Chapter 3.13 Other cultivated plants

Malcolm Ross

1. Introduction⇫

In his study of the uses of plants in central Pacific societies, Thaman (1994) shows that a number of plants are cultivated for purposes other than food production, an observation that equally applies to NW Island Melanesia. There are a number of such plants that have probably been cultivated in the Bismarcks since Proto Oceanic (POc) times. The evidence of longtime cultivation lies in the domesticated varieties of certain plants that only occur under cultivation (sugarcane, Morinda citrifolia, Codiaeum variegatum, Cordyline fruticosa, Dracaena angustifolia and Heliconia indica) and in the fact that some plants (sugarcane, some bamboo species, kava and Broussonetia papyrifera) only reproduce through the planting of cuttings and have no immediate wild relatives. Evidence that they have quite possibly been cultivated continuously since POc times is provided by the POc reconstructions in this chapter.

The plants treated in this chapter are used in a variety of ways. First there are those which are consumed by human beings but do not readily fall under the headings of earlier chapters: sugarcane is sucked to extract the sugar, betelnut is used as a stimulant, kava as a narcotic (§2). Then come plants which provide materials for making things: bamboo, the candlenut for illumination and black dyes, the paper mulberry for making bark cloth, and the Indian mulberry for yellow and red dyes (§3). Fish poisons are treated in §4, turmeric and ginger, which have various uses, in §5, plants with large colourful leaves used for decoration and for magic in §6 and cucurbits in §7. The pumpkin and the cucumber are of course foods, but cucurbits are discussed here, at the end of the volume, because (i) there is a question regarding the dating of the pumpkin’ s introduction(s) into Oceania and (ii) they do not readily belong in any of chapters 9 to 12.

2. Plants grown for human consumption⇫

2.1. Saccharum officinarum, sugarcane, TP suga (Poaceae)⇫

Sugarcane, Saccharum officinarum belongs to the family Poaceae, i.e., it is a grass, and is closely related to the wild species Saccharum spontaneum (ch.8, §3.4).

Sugarcane is a domesticate, only planted from cuttings, and numerous varieties are found in gardens all over lowland and highland areas of NW Melanesia. Sugarcane prospers especially where the soil is moist and rich (Paijmans 1976: 125-126, Kwa’ioloa & Burt2001: 204, Hviding 2005: 148). The jointed, fibrous stalks contain sucrose, obtained by cutting off a stem and chopping it into convenient lengths which are sucked and chewed as a refreshing snack. When the sugar has been sucked out, the rubbish is spat out.

The POc term *topu ‘sugarcane, Saccharum officinarum’ continues PAn *tebuS and is continued throughout Oceanic except in Polynesian, where reflexes attest to PPn *to ‘sugarcane’ rather than expected †*tofu .

PAn		*tebuS	‘sugarcane’ (Blust 1969)
POc		*topu	‘sugarcane, Saccharum officinarum’ (Capel11943)
Adm	Mussau	tou	‘sugarcane’
Adm	Seimat	topu	‘sugarcane’
Adm	Pak	tuo	‘sugarcane’
NNG	Mengen	tau	‘sugarcane’
NNG	Tuam	tov	‘sugarcane’
NNG	Lukep	to	‘sugarcane’
NNG	Gedaged	tou	‘sugarcane’
NNG	Manam	tou	‘sugarcane’
PT	Wedau	tom	‘sugarcane’
PT	Bwaidoga	tovu	‘sugarcane’
PT	Duau	tohu	‘sugarcane’
PT	Saliba	tou	‘sugarcane’
PT	Motu	tohu	‘sugarcane’
MM	Vitu	tovu	‘sugarcane’
MM	Nakanai	tovu	‘sugarcane’
MM	Tangga	tuf	‘sugarcane’
MM	Siar	tu	‘sugarcane’
MM	Mono-Alu	tohu	‘sugarcane’
MM	Marovo	tovu	‘sugarcane’
TM	Buma	luro	‘coconut’
TM	Nebao	na-nə	‘coconut’
TM	Asuboa	u-ñio	‘coconut’
TM	Tanibili	no-ñio	‘coconut’
SES	Gela	tovu	‘sugarcane’
SES	Kwara’ae	ufu	‘sugarcane’
SES	Arosi	ohu	‘sugarcane’
NCV	Mota	tou	‘sugarcane’
NCV	Mwotlap	tōw	‘sugarcane’
NCV	Ambae	tovu	‘sugarcane’
NCV	Araki	rovu	‘sugarcane’
NCV	Tamambo	tovu	‘sugarcane’
NCV	Nokuku	tov[u]	‘sugarcane’
NCV	Kiai	tovo	‘sugarcane’
NCV	Raga	toi	‘sugarcane’
NCV	Naman	ni-cəv	‘sugarcane’
NCV	Neve’ei	netev	‘sugarcane’
NCV	Larëvat	n-sev	‘sugarcane’
NCV	Uripiv	na-tiv	‘sugarcane’
NCV	Port Sandwich	na-roev	‘sugarcane’
NCV	Paamese	a-tehi	‘sugarcane’
NCV	Lewo	(puru)tevi	‘sugarcane’
NCV	Namakir	tov	‘sugarcane’
SV	North Tanna	nə-təp	‘sugarcane’
SV	Lenakel	nə-ruw	‘sugarcane’
SV	Southwest Tanna	nə-tukʷ	‘sugarcane’
SV	Anejom̃	ne-to	‘sugarcane’
NCal	Iaai	(aa)kü	‘sugarcane’
NCal	Xârâcùù	de	‘sugarcane’
Mic	Kiribati	tou	‘fruit of the pandanus, chewed like sugarcane’
Mic	Kosraean	tʌ	‘sugarcane’
Mic	Ponapean	sēw	‘sugarcane’
Mic	Woleaian	wōw	‘sugarcane’
Mic	Puluwatese	wōw	‘sugarcane’
Mic	Chuukese	wōw	‘sugarcane’
Fij	Wayan	tovu	‘sugarcane’
Fij	Bauan	dovu	‘sugarcane’

cf. also:

Mic

Marshallese

taw

‘sugarcane’ (probably a loan, according to Bender et al. 2003)

2.2. Stimulants and narcotics⇫

One traditional stimulant, the betelnut, Areca catechu (§2.2.1 ), and one traditional narcotic, kava, Piper methysticum (§2.2.3) are in more or less complementary distribution in Oceania. Betelnut is chewed throughout New Guinea and NW Island Melanesia, whilst kava is consumed in Vanuatu, Fiji, Tonga and Samoa. The one point where the betelnut and kava domains overlapped was Vanikoro, to the east of the main Solomons archipelago (Darrell Tryon, pers. comm.). There were also locations in the small islands of the Admiralties and along the coast of the Madang Province of Papua New Guinea, i.e. in the betelnut domain, where kava was also used, having apparently found its way there from its place of domestication in Vanuatu.

2.2.1. Areca catechu, betel palm, betelnut, TP buai, P bilnat, natnat (Arecaceae)⇫

Areca catechu is a slender palm which grows up to 30 m. Its crown is smaller relative to its height than that of a coconut, and leaflets are less densely spaced. A many-branched flower cluster develops below the leaf spathe, and the fruit develop from the flowers. The resulting branched cluster of fruit is a familiar sight in markets in lowland Papua New Guinea. The fruit consists of a seed enclosed in a thin green skin and a hard fibrous pericarp, i.e. husk (Figure 13.1, left).

Figure 13.1: Left: Areca catechu, betelnut: A, tree; B, portion of inflorescence; C, portion of fruit cluster; D, fruit: longitudinal section showing fibrous exocarp and inner chewed ‘nut’. Right: Piper betle: E, plant; F, G, H, flowering shoots of three varieties.

Chewing betelnut as a stimulant is widespread in Papua New Guinea and the Solomon Islands, but not in other parts of the Pacific. Palms are grown in village groves or singly near houses. The seed may be chewed alone, but usually people chew a quid consisting of the seed, lime and a catkin or leaf of Piper betle (§2.2.2) (Paijmans 1976: 135).¹ Chewing the seed induces salivation, and if lime is present it turns the chewed mass bright red. Some people swallow all but the initial burst of saliva, whilst others spit out the red masticate. Initially, chewing leads to a very short-lived dizziness, folllowed by a sense of renewed wakefulness. In Papua New Guinea and parts of the Solomons chewing betelnut is a social ritual when people meet. Convention requires that the host offer betelnut to visitors (Henderson & Hancock 1988: 146). The husk of the fruit is sometimes used to clean the teeth after chewing, as the quid causes uncleaned teeth to turn dark brown.

Like other palms, the trunk of the betel pallm can be split and the outer wood used for walling or flooring. In the western Solomons, liquid is squeezed from the husk into the eyes of conjunctivitis sufferers.

POc *buaq ‘Areca catechu’ is unusual, as both it and POc *puaq ‘fruit’ evidently reflect a single etymon, PMP *buaq, which had many meanings. Blust (ACD) gives the following glosses:

fruit; areca palm and nut; grain; berry; seed; nut; endosperm of a sprouting coconut; kidney; heart; finger; calf of the leg; testicle; various insects; scar tissue; roe; bud; flower; blossom; bear fruit; words, speech, or songs; meaning, contents of discussion; numeral classifier for roundish objects.

He makes the following observations:

1. PMP *buaq formed the head of many head-attribute constructions which functioned as the names of particular fruits (*buaq kuluR ‘breadfruit’, *buaq niuR ‘coconut’, *buaq pahuq ‘mango’, *buaq punti ‘banana’, etc.) … .
2. PMP *buaq almost certainly functioned in addition as a numeral classifier which applied to roundish or fruit like objects … .
3. Most WMP and CMP reflexes support an inference that the meanings ‘fruit’ and ‘areca nut’ were associated with the same morpheme, but the Oc[eanic] evidence suggests that there were two homophonous morphemes. I assume that one morpheme expressed both meanings, but that in POc these meanings become disassociated and connected respectively with oral and nasal grade reflexes of PMP *buaq (POc *puaq ‘fruit’, *mpuaq [our POc *buaq] ‘betel [areca] nut’) … Quite how this disassociation occurred is not clear, as no other similar cases are known. In vol. 1 (ch.2, §2.4), I explained that PMP *band *p merged then split again into POc *b and *p. As a result, one cannot predict whether a PMP etymon with initial *b- will begin with POc *b- or with POc *p–but PMP *buaq is the only etymon known to have both outcomes, in POc *buaq ‘betel [areca] nut’ and POc *puaq ‘fruit’ .

Frantisek Lichtenberk (1998) examines the question of whether POc speakers chewed betelnut, concluding that they did, since a suite of terms associated with betel chewing can be reconstructed. Certainly there are no grounds in the set below for thinking that POc *buaq was diffused across Oceanic languages rather than continued from POc. The easternmost limit of betelnut-chewing is in the Temotu Province of the Solomons, but *buaq is reflected in Southern Vanuatu, where it is applied to palms of other kinds (John Lynch, pers. comm.).

PAn		*buaq	‘fruit’ (ACD; Dempwolff 1938)
PMP		*buaq	‘roundish or fruit-like object, including betelnut, Areca catechu’ (see above for the full ACD gloss)
POc		*buaq	‘betelnut, areca nut, palm, Areca catechu’ (Capell1943: *pua(q))
Adm	Mussau	(ai) bua	‘Areca catechu’
Adm	Loniu	bue	‘Areca catechu’ (Nevermann 1934)
Adm	Nyindrou	puii	‘Areca catechu’
SJ	Sobei	pue	‘Areca catechu’
NNG	Tami	bu	‘Areca catechu’
NNG	Kove	vua	‘Areca catechu’
NNG	Bariai	bua	‘Areca catechu’
NNG	Lukep	bu	‘Areca catechu’
NNG	Yabem	buʔ	‘Areca catechu’
NNG	Manam	bua	‘Areca catechu’
PT	Gabadi	bua(kau)	‘Areca catechu’
PT	Motu	bua(tau)	‘Areca catechu’
PT	Kilivila	bua	‘Areca catechu’
MM	Vitu	bua	‘Areca catechu’
MM	Bola	bua	‘Areca catechu’
MM	Nakanai	bua	‘Areca catechu’
MM	Tolai	bua	‘Areca catechu’
MM	Notsi	bua	‘Areca catechu’
MM	Tabar	bua	‘Areca catechu’
MM	Tangga	bu	‘Areca catechu’
MM	Tomoip	bu	‘Areca catechu’
MM	Selau	boko	‘Areca catechu’
MM	Taiof	bok	‘Areca catechu’
MM	Tinputz	poe	‘Areca catechu’
MM	Banoni	buɣava	‘Areca catechu’ (metathesis of *bu(v)aɣa)
MM	Uruava	bua	‘Areca catechu’
MM	Torau	buka	‘Areca catechu’ (metathesis of *buak)
TM	Buma	buioe	‘Areca catechu’
TM	Tanema	boie	‘Areca catechu’
SES	Gela	bua	‘Areca catechu’
SES	Arosi	bua	‘Areca catechu’
SES	West Guadalcanal	bua	‘Areca catechu’
SES	Talise	bua	‘Areca catechu’
SES	Kwaio	bua	‘Areca catechu’
SES	’Are’are	pua	‘Areca catechu’
SV	Lenakel	na-pʷo(k)	‘palm sp.’
SV	Kwamera	na-puei	‘coconut’

2.2.2. Piper betle (syn. Chavica betle), betel pepper, TP daka (Piperaceae)⇫

The betel pepper vine, Piper betle, is a woody climber with catkins, pendulous spikes of berries in a crowded mass forming a cylindrical syncarp (a composite fruit) (Figure 13.1 , right). It climbs on Erythrina indica and Artocarpus incisa. The catkins and/or leaves are chewed with betelnut, Areca catechu (§2.2.1), and lime throughout the Bismarck Archipelago and the Solomons (Peekel 1984: 124, Kwa’ioloa & Burt 2001:226, Hviding 2005: 115).

The reconstruction of POc *[pu-]pulu ‘betel pepper, Piper betle’ is taken from Lichtenberk’s (1998) discussion of whether POc speakers chewed betelnut. His reconstruction of POc *pulu is based partly on data assembled for the present project, with the addition of Chamorro pu-pulu ‘Piper betle’, supporting the: PMP reconstruction, and a Manam reflex. I add Mussau ulo.

PMP		*pu-pulu	‘Piper betle’ (Frantisek Lichtenberk 1998)
POc		*[pu-]pulu	‘betel pepper, Piper betle’ (Frantisek Lichtenberk 1998: *pulu)
Adm	Mussau	ulo	‘betel pepper, Piper betle’
Adm	Loniu	pun	‘betel pepper, Piper betle’
Adm	Bipi	pun	‘betel leaf’
Adm	Bipi	(pue)pun	‘betel pepper’
NNG	Lukep	ul	‘betel pepper, Piper betle’
NNG	Takia	ful	‘betel pepper, Piper betle’
NNG	Gedaged	fu	‘betel pepper, Piper betle’ (expected final-/ missing)
NNG	Manam	ulu(salaga)	‘big variety of betel vine’ (salaga ‘be long’; Frantisek Lichtenberk 1998)
SES	Bugotu	vu-vulu	‘betel pepper, Piper betle’ (Henderson and Hancock 1988)

PMM		*siqa(r,R)(a)	‘betel pepper, Piper betle’
MM	East Kara	sie	‘Piper betle’
MM	Tabar	sia	‘Piper betle’
MM	Lihir	sie	‘Piper betle’
MM	Patpatar	sier	‘Piper betle’
MM	Tolai	ier	‘Piper betle’
MM	Nehan	hiara	‘Piper betle’
MM	Petats	sil	‘Piper betle’
MM	Teop	hia(kuru)	‘Piper erectum’
MM	Tinputz	(ta)sian	‘Piper betle’
MM	Banoni	siɣana	‘betel pepper catkin’

cf. also:

MM

Marovo

hirata

‘Piper betle’

2.2.3. Piper methysticum, kava (Piperaceae)⇫

Figure 13.2: Leaves, stem and plant habit of kava, Piper methysticum

The kava plant, Piper methysticum, is a many branching plant with rounded green leaves. The plant is grown, usually near houses, exclusively from cuttings, and a narcotic is made from it in parts of Remote Oceania. Traditionally kava is consumed as a drink. The root is first reduced to small fragments by chewing, grinding or pounding. The fragments are deposited in a bowl, mixed with water and strained through the cloth-like fibre of a coconut spathe (ch.12, §5.3) to give a cloudy grey liquid (Paijmans 1976: 135). In Fiji, Tonga and Samoa the liquid is made from mature roots, is of low strength and plays a part in various ceremonies. In Vanuatu it is made from the roots of green plants and often has a much greater narcotic effect. Initially it causes the blood vessels in the lips and tongue to contract with a certain numbing effect. The drinker then senses some degree of euphoria, followed by a sense of calm well-being and clear thinking and a general relaxation of the muscles.

Kava is also consumed in scattered areas of New Guinea and the Bismarcks. In fact the only Oceanic-speaking areas in this region where it is or was drunk are the small southeastern islands of the Admiralties (Lou, Pam and Baluan; Brunton 1988) and among the Takia ofKarkar Island, where in the 1980s some older men told me that it had been drunk within their lifetimes. The Takia were traditionally in contact with Papuan-speaking villages on the Rai Coast (the north coast of the mainland east of Madang) where Mikloucho-Maclay (Mikloucho-Maclay 1886, 1975) reported that the kava root was chewed by some inhabitants of just a few villages. Bourke (1990) considers that the limited distribution of kava indicates recent introduction, and Miklouho-Maclay noted in 1886 that it had only recently been introduced to the Rai Coast.

Pawley & Green (1973) proposed a POc reconstruction *kawa ‘Piper methysticum’. However, the facts that Piper methysticum only occurs in domesticated form and that it is not generally consumed in Oceanic-speaking communities in the Bismarcks imply that it was not present there in POc times, and that we should not expect to be able to reconstruct a POc term for it. The origin of kava has attracted a good deal of debate, both botanical and linguistic, over the past 35 years, partly because of its presence at New Guinea locations (Brunton 1988, Lebot 1989, Lebot et al. 1992, T. Crowley 1994, Lynch 2002f).

The best supported hypothesis appears to be that Piper methysticum is a domesticated variety of Piper subbullatum (syn. Piper wichmannii),² a plant of similar appearance which grows wild in the Solomons (Kwa’ioloa & Burt 2001: 206), and that this domestication took place in northern Vanuatu. The linguistic evidence for this position is presented by Lynch (2002f). He reconstructs a POc term *kawaRi, which he glosses as ‘root with special properties: one or more of Zingiber zerumbet, Piper subbullatum, fish-poison plants’, i.e. as a generic term for what in the title of his article he calls ‘potent roots’. He points out that none of the apparently directly inherited reflexes of *kawaRi outside Vanuatu, Fiji and Polynesia means ‘kava’ or ‘Piper methysticum’. Only reflexes with a form that suggests borrowing have this meaning. The evidence suggests that POc *kawaRi is reconstructable, but only with the ‘potent roots’ meaning. Only in northern Vanuatu and regions settled from it (the rest of Vanuatu, Fiji and Polynesia) do we find directly inherited reflexes denoting ‘kava’ or’ Piper methysticum’. The reconstructions and directly inherited reflexes below are from (Lynch 2002f), except where shown.³

POc		*kawaRi	‘root(s) with special properties: one or more of Zingiber zerumbet, Piper subbullatum, and various fish-poison plants’ (Lynch 2002f) 4
Adm	Baluan	kau	‘Piper subbullatum’ (Ambrose 1991)
NNG	Sissano	(una)kaw	‘k.o. ginger’
PT	Muyuw	ka-kawowa	‘Piper sp.’
PT	Gapapaiwa	kaware	‘k.o. ginger’
PT	Tubetube	kalava	‘Piper subbullatum (?)’ (metathesis)
MM	Maringe	kʰori	‘(use) fish poison’
PROc		*kawa	‘kava’; ‘ginger, fish-poison plants’ (presumably)
PNCV		*kea	‘kava; sour, bitter’ (Lynch 2004a)
NCV	Lehali	n-ɣa	‘kava’
NCV	Mota	ɣea	‘kava’
NCV	Mwotlap	na-ɣa	‘kava’
NCV	Vera’a	ɣie	‘kava’; ‘kava; sour, bitter’
NCV	Mwesen	ɣe	‘kava’
NCV	Vera’a	ɣie	‘kava’; ‘kava; sour, bitter’
NCV	Vurës	ɣe	‘kava; sour, bitter’
NCV	Araki	hae	‘kava; sour, bitter’
NCV	Lewo	(wi)kawa	‘brackish water’ (wi ‘water’)
PCP		*kawa	‘kava, Piper sp., fish-poison plants; sour, bitter’; ‘Zingiber spp.’ (probably also)
Fij	Bauan	(wā)gawa-gawa	‘Piper betle’
Fij	Nadrogā	kawa	‘fish-poison tree, Euphorbia sp. (?)’
Fij	Lautoka	kawa	‘fish-poison vine, Derris sp.’
Fij	(many dialects)	kawa	‘k.o. round fish-trap with mouth on top, made from Derris vines’
PPn		*kawa	‘kava; sour, bitter’
Pn	Tongan	kava	‘kava’
Pn	West Futunan	kava	‘kava’
Pn	Samoan	ʔava	‘kava’
Pn	Samoan	ʔa-ʔava	‘pungent, acrid’
Pn	Hawaiian	ʔava	‘kava; sour, bitter’
Pn	Rarotongan	kava	‘kava; sour’
Pn	Māori	kawa	‘sour, bitter; perform certain kinds of ceremony’
Pn	Māori	kawa-kawa	‘Macropiper excelsum’

The denotations of PCP *kawa as ‘Piper sp.’, ‘Zingiber spp.’ and ‘fish-poison plant’ are continued as compounds in PPn:

PNPn		*kawa-kawa qatua	‘a shrub or vine, Piper sp.’
Pn	East Futunan	kava-kava atua	‘a climbing vine, Piper vaupelii’
Pn	Marquesan	kava-kava atua	‘Piper latifolium, Piper tristachyon’
Pn	Rarotongan	kava-kava atua	‘a shrub, Piper latifolium’

PPn		*kawa-sasa	‘a creeper used to poison fish’
Pn	Tongan	kava-haha	‘Derris trifoliata, used as a fish poison’
Pn	East Futunan	kava-sasa	‘a vine’
Pn	Samoan	ʔava-sa	‘shrub sp., Tephrosia purpurea sp., used to poison fish’

PPn		*kawa-susu	‘shrub sp., Tephrosia sp., used to poison fish’
Pn	Niuean	kau-huhu	‘a plant used to stupefy fish’
Pn	East Uvean	kava-huhu	‘shrub sp., Tephrosia piscatoria’
Pn	East Futunan	kava-susu	‘shrub sp., the leaves of which are used to poison fish’
Pn	Nukuoro	gava-usu	‘Barringtonia asiatica, used to poison fish’

PNPn		*kawa-pui	‘a plant, Zingiber sp.’
Pn	Tikopia	kava-pui	‘plant of the ginger family’
Pn	East Uvean	kava-pui	‘Zingiber sp.; white ginger, Hedychium coronarium’5
Pn	Anutan	kava-pui	‘Alpinia sp.’
Pn	Samoan	ʔava-pui	‘a herb, Zingiber zerumbet’
Pn	Hawaiian	ʔawa-puhi	‘wild ginger, Zingiber zerumbet’
Pn	Tahitian	ava-puhi	‘an odiferous plant’

3. Plants which provide materials for making things⇫

3.1. Bambusa, Schizostachyum and_Nastus_ spp., bamboo, TP mambu, Bbambu (Poaceae)⇫

Three genera of bamboos are represented in NW Island Melanesia: Bambusa, Schizostachyum and Nastus. They are treated together here, since only one of the reconstructable terms can be related to a particular species. A number of species are often cultivated.

Three species of Bambusa, all thick-walled, are reported from the region. The most common in the Bismarcks and the largest in the Solomons is Bambusa vulgaris, with a culm (walled stem) about 5-10 m high and 10-15cm in diameter and a short internode distance around 30-50 cm. However, Bambusa vulgaris was introduced after European contact (Rhys Gardner, pers. comm.) and was not among the denotata of the POc terms below. B. blumeana is very similar and apparently important only in Malaita. B. forbesii, reported from the Bismarcks, is smaller, with a culm 2-4 m high and 1-4cm diameter, and has broad leaves (Peekel 1984: 55-56, Henderson & Hancock 1988: 203-205).

Bambusa bamboos are used in housing construction for beams and rafters, as well as to make containers for lime and water and tongs for lifting the hot stones of a stone oven (Henderson & Hancock 1988: 204–205). At Marovo they are used to make large traditional tuna-fishing rods and a whole stem serves as a ‘ladder’ to climb sago palms in order to cut its fronds (Hviding 2005: 117).

Two species of the thin-walled genus Schizostachyum are reported from the region. No comparative account is available, but the two appear similar. Schizostachyum lima, found in the Bismarcks, has a culm about 4–8 m high and 4–10cm in diameter, with internode lengths of 50-80 cm (Peekel 1984: 55). S. tessellatum, found in the Solomons, is described as tall and very thin, growing wild in many habitats in clumps 10 m tall. It is the only bamboo amongst the three genera that is observed to flower regularly

Schizostachyum species are a source of very straight, lightweight poles with many purposes: as battens from which to hang sago leaf thatch, as yam poles, as small fishing rods and as spears. Pieces of Schizostachyum are used to make traditional panpipes and coconut scrapers. They are also used as fencing material (Henderson & Hancock 1988:202-203, Hviding 2005: 107, Kwa’ioloa & Burt 2001: 202).

Three species of the thin-walled genus Nastus are reported from the region. One is Nastus productus, a small, often drooping bamboo. A second is Nastus obtusus, a very slender bamboo which is almost always cultivated and attains heights of over 20 m with stem diameters of 8-11cm and internode lengths of 50 em to a metre. A third Nastus species is not given a scientific name,⁶ but sources describe it as pliant and behaving as a tree or ground creeper. It has very short internode lengths and small, narrow leaves (Peekel 1984: 55, Henderson & Hancock 1988: 177, Kwa’ioloa & Burt 2001: 200-201).

The first two Nastus species, and especially Nastus obtusus, have similar uses to those of Schizostachyum, but the third Nastus species is considered by the Kwara’ae to be useless (Henderson & Hancock 1988: 176, 199-201, Kwa’ioloa & Burt 2001:200–202, Hviding 2005: 112).

Figure 13.3: Nastus obtusus

The term reconstructed as *qauR ‘bamboo spp.’ below was probably the generic term for bamboos of the three genera mentioned above, as it still is in some modern languages. In Tolai, for example, we find the following:

MM	Tolai	kaur	‘Bambusa vulgaris’
MM	Tolai	kaur lubaŋ	‘Bambusa vulgaris, larger variety’
MM	Tolai	kaur goragoro	‘Schizostachyum lima’
MM	Tolai	kaur laur	‘thin-walled bamboo species’

PAn *qauR ‘bamboo’ is attested in Fom1osan languages, PMP *qauR ‘bamboo’ in languages across Indonesia (Blust, ACD ), and a number of Oceanic languages have reflexes indicating POc *qauR. However, there are als:o a number of languages that reflect initial *k- rather than *q-, and I show below the POc segments seemingly reflected by the consonant(s) of each form in order to illuminate the discussion here. In those Remote Oceanic languages which reflect POc *k- and *q- differently from one another the form indeed reflects initial *q-. However, in Meso-Melanesian languages which reflect POc *k- and *q- differently and in Temotu languages the form reflects *k-, not *q-.

On the basis of Meso-Melanesian reflexes POc *kauR ‘bamboo’ was reconstructed as in vol.1 (ch.4, §6.2). However, non-Oceanic and Remote Oceanic data support *qauR.⁷

Among languages which retain a reflex of the final consonant, Mussau and Petats reflect *d or *dr, rather than *R, and Nakanai reflects *s or *c.⁸ The Bilibil, Gedaged and Sio forms reflect final *-R where the final consonant is normally lost.

Blust (1984) reconstructed POc *kaudru⁹ ‘bamboo sp.’ to account for the Mussau, Petats and Selau forms.¹⁰ However, it does not account for the other irregularities below, and its putative reflexes are here treated as (albeit irregular) reflexes of POc *qauR. Instead, the simplest linguistic explanation of these irregularities is that the forms reflect borrowings among neighbouring Oceanic languages. Indeed, it is possible that some of the forms with regular reflexes also result from borrowings, but happen to have regular reflexes of the consonants. However, it is not easy to see why such borrowings might have occurred, unless much of Near Oceania lacked useful bamboo species, and new species were imported by Oceanic speakers. Hviding (2005: 107, 112, 117) hints at this when he writes that old people say that Bambusa vulgaris, B. blumeana, Nastus obtusus and a Schizostachyum species were all long ago introduced to Marovo from elsewhere. It is just possible that some of these borrowings were associated with the introduction of Bambusa vulgaris after contact with Europeans.

PAn		*qauR	‘bamboo sp.’ (ACD)
PMP		*qauR	‘type of large bamboo’ (ACD)
POc		*qauR	‘bamboo spp.’
Adm	Mussau	kauru	‘large bamboo from which combs are made’ (*k~q *d~dr)
NNG	Mangap	kau-kau	‘bamboo (big)’ (*k~q)
NNG	Bilibil	aur	‘bamboo’ (*k~q *r~R)
NNG	Gedaged	auɬ	‘bamboo, tall with thick walls’ (*k~q *r~R)
NNG	Sio	kaul(a)	‘bamboo’ (*k~q *r~R)
NNG	Lukep	kau-kau	‘bamboo’ (*k~q)
NNG	Atui	kaur	‘bamboo’ (*k~q *rRd~dr)
NNG	Akolet	e-kaur	‘bamboo’ (*k~q *rRd~dr)
MM	Bali	kaura	‘bamboo’ (*k *r~R)
MM	Nakanai	kauru	‘large bamboo’ (*k *s~c)
MM	Ramoaaina	kauru	‘bamboo’ (*k~q *r~R)
MM	Tolai	kaur	‘bamboo, generic’ (*k~q *r~R)
MM	Patpatar	kor	‘bamboo, generic’ (*k~q *r~R)
MM	Petats	kahur	‘bamboo’ (*k *d~dr)
MM	Selau	kawur	‘bamboo’ (*k *rRd~dr)
MM	Mono-Alu	aulu	‘bamboo sp.’ (*k *rRl)
SES	Gela	ɣau	‘a bamboo, bamboo knife’ (*k~q)
SES	Bugotu	ɣau	‘bamboo; bamboo sliver used as a knife’ (*k~q)
SES	West Guadalcanal	ɣau-ɣau	‘bamboo’ (*k~q)
SES	Lau	ʔau	‘bamboo’ (*k~q)
SES	Kwara’ae	kaʔo	‘bamboo (generic)’ (*k~q)
SES	Kwaio	ʔau	‘bamboo (generic)’ (*k~q)
SES	Sa’a	äu	‘bamboo’ (*k~q)
SES	Arosi	ʔau	‘a bamboo’ (*k~q)
NCV	Mota	au	‘the bamboo, generic’ (*k~q)
NCV	Southeast Ambrym	eou	‘bamboo’ (*k~q)
NCV	Paamese	eau	‘bamboo’ (*k~q)
NCV	Lewo	(pla)yu	‘bamboo, as knife’ (*k~q)
NCV	Port Sandwich	n-au	‘arrow’ (*k~q)
NCV	Namakir	ʔo	‘bamboo’ (*q)
NCV	Nguna	na-au	‘wild cane, reed; flute, mouth organ’ (*k~q)
SV	Sye	n-au	‘bamboo’ (*q)
SV	Ura	n-au	‘spear’ (*q)
SV	North Tanna	n-ao	‘bamboo’ (*q)
SV	Whitesands	n-au	‘bamboo’ (*q)
SV	Lenakel	n-au	‘bamboo’ (*q)
SV	Kwamera	n-au	‘bamboo’ (*q)
SV	Anejom̃	n-au	‘bamboo’ (*q)
NCal	Nêlêmwa	ŋ-ga	‘bamboo’ (*q)
NCal	Nyelâyu	ŋ-gao	‘bamboo’ (*q)
NCal	Jawe	ŋ-go	‘bamboo’ (*q)
NCal	Pije	(du)ko	‘bamboo’ (*q)

Reflexes of POc *bʷau ‘bamboo’ are largely complementary to those of *qauR above, with overlaps on the mainland coast of the Vitiaz Strait and in New Caledonia. The Manam and Kaiep reflexes provide a weak indication that *bʷau may have denoted bamboo as a construction material.

POc		*bʷau	‘bamboo’
SJ	Ormu	bau	‘bow’
NNG	Tami	bʷai	‘bamboo’
NNG	Malasanga	boa	‘bamboo’ (irregular loss of *u)
NNG	Bing	buau	‘bamboo’
NNG	Manam	buau-buau	‘(house) wall’
NNG	Kaiep	ba-boi	‘(house) wall’
PT	Molima	bau-bau	‘bamboo tube, smoking pipe made therefrom’ (A. Chowning, pers. comm.)
PT	Dobu	bau-bau	‘bamboo’
PT	Duau	bau-bau	‘bamboo’
PT	Wedau	bau-bau	‘bamboo’
PT	Tawala	bau-bau	‘bamboo’
PT	Tubetube	bau-bau	‘bamboo’
PT	Motu	bau	‘bamboo’
Mic	Marshallese	bʷae	‘bamboo’
Mic	Woleaian	pʷāɨ	‘bamboo’
Mic	Chuukese	pʷāw	‘bamboo’
Mic	Puluwatese	pʷāy	‘bamboo’
NCal	Nyelâyu	gao	‘bamboo’

Five more reconstructions are listed below: POc *bitu(ŋ) ‘bamboo sp., probably Schizostachyum glaucifolium’, POc *botu(ŋ), POc *kopu, POc *bʷele, POc *bue ‘(made of) bamboo’. Except for the first and last, the glosses of the supporting data do little to help elucidate their denotata, some of which were presumably taxa made up of one or more of the species discussed above. The gloss of Nakanai hele ‘small cultivated bamboo used for thatching rods’ suggests that it denotes a Schizostachyum species, but this is insufficient information on which to base a gloss of POc *bʷele.

The two reconstructions below are better supported by non-Oceanic than by Oceanic reflexes. Since Bambusa vulgaris is a recent import, the Fijian and Polynesian glosses indicate that PCP *hitu probably denoted Schizostachyum glaucifolium. This may also be true of the POc reflex, but this is uncertain.

PMP		*bituŋ, *pituŋ	‘bamboo sp.’ (ACD)
POc		*bitu(ŋ)	‘bamboo sp.’ (ACD)
PCP		*bitu	‘bamboo sp., probably Schizostachyum glaucifolium’ (ACD)
Fij	Wayan	bitu	‘generic for two bamboo species, Schizostachyum glaucifolium and Bambusa vulgaris’
Fij	Bauan	bitu	‘bamboo, Bambusa sp. or Schizostachyum glaucifolium’
Pn	Tongan	pitu	‘bamboo, Bambusa vulgaris, variety with yellow stems’

PAn		*betuŋ	‘bamboo of very large diameter, probably Dendrocalamus sp.’ (ACD)
POc		*botu(ŋ)	‘large bamboo, presumably Bambusa sp.’ (ACD: *potuŋ)
Adm	Lou	pot	‘large thick bamboo variety’

POc		*kopu	‘bamboo sp.’
Adm	Drehet	ɔp	‘bamboo sp.’
Adm	Nyindrou	op	‘bamboo sp.’
MM	Tinputz	kop	‘bamboo sp.’
MM	Teop	kopu	‘bamboo sp.’

The only non-WOe evidence for POc *bʷele ‘bamboo’ is from Lonwolwol (NCV), where bʷele-bo consists of bʷele- ‘hollow vessel’ and bo, apparently reflecting PNCV *bue ‘made of bamboo’ (see below). A possible inference is that *bʷele denoted bamboo as a household utensil, e.g. a container for water.

POc		*bʷele	‘bamboo sp.’
NNG	Mapos Buang	a-pɛl	‘bamboo sp.’
NNG	Patep	pɛel	‘bamboo sp.’
MM	Bulu	bele	‘bamboo sp.’
MM	Nakanai	bele	‘small cultivated bamboo used for thatching rods’ (Goodenough 1997)
MM	Tiang	bele	‘bamboo sp.’
NCV	Lonwolwol	bʷele(bo)	‘bamboo’ (bʷele- ‘hollow vessel’; -bo <POc *bue ‘(made of) bamboo’)

Hus (Adm) ʙu ‘bamboo’ corresponds regularly with the NCV items below, implying that it and PNCV *bue both reflect a POc *bue ‘(made of) bamboo’.

POc		*bue	‘(made of) bamboo’
Adm	Hus	ʙu	‘bamboo’ (Nevermann 1934)
PNCV		*bue	‘(made of) bamboo’ (Clark 1996)
NCV	Mota	pue	‘bamboo water-carrier’
NCV	Raga	bua	‘bamboo (generic); knife’
NCV	Nokuku	pue	‘water-pot’
NCV	Kiai	pue	‘bamboo’
NCV	Araki	(vi)pue	‘bamboo tree’
NCV	Tamambo	(vu)bue	‘bamboo tree’
NCV	Uripiv	na-ʙu	‘bamboo’
NCV	Port Sandwich	na-mbu	‘bamboo; bamboo water container; bamboo knife’
NCV	Port Sandwich	mbʷe(var)	‘hard bamboo’ (na-var ‘stone’)
NCV	Labo	(na-na)mbuo	‘bamboo’
NCV	Lonwolwol	(bʷele)bo	‘bamboo’ (bʷele- ’hollow vessel, empty shell of)
NCV	Lewo	(pila)pʷe	‘bamboo (used for walls of house); length of thatch woven on bamboo’

3.2. Aleurites moluccana (syn._Aleurites triloba_), candlenut, P kandoltri, Bkandeltri (Euphorbiaceae)⇫

Aleurites moluccana is a tree which grows to between 10 and 35 m in height, depending on its location. Its young leaves and inflorescence are dusted with grey to rust-brown scurf, and the mature fruit is chestnut brown, 3-7cm in diameter. It has little pulp and a thick rind that encloses one or two very large seeds, the candlenuts. There appears to be considerable intra-species variation across its range, from India to eastern Polynesia, as well as a striking variety of uses. The seeds of some varieties are toxic, but others can be eaten without a problem. Even toxic seeds can often be eaten in small quantities, and the toxicity is reduced by roasting (Henty 1982, Walter & Sam 2002: 87).

In Vanuatu Aleurites moluccana is not cultivated, as planted seeds do not germinate well, but people protect new seedlings (Walter & Sam 2002: 88). In Fiji the tree occurs close to villages but not in the wild, implying that it is planted.

Figure 13.4: Aleurites moluccana: A, tree; B, leaves and inflorescence; C, whole fruit; D, kernel (candlenut).

The candlenut owes its name to the fact that in earlier times the nuts were threaded onto the midrib of a coconut palm leaf and lit, burning slowly one after another and giving a feeble light. The smoke given off is also a good insect repellant (Sperlich 1997). Charred seeds are used in New Ireland to make a sooty paint for blackening the face in mourning (Peekel 1984: 313) and by Ponapeans for making a black or brown dye. The oil extracted from them is also used as a paint base in New Britain (Powell 1976) and to polish wood in Fiji, where an extract of the seed is also used to scent the oil (Gardner & Pawley 2006). The products of Aleurites moluccana have numerous medicinal uses (Walter & Sam 2002: 89).

POc *tuRi-tuRi ‘Aleurites moluccana’ is reconstructed below. It is quite widely supported, but it is by no means certain that its denotatum was Aleurites moluccana. Blust (ACD) reconstructs PEMP *tuRi-tuRi ‘Aleurites moluccana’ on the basis of Biak kuker ‘tree with edible nut’ and Central Pacific reflexes. However, Geraghty (2004: 72) believes that PPn *tuitui ‘Aleurites moluccana’ was identical with *tuitui ‘strung together’, reflecting the fact that candlenuts are strung together for use as torches or sources of dye. He does not discuss Blust’s reconstruction, but if Geraghty is right, the Biak, Motu and Roro terms must be dismissed as chance resemblances (none has the specific gloss Aleurites moluccana).

PEMP		*tuRi-tuRi	‘candlenut tree, Aleurites moluccana’ (?) (ACD)
POc		*tuRi-tuRi	‘candlenut tree, Aleurites moluccana’ (?)
PT	Motu	turi-turi	‘Cordia subcordata’
PT	Roro	curi-curi	‘tree sp., with wood used for making drums’
PCP		*tui-tui	‘candlenut tree, Aleurites moluccana’
Fij	Bauan	tui-tui	‘candlenut, Aleurites moluccana’
Pn	Tongan	tui-tui	‘candlenut, Aleurites moluccana’
Pn	Niuean	tui-tui	‘candlenut, Aleurites moluccana’
Pn	East Uvean	tui-tui	‘candlenut, Aleurites moluccana’
Pn	East Futunan	tui-tui	‘candlenut, Aleurites moluccana’
Pn	West Futunan	tu-tui	‘candlenut, Aleurites moluccana’
Pn	Emae	tui-tui	‘candlenut, Aleurites moluccana’
Pn	Tahitian	tu-tui	‘candlenut, Aleurites moluccana’
Pn	Tuamotuan	tui-tui	‘candlenut, Aleurites moluccana’
Pn	Hawaiian	ku-kui	‘candlenut, Aleurites moluccana’
Pn	Rarotongan	tui-tui	‘candlenut, Aleurites moluccana’

3.3. Broussonetia papyrifera, paper mulberry, TP burua (Moraceae)⇫

Broussonetia papyrifera is native to Japan and Taiwan and is an ancient introduction into the rest of the Pacific. A shrub or small tree 3-5, and sometimes up to 12, metres high, it is fertile in its native range, but the plants found throughout the Pacific are all male clones, transported and planted as rootstock. It is thus subject to deliberate propagation by human agency (Whistler & Elevitch 2006a ).

As Osmond & Ross noted in vol.1 (ch.4, §5.1), Kooijman (1972:446–453) believes from descriptions of manufacture and an examination of museum pieces that bark cloth in New Guinea was made from Ficus and Artocarpus species rather than from Broussonetia papyrifera, from which bark cloth is made in other parts of the Pacific, e.g. Fiji (Gardner & Pawley 2006). However, the glosses of the items listed below suggest fairly strongly that speakers of Oceanic languages in widely separated New Guinea locations were using Broussonetia papyrifera to make bark cloth at European contact.

Figure 13.5: Broussonetia papyrifera, paper mulberry

POc *malo probably denoted Broussonetia papyrifera, the tree which throughout much of the Pacific provides the bast from which barkcloth is made. If this is so, then Broussonetia papyrifera must have been introduced into the Pacific islands by early Oceanic speakers. Matthews (1996), however, notes its absence from the Philippines and Borneo, and thinks it possible that it did not arrive in Polynesia with the earliest colonisers. POc *malo also denoted the cloth and the male genital covering made from it, and it is just possible that these were its primary meanings. It is also possible that some of the reflexes below reflect early Pacific Pidgin malo or maro ‘loincloth’. PMP *mal(u,aw) apparently denoted a species of tree which provided bast for clothmaking, but it is unclear which species this was, as it is reconstructed on the basis of POc *malo and Kaili (WMP, Sulawesi: Parigi dialect) malo ‘old term for Trema amboinensis, the tree whose bast is most commonly used for barkcloth in Sulawesi’ (Adriani & Kruijt 1901: 140, note 5, cited by R. Kennedy 1934: 242).

PMP		*mal(u,aw)	‘tree whose bast is used for barkcloth’11
POc		*malo	‘paper mulberry, Broussonetia papyrifera; barkcloth, loincloth’ (Milke 1968)
NNG	Kove	malo	‘male genital covering’ (A. Chowning, pers. comm.)
NNG	Gedaged	mal	‘tree, bark used for G-strings and blankets; loincloth made from this’
NNG	Takia	malu	‘tree sp., loin cloth made from pounded bark’
NNG	Manam	malo	‘barkcloth belt from dodoli tree, given to a boy at first initiation’ (Böhm 1983: 81)
MM	Patpatar	māl	‘cloth, clothing’
MM	Tolai	mal	‘Broussonetia papyrifera; native cloth made from bark of this tree’
MM	Teop	maro	‘cloth, clothing’
SES	Arosi	maro	‘Broussonetia papyrifera; beaten cloth of the maro tree’
NCV	Raga	malo	‘men’s loincloth’
Fij	Bauan	malo	‘Broussonetia papyrifera; hence the native cloth made from it and the former native male dress, passed between the thighs and fastened with a girdle’
PPn		*malo	‘barkcloth loin garment’ (POLLEX)
Pn	Emae	maro	‘barkcloth’
Pn	Samoan	malo	‘loincloth’
Pn	Hawaiian	malo	‘loin garment’

POc		*(m,mʷ)ase	‘wild mulberry, paper mulberry, Broussonetia papyrifera’ (Ross 1996d)
NNG	Mapos Buang	ñɛ̃s	‘Broussonetia papyrifera’
MM	Tolai	mae	‘Broussonetia papyrifera’
Fij	Wayan	masi	‘Broussonetia papyrifera’
Fij	Rotuman	mɔsi	‘tree sp. used to make bark cloth’

3.4. Morinda citrifolia, Indian mulberry, yellow wood, P ialotri, Byelotri, noni (Rubiaceae)⇫

A small tree, 5-6 min height, and occasionally more, up to 15m, Morinda citrifolia (Figure 13.6, left) grows throughout Oceania behind the beach and in dry soils exposed to the sun. Barrau (1962: 188) comments on the large shiny elliptical leaves about 20cm in length and 10cm broad. Peekel (1984: 539) recognises two varieties, one with green, yellow, white or variegated leaves, the other (var. bracteata) with enlarged calyx lobes. The small white flowers grow on an oval or cone-shaped structure, the syncarpium, which later swells to become the syncarp, ‘a heavy, pungent smelling-sometimes repulsive-succulent fruit’ (Henderson & Hancock 1988: 52). The fruits of wild varieties are spherical or elongated syncarps of 3-6cm and are soft and straw yellow at maturity, with numerous protuberances (French 1986:272 describes them as ‘warty looking lumps’). Cultivated forms have paler, larger fruit with no protuberances, but are rarely found within village areas because of the strong smell when their fruit fall and rot (Walter & Sam 2002: 204-205).

Morinda citrifolia is a tree with many uses in Oceanic speaking societies. Wild young trees are well known as a source of dyes, red and yellow, extracted by boiling the root-bark (Floyd 1954, Gowers 1976: 99). In Marovo the dyes are applied to cocount-frond baskets, in Vanuatu to vegetable fibres used in items of apparel. S. Foale (2001) describes it as an all-purpose medicine on Lihir, and Walter & Sam (2002: 206–207) refer to research indicating that the Indian mulberry contains immuno-stimulant substances. This would explain its numerous medicinal uses. In Marovo, Manus and Tonga young leaves softened over the fire are applied to infected wounds and boils to draw out pus (O’Collins & Lamothe 1989, Hviding 2005: 133, Walter & Sam 2002: 207). In Vanuatu the raw fruit is crunched and eaten to treat an enlarged spleen. However, the plant is also cultivated and the fruit of the cultivated variety eaten, raw or cooked, on small islands scattered across Papua New Guinea, on the Temotu islands of the Solomons and in the Banks and Torres Islands of Vanuatu. Elsewhere the fruit of the wild variety is a famine food (Walter & Sam 2002: 205-207). In parts of Papua New Guinea and the Solomons the young leaves, which are a rich source of carotene, are cooked as a vegetable or consumed for medicinal reasons (Henderson & Hancock 1988:54, French 1986: 272).

Figure 13.6: Left Morinda citrifolia, Indian mUtlberry: A, mature tree; B, shoot bearing leaves, fruit and flowers. Right Derris species: A, climbing plant; B, base of mature vine; C, woody vine stem; D, vine with leaves and pods; E, stem with inflorescence; F, immature leaf.

There were apparently two POc labels for Morinda citrifolia, *fiofiu and *kurat. If Verheijen (1990: 86) has correctly identified the referent of Bima (CMP) nonu as Morinda citrifolia (he marks it with a question mark), then *fiofiu is reconstructable for PCEMP and was inherited into POc. If Tagalog, Bisayan, Tausug nino Morinda citrifolia (Madulid 2001a: 527) are also cognate with POc *fiofiu, then PMP *fiefiu is reconstructable. Milke reconstructed this etymon as POc *fiofium with final *-m, but the only reflexes with this segment are Gedaged nonom and Takia nom. Both languages lose POc final *- VC, and it is more likely that -m in both languages irregularly reflects medial POc *-fi-, with reduplication in Gedaged.

Geraghty (1993) speculates that *fiofiu (his *fiofium) was the name of the plant and *kurat the name of the dye produced from it. In Geraghty (2004: 91) he finds support for this speculation in evidence provided by Mahdi (1994: 192-193). Papuan languages of the North Halmahera family show forms such as Temate guraci ‘turmeric’, Tidore guraci ‘gold’, guraci ‘yellow’. These are the source of loans in Malaya-Polynesian languages of the South Halmahera family such as Buli *guraci ‘gold’. Geraghty suggests that such a form could also have been borrowed into languages of the New Guinea region and thence into early Oceanic. This proposal seems very plausible and leads to a speculation of my own. As Geraghty points out, the formal correspondence of the Halmahe:ra and Oceanic forms is perfect. In the light of this, it is possible that POc *kurat reflects a PEMP form *gurati meaning ‘yellow dye’.

The distributions of POc *ñoñu and *kurat are complementary: *kurat is reflected solidly through Melanesia from New Ireland (Lihir, Tangga) through NW Solomonic (Nehan, Roviana), SE Solomonic, North/Central and Southern Vanuatu, New Caledonia and Fiji, while *ñoñu occurs in the Admiralties, North New Guinea, Papuan Tip, Micronesian and Polynesian. These distributions suggest that POc *ñoñu was in some sense the default term for Morinda citrifolia and that it was then replaced by *kurat in a solid Melanesian block from New Ireland to New Caledonia and Fiji. However, the reflexes of *kurat are generally regular, suggesting that replacement took place very early in the history of Oceanic.¹²

POc		*ñoñu	‘Morinda citrifolia’ (Blust 1978b; Milke 1965: *nonum)
Adm	Seimat	naun	‘Morinda citrifolia’ (*-o- > Seimat -au-) (Sorensen 1950)
Adm	Leipon	ñoñ	‘Morinda citrifolia’
Adm	Bipi	ñoy	‘Morinda citrifolia’
NNG	Gitua	nono	‘Morinda citrifolia’
NNG	Gedaged	no-nom	‘Morinda citrifolia’
NNG	Takia	nom	‘Morinda citrifolia’
NNG	Manam	noŋ	‘tree sp.’
NNG	Wogeo	ñoñ	‘Morinda citrifolia’
PT	Bwaidoga	nono	‘Morinda citrifolia’
PT	Motu	nonu	‘Morinda citrifolia’
Mic	Kiribati	non	‘Morinda citrifolia’
Mic	Marshallese	nen	‘Morinda citrifolia’
Mic	Puluwatese	nēn	‘Morinda citrifolia’
Mic	Woleaian	lēli	‘Morinda citrifolia’
Pn	Tongan	nonu	‘Morinda citrifolia’
Pn	Tikopia	nonu	‘Morinda citrifolia’

POc		*kurat	‘the dye produced from Morinda citrifolia’ (Geraghty 2004)
MM	Lihir	ulet	‘Morinda sp.’
MM	Tangga	urat	‘Morinda citrifolia’
MM	Nehan	rata	‘Morinda citrifolia’
MM	Mono-Alu	urati	‘Morinda citrifolia’ (W. McClatchey, pers. comm.)
MM	Varisi	ku-kure	‘Morinda citrifolia’ (W. McClatchey, pers. comm.)
MM	Roviana	ɣurata	‘Morinda citrifolia’
PEOc		*kurat	‘M orinda citrifolia’
SES	Longgu	ʔura	‘Morinda citrifolia’
SES	Santa Ana	ɣura	‘Morinda citrifolia’
NCV	Mwotlap	wʋy	‘Morinda citrifolia’
NCV	Mota	wura	‘Morinda citrifolia’
NCV	Ambae	hure	‘Morinda citrifolia’
NCV	Araki	hura	‘Morinda citrifolia’
NCV	Raga	ɣuresi	‘Morinda citrifolia’ (Walsh 2004) 13
NCV	Uripiv	na-ur	‘Morinda citrifolia’
NCV	Paamese	o-ulo	‘Morinda citrifolia’
NCV	Lewo	(pur)kula	‘Morinda citrifolia’
NCV	Nguna	na-kura	‘tree sp.’
NCV	South Efate	na-kur	‘Morinda citrifolia’
PSV		*na-ɣura(t,c)	‘Morinda citrifolia’ (Lynch 2001c)
SV	Sye	noɣrat	‘Morinda citrifolia’
SV	Lenakel	nauias	‘Morinda citrifolia’
SV	Kwamera	noueis	‘Morinda citrifolia’
SV	Anejom̃	na-uras	‘Morinda citrifolia’
NCal	Nyelâyu	yelac	‘Morinda citrifolia’
NCal	Iaai	hulak	‘Morinda citrifolia’
NCal	Dehu	xelek	‘Morinda citrifolia’
Fij	Wayan	kura	‘Morinda citrifolia’
Fij	Rotuman	ʔurɔ	‘Morinda citrifolia’

4. Fish poisons⇫

One method of catching fish was to put a toxic plant substance into an area of water surrounded by rocks or reef so that fish were stunned or killed and could then simply be collected by hand. As well as the plants described bellow, the seeds of the tree Barringtonia asiatica (ch.5, §5.2) were also used for this purpose.

4.1. Derris spp., derris root, fish poison plant, TP vut (Fabaceae)⇫

The use of climbing shrubs of the genus Derris to stun fish for an easy catch is known from the Bismarcks to Fiji (see vol.1, ch.8, §7). However, of fifteen Derris species that Verdcourt (1979: 314–331) identifies in Papua New Guinea, only three are said to be used for fishing: Derris elliptica, Derris elegans (syn. D. rufula, Dillenia salomonensis and perhaps Derris heterophylla) and Derris malaccensis. We can evidently add a fourth, as Peekel mentions the use of Derris trifoliata (syn. D. uliginosa) to stun fish, a claim confirmed by Hviding (2005: 105).

In its wild form Derris elliptica is a climbing shrub up to 10m high which grows in coastal locations (Figure 13.6, right). Cultivated plants are mostly low and tangled, and produce thick fleshy roots which are scraped or crushed, then stirred into shallow water to benumb fish so that they can be easily caught.

Chewing derris root and then swallowing copious amounts of water was the traditional means of suicide in NW Island Melanesia.

Derris trifoliata is a smaller species but has a thicker stem, larger leaves and a weaker fish- stunning effect than Derris elliptica (Peekel 1984: 243). Hviding (2005: 105) reports that in Marovo Derris trifoliata is indigenous, and Derris elliptica has been introduced from New Guinea.

It seems likely that POc *tupa referred to Derris creepers in general, or at least to those used in fishing, and perhaps also to Derris elliptica in particular, as this was the source of the strongest poison. Also reconstructable are

• POc *puna(t) ‘vine used for fish poison, probably Derris elliptica’
• PWOc *m(ʷ)ali ‘Derris sp.’
• PROc *vuba ‘k.o. vine, probably Derris elliptica’

but I am unable to determine how they differed in meaning from each other or from *tupa.

PMP		*tuba	‘Derris fish poison’ (Dempwolff 1938)
POc		*tupa	‘climbing shrubs, Derris spp.’ (Capell 1943)
Adm	Seimat	tu[hi]	‘Derris sp.’ (Sorensen 1950)
NNG	Aria	tuva	‘derris root’
NNG	Kove	tuva	‘derris root’
PT	Molima	tuva	‘derris root’
PT	Kilivila	tuva	‘poisonous root used for fishing’
PT	Motu	tuha	‘derris root’
MM	Vitu	tuva	‘derris root’
MM	Tigak	tua	‘derris root’
MM	Teop	suva	‘derris root’
MM	Mono-Alu	tuha	‘Derris heterophylla’ (W. McClatchey, pers. comm.)
MM	Marovo	tuva	‘Pongamia pinnata’
SES	Gela	tuva	‘derris root’
SES	Sa’a	uha, uhe	‘derris root’
NCV	Mota	tua	‘a creeping plant used to poison fish’
NCV	Uripiv	na-tuv	‘fish poison vine’
PSV		*(i)tuv	‘Derris sp. usually trifoliata’ (Lynch 2004a)
SV	Sye	(nos)(i)tup	‘Derris sp. usually trifoliata’ (nos ‘vine’)
NCal	Nyelâyu	(ñale)jep	‘derris root’
NCal	Xârâcùù	(kʷa)dɨ	‘derris root’
Mic	Woleaian	sūpe	‘fish poison (root)’ (-p- for †-f-)
Fij	Wayan	tuva	‘generic for Derris spp.’
Fij	Rotuman	fuha	‘Derris spp., used to stun fish’

PMP *bunat is reconstructed on the basis of the Oceanic data here and Botolan Sambal bunat ‘Derris elliptica’.

PMP		*bunat	‘Derris elliptica’
POc		*puna(t)	‘vine used for fish poison, probably Derris elliptica’
Adm	Lou	pun	‘vine used for fish poison’
PT	Sudest	vun	‘poison fish with derris’
MM	Tolai	vun	‘Derris elliptica, root with which fish are poisoned; to kill or benumb fish with poison of this name’
MM	Nduke	buna	‘Derris heterophylla, a poison-leaf vine, crushed and thrown into rock-holes to stun reef fish’
MM	Roviana	buna	‘littoral vine (macerated and thrown into rock pools, it stupefies fish)’
MM	Marovo	buna	‘climber with poisonous leaves, fish poison’
SV	Lenakel	no-un	‘fish poison’

PWOc *(m,mʷ)ali (incorrectly given as *maRi in vol. 1, ch.8, §7) is reconstructed on the basis of North New Guinea and Meso-Melanesian languages:

PWOc		*(m,mʷ)ali	‘Derris sp.’
NNG	Gitua	(waro)mali(ŋ)	‘Derris root’
MM	Nalik	mal-mal	‘Derris root’
MM	East Kara	mal	‘Derris sp.’
MM	Sursurunga	mel	‘a vine used to poison fish or humans’
MM	Marovo	moli	‘a coastal creeping vine’

Bender et al. (2003) offer two Proto Micronesian reconstructions with the gloss ‘fish poison’ or ‘Derris sp.’. One is PMic *(t,T)upa, continuing POc *tupa and reflected only in Woleaian sūpe ‘fish poison (root)’. Since sūpe has -p- for †-f and PMic *(t,T)upa is reconstructed for expected †*tufa, the reconstruction is suspect. The other PMic reconstruction is *upa ‘derris vine’, which is well supported in Micronesian languages and appears to have cognates in Southern Vanuatu, permitting reconstruction of PROc *vuba.

PROc		*vuba	‘k.o. vine, probably Derris elliptica’
PSV		*na-vup	‘k.o. vine’ (Lynch 2001c: 236)
SV	Sye	na-vup	‘k.o. vine, probably Derris elliptica’
SV	Anejom̃	no-hop(ɣev)	‘k.o. vine, probably Derris elliptica’
PMic		*upa	‘Derris vine’ (Bender et al. 2003)
Mic	Kosraean	op	‘plant used as fish poison’
Mic	Marshallese	wep	‘a tree, Barringtonia asiatica, seeds used for fish poison’
Mic	Ponapean	ūp	‘Derris elliptica’
Mic	Ponapean	ūpa-wp	‘to poison fish’
Mic	Chuukese	wɨ̄p, wɨpe(n)	‘Derris elliptica’

5. Multi-purpose roots⇫

Turmeric and ginger are both cultivated for their roots, which have a variety of uses. Among these are magical applications, shared with the first three plants in §6. Like the candlenut (§3 .2) and Indian mulberry (§3.4) turmeric also produces a dye.

5.1. Curcuma longa (syn. Curcuma domestica), turmeric (Zingiberaceae)⇫

The turmeric: plant, Curcuma longa, is much smaller than Alpinia species, being only 50-120cm tall, but it is treated by at least speakers of Marovo and Kwara’ae as belonging to the same taxon as Alpinia (Kwa’ioloa & Burt 2001:193-194, Hviding 2005: 131).¹⁴ It has long green leaves and pale yellow flowers. The rhizome is lumpy and an intense yellow (Peekel 1984: 109).

Although it grows readily in the wild in locations where there is light, the Marovo and Kwara’ae sources also report that it is often cultivated. The rhizome provides spice. It is also a source of yellow dye, but the colour fades easily (Floyd 1954, Peekel 1984: 109). At Marovo the aromatic leaves are used for parcelling fish for the stone oven.

Turmeric also has ritual significance. At Marovo it is planted to keep evil spirits away from gardens, and in both Marovo and Kwara’ae the roots are chewed and spat out of the door or window of the house to fend off evil spirits (Hviding 2005: 131). The Kwara’ae also chew it both with betelnut and alone, the latter both as a snack and medicinally.

Figure 13.7 Curcuma longa

There are two reconstructions, POc *yaŋo and PEOc *re(ŋ,ŋʷ)a. There is some evidence in the glosses of reflexes that the latter meant yellow material, including prepared turmeric and perhaps the yolk of an egg.

POc		*yaŋo	‘turmeric, Curcuma longa’ (Milke 1968)
NNG	Mangap	(n)aŋgo-ŋgo(ŋana)	‘a plant, ginger type, yellow when crushed’ (-ŋana ‘nominaliser’)
NNG	Adzera	yaŋa(n)	‘ginger’
MM	East Kara	ioŋ	‘turmeric’
MM	Patpatar	iaŋ	‘turmeric’
MM	Kubokota	aŋo	‘plant similar to ginger, of various colours (white and yellow); it has various medicinal and magical uses in healing and cursing’
MM	Nduke	aŋo	‘turmeric’
MM	Roviana	aŋo	‘turmeric’
NCV	Mota	aŋo	‘turmeric’
Mic	Ponapean	ɔ̄ŋ	‘turmeric’
Mic	Mokilese	ɔŋ	‘turmeric’
Mic	Woleaian	yāŋa	‘ginger’
Fij	Bauan	ðaŋo(laya)	‘wilid ginger, Zingiber zerumbet’
Fij	Bauan	ðaŋo	‘turmeric’
Pn	Tongan	aŋo	‘turmeric’
Pn	Tongan	aŋo-aŋo	‘shampoo ginger, Zingiber zerumbet’
Pn	Samoan	aŋo	‘turmeric’

The root above was the source of the POc term for ‘yellow’, *yaŋo-yaŋo:

POc		*yaŋo-yaŋo	‘yellow’ (Grace 1969)
NNG	Kove	eaŋ-eaŋo	‘yellow’
NNG	Tami	yaŋo-yaŋo	‘yellow’
NNG	Kairiru	yaŋ-yaŋ	‘yellow’
NNG	Manam	aŋo-aŋo	‘yellow’
PT	Molima	yawo-yawo(na)	‘yellow’
MM	Nakanai	ial-alo	‘yellow’
SES	Gela	aŋo-aŋo	‘yellow’
NCV	Mota	aŋo-aŋo	‘yellow’
Mic	Woleaian	yaŋo-yaŋo	‘yellow colour of ginger’
Fij	Wayan	aŋo-aŋo	‘yellow’
Fij	Bauan	ðaŋo-ðaŋo(a)	‘reddish or orange like the turmeric plant’

PEOc		*[re]reŋʷa	‘yellow material, prepared turmeric(?)’ (Biggs 1965: *reŋa)
SES	Sa’a	reŋa	‘decorate with black, white and red designs; beautiful’
NCV	Loh	eŋ	‘turmeric’
NCV	Mota	re-reŋa	‘yolk of an egg; yellow colour’
Mic	Kiribati	reŋa	‘Curcuma tonga’15
Fij	Wayan	re-reŋʷa	‘Curcuma longa, turmeric’
Fij	Bauan	re-reŋa	‘turmeric’
Pn	Tongan	eŋa	‘turmeric’
Pn	Rennellese	ŋeŋa	‘prepared turmeric’
Pn	Samoan	leŋa	‘yellow dye from turmeric’ (Whistler 2000: 177)
Pn	Tokelauan	leŋa	‘yolk of egg, turmeric’
Pn	Tahitian	reʔa	‘ginger, turmeric’
Pn	Hawaiian	lena	‘yellow’
Pn	Māori	reŋa-reŋa	‘a large: herb with curcuma-like leaves and short rhizome, Arthropodium cirratum’ (R. Gardner, pers. comm.)

5.2. Zingiber spp., ginger, TP kawawar (Zingiberaceae)⇫

Zingiber officinale (common ginger), 50-120cm tall, cultivated and wild, serves as a spice plant, medicine and magic (Peekel 1984:100). It appears to have arrived in Near Oceania a long time ago (R.M. Bourke, pers. comm.), but does not seem to have been carried into (at least parts of) Remote Oceania until European contact, where the only species of ginger was Zingiber zerumbet (‘wild ginger’, ‘shampoo ginger’, ‘pinecone ginger’), a woody shrub, 80-120 cm high with long narrow green leaves and magnificent red flowers which resemble a pinecone in shape. It is widely cultivated, with numerous cultivars.

If it is true that Zingiber officinale did not find its way into Remote Oceania with Oceanic speakers, then one must ask whether it had in fact arrived in NW Island Melanesia by POc times. This is a matter which seems to require further research.

In Marovo wild ginger was associated with magic and sorcery. The leaves of particular varieties were important in calling on ancestor spirits, and the roots were chewed for magical purposes. In earlier times, one was planted at each comer of a garden to protect it against destructive magic. The roots of some varieties are chewed for their healing properties. Because of these associations, ginger is rarely used in ordinary cooking (Hviding 2005: 130).¹⁶ Peekel (1984: 101) in any case reports that the rhizome of Zingiber zerumbet is less tasty than those of other Zingiber species.

Two POc terms are reconstructed below. The gloss of the first, *laqia, remains a little doubtful because of the difficulty in dating the arrival of Zingiber officinale in NW Island Melanesia. POc *para(k) evidently referred to Zingiberaceae species but just how large a taxon it denoted is currently unknown.

PMP		*laqia	‘ginger, Zingiber officinale’ (Headlland & Healey 1974)
POc		*laqia	‘ginger, Zingiber officinale (?)’ (French-Wright 1983)
Adm	Mussau	laia	‘ginger’
Adm	Titan	lei	‘ginger’
Adm	Lou	lei	‘ginger’
NNG	Kove	haia	‘ginger’
NNG	Tami	lagi	‘ginger’
NNG	Adzera	rakia	‘ginger’
NNG	Mumeng	lɛhaʔ	‘ginger’
NNG	Kairiru	lakea	‘Curcuma spp.’ (Borrell 1989: 42)
PT	Wedau	naia	‘ginger’
PT	Iduna	naiya	‘ginger’
PT	Motu	aɣi	‘ginger’
MM	Bulu	laɣia	‘ginger’
MM	Bola	lahia	‘ginger’
MM	Nakanai	lahia	‘ginger’
MM	Tangga	lae	‘ginger’
MM	Nehan	laia	‘ginger’
NCV	Tape	ləɣ-ləɣ	‘wild ginger’
SV	Kwamera	nə-re	‘ginger’
Fij	Bauan	(ðaŋo)laya	‘ginger, Zingiber zerumbet’

PSES *ria ‘ginger’ (Gela ria ‘ginger’, W Guadalcanal, Arosi ria ‘turmeric’, Kwaio lia ‘turmeric’) is evidently derived from POc *laqia by borrowing rather than by direct inheritance, as the expected PSES form is †*laɣia.

PMP *badak is tentatively reconstructed below on the basis of the Tolai and Wayan Fijian items and of Tagalog barak ‘Curcuma zedoaria, Zingiber zerumbet’, Kuyunon barak ‘Curcuma domestica, Globba marantina’ (Madulid 2001a).

Figure 13.8 Codiaeum variegatum

PMP		*badak	‘Zingiberaceae spp. with edible rhizomes’
POc		*para(k)	‘Zingiberaceae spp. with edible rhizomes’
MM	Tolai	va-var	‘Curcuma longa’
MM	Tolai	(ka)va-var	‘Zingiber officinale’
Fij	Wayan	va-va	‘Alpinia boia’

6. Plants with large colourful leaves⇫

Here four plants are discussed. The first three, Codiaeum variegatum, Cordyline fruticosa and Dracaena angustifolia, are described separately but the reconstructions are presented together in §6.4 because there is evidence of a POc taxon, *jiRi, which included ’Cordyline fruticosa and Dracaena angustifolia and of a PWOc taxon, _*mʷa(r,R)e, which included Codiaeum variegatum and Cordyline fruticosa. All three species are cultivated for their brightly coloured glossy leaves and are used decoratively, ceremonially and ritually.

6.1. Codiaeum variegatum, croton, TP purpur, Bkala lif (Euphorbiaceae)⇫

Codiaeum variegatum usually takes the form of a non-woody shrub a metre or more in height, consisting of a clump of large leathery leaves, at their largest 30cm by 9 em. In the wild the leaves are green and the plant may grow into a small tree up to 5 m tall. Under cultivation the leaves vary greatly in shape and colour, and traditionally, Codiaeum variegatum was cultivated for its brilliant colours, although it grows in the wild in open locations with access to sunlight (Peekel 1984: 313,Wheatley 1992:89, 91). Kwa’ioloa & Burt (2001: 175) names five cultivars, four of them distinguished by colour (green, yellow, red and dark red).

Codiaeum variegatum shrubs were cultivated around the edges of villages for their decorativeness and as boundary markers and markers of taboo sites like graves. They also provided a source ofleaves to decorate men’s bodies during dances (Floyd 1954, Kwa’ioloa & Burt 2001: 175, Wheatley 1992: 91, Hviding 2005: 117, Gardner & Pawley 2006).¹⁷

On New Britain the bark or the leaves were rubbed on the skin to cure skin diseases (Powell 1976). On Waya Island the leaves are used in the treatment of high blood pressure (Gardner & Pawley 2006).

6.2. Cordyline fruticosa (syn. Cordyline terminalis, Taetsia fruticosa), cordyline, victory leaf, ti plant, TP tangget, Bnanggaria (Agavaceae)⇫

Cordyline fruticosa has a woody stem usually 1-2m high with no branches; the leaves issue from the top of the stem. There is a great variety of leaf shapes from long and narrow to broadly oval and a great variety of leaf colours: green with white: or red stripes, white with green or red stripes, rose-red, rose-red and dark-red stripes, and dark-redl to blackish purple (Peekel 1984: 81-83).

The cordyline has a tuber which is edible and occasionally eaten in Papua New Guinea, although it may have been used more extensively for food in the past. Its young leaf shoots are sometimes cooked and eaten (Barrau 1965, Bourke 1982: 60, French 1986: 335). The Saliba eat the flowers, cooked with leaves of Gnetum gnemon (ch.10, §2.3) (Margetts 2005b). The Molima simply add them to the cooking pot (A. Chowning, pers. comm.). Cordyline fruticosa is used for plot markers, and tuberous root as an item of clothing, and as a warning to thieves (Bourke 1982).

Figure 13.9 Cordyline fruticosa: plant

Leaves of C.fruticosa are worn by dancers in New Britain (Arentz et al. 1989: 94). Bourke comments on the spiritual significance of C.fruticosa in some Papua New Guinea societies. The Marovo believe that cordyline holds spiritual power, and different cultivars of cordyline are used for different kinds of magic, often exercised by holding the leaf itself (Hviding 2005: 118). It also functions as a charm against magic among the Kwaio and the Kwara’ae, and the latter is the one place where a non-decorative, non-ritual use is mentioned: it is used to wrap lizards for cooking (Kwa’ioloa & Burt 2001: 180). On Waya Island it is usually planted in villages and at taboo sites, and today often at graves. Different varieties were used for different kinds of magic (Gardner & Pawley 2006).

6.3. Dracaena angustifolia (Liliaceae)⇫

In its mature form Dracaena angustifolia is a woody shrub growing to 3-6 m high. It has no bole and splits into many branches at ground level. Each branch splits again recursively, and at the end of each branch, issuing directly from it, is a plume of long narrow leaves. In cultivated varieties these leaves are often brightly striped. Dracaena angustifolia receives far fewer mentions in the literature than croton or cordyline: Hviding, for example, does not mention it. This may be because non-botanist observers have confused the species or because Dracaena angustifolia has less magical and taboo significance. Kwa’ioloa & Burt (2001: 181) say that it is used for living fences. However, the Arosi dictionary describes it as a very sacred plant which is planted on burial grounds and on gardens to protect crops, is used in divination, and is waved at the annual harvest ceremony (Fox 1978). Sa’a speakers use the leaves in incantations, in bonito ceremonials and in malevolent magic (Ivens 1929). On Ulawa the priest uses a Dracaena branch bound with a climbing fern to draw out the yam beetles when a garden is dedicated for planting. Such a branch is then planted at the entrance stile to keep out disease (Ivens 1927:360, 362).

6.4. Reconstructions⇫

Cordyline and Dracaena varieties are both used as decorative plants and for making leaf skirts, and many Oceanic languages treat them as a single taxon. In E Kara (MM), for example, si denotes Cordyline fruticosa, and the binomial si tavul is Dracaena angustifolia. In Kwara’ae Cordyline fruticosa is dili and Dracaena angustifolia is mala-dili ‘resembling cordyline’ (see ch.2, §7 .1.4 ). Each of the cognate sets below spans plants of both genera, and it is probable that many more of the glosses should refer to both Cordyline and Dracaena: the full span of the denotation was missed when the data were recorded.

Chowning (1963, 2001: 81) suggests that the ‘Prato-Melanesian’ terms *babaka,¹⁸ *dili (my POc *jiRi ‘Cordyline sp., Dracaena sp.’) and *male (my PWOc *mʷa(r,R)e ‘Codiaeum variegatum; Cordyline fruticosa’) were each used collectively for three species. Her first two species are Codiaeum variegatum and Cordyline fruticosa. Her third, however, is not Dracaena angustifolia, but Cycas rumphii. She writes that all three are ‘used primarily for decorative and magico-religious purposes’ (Chowning 1963).¹⁹

It is quite possible that PWOc *mʷa(r,R)e indeed denoted a taxon including Codiaeum variegatum and Cordyline fruticosa, and perhaps other decorative plants (note Nakanai mamale ‘Cananga odorata’), but I have found no evidence of a term that denoted plants of all three species. Possibly Chowning considered some of the items collected together under PEOc *mʷa(q)ele ‘cycad’ (ch.9, §5.2) to belong to the same cognate set as the items under PWOc *mʷa(r,R)e below. Despite the formal similarity between the two reconstructions, however, the two cannot be reconciled: reflexes of the former agree in reflecting *-l-, the latter in reflecting *-(r,R)-. The two sets can be united only by positing irregular sound change or borrowing.²⁰

Three other etyma are reconstructed below: POc *jiRi, *kaRi(q)a and *jajal. POc *jiRi indeed introduces a third species into the discussion, but it is Dracaena angustifolia, not Cycas rumphii. The distribution of the reflexes of POc *kaRi(q)a ‘taxon of decorative plants’ leaves us unable to determine which of the three species it denoted. POc *jajal probably denoted a particular variety of Codiaeum variegatum.

PWOc		*mʷa(r,R)e	‘taxon including Codiaeum variegatum and Cordyline fruticosa’ (Chowning 1963: *male; Ross 1996d)
NNG	Kove	mohe	‘Cordyline sp.’ (Chowning 1996: 17)
NNG	Yabem	(ka)maʔ	‘Cordyline sp.’
NNG	Takia	mra-mor	‘Cordyline fruticosa’
NNG	Kairiru	(moel) morie(p)	‘thin-leaved, green variety of Codiaeum variegatum’
PT	Roro	mare	‘plant with yellow leaves’
MM	Vitu	mare	‘Cordyline sp.’
MM	Bulu	mara	‘Cordyline sp.’
MM	Nakanai	male-male	‘Cordyline sp.’ (Chowning 1996: 17)
MM	Nakanai	ma-male	‘Cananga odorata (like Cordyline used for personal adornment)’ (Floyd 1954)
MM	East Kara	ma-mara	‘Codiaeum variegatum’
MM	Patpatar	mora-mora	‘Codiaeum variegatum’
MM	Kandas	muro	‘Codiaeum variegatum’
MM	Roviana	mar-mar	‘Codiaeum variegatum’

cf. also:

SV	Anejom̃	na-mraθ	‘Codiaeum variegatum’
Fij	Wayan	ŋʷali	‘croton, Codiaeum variegatum’

PMP		*siRi	‘Cordyline sp., Dracaena sp.’ (Blust 1989)
POc		*jiRi	‘taxon consisting of Cordyline fruticosa and Dracaena angustifolia’ (Milke 1968: *diRi; Ross 1988)
Adm	Mussau	sii	‘banyan, Ficus sp.’
Adm	Kurti	siy	‘Dracaena sp.’
Adm	Wuvulu	ti	‘Dracaena sp.’
Adm	Aua	ti	‘Dracaena sp.’
Adm	Ere	siy	‘Dracaena sp.’
NNG	Malasanga	sir	‘grass skirt’
NNG	Mindiri	(da)dir	‘grass skirt’
NNG	Mapos Buang	sĩ	‘tree sp., Euphorbeaceae’
NNG	Wampur	riciʔ	‘cordyline’
NNG	Adzera	ji-ji	‘cordyline’
NNG	Kairiru	jir	‘small pandanus sp.’
PT	Wedau	diri	‘Dracaena sp.’
PT	Tawala	diri	‘Dracaena sp.’
PT	Molima	dili	‘red cordyline’
MM	Tiang	si	‘cordyline’
MM	East Kara	si	‘Cordyline fruticosa’
MM	East Kara	si(-tavul)	‘Dracaena angustifolia’
MM	Nalik	sir	‘cordyline’
MM	Tabar	cir	‘cordyline’
MM	Lihir	cir	‘Cordyline terminalis/Cordyline fruticosa’
MM	Notsi	cil	‘cordyline’
MM	Konomala	si	‘cordyline’
MM	Patpatar	suri(ah)	‘Dracaena angustifolia’
MM	Tolai	ir	‘leaf of the croton’ (ACD)
MM	Ramoaaina	ir-ira	‘croton leaf, Dracaena’
MM	Petats	rin	‘cordyline’
MM	Taiof	rir	‘cordyline’
MM	Varisi	zili(para)	‘Cordyline fruticosa’ (W. McClatchey, pers. comm.)
MM	Marovo	ji(polo)	‘Cordyline fruticosa’
MM	Roviana	zi(polo)	‘shrub, Dracaena sp.’
SES	Gela	dili	‘Dracaena’
SES	Lau	dili	‘Dracaena with thin green leaves’
SES	Kwara’ae	dili	‘Cordyline fruticosa’
SES	Kwaio	dili	‘Cordyline fruticosa’
SES	’Are’are	siri	‘herb with red stem and leaves, used for magical purposes’
SES	Sa’a	dili	‘Draecena’
SES	Arosi	diri	‘Draecena’
NCal	Nyelâyu	di	‘Cordyline fruticosa’
Mic	Chuukese	tĩ-n	‘Cordyline fruticosa’
PCP		*jĩ	‘taxon consisting of Cordyline fruticosa and Dracaena angustifolia’
Fij	Rotuman	jĩ	‘plant with ornamental leaves and a sweet root: Cordyline fruticosa’
Pn	Tongan	sĩ	‘Cordyline fruticosa’
Pn	East Futunan	tĩ	‘Cordyline fruticosa’
Pn	Rennellese	tĩ	‘Cordyline fruticosa’
Pn	West Uvea	tĩ	‘Draecena’
Pn	Emae	tĩ	‘Cordyline sp.’
Pn	Samoan	tĩ	‘Cordyline fruticosa’
Pn	Tikopia	tĩ	‘Cordyline sp.’
Pn	Tuamotuan	tĩ	‘Cordyline fruticosa’
Pn	Hawaiian	kĩ	‘Cordyline fruticosa’
Pn	Marquesan	tĩ	‘Cordyline sp.’
Pn	Tahitian	tĩ	‘Cordyline fruticosa’
Pn	Māori	tĩ	‘Cordyline sp.’

Medial parenthesised *-q- is reconstructed in POc *kaRi(q)a on the assumption that Vitu kariɣa ‘rattan’ is cognate, but the difference in denotation calls this into doubt.

POc		*kaRi(q)a	‘taxon of decorative plants’ (Geraghty 1990: PEOc *gaRi(a))
NNG	Tami	kali-kali	‘Codiaeum variegatum’ (Dempwolff 1902)
SES	Kwara’ae	ka-kali	‘Hornstedtia lycostoma’ (Henderson and Hancock 1988)
NCV	Mota	karia	‘Dracaena sp.’
NCV	Ambae	karie	‘Cordyline sp.’
NCV	Raga	ɣaria	‘generic for Cordyline spp.’
NCV	Atchin	kari	‘Dracaena sp.’
NCV	Uripiv	gari	‘Cordyline sp.’
NCV	Naman	na-ɣari	‘Cordyline sp.’
NCV	Neve’ei	na-ʔari	‘Cordyline sp.’
NCV	Larëvat	na-ɣri	‘Cordyline sp.’
NCV	Port Sandwich	xari	‘crotons’
NCV	Lewo	(puru)kalie	‘Cordyline sp.’
NCV	Namakir	kari	‘Cordyline sp.’
NCV	Nguna	na-karie	‘palm-lily plant, Dracaena sp.’
SV	Sye	(tana)ŋklai	‘Cordyline sp.’
Fij	Bauan	gai	‘Cordyline sp.’

cf. also:

MM

Vitu

kariɣa

‘rattan’

POc		*jajal	‘croton, Codiaeum variegatum’
NNG	Yabem	(ka)dada	‘a grassland shrub, the sap of which is used to blacken teeth’ (probably Rhus taitensis-MDR)
MM	Nehan	dedel(am)	‘Codiaeum variegatum, yellow and green variety’
MM	Varisi	zazala (piru)	‘Codiaeum sp.’ (W. McClatchey, pers. comm.)
MM	Babatana	jajala	‘Codiaeum variegatum’ (W. McClatchey, pers. comm.)
MM	Nduke	zazala	‘Codiaeum variegatum’
MM	Roviana	zazala	‘Codiaeum variegatum’
MM	Marovo	jajala	‘Codiaeum variegatum’
NCV	Mwotlap	na-sas	‘Codiaeum variegatum’
NCV	Mota	sasa	‘Codiaeum variegatum’
NCV	Ambae	sasa	‘Codiaeum variegatum’
NCV	Lonwolwol	ha	‘Codiaeum variegatum’
NCV	Raga	hahali	‘Codiaeum variegatum’
NCV	Apma	sasli	‘Codiaeum variegatum’
NCV	Uripiv	na-jej	‘Codiaeum variegatum’
NCV	Rerep	ne-jaj	‘Codiaeum variegatum’
NCV	Neve’ei	ne-nsah	‘Codiaeum variegatum’
NCV	Tape	cec	‘Codiaeum variegatum’
NCV	Avava	a-sah	‘Codiaeum variegatum’
NCV	Lonwolwol	ha	‘Codiaeum variegatum’
Fij	Bauan	ðaða	‘Codiaeum variegatum’ (Geraghty 2004: 79)
Fij	Rotuman	sasa	‘Codiaeum variegatum’ (Geraghty 2004: 79)

6.5. Heliconia indica, heliconia, Blif laplap (Heliconiaceae)⇫

Two species of Heliconia grow in NW Island Melanesia, Heliconia indica in eastern mainland New Guinea and the Bismarcks and Heliconia solomonensis in Bougainville and the Solomons (R. Gardner, pers. comm.). The two appear to be very· similar (Figure 13.10, left).²¹ Numerous short stems formed from leaf sheaths rise from an underground rhizome to a height of 5 or 6 m. The leaves resemble banana leaves, and under cultivation assume various colours and patterns: yellow, yellow-and-green striped, bright red or dark red. In the wild heliconias grow in damp shady environments in primary and secondary forest.

The leaves are used for covering and sealing the stone oven because of their thick, waxy cuticle and very large leaf surfaces. Thanks to their size, fewer are needed, reducing the labour of cutting and carrying (Henderson & Hancock 1988:239, Kwa’ioloa & Burt 2001: 196, Hviding 2005: 120).

POc		*paqo	‘Heliconia sp’
MM	Nduke	vaɣo	‘a large: perennial herb with banana-like leaves, Heliconia solomonensis’
MM	Marovo	vaɣo	‘tree sp. with large banana-like leaves, Heliconia sp.’
SES	Gela	vao-vao	‘shrub with large leaves; wild banana’
SES	Santa Ana	fao	‘Heliconia solomonensis’
NCV	Mwotlap	(no-yo)va	‘Heliconia indica’ (yo- ’leaf)
NCV	Mota	vao	‘a heliconium’
Fij	Wayan	vā-vā	‘Heliconia sp.’
Fij	Bauan	va-vao(a)	‘a plant, Bleekeria elliptica’

PEOc		*rako	‘Heliconia sp., usually Heliconia indica’ (Lynch 2004a: PSOc *rau)
SES	Kwara’ae	rako	‘Heliconia indica’
NCV	South Efate	n-rau	‘Heliconia sp.’
SV	Sye	n-rau	‘Heliconia sp.’
SV	Ura	lau	‘Heliconia sp. (generic)’

6.6. Crinum asiaticum, spider lily (Amatyllidaceae)⇫

Crinum asiaticum is a large lily with long erect leaves that are arranged in a spiral rosette to form impressive clumps up to 1.5 m in height and 2 m wide (Figure 13 .10, right). The leaves may be a metre long and 10cm wide. They emerge from huge bulbs that may weigh as much as 9 kg. Its white flowers, with thick succulent stems, are shaped like tubes that flair open into a crown of narrow petals.²² Hviding (2005: 103) distinguishes between Crinum asiaticum, which is cultivated, and Crinum pedunculatum, which grows wild in sandy areas near beaches and in other damp locations. Kwa’ioloa & Burt (2001: 216) distinguish between a green-leaved wild variety of Crinum asiaticum and a yellow-leafed variety cultivated for decoration in villages.

Figure 13.10: Left: Heliconia solomonensis: A, plant; B, mature inflorescence, open and showing fruit. Right: Crinum asiaticum, spider lily.

At Tinputz (Bougainville) the pounded roots of this plant were mixed with red iron oxide and the mixture was rubbed onto coconut palm trunks to ensure that they bore well (Blackwood 1935: 311). The Nakanai of New Britain and the Marovo use the light-coloured leaf bases as trolling lures to catch large fish, in Marovo barracuda and Spanish mackerel (Hviding 2005: 103, Floyd 1954). The Nakanai also use the leaves for making women’s leaf skirts. In Marovo the leaves are used to treat bruises and fractures.

The forms listed below are clearly cognate, but it is not clear whether or how many of the parenthesised segments of the Kairiru and Kilivila forms reflect part of the protoform: hence the question mark against *mʷalak.

POc		*mʷalak	‘spider lily, Crinum asiaticum’
NNG	Kairiru	mlak(ap)	‘Crinum asiaticum’
PT	Kilivila	mola(bau)	‘a lily, Crinum asiaticum’
PSOc		*mʷalak	‘spider lily, Crinum asiaticum’ (Lynch 2004a)
NCV	Mwotlap	mʷ[a]lak	‘spider lily, Crinum asiaticum’
SV	Anejom̃	no-(hos)meleɣ	‘Crinum sp.’

7. Cucurbits⇫

The four plants discussed below, namely the bottle gourd (Lagenaria siceraria), the wax gourd (Benincasa hispida), the pumpkin (Cucurbita moschata) and the cucumber (Cucumis sativus) have in common not only that they all grow on vines but, more relevant here, that the dates and directions of their introductions into Oceania have been objects of controversy. The linguistic evidence supports the controversial claim that the pumpkin, Cucurbita moschata, was absent from the world of POc speakers but arrived soon after the break-up of POc.

7.1. Lagenaria siceraria (syn. Lagenaria vulgaris, Lagenaria leucantha, Cucurbita siceraria, Cucurbita lagenaria), bottle gourd, TP kambang (Cucurbitaceae)⇫

The bottle gourd, Lagenaria siceraria, is grown throughout much of the Pacific, mainly for its value as a container and occasionally for its food value (French 1986: 107). The plant originated in Africa, but had spread across much of the world in pre-Columbian times. Barrau (1962: 189) regards it as an ancient plant in Melanesia, but no POc form can be reconstructed, as noted by Ross (1996d).

Clark (1996) reconstructs PNCV *tavaya (Raga tavai, Mota wo-tavae), which, together with Bauan tavaya ‘bottle’, implies PEOc *tapaya. But no cognates of the latter nor alternative cognate sets have been found in Western Oceanic or Admiralties languages, implying that the bottle gourd was not known to POc speakers. Surveying recent findings, Green (2000) suggests quite strongly that this was the case. The gourd may well have reached Oceania from two directions, arriving in Melanesia from the Indo-Malaysian region and much later in eastern Polynesia from South America. There is good evidence that the Polynesians did not carry the gourd with them into eastern Polynesia (Whistler 1990, 1991). It is thus possible - and on the linguistic evidence likely - that the bottle gourd first reached the Bismarck Archipelago after the break-up of POc.

7.2. Benincasa hispida, wax gourd, white gourd, winter melon (Cucurbitaceae)⇫

Golson (2002) also provides a survey of archaeological evidence which shows that allegedly pre-Oceanic dates for the bottle gourd in Oceania are the result of misidentification of remains of the wax gourd, Benincasa hispida. Whistler (1990) similarly shows that reports of the bottle gourd in Polynesia reflect incorreclt identifications of the wax gourd. The latter is a native of SE Asia and archaeological evidence indicates that it was present at least on mainland New Guinea when the Austronesian speaking ancestors of POc speakers arrived there. There are terms for it in non-Oceanic languages: terms from Philippine languages listed by Madulid (2001b: 42) together with Malay kundur point to a possible PMP *kundur, but this may be an outcome of borrowings across island SE Asia (Wolff 1994). Verheijen (1990: 195) cites terms that point to a possible PCMP *kelas. As Golson points out, the botanical literature indicates that Benincasa hispida is at least occasionally present at locations on the mainland and in the Bismarcks where Oceanic languages are spoken today. In support of this he cites Peekel (1984: 547-548) and Borrell (1989: 66). French (1986: 108) also indicates that Benincasa hispida is grown in Papua New Guinea, but does not specify locations.

Whistler provides names for Benincasa hispida, listed below, which allow us to reconstruct Proto Central Pacific *vaŋ(o,u). I have accepted his arguments for glossing the Fijian terms as Benincasa hispida: his sources (H. B. R. Parham 1943, Capell 1941) gloss them as species of Lagenaria. Despite the obvious presence of Benincasa hispida further west I have been unable to reconstruct an earlier term for it, but not for the usual reasons. Usually, a term is not reconstructable because there are no cognates in the data or because cognates are insufficiently widespread. In the case of Benincasa hispida, however, there are no data from outside Fiji and Polynesia except Peekel’s Patpatar hulhul paraho. This can only indicate that, even if Benincasa hispida was present in early Oceanic times, its presence and significance in the agricultural suite of Oceanic-speaking societies has faded almost to zero.²³

PCP		*vaŋ(o,u)	‘wax gourd, Benincasa hispida’
Fij	Lau (Eastern Fijian)	vaŋo	‘Benincasa hispida’
Fij	Bauan	vaŋo	‘Benincasa hispida’
Pn	Tongan	faŋu	‘Benincasa hispida’
Pn	Niuean	faŋu	‘bottle’
Pn	Samoan	faŋu	‘Benincasa hispida’
Pn	East Uvean	faŋu	‘Benincasa hispida’
Pn	East Futunan	faŋu	‘Benincasa hispida’

7.3. Cucurbita moschata, pumpkin, TP pamken (Cucurbitaceae)⇫

Barrau (1962: 190) declares that the pumpkin²⁴ is a European introduction to Melanesia, and the linguistic data largely support his assertion. A number of terms for it are borrowings from English (MM: E Kara baniyin, Tinputz banken, NCV: Paamese vamuken, Lewo pamken) or from elsewhere (PT: Motu mausini, Roro mauteni, apparently from Samoan mau- tini ‘gourd’).

The set of seeming cognates below provides evidence, on the other hand, of an early Oceanic etymon, *waluq or *[w]aruq, thatmusthave denoted Cucurbita moschata or a similar plant. I write ‘seeming’ because, however one reconstructs the term, there are irregular phonological developments due to borrowing. Dempwolff (1938) reconstructed *baluq ‘bottle gourd, Lagenaria siceraria’. However, the Oceanic forms below, together with Sundanese, Old Javanese, Balinese waluh, Buru walu ‘bottle gourd’, point to PMP *waluq ‘bottle gourd’. Whether PMP actually had such a form or whether this is a series of borrowings postdating the break-up of PMP and perhaps even of POc (cf §7.1) is not relevant here. What is relevant is that a form related to these entered early Oceanic, but was applied to the pumpkin, Cucurbita moschata, or a related species.

If we assume that the Oceanic etymon was *waluq, then the consonants are regularly reflected in Kela, Dami and Takia. Sio reflects *waruq, and Kela may also do so, as *-l- and *-r- are neutralised as Kela -r-. Hote, Mapos Buang and Patep reflect *ruq-aruq and Lenakel and Anejom reflect *ruq-ruq.²⁵ Gapapaiwa and Misima bonu-bonu (and similar forms in other PT languages) reflect *bʷalu-bʷalu. If we assume instead that the Oceanic etymon was *waruq or *aruq, we are still left with irregularities.

The strong likelihood, then, is that the term was borrowed from community to community sometime very soon after the break-up of POc. That is, Cucurbita moschata or a similar plant must have found its way into NW Melanesia not much less that 3000 years ago. The genus Cucurbita is indigenous to the tropical zone of the Americas, where domestication is known to have occurred by 4000 years ago (Sauer 1993).

This finding is not at odds with the historical fact that the modern pumpkin was introduced to NW Melanesia by Europeans after 1870. It simply indicates that a variety of Cucurbita moschata or a similar plant was already present, and had been present for a long time, albeit perhaps as a very minor crop in a few areas.

Early Oceanic		*waluq, *[w]aruq	‘pumpkin, Cucurbita moschata ?’
NNG	Sio	waru	‘pumpkin’
NNG	Kela	waru-waru	‘pumpkin’
NNG	Hote	lu-alu	‘pumpkin’
NNG	Mapos Buang	ruχ-arūχ	‘pumpkin’ (-r- for †-l-)
NNG	Patep	luʔ-əlu	‘pumpkin’
NNG	Dami	olu	‘pumpkin’
NNG	Takia	walu	‘pumpkin’
PT	Gapapaiwa	bonu-bonu	‘pumpkin’
PT	Misima	bonu-bonu	‘pumpkin’
PSV		*na-r(o,u)r(o,u)(q)	‘pumpkin, gourd’
SV	Lenakel	(noua)ne-lulu	‘pumpkin’
SV	Anejom̃	ne-rero	‘gourd’

7.4. Cucumis sativus, cucumber (Cucurbitaceae)⇫

Because of the varieties that have been introduced since 1870, it is sometimes assumed that the cucumber is a crop introduced by Europeans. However, Bourke & Allen (forthcoming) write that it was probably introduced via SE Asia thousands of years ago. The cognate set supporting the reconstruction of POc *[ka]tim(o,u)n ‘cucumber, Cucumis sativus’ below con- firms that the cucumber was part of the world of POc speakers, even though it is not indigenous to the Pacific. The genus Cucumis has its origins in Africa, and the cucumber is one of the few African species known to have been domesticated outside Africa. Until recently it was thought that it was part of a suite of plants that were domesticated in the Fertile Crescent of SW Asia or in India, but recent work in molecular biology has cast doubt on this, placing the cucumber instead in an Asian-Australian clade of the genus Cucumis (Renner et al. 2007). There is thus no biogeographic challenge to the proposal that the cucumber was known to both PMP and POc speakers.

Blust (ACD) produces strong evidence to suggest that the denotatum of POc *[ka]tim(o,u)n was the cucumber, but the reflexes below suggest that it may also have been used of other species of Cucumis.

The initial syllable of *[ka]tim(o,u)n is reflected only in Western Oceanic languages , whereas Central Pacific reflexes point to *timun. This distribution justifies the reconstruction of POc forms with and without initial *ka-, which may represent or have been reanalysed as the prefix *ka- ‘tree’ reflecting earlier *kayu ‘tree’ (ch.2, §7.1.2).

Shown in parentheses below are irregular phonological developments, implying that al- though the term was inherited into POc, it was sometimes borrowed from one language into another.

PMP		*[ka]timun	‘cucurbit (generic); cucumber, Cucumis sativus’ (Dempwolff 1938: *timun)
POc		*[ka]tim(o,u)n	‘Cucumis spp. (generic?); cucumber, Cucumis sativus’ (ACD)
NNG	Sengseng	tamun	‘Cucumis sativus’ (A. Chowning, pers. comm.)
NNG	Sengseng	katim	‘small cultivated cucurbit with dark green skin and sweet orange flesh’ (A. Chowning, pers. comm.; probably borrowed)
NNG	Gitua	karimon	‘cucumber’ (ACD)
NNG	Mapos Buang	qatimŋ	‘cucumber, Cucumis sativus’
NNG	Patep	kətima	‘cucumber, Cucumis sativus’ (-a- for †-(o,u)-)
NNG	Numbami	katimana	‘cucumber’ (-a- for †-(o,u)-)
PT	Motu	asemo	‘small wild cucumber, Cucumis sp.’ (ACD)
MM	Sursurunga	katmur	‘vegetable like pumpkin or cucumber’ (-r for †-n)
MM	Tolai	katimur	‘Pacific melon, Cucumis melo’ (-r for †-n)
MM	Teop	asimuru	‘a vegetable like a cucumber’ (-r for †-n)
MM	Tinputz	asimū	‘cucumber’
Fij	Bauan	timo	‘a plant, Cucumis acidus’
Pn	East Futunan	timo	‘melon’

Notes⇫