| PT | Dobu | kʷaleya | ‘yam’ |
| PT | Iduna | kʷavi-kʷavi | ‘greater yam variety’ |
| POc | *paji | ‘cut yams for planting’ | |
| PT | Dobu | yadi | ‘cut seed yams for planting’ |
| Fij | Wayan | vāði, vaðiti | ‘snatch or grab at s.t.; pluck or break off s.t.’ |
| Fij | Bauan | vaði | ‘cut yams for planting; a piece of yam cut for planting’ |
Proto Oceanic *sopu referred to yam planting. This much is clear from the Muyuw and Gela glosses. It is impossible to know from the available reflexes, however, whether the activity referred to was specific or general.
| POc | *sopu | ‘prepare yams for planting’ | |
| PT | Kilivila | sopu | ‘the whole activity of planting, including preparing the hole, planting the seed or tuber, covering and mounding the soil’ (Malinowski 1935: 132) |
| PT | Muyuw | sopʷ | ‘planting (yams)’ |
| SES | Gela | sovu | ‘cut off slices of yams for planting’ |
| SES | Lau | tofu(a) | ‘chop’ |
| SES | Kwaio | tofu(a) | [v] ‘cut things up’ |
| SES | ’Are’are | tohu(a) | ‘fell, cut, chop’ |
| SES | Sa’a | tohu(a) | ‘chop down, fell’ |
| SES | Arosi | tohu | ‘a bit, portion’ |
| PMP | *hasek | ‘dibble, plant seeds with a dibble stick’ (ACD) | |
| POc | *asok | ‘plant in holes in the ground’ (ACD) | |
| MM | Notsi | soka | ‘plant (sweet potato +)’ |
| MM | Sursurunga | soi | ‘plant (sweet potato +)’ |
| MM | Patpatar | soh | ‘plant (sweet potato +)’ |
| SES | Sa’a | ato(taha) | ‘throw the first yam set into the hole, of priest’ |
| SES | Ulawa | ato(ni waʔa) | ‘throw the yams into the holes’ |
| SES | Ulawa | ato-ato | ‘up and down dividing sticks in a yam garden’ |
| SES | Arosi | ato | ‘distribute yams in holes for planting’ |
| SES | Arosi | ato-ato | ‘lay in rows, mark out thus, as a garden with horizontal poles’ |
The following verb seems to have resulted from affixing *pa- ‘causative’ to *asok above.
| PWOc | *pasok, *pasok-i | ‘plant (tuber +)’ | |
| NNG | Malai | vazogi | ‘plant (tuber +)’ |
| NNG | Gitua | va-vazok | ‘plant (tuber +)’ |
| NNG | Hote | vaðo | ‘plant (by making hole in the ground)’ |
| PT | Hula | varo | ‘plant (tuber +)’ |
| PT | Motu | hado | ‘plant (tuber +)’ |
| MM | Bali | vazoɣi | ‘plant (tuber +)’ |
| MM | Bola | varo | ‘plant (tuber +)’ |
| MM | West Kara | fasu | ‘plant (tuber +)’ |
| MM | Nalik | fasu | ‘plant (tuber +)’ |
Despite the formal resemblance of the following set to the one above, it is probable that they are not historically related.
| PEOc | *pasi | ‘plant yams’ | |
| SES | Gela | vahi | ‘plant yams’ |
| SES | Tolo | hasi | ‘plant (s.t.)’ |
| SES | Lau | fasi | ‘plant (s.t.)’ |
| SES | Arosi | hasi | ‘plant (s.t.)’ |
| Fij | Bauan | vasi | ‘plant yams in the same ground two or more years in succession’ |
| Adm | Mussau | pasa | ‘plant (s.t.)’ |
Other terms are reflexes of POc *(su)suk-i ‘pierce, sew’ (Ch. 9, §4.1):
| SES | Arosi | suʔi | ‘transplant, put in a seedling’ |
| NCV | Mota | sug | ‘dig up, transplant’ |
| Fij | Bauan | ðuki | ‘dig up ground, loosen ground with a stick’ |
| POc | *tanum, *tanum-i- | ‘bury, plant (tuber)’ | |
| NNG | Bing | tan | ‘plant (sweet potato +)’ |
| NNG | Manam | -tano | ‘plant (sweet potato +)’ |
| NNG | Bam | -tani | ‘plant (sweet potato +)’ |
| NNG | Wogeo | -tano | ‘plant (sweet potato +)’ |
| SJ | Bongo | tni(ei) | ‘bury’ |
| PT | Ubir | tanum | ‘plant’ |
| MM | Bulu | tanu | ‘bury’ |
| MM | Taiof | tanum | ‘garden’ |
| SV | Lenakel | renəm | ‘bury, plant’ |
| Pn | Māori | tanumi(a) | ‘bury’ |
| MM | Nakanai | galu | [v] ‘plant’ |
| MM | Meramera | danu | [V] ‘plant’ |
Barrau (1955: 56) describes cultivation of yams in Melanesia:
As a general rule, yams flourish on light, deep and well-drained soils. These requirements have led the natives to perfect cultural techniques which are often most ingenious. The simplest consists of carefully earthing up the plant. Sometimes a large ridge of well-worked soil is made… surrounded by draining ditches.
The following set suggests related reconstructions for ‘mound up (earth) for yams’ and ‘yam mound’.
| POc | *(p,b)uk(i,e) | ‘mound up (earth) for yams’ | |
| POc | *ta-(p,b)uk(i,e) | ‘yam mound’ | |
| MM | Petats | tahui | ‘earth heaped up for planting yams in’ |
| SES | Arosi | hu-huʔi | ‘dig, mound up’ |
| NCV | Nguna | tāki | ‘yam mound’ |
| NCV | Atchin | tawu | ‘pile of earth’ |
| Mic | Kiribati | tabuki | ‘hill, hillock’ |
| Fij | Wayan | buke | ‘mound on which yams + are planted’ |
| Fij | Bauan | buke-buke | ‘mound of earth for planting yams’ |
| Pn | Rennellese | puke | ‘hill, as for tubers’ |
| Pn | Hawaiian | puʔe | ‘mound, hill, heap’ |
| PMP | *kambuR | ‘sprinkle, scatter (seed +)’ (ACD) | |
| POc | *kabu(R) | ‘sow or scatter small seeds’ | |
| Fij | Bauan | kabu | ‘sow or scatter small seeds’ |
Only a single Oceanic reflex has been located for this reconstruction, from the most easterly extension of Melanesia. This raises two questions. Firstly, why are there not WOc cognates? If the term reached this point, retaining its PMP meaning, it must have been a familiar one in the intervening regions. This situation arises often enough in linguistic reconstruction, and may be due to any number of sociolinguistic variables or perhaps to shortfalls in available word lists. It is of course possible that our searches have simply not located cognates that exist elsewhere. But the fact remains that most of the cultivated plants for which we have POc reconstructions are propagated by planting parts of the plant other than its seeds. Secondly, are we overlooking the existence of plants that are cultivated from seed but are seldom recorded in word lists, perhaps being of minor importance in the diet, or subsumed under a more generalised term such as ‘greens’. One such plant may be an Amaranthus (Amaranthus tricolor), recorded as an indigenous food plant by Quiros on his visits to Santa Cruz and Espiritu Santo at the beginning of the seventeenth century (Markham 1904:50). Another pre-European plant in New Guinea whose leaves are eaten is a nasturtium species found in the lowlands (Nasturtium hybospermum). It is a widespread practice today for bunches of these plants, with seed pods, to be hung up to dry until such time as a new garden is prepared. The seeds are then broadcast into the soil surface. Other contenders may be legumes, such as the winged bean (Psophocarpus tetragonolobus)8 and Lablab purpureus, both of which may date from pre-contact times (pers.comm. Michael Bourke). The association of these or similar plants with POc speakers will, however, depend on our reconstruction of appropriate terms.
| POc | *tupul | ‘to send out new growth’ (French-Wright 1983: 78) | |
| PT | Motu | tuhu-tuhu | ‘young shoot’ |
| MM | Roviana | tuvulu | ‘send out new growth, of trees that have been cut down’ |
| Fij | Bauan | tuvu | ‘shoot up, of a tree’ |
| POc | *polok | ‘(plant) grow (tall)’ | |
| NNG | Sengseng | polok, plok | ‘sprout (of plants, new tooth)’ |
| MM | Lavongai | polok | ‘(plant) grow’ |
| MM | Tigak | polok | ‘(plant) grow’ |
| SES | Arosi | horo | ‘begin to make stem, of a tree; the beginning of a stem in a coconut’ |
| PWOc | *pusuk | ‘(plant) grow’ | |
| NNG | Bing | pus | ‘(plant) grow’ |
| MM | Konomala | pusa | ‘(plant) grow’ |
| MM | Solos | pusuk | ‘(plant) grow’ |
| PMP | *tu(m)buq | ‘grow, thrive; swell’ (Blust 1986) | |
| POc | *tubuq | ‘grow, swell’ | |
| NNG | Numbami | tubu | ‘grow, fatten’ |
| NNG | Roinji | tubu | ‘(plant) grow’ |
| NNG | Kove | tuvu-tuvu | ‘grow’ (tuvu ‘physique’; pa-tuvu ‘grow a child’) |
| PT | Motu | tubu | ‘grow; fennent; swell’ |
| PT | Bwaidoga | tubuga | ‘grow large, swell’ |
| MM | Ramoaaina | tubu | ‘grow; principally of men and animals not trees’ |
| MM | Nakanai | tubu | ‘be fat; grow’ |
| MM | Teop | subu | ‘swell’ |
| SES | Bugotu | tubu | ‘swell’ |
| SES | Sa’a | upu | ‘swell’ |
| SES | Arosi | ubu | ‘swell’ |
| Fij | Wayan | tubu | ‘grow’ |
| Fij | Bauan | tubu | ‘grow, increase, spring up, of plants’ |
Chapter 6, §3.9 contains two reconstructions, POc *ma-osak ‘ready to be eaten (because ripe or cooked)” and POc *[ma-]noka ’be in good condition for eating; nicely ripe, well-cooked, soft’.
Although any plant may become dry and droop through lack of water, it is a feature of yam vines that the leaves turn yellow and drop off when the tubers are ready to harvest.
PCEMP *[ma]Raŋaw below leads us to expect POc *[ma]Raŋo, but the cognate set apparently reflects POc *[ma]raŋo, as the Bugotu, Sa’a and Tape items unambiguously reflect *r, not *R. However, the Niuean and Samoan items under ‘cf also’ reflect POc *maRaŋo (*R is lost in Central Pacific languages but *r isn’t). This implies that *[ma]Raŋo and an idiosyncratic *[ma]raŋo co-occurred in early Oceanic. As Mussau, NNG and MM languages merge POc *r and *R, they may reflect either form. Hence reconstructed *[ma][R,r]aŋo.
| POc | *[ma-]raŋo | ‘become withered (of vegetation)’ | |
| NNG | Manam | ma-raŋo | ‘dry, arid’ |
| MM | Tolai | ma-raŋa | ‘withered, dry (leaves, husk, tree)’ |
| MM | Selau | raŋo | ‘dry’ |
| SES | Bugotu | raŋo | ‘wither (leaves, yam vines)’; ‘wither (leaves, ŋam vines)’ |
| SES | Sa’a | raŋo | ‘be withered, dry (esp. yams when vine withers)’; ‘be withered, dry (esp. ŋams when vine withers)’ |
| SES | Arosi | raŋo | ‘withered, dead (of grass, green boughs +)’ |
| NCV | Mota | raŋo | ‘become dried up in the course of nature’ |
| PCEMP | *[ma]Raŋaw | ‘dry’ (ACD) | |
| POc | *[ma]raŋo | ‘become withered (of vegetation)’ | |
| Adm | Mussau | malaŋo | ‘dry’ |
| NNG | Manam | maraŋo | ‘dry, arid’ |
| MM | Patpatar | maraŋa | ‘dry coconut’ |
| MM | Tolai | maraŋa | ‘withered, dry (leaves, husk, tree)’ |
| MM | Selau | raŋo | ‘dry’ |
| SES | Bugotu | raŋo | ‘wither (leaves, yam vines)’; ‘wither (leaves, ŋam vines)’ |
| SES | Sa’a | raŋo | ‘be withered, dry (esp. yams when vine withers)’; ‘be withered, dry (esp. ŋams when vine withers)’ |
| SES | Arosi | raŋo | ‘withered, dead (of grass, green boughs +)’ |
| NCV | Mota | raŋo | ‘become dried up in the course of nature’ |
| NCV | Tape | məraŋ | ‘(of wood, leaves) dry’ (J. Lŋnch, pers. comm.) |
| SV | Kwamera | (kahi)mareŋi | ‘turning brown’ |
| Pn | Niuean | maŋo | ‘dry (of wood, trees)’ |
| Pn | Samoan | maŋo | ‘dry up; be dry (of wood, clothes)’ |
| POc | *malai | ‘withered, faded’ | |
| MM | Nakanai | malai | ‘be withered or drying up (leaves, fruit, copra)’ |
| Fij | Wayan | mālai | ‘withered, faded (living things)’ |
| Fij | Bauan | malai | ‘withered, faded’ |
| NNG | Kove | malai | ‘tired, feeling lazy’ |
| PEOc | *palo | ‘wither, wilt’ (ACD) | |
| SES | Sa’a | halo | ‘wither, of full-grown taro plants’ |
| Pn | Māori | paro | ‘shrunk, wilted’ |
We have not identified a reconstruction specifically for the digging up of tubers, either with or without the use of a digging stick, although *kali/*keli (§5.4) is apparently a general term whose meaning includes this activity. It is possible that terms included under the heading ‘weeding’ were sometimes used to refer to pulling any plant out of the soil. Ann Chowning (pers.comm.) points out that, density of soil permitting, taro is often harvested by being pulled out directly by its thick stem, whereas yams have to be dug up. The reconstructions below relate to the gathering of fruit and nuts and sometimes leaves. They carry distinctions made on the basis of whether the item is plucked by hand, or with the aid of a stick or hook.
| PMP | *kawit | ‘hook’ (Dempwolff 1938) | |
| POc | *kawi(t), *kawit-i- | [V] ‘hook, catch hold of; fruit crook’ (French-Wright 1983) | |
| NNG | Manam | ʔaut | ‘pluck fruit with a fruit crook’ |
| MM | Tolai | kait | ‘catch, as clothes on thorns’ |
| MM | Nakanai | kau | [V] ‘hook, as in hooking fruit down from a tree’ |
| SES | Gela | kauti | ‘drag off fruit with a hook’ |
| SES | Sa’a | i-kau | ‘hooked stick for fruit-picking’ (i- < POc *i INS) |
| SES | Arosi | kau | ‘catch and hold, as a shirt in a nail; a crook for pulling down fruit’ |
| SES | Arosi | ʔawi | ‘hook’ |
| NCV | Mota | kaut | ‘catch hold and pluck, twitch’ |
| NCV | Mota | i-kau | ‘the cleft bamboo used to twitch off almonds, breadfruit +’ (i- < POc *i INS) |
| Mic | Kosraean | kai | ‘catch with a hook’ |
| Fij | Wayan | kauti | ‘hook s.t., catch s.t. on a hook’ |
For the implement, Polynesian languages use terms derived from PPn *lohu ‘fruit-plucking pole, hook something with a pole’ (Biggs & Clark 1993).
The set above bears a formal and semantic resemblance to PMP/POc *kawil ‘hook, fish hook’ (Ch. 8, §4), but it is unclear what historical connection (if any) exists between them.
The following sets, all relating to plucking by hand, have formal similarities, and it is sometimes hard to know which cognate set a given reflex belongs to. For discussion of the first three items below, see Ch. 9, §6.1.
| PAn | *buCbuC | ‘pull up (weeds +), pluck (feathers +)’ (ACD) | |
| POc | *pupu(t) | ‘pick (fruit +), pluck (feathers +)’ | |
| PT | Motu | huhu- | ‘break off bananas singly’ |
| PT | Dobu | (lo)pupu | ‘pluck feathers from a bird’ |
| PT | Gapapaiwa | pu(i) | ‘pluck feathers from a bird’ |
| PT | Tubetube | pupu | ‘pull off (leaves from tree +)’ |
| PT | Sudest | vu | ‘pick; harvest a fruit’ |
| SES | Lau | fufu | ‘pick fruit’ |
| SES | Arosi | huhu | ‘pluck fruit’ |
| SES | Sa’a | huhu | ‘pluck, pick off’ |
| Fij | Bauan | vuvu | ‘root up entirely’ |
| Pn | Niuean | fufu | ‘strip off (leaves, bark +)’ |
| POc | *sapu(t), *saput-i- | ‘pull out, pull up, pluck (fruit, nuts)’ | |
| PT | Molima | sabu | ‘pull up taro or grass’ |
| SES | Arosi | tahu | ‘take by force’ |
| NCV | Raga | havusi | ‘pluck, as a fowl’ |
| NCV | Tamambo | sabuti | ‘pluck, pull out (plant, tooth +)’ |
| Fij | Bauan | ðavu | ‘pull up, eradicate’ |
| Fij | Bauan | ðavut(a) | [VT] ‘pull up, eradicate’ |
| Fij | Wayan | ðavu | ‘(tooth, root +) be pulled out, extracted, removed from a fixed position’ |
| Fij | Wayan | ðavuti- | ‘pull s.t. out, remove s.t.’ |
| Pn | Tongan | hafu(le) | ‘strip the dry leaves from sugarcane, pandanus, banana and plantain plants’ |
| MM | Roviana | zapu | ‘pull coconuts from a tree’ |
| POc | *tapu(t), *taput-i- | ‘strip (crops), pull off’ (French-Wright 1983: *tapu) | |
| NNG | Lukep | tau(rai) | ‘pick’ (-rai < POc *-(r,R)-aki) |
| PT | Bwaidoga | tavu(na) | ‘harvest bananas’ |
| PT | Motu | tapusi | ‘pull strongly, with jerk of a string, and possibly break it’ |
| MM | Nakanai | tavu | ‘grasp, capture’ |
| MM | Simbo | tapu | ‘pull off, as husk off canarium nut’ |
| SES | Arosi | ahu | ‘(coconut +) fall; strip completely (garden of food); gather fruit’ |
| Fij | Bauan | tavu | ‘knock down and beat’ |
| POc | *sapaki- | ‘pluck off, break off (leaves) with the hand’ (French-Wright 1983) | |
| NNG | Manam | sapaʔ | ‘pluck off’ |
| MM | Tolai | apak | ‘break off leaves from a tree, as for cooking or ornament’ |
| Mic | Woleaian | tepagi | ‘cut (leaves)’ |
| PEOc | *sapi[-] | ‘strip (leaves); pluck (fruit, nuts)’ | |
| SES | Gela | sapi | ‘pluck fruit from a bunch; strip off leaves’ |
| SES | Lau | tāfi | ‘lop off; take off midrib of sago palm leaf’ |
| SES | ’Are’are | tahi | ‘strip off leaves; cut into slices’ |
| SES | Sa’a | tahi | ‘pluck hanging vines’ |
| SES | Arosi | tahi | ‘cut, cut off; strip off’ |
| NCV | Mota | sav | ‘pluck (hair, feathers)’ |
| PCP | *paki | ‘pluck, break off (leaves) with the hand’ | |
| Fij | Rotuman | haʔi | ‘pluck (feathers), pull out’ |
| Pn | Tongan | faki | ‘pick, pluck, esp. banana, coconut’ |
| Pn | Samoan | faʔi | ‘break off, snap off, pick’ |
| Pn | Tikopia | faki | ‘gather (breadfruit +)’ |
| NNG | Sengseng | pak | ‘collect bedpoles by breaking off long straight branches or trunks’ |
| PT | Motu | baki | ‘break, of bread, sago +’ |
| Pn | Tongan | paki | ‘break or break off, esp. with the hand; pick or pluck’ |
| POc | *paRi | ‘cut or lop off branches’ (ACD) | |
| NNG | Mangseng | var | ‘cut, clip’ |
| MM | Nakanai | vali | ‘cut, as wood or a leaf from a tree; remove all the limbs from a tree’ |
| SES | ’Are’are | hari | ‘lop off branches, cut off a bunch of bananas, betel nuts’ |
| SES | Sa’a | hali | ‘lop off branches’ |
| SES | Arosi | hari | ‘tear, tear off, pull off a cluster of fruit’ |
| Fij | Rotuman | fai | ‘cut or chop down (tree or branch)’ |
The yam is the only tropical tuber which is easily preserved, provided it is placed in an airy, dry, dark and cool spot. It was a community’s safeguard at times when the regular supply of foodstuffs was low. From New Guinea to New Caledonia, in areas where yam is the staple food, special huts are found in which they are preserved (Barrau 1955:58). Malinowski (1935: 218) reports that in the Trobriands such structures (bwema) were given a more prominent position in the village than the family dwellings, and were more meticulously decorated and maintained. POc *pale ‘open-sided building’, has been reconstructed (Ch. 3, §3.3), reflexes of which in some instances refer to yam storehouses (NNG: Lukep (Pono) para ‘yam house’; SES: Arosi (E. dialect) hare ‘shed for yams’). No term has been reconstructed that refers specifically to a hut for storing yams. In Mapos Buang (South Huon Gulf), yam houses are called jok. In Nengone, Loyalty Islands, methuma is the term for a harvest hut for yams.
Water control has two opposing functions: (1) to aid drainage in wet areas, and (2) to carry water to drier areas. The New Guinea area apparently concentrated its developments towards the former, particularly in yam gardens, through such means as mounding and ditching (Yen 1973: 83). In areas with lower or less reliable rainfall, it has long been the practice to irrigate—particularly the water-loving taro crops, whose yield can be considerably increased. For various reasons, including the fact that a more restricted environment limits crop choice, irrigated taro fields are more a feature of eastern Oceanic than of the west. Kirch and Lepofsky (1993: 185) have written:
As archaeological and ethnobotanical data on the distribution of Oceanic irrigation technology accumulated, it became evident that the widespread distribution of both pondfield and raised-bed agricultural technologies in Polynesia was not matched by an equally extensive distribution in Melanesia… Although there are ‘pockets’ of highly intensive taro pondfield irrigation in Melanesia (notably in the New Georgia group of the western Solomons, on Anejom in Vanuatu (Spriggs 1981), and in New Caledonia (Barrau 1956)), on most of the intervening islands (including the large island of New Guinea [but presumably excluding the Western Highlands systems of the Papuans mentioned below (m.o.)]) irrigation is absent.
The question to be resolved is whether the early Oceanic gardeners were familiar with what might be called more sophisticated irrigation techniques, or whether the knowledge was acquired later. Proto Oceanic speakers were generally coastal and island dwellers, far removed geographically from Papuan-speaking peoples of the New Guinea Highlands who may have been associated with ditch systems in the growing of taro as long ago as 6000 BP (Golson 1990, Swadling & Muke 1998). More feasible is the possibility that POc knowledge was passed down from Southeast Asian forebears. Although Southeast Asian paddies were used predominantly for rice growing, taro is reportedly grown in similar fashion in parts of the northern Philippines, south-west Sulawesi and Java (Spriggs 1981: 173).
Kirch and Lepofsky (1993: 196) considered both linguistic and archaeological evidence and concluded that:
neither body of evidence lends support to the… model for the origins of Polynesian irrigation technology, [which proposes] a direct transfer or diffusion from Southeast Asia via the migrations of Austronesian-speaking peoples. Rather, all of the evidence suggests that there have been multiple local sequences of innovation and development of complex irrigation or swamp drainage technology, or both, for Colocasia taro cultivation within eastern Melanesia and Polynesia.
Working from our own linguistic sources, which include possibly more data from Western Oceanic languages than were available to Kirch and Lepofsky, we have not been able to add a single cognate set to those already listed by the latter.9 No reconstructions have been possible for such concepts as ditch or water channel or irrigated garden. A single reconstruction is possible to PCP level.
| PCP | *vusi | ‘taro swamp; taro bed’ (from French-Wright) | |
| Fij | Bauan | vuði | ‘taro garden under wet cultivation’ |
| Pn | Samoan | fusi | ‘raised bed’ |
| Pn | East Futunan | vusi(ga) | ‘pondfield’ |
| Pn | East Uvean | fuhi | ‘raised bed’ |
FIXME: In vol.1,139 the form PCP *pusi was erroneously given for *vusi (POc *p split into PCP *p and *v).
A second reconstruction is open to conjecture. Both Spriggs and Kirch and Lepofsky drew attention to the Hawaiian and Rapa (Austral Islands) terms for a pondfield, loʔi and roki. We have a tentative reconstruction POc *logi meaning something like ‘partition, partitioned area’ with a reflex from Hote in the Huon Gulf (NNG), lok ‘(garden) fence; wall of house’ (Ch. 3, §3.5).
| POc | *logi | ‘partition, partitioned area’ | |
| NNG | Hote | lok | ‘(garden) fence; wall of house’ |
| Mic | Kiribati | roki | ‘screen (of mats +); curtain; bathroom; closed with screen’ (possibly borrowed from Polynesian) |
| Fij | Rotuman | loki | ‘the end of the interior of a house’ |
| Fij | Bauan | loqi | ‘inner or private part of a house, used as a bedroom and for keeping valuables’ |
| Pn | Tongan | loki | ‘inner room’ |
| Pn | Samoan | loʔi | ‘inner room’ |
| Pn | Rapa | roki | ‘irrigated terrace’ |
| Pn | Tahitian | roʔi | ‘bed’ |
| Pn | Hawaiian | loʔi | ‘irrigated terrace for taro’ |
| Pn | Tuamotuan | roki | ‘wall of a temple; bed, couch’ |
| SES | Gela | voki | ‘room, partition in a house’ |
The Polynesian reflexes, apart from the Rapa and Hawaiian terms, have the sense of ‘inner room, room for sleeping’ or similar. However, it is conceivable that the POc term did have a more generalised meaning and could be applied to any partitioned area, whether house or garden. Its reference to an irrigated partitioned area remains a localised one.
In spite of remarkable similarities of technique commented on by Matthew Spriggs (pers.comm.) in islands throughout eastern Oceania, including New Georgia, Guadalcanal, Maewo, Santo, New Caledonia, Fiji and a number of Polynesian islands, and possibly also New Hanover in the Bismarck Archipelago, archaeological findings to date would seem to support the theory that more complex irrigation systems are not as old as Oceanic settlement. Although many of these systems have fallen into disuse since European contact (largely as the result of dramatic decline in population that came about with exposure to European diseases), our best evidence for their existence is in the physical traces left by their water channels, sometimes stone-lined, and their banks and terraces. More recently, archaeologists have, under certain conditions, been able to identify soil strata that have at some time in the past been subjected to pondfield irrigation (Spriggs 1997b: 84-85). Spriggs reports on one such site in Hawaii now dated to 1500 BP, the earliest date yet for evidence of pondfield irrigation (p.89). In New Caledonia, a buried taro pondfield terrace at Col de Piroque has been dated back to 1150-1100 BP (p. 184). Other evidence points to more recent dating. On Aneityum, Spriggs estimates that the large-scale canal-fed irrigation systems of the valley flats have been constructed within the period since about 500-600 BP (Spriggs 1986: 11). Oldest evidence for the Solomons remains the eye-witness report of the explorer Mendafia in 1568 (Amherst & Thomson 1901: 306).
Other concepts for which we currently lack reconstructions include a term for the stakes or supports on which the yam vines were trained. French-Wright (1983: 64-65) lists a variety of ways in which vines can be supported. In addition to stakes, they include, in the Admiralties and the Trobriands, saplings left standing after the garden has been cleared; in the centre of Guadalcanal in the Solomons the vines trail over felled trees; in the Loyalty Islands an arrangement rather like an upturned basket is used, and on Efate, Vanuatu, a trellis-like structure is built over the yam mounds. French-Wright suggests that perhaps the lack of cognates denoting this basic feature of yam cultivation is due to the wide variety of support systems used, in other words, that each cultural adaptation has led to changes in terminology.
Gardens in the Trobriands (Malinowski 1935: 55, 121) and in Malaita, Southeast Solomons (Ivens 1927 (reissued 1972):356), and elsewhere, are divided into sub-plots by means of criss-crossing logs laid on the ground, but we have been unable to reconstruct terms for either the plots or the dividing logs, although these are named in their own communities.
A term must surely have existed for the period or activity of yam harvesting, which was frequently carried out by the whole community at the one time and typically accompanied by certain rituals. It does seem, however, that POc speakers associated the period of yam harvest with the conclusion of an annual cycle. In a number of widespread localities, past and future events are placed in time by referring to yam crop cycles, thus according the terms a meaning equivalent to our concept of ‘year’ (PT: Kilivila tetu ‘small yam (the main staple); year’; NNG: Mapos Buang ta ‘year; a complete cycle of yam growing’; NCV: Paamese ouh ‘a yam; a year’). Similarly, Ivens (1927:358) writes, “Our [English] word year would be rendered in a Melanesian language (Sa’a halisi, Lau, Kwaio falisi ‘yam harvest, yam season, year’] by whatever is equivalent to the yam crop.”
Archaeological evidence in support of POc horticultural practices comes, not from discovery of items of material culture, but in the form of the products of horticulture, and, less directly, from apparent sudden and dramatic changes in land use following settlement by POc speakers.
Direct archaeobotanical evidence for horticulture, in the form of preserved plant remains, largely eluded prior generations of Oceanic archaeologists… Then, during the 1986 excavation season at the waterlogged Talepakemalai site in the Mussau Islands, we were suddenly confronted with an unprecedented array of well-preserved (non-carbonised) plant remains. Sealed in an anaerobic, wet environment, the literally thousands of seed cases, endocarps, husks, syncarps, and other plant material, along with wood (including adzed wood chips) opened up a new window on Lapita relations to the plant world. Subsequently, a similar assemblage of preserved plant parts was also recovered by Chris Gosden … at his Arawe Island sites. The plant parts preserved… are primarily woody or fibrous, such as seed cases, fibrous husks, and wood. Soft, fleshy materials such as tubers or the soft flesh of fruits have not survived.
Among the food plants identified from archaeological sites are Canarium almond (Canarium indicum), Indian almond (Terminalia catappa), the New Guinea walnut (Dracontomelon dao), Pometia pinnata, Burckella obovata, Tahitian chestnut (Inocarpus fagifer) and vi-apple (Spondias dulcis) (Kirch 1997: 206-207).
The second source of evidence, modification of the landscape, supports the POc horticultural practice of swidden cultivation. Spriggs (1997a:85-86) describes the dramatic change in land use that coincides with the influx of Lapita people, a change particularly evident in the previously uninhabited islands east of the Solomons. Spriggs postulates initial settlement with large-scale clearance for agriculture, as indicated in the pollen record which shows suddenly high carbonised particle counts resulting from burning. This is typically followed by environmental degradation such as increased erosion rates, and often the extinction of endemic species of birds and reptiles resulting mainly from habitat alteration. Such changes have been documented for Aneityum in Vanuatu occurring about 2,900 years ago and for Fiji perhaps by 1200 BC (Kirch 1997:222).
A third potential source of evidence would lie in physical traces of irrigation systems and associated terracing. To date, however, neither linguistic nor archaeological evidence can support the association of sophisticated irrigation practices with POc speakers (see § 12).
Proto Oceanic speakers cultivated their gardens with a minimum of implements other than their hands. Apart from the axe (*kiRam) and smaller cutting implements of bamboo or shell such as *piso or PEOc *sele (Ch. 4, §4.1.2) used in clearing their plots, the only tools were a digging stick for tillage (*waso), possibly a dibble stick for planting (for which we lack a reconstructed term other than that for digging stick) and a hooked stick (*kawit) for picking fruit and nuts. The POc reconstructions to do with horticulture refer primarily to gardening activities: clearing, burning, weeding, digging, planting, picking and gathering. Also represented are the processes undergone by the plants themselves—sprouting, growing, swelling, ripening, withering.
We have no evidence that any form of fertilising was used, apart from the occasional spreading of ash from the burning off of new garden plots. There is no tradition of composting the soil with human waste, for instance. Mulching and digging-in of decomposed vegetable matter have been recorded in places as traditional gardening practice, but we have no specific reconstructions for these activities. Nor has linguistics shed any light on ways in which they may have attempted to control pests or diseases. Weight was given rather to magic, as a means of ensuring healthy crops.
Finally, we have no evidence from reconstructions that Proto Oceanic speakers irrigated their crops in any systematic way. Current archaeological findings support the view that elaborate systems are a more recent, independent development.
Brookfield and Hart (1971: 81) observe that “cultivation techniques move less easily than crops, for while the latter can be transmitted from hand to hand, the former generally require migrants to carry their skills and adapt them to a new environment”. It is likely that the horticultural knowledge carried into the region by the first POc speakers would have included at the very least, awareness of preferred conditions for the cultivation of various kinds of taro and yams, together with related knowledge of growth stages, propensity for storage and propagation techniques for these and other garden plants such as sugar cane and bananas.