Chapter 1.8 Fishing and Hunting Implements

Meredith Osmond

An earlier version of this chapter is included in Lynch and Pat, eds, (1996). I am grateful to Andrew Pawley and Malcolm Ross who have contributed substantially to the preparation of this chapter, to Atholl Anderson for his comments and to Jean Kennedy for drawing my attention to P.H. (Holly) McEldowney’s (1995) thesis (1995), which contains a wealth of information on fishing techniques in the Admiralties.

1. Introduction⇫

This chapter takes a close look at terms for fishing and hunting implements and processes attributable to Proto Oceanic (POc), with a view to reconstructing a small portion of the culture of POc speakers.[^1] There is ample linguistic evidence that these people were fishermen and seafarers. The Oceanic Lexicon Project has collected cognate sets supporting POc reconstructions for over eighty fish names and an additional forty terms for shellfish (see vol. 3). This paper adds a further twenty terms for various fishing techniques used, both for deep-sea fishing requiring line and hook capture and for coastal, reef or river fishing where netting, spearing, poisoning or fish-trapping could suffice. We also know that a number of animals, birds and reptiles were hunted for food, and terms for many of these have been reconstructed. Here we attempt to establish terms for the hunting implements and techniques used.

2. Nets⇫

Most Oceanic languages have a wide range of terms for different kinds of nets. These include terms for hand nets, casting nets, seine nets, long-handled nets for catching flying fish, and so on. Many languages have terms for nets for catching birds and occasionally larger animals, such as pigs and wallabies. Presumably there were POc terms for a wider range of nets than the four we have reconstructed.

PMP		*puket	[N] ‘dragnet’; [V] ‘surround, engulf’ (Blust 1972b)
POc		*pukot	‘fishing net, seine’
Adm	Andra	hu	‘large rectangular net, used in pairs’ (McEldowney 1995)
NNG	Kove	puo	‘seine net’
PT	Gumawana	uwosi	‘net for fishing and for trapping turtle’ (final -i suggests a borrowing from Suauic)
PT	Motu	huo	‘kangaroo net’
MM	Nakanai	vuo	‘seine net’
MM	Bali	vuɣoto	‘fishing net’
SES	Arosi	huʔo	‘large net, seine net’
SES	Sa’a	huʔo	‘seine net’
Mic	Kiribati	ikot-	‘bring together, gather, collect’
Mic	Mokilese	uk	‘round net’
Mic	Woleaian	uxo	‘fish net’

Reflexes of *pukot meaning specifically ‘seine’ occur in both WOc (Kove, Nakanai) and EOc (Arosi, Sa’a) witnesses. This suggests the strong possibility that *pukot denoted ‘seine’ in POc. The more generalised meanings of other witnesses do not rule out this implication.

POc		*reke	‘fishing net’
PT	Balawaia	leke	‘fishing net’
PT	Motu	reke	‘fine fishing net, seine’
PT	Roro	reʔe	‘fishing net’
MM	Vitu	neke	‘fishing net, fish trap’ (n- for expected **r-)
PCP		*dreke	‘recess, cavity, pocket of a seine net’ (Biggs 1965)
Fij	Wayan	dreke	‘hold of a boat’
Fij	Bauan	dreke	‘hollow or cavity in a thing’
Fij	Rotuman	reke	‘pocket of a seine net’
Pn	Tongan	leke	‘small room or recess’
Pn	Rarotongan	reke	‘end of a net’

*reke perhaps also referred to a seine net, but its Eastern Oceanic reflexes imply that it may have referred to an end or pocket of the net.

PMP		*lawa(n,q)	‘k.o. fishnet’ (ACD)
POc		*lawa((n,q))	‘k.o. fishnet’
Adm	Loniu	law	‘k.o. long narrow fishnet’
Adm	Andra	lau	‘long medium-mesh barrier net, held upright by floats and weights and secured by men at intervals’ (McEldowney 1995)
SES	Arosi	rawa	‘small net’
SES	Sa’a	lawa	‘the name of a creeper from which twine for nets is made’
Fij	Bauan	lawa	‘fishing net’
Fij	Wayan	lawa	‘fishing net’

POc		*ku(p,pʷ)ena	‘fishing net’ (Blust 1981a)
Adm	Mussau	uena	‘fishing net’
Adm	Loniu	kupʷen	‘long rectangular fish net’
MM	Tolai	ubene	‘fish net; any net or net-like thing; spider’s web’
SES	Arosi	ʔubena	‘large net’
NCV	Nguna	kupʷena	‘fishing net’
SV	Lenakel	na-kapun	‘fishing net’
Pn	Tuvalu	kupeŋa	‘large heavy net for communal fishing’1
Pn	Tongan	kupeŋa	‘fishing net: generic’2
Pn	Rennellese	kupeŋa	‘net: generic’ (usually refers to fishing net)3

The fourth term, *ku(p,pʷ)ena, is the most widespread term of the four, and arguably the best bet for a POc generic term for fishing nets. Our files list more than forty cognates from all major subgroups. In Tuvalu (Koch n.d.: 30) the kupeŋa is “a primitive kind of net … large and heavy and knotted from rolled coconut fibre twine … affixed to two poles tied together at one end forming a n acute angle; large cowrie shells are fixed under the edge of the third side”. Two men hold the poles, and the net is tilted so the fish can swim into it. In parts of Polynesia, the term is used as a generic. On Niuatoputapu (Tonga), kupeŋa is the term used for all netting techniques, as opposed to tau (angling) or uku (diving) (Dye 1983: 252-254). Dye lists kupeŋa fakamamaha ‘netting with the ebbing tide’ ; kupeŋa ʔava ‘netting for ʔava (milkfish)’; kupeŋa hokohoka, in which a handled net is used in rough surf; kupeŋa sili pulou, where the common throw net, about three metres across, is used to catch bait fish; and so on. In Rennell, the term evidently subsumes all kinds of nets. Although Elbert (1975) defines kupeŋa as ‘fine-meshed fishing net’, he includes kupeŋa tape peka ‘flying fox snaring net’, as well as six kinds of fishing nets labelled with compound terms beginning with kupeŋa. The same general term kupeŋa ‘net, traditionally of hibiscus fibre’ is found in Tikopia, with particular types named as compounds (kupeŋa tā save ‘pole net for flying fish’, kupeŋa fukifuki ‘pole net for reef work’) and mata kupeŋa referring to net mesh.

The techniques that involve large nets usually require use of floats and sinkers. We have a reconstruction for net float:

POc		*utoŋ	‘float of fishing net’
NNG	Gitua	uton	‘float of fishing net’
PT	Bwaidoga	utoɣa	‘float of fishing net’
MM	Nakanai	uto	‘handle or stick of fish net’
SES	Lau	uo	‘float of fishing net’
NCV	Mota	uto	[V] ‘come above the surface in water’
Mic	Ponapean	ūs	‘float of fishing net’
Mic	Satawalese	wūs, wuso-	‘any kind of wood that floats well’
Fij	Bauan	uto-uto	‘float of fishing net’
Fij	Rotuman	ut	‘float of fishing net’
Pn	Tongan	uto	‘float of fishing net’
Pn	Rarotongan	uto	‘float of fishing net’

There is also a competing form that co-exists with *utoŋ in the Central Pacific, PCP *vuta (Rotuman hufa ‘float on a fishing net’, Tikopia futa ‘net float’).

The only terms I can locate for sinker are reflexes of POc *patu ‘stone’, or the name of the cowrie shell which is sometimes used as a sinker:

PMP		*buliq	‘cowrie shell’ (Blust 1980b; ACD)
POc		*buli(q)	‘cowrie shell; cowrie shell used as net sinker’
SES	Sa’a	puli	‘cowrie shell; cowrie shell used as net sinker’
Pn	Tuvalu	pule	‘cowrie shell; cowrie shell used as net sinker’

Other reconstructions that can be included within the vocabulary of nets are:

POc		*sika	‘netting needle’
PT	Kilivila	(va)sia	‘needle’
Mic	Kiribati	rika	‘netting needle’
Fij	Rotuman	siʔa
Fij	Bauan	sika (ni lawa)	‘netting needle’
Pn	Tongan	hika	‘mesh needle, used in making nets’
Pn	East Futunan	sika	‘shuttle used for net-making’
Pn	Hawaiian	hiʔa	‘shuttle or needle for making nets’

POc		*mata-	‘mesh of net’
NNG	Kove	mata	‘net gauge’
PT	Molima	mata-	‘net gauge’
SES	Arosi	mā	‘hole; opening; mesh of net’
SES	Kwaio	mā	‘k.o. fish weir; space through which birds frequently fly, where traps are set up’
Pn	Māori	mata	‘mesh of net’
Pn	Hawaiian	maka	‘mesh of net’

POc *mata- and its reflexes in many contemporary languages extend to a range of concepts that carry the idea of an opening, a doorway, something allowing access, and thus the mesh of a net. Evidence is that it is used in connection with nets in both Western Oceanic, as ‘net gauge’, and Eastern Oceanic, as ‘mesh of net’. Presumably the POc term embraced the idea of ‘mesh of net’ in both places, but in WOc languages reflexes now refer to the instrument used to maintain uniform mesh when net-making.

Another term for net gauge is found in Eastern Oceania:

PCP		*qava	‘net gauge’
Fij	Bauan	yava (ni lawa)	‘net gauge’
Pn	Samoan	afa	‘mesh stick’
Pn	Rennellese	ʔaha	‘net gauge’

3. Scareline⇫

Some communities use a scareline for fishing, in which leaves are tied to a long rope which is drawn through the water. A term for this has been reconstructed to PCP level, although it is possible that cognates are derived from POc *raun ‘leaf’:

PCP		*rau	‘dragline, scareline, made from rope and coconut leaves’
Fij	Wayan	rau (sole)	‘barrier of leaves used in a sole rau or ara-rau fish drive’
Pn	Tongan	au	‘long fishing net made of rope and coconut leaves’
Pn	Tikopia	rau	‘sweep with a net’

However, in Andra (Admiralties) a leaf sweep is called you, pointing to POc *Rau(C) rather than *raun.

4. Angling implements⇫

Use of fishhooks was evidently not an automatic ‘given’ in western Oceanic communities. Ann Chowning (pers.comm.) has commented that

at the time of European contact, a number of societies did not use fish hooks, even though the archaeological evidence makes it clear that they can be attributed to POc culture. The Nakanai were a case in point. In some other societies, such as Sengseng, line and bait fishing were done only by using a gorge, pointed at both ends, or sometimes, as I have seen the Molima do, bait was simply tied to a line and the fish pulled out when it swallowed the bait. In many places fishing with hook and line is a post-contact phenomenon.

Nonetheless, the linguistic evidence leaves us in no doubt that POc speakers were familiar with the technique.

PMP		*hapen	‘fishing line’ (ACD)
POc		*apon	‘fishing line’ (ACD)
Adm	Drehet	cap	‘fishing line’
MM	Petats	ahon	‘string’
SES	Gela	(rau ni) avo(lo)	‘leaf kite for fishing’
Mic	Kiribati	ao	‘fibre of coconut husk, fishline, twine’
Mic	Kosraean	æ	‘string, fishing line, rope, thread, cord’
Fij	Bauan	ðavo	[v] ‘fish with rod (and line)’
Pn	Tongan	afo	‘cord, fishing line’
Pn	Samoan	afo	‘fishing line’

PMP		*kawil	‘hook’ (Blust 1972a) ; ‘fish hook’ (Dahl 1973)
POc		*kawil	[N] ‘hook; fishhook’
Adm	Baluan	kow	‘hook; fishhook’
NNG	Kairiru	qawil	‘hook; fishhook’
MM	Tangga	auil	‘hook; fishhook’
MM	Roviana	gaili	‘fishhook made from pearlshell and turtle shell, used in trolling’
SES	Arosi	ʔawi
NCV	Mota	gau	‘fishhook’
SV	Anejom̃	in-ɣowoc	‘hook; fishhook’
Fij	Wayan	kau	‘fishhook of any kind’
Pn	Tuvalu	kau	‘fishhook, generic’

POc *kawil is widely attested. I have over fifty reflexes containing specific reference to fishhooks, occurring in all major subgroups. In Tuvalu it is used as a generic, followed by the name of the fish for which it is appropriate—kau galata, kau palu, etc. It has an obvious formal and semantic relationship with *kawi(t), *kawit-i- (V) ‘hook, catch hold of; fruit crook’ (Ch. 5, §10), but the origin of this relationship is not understood.

POc		*ta(g,k)o	‘(barbless ?) fishhook’
MM	Harua	toga	‘fishhook’
MM	Maringe	tʰaɣo	‘fishhook’
SES	Bugotu	taɣo	‘fishhook’
SES	Lau	aʔo	‘fishing rod’
Pn	Rarotongan	toko	‘an old type of wooden fishing hook (no barb)’

The extension of a meaning from a material to something made from that material is very common across languages (viz. English glass/a glass, cork/a cork). Examples occur in Oceanic languages as well (Bauan gasau ‘a reed; an arrow’, Tolai vat ‘a stone, a sinker’). POc speakers evidently made fishhooks from a variety of shells, and in some daughter languages the term for a particular shell has become the term for a fishhook (made from that shell?). For instance, POc *kima ‘clamshell’ is reflected in some Central Papuan languages as kimai ‘fishhook’ (Ross 1994a: 404). A similar example is:

PMP		*qunap	‘scales’ (Dempwolff 1938)
POc		*qunap	‘turtle shell, fishhook’
NNG	Rauto	a-una	‘fishhook’
MM	Tinputz	unā	‘hook; fishhook’
SES	Longgu	ūnai	‘fishhook’
SES	Ghari	unua-na	‘turtle scales (not turtle shell)’
SES	Arosi	una-na	‘tortoiseshell’ (una ‘fish scales’)
Pn	Rennellese	ʔuna	‘outer shell (of turtle +)’
Pn	Samoan	una	‘scale; scale of hawk’s bill turtle, i.e. tortoise-shell’
Pn	Tuvalu	una	‘turtle shell’
Pn	Tikopia	una	‘carapace of marine turtle’
Pn	Hawaiian	una	‘shell of turtle’

Final *-p is reconstructed because this item is almost certainly identical to POc *qunap ‘fish scales’. The Arosi (SES) verb derived from this term, unahi- ‘remove shell from turtle, scale a fish’ (see Ch. 6, §5.4), reflects *-p- and refers to the removal of both turtle shell and fish scales.

,

We have evidence, both archaeological and linguistic, that trolling lures were also widely used. This technique of dragging a line through the water, with a lure of mother-of-pearl or similar bright shell, is used to catch bonito and other pelagic fish from large paddling or sailing canoes.

POc		*bayan	‘fish bait, trolling lure, trolling hook’
MM	Ramoaaina	bain	‘bait’
MM	Teop	beana	‘bait’
MM	Mono-Alu	beana	‘bait’
SES	’Are’are	pasa	‘a barbless bonito fishhook’
SES	’Are’are	pā	‘fish bait’
SES	Sa’a	pasa	‘bonito lure of clamshell’
SES	Sa’a	pā	‘bait’
SES	Arosi	bā	‘bait for fish, food to entice into a trap’
NCV	Mota	pea	‘bait; to entice by a bait’
SV	Lenakel	nə-pien	‘bait’
SV	Kwamera	nə-piien	‘bait’
SV	Anejom̃	ne-pyañ	‘bait’
Fij	Wayan	bā	‘trolling line with lure’
Fij	Wayan	baya	‘earthworm (earthworms are used for bait)’
Fij	Bauan	bā	‘Tongan variety of fishhook’ (vakasavu bā ‘troll’)
Fij	Bauan	baða	‘worm (hence bait for fishing)’
Pn	Tongan	pā	‘fish hook, especially for catching bonito’
Pn	Samoan	pā	‘pearlshell lure; spinner including lure and hook’
Pn	Tuvalu	pā	‘bonito lure, generally made from mother-of-pearl shell’
Pn	Tokelauan	pā	‘generic term for trolling hook (pa si malau, pa si aseu etc.). All are made with pearlshell shanks and turtle-shell points’

It would appear that in some Southeast Solomonic languages and also in Fiji, reflexes of POc *bayan have split into doublets, evidently to distinguish ‘bait’ from ‘trolling lure’. Because there is crossover of meaning, with the long-vowel form referring to bait in one region and trolling lure in the other, it is assumed that the two splits occurred independently. Nor can the possibility of borrowing be discounted.

John Lynch (pers.comm.) points out that the final -ñ of Anejom ne-pyañ reflects earlier *-ni, implying POc **bayani. However, a number of the forms above reflect a loss of final *-n which would not have occurred if the form had been **bayani. It is possible, however, that Anejom ne-pyañ reflects a conflation of *bayan and *bani (below), forms which are similar but, it appears, not cognate.

PCEMP		*paniŋ	‘bait; fodder’ (ACD)
POc		*bani	‘bait’
NNG	Gitua	bani	‘bait’
PT	Tawala	bani	‘bait’
PT	Molima	bani	[v] ‘fish with hook and line’4
MM	Ramoaaina	ban	‘bait; use for bait’
Mic	Ponapean	pān	‘bait, lure’

A form PMP/POc *baŋi ‘bait’, is reconstructable on the basis of Javanese baŋi ‘bait’ and the obsolete Samoan paŋi ‘coconut bait for flying fish’. Its relationship to the reconstructions above is unclear.

5. Fish trap⇫

PAn		*bubu	‘conical bamboo basket trap for fish’ (ACD)
POc		*pupu	‘basketry fish trap’
Adm	Lou	pup	‘bamboo basket trap for fish’
MM	Nakanai	vuvu	‘k.o. fish trap’
SES	Arosi	huhu	‘large wickerwork eel trap’
Mic	Kiribati	ū	‘trap for moray eel’
Mic	Puluwatese	wū	‘fish trap’
Fij	Bauan	vuvu	‘long narrow fish trap made of bamboo’

The term is widely attested. It is noteworthy that we have no Polynesian cognates. Instead we find PPn *fīnaki (Tongan finaki, Rarotongan ʔinaki ‘fish trap’; Hawaiian hīnaʔi ‘k.o. basket fish trap’).

6. Fish weir⇫

POc		*baRa	‘fence, wall, enclosure’
Adm	Mussau	bala-bala	‘fence’
Fij	Wayan	bā (ni ika)	‘fish weir’
Pn	Tongan	pā	‘fence, wall, enclosure, especially for fish trap, made of stone or sticks’
Pn	Rarotongan	pā	‘fish weir of stone walls’

The POc term *baRa ‘fence, wall, enclosure’ (Ch. 3, §3.6) has evidently acquired an additional specialised meaning in the Central Pacific, where it refers to a fish trap made of stone or sometimes of sticks. (The Lou (Admiralties) term pas ‘stone fish corral’ has unexplained -s, and at this stage cannot be accepted as a cognate.) In Oceania, walls of stone or coral are constructed across channels in the reef to catch fish on a falling tide. In the Cook Islands (Rarotonga), pā has become a generic. Buck (1927:298) writes that:

Most of the pa are very old, having been laid down far back in pre-European times. The channels were studied and the course taken by fish observed. The lines of the walls were laid down with such skill and accuracy that any departure from them ends in failure. The walls are made of loose coral rock. The most important weirs are named, and are owned by particular families. No outsider can use a weir without permission from the hereditary owners.

He describes various types of weirs, for example, pa kiokio (roughly Z-shaped), pa tute (temporary), and pa tuakirua (V-shaped with opening towards the sea).

7. Fish poison⇫

PMP		*tuba	‘Derris fish poison’ (Dempwolff 1938)
POc		*tupa	‘Derris fish poison’
NNG	Aria	tuva	‘Derris fish poison’
NNG	Kove	tuva	‘Derris fish poison’
PT	Molima	tuva	‘Derris fish poison’
PT	Kilivila	tuva	‘poisonous root used for fishing’
PT	Motu	tuha	‘Derris fish poison’
MM	Vitu	tuva	‘Derris fish poison’
MM	Tigak	tua	‘Derris fish poison’
MM	Teop	suva	‘Derris fish poison’
SES	Gela	tuva	‘Derris fish poison’
SES	Sa’a	uha, uhe	‘Derris fish poison’
NCV	Uripiv	na-tuv	‘fish poison vine’
Fij	Wayan	tuva	‘Derris fish poison’

POc		*puna	‘vine used for fish poison’
Adm	Lou	pun	‘vine used for fish poison’
MM	Tolai	vun	‘root with which fish are poisoned; to kill or benumb fish with poison of this name’
MM	Roviana	buna	‘littoral vine (macerated and thrown into rock pools, it stupefies fish)’
SV	Lenakel	no-un	‘fish poison’

From North New Guinea and New Ireland we have been able to reconstruct a lower-level term:

PWOc		*(m,mʷ)ali	‘Derris root’
NNG	Gitua	(waro)mali(ŋ)	‘Derris root’
MM	Nalik	mal-mal	‘Derris root’

FIXME: 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.⇫

The method of stunning fish by throwing pounded Derris root into pools is widespread, and the three terms seem to have identical reference, although Derris elliptica is not mentioned specifically in relation to *puna. Describing the technique used by Sa’a speakers, a Southeast Solomons language, Ivens (1927 (reissued 1972):389) writes that:

… fish in streams are poisoned by a preparation. The bark of the edible Barringtonia tree, the one with red flowers, is stripped off and heated in the fire to bring out its bitter qualities. It is then beaten into shreds with stones in water and thrown into the place chosen. Along with it they use the grated nuts of the barringtonia speciosa [sic.], a littoral tree. A third ingredient is made from pounded sections of a creeper called uhe, the juices of which are very bitter.

The last ingredient is obviously a variant of uha, a reflex of POc *tupa.

The kernel of the fruit of Barringtonia asiatica (POc *putun) is used for poisoning water in Samoa, while the Fijians use the outer portion of the same fruit (Blackwood 1935:354-355), but I have not been able to reconstruct a term for this.

8. Torch fishing⇫

PMP		*damaR	‘resin, torch, light’ (Dempwolff 1938)
POc		*(d)rama(R)	[N] ‘torch’; [V] ‘fish at night with torch’
Adm	Lou	(ka)ram	‘torch’
Adm	Lou	ram-ram	‘fish at night by torchlight’
Fij	Bauan	rama-rama	‘lamp of coconut shell filled with oil’
Pn	Tongan	ama	[V] ‘fish at night by torch-light; torch made of coconut spathes bound together’
Pn	Samoan	lama	‘torch (made of dry coconut leaflets +); fishing with torches’
Pn	Māori	rama	‘torch; eeling with torches’

Night-fishing is still a popular and widespread activity throughout the region. To attract the fish, a torch is used which consists of dried coconut fronds or spathes bound together to burn slowly over a long period. In many languages the same term refers both to the torch and to the activity. In Tonga, lama can simply be prefixed to the names of other fishing techniques to indicate that it is carried out at night (e.g. lama fakasiosio ‘go spear-fishing at night’). (The ambiguity of the initial consonant *(d)r- is touched on in Chapter 2, §3.1.3.)

9. Pointed weapons⇫

Most Oceanic languages contain a number of terms for spear, and it is probable that POc was no exception. There are various ways in which spears can be physically distinguished. There may be a distinction between fishing, fowling and fighting spears. Often, but not always, fishing spears have three or more prongs. Shorter spears can be used for thrusting, longer spears for hurling. Spears can be multi-barbed, have a single barb, or be barbless. They can be made with a head that separates from the shaft. Every language community will name its spears according to combinations of these and possibly other properties. Although we have been able to reconstruct perhaps six POc terms, it has been difficult to distinguish between them, other than to single out *tara and *kuj(u,i)r as probably fish spears. At the generic level-and that is itself a flexible category which can include arrows, and possibly other pointed weapons such as darts and spikes—*qio(r,R) seems to have the widest distribution, although it does not occur east of Vanuatu. For the Central Pacific, *sao(t) seems the most likely generic term. In POc times, some spears would have had obsidian heads, and the term used for these was evidently *koto ‘obsidian head of spear’ (Ch. 4, §4.1.3).

PMP		*saet	[N, v] ‘spear’ (ACD)
POc		*sao(t)	[N] ‘spear’
SV	Sye	sau	‘spear’
Fij	Wayan	sā	‘spear; harpoon’
Fij	Rotuman	jao	‘spear’
Pn	Tongan	tao	‘spear; javelin’
Pn	Rennellese	tao	‘spear or dart for fishing, fighting, fowling’
Pn	Hawaiian	kao	‘dart; fish spear; javelin; spike as on the tail of a stingray’
Pn	Rarotongan	tao	‘short throwing spear’

POc		*qio(r,R)	‘spear, arrow’ (Ross 1994a: 464)
NNG	Lukep	yu	‘arrow; spear traditionally used for pig hunting and fighting’
NNG	Kaiwa	ī	‘arrow’
NNG	Manam	io	‘long lance, whose wooden spear point has prongs on four sides’
NNG	Uvol	io	‘arrow’
PT	Anuki	io	‘spear’
PT	Bwaidoga	ɣio	‘spear’
PT	Kalokalo	gio	‘spear, arrow’
PT	Tawala	iyola	‘fish spear’
PT	Balawaia	gio	‘spear’
PT	Motu	io	‘spear’
MM	Konomala	iu	‘shoot; arrow’
MM	Tangga	iu	‘fish spear’
SES	Lau	īo	‘a war arrow; a poisoned spear’
NCV	Nguna	na-io	‘spear’

As a subset of the above, certain NNG languages have inserted a medial consonant derived from epenthetic glides: Kove iðo ‘spear’; Bariai ido ‘arrow’; Gitua izo ‘spear’; Wogeo iwo ‘spear’. This appears to be a local feature, but not attributable to a common protolanguage.

PMP		*baŋkaw	‘barbless spear’ (ACD)
POc		*bako	‘spear’
MM	Hoava	ba-bao	‘spear’
SES	Malango	bao	‘spear’
SES	West Guadalcanal	bao	‘spear’

(In all these languages, *k is regularly lost.)

POc		*kuj(u,i)r	‘fish spear’ (based on Ross 1994a)
PT	Kalokalo	kudila	‘fish spear’
PT	Iduna	hudila	‘fish spear’
PT	Motu	udi	‘fish spear’
MM	Nakanai	kusi	‘multi-pronged fish spear’
MM	Notsi	kucil(a)	‘arrow’
MM	Siar	kusur	‘fish spear’
MM	Meramera	kusul(u)	‘fish spear’
MM	Kandas	(kabo)kusur	‘fish spear’
MM	Laghu	kuroho	‘spear’
SV	Kwamera	kahar-kahar	‘type of arrowhead made of three prongs, used for spearing fish’
Pn	Tuvalu	koho	‘thrusting spear’ (Nanumaga dialect; u > o unexpected)

There is sufficient consistency among the glosses of this set to suggest that the POc form indeed referred to a multi-pronged fish spear.

POc		*tara	‘fish spear’
Adm	Mussau	tala(ŋai)	‘fish spear’
PT	Molima	tala(beya)	‘spear with double or triple point’
PT	Molima	tala(dabadaba)	‘k.o. spear’
MM	Tolai	ta-tara	‘fish spear’
Pn	Niuean	tala	‘sharp-pointed object; barb; prong’
Pn	Rapanui	tara	‘hom; thorn; spur’
Pn	Mangarevan	tara	‘spine; hom; thorn’
Pn	Anutan	tara	‘fishhook barb’

It seems from Western Oceanic and Mussau glosses that the original POc meaning of *tara is probably ‘fish spear’. Pn glosses indicate a later narrowing of meaning to ‘barb’ or ‘prong’.

POc		*sua	[N, V] ‘spear (weapon retained in the hand)’
NNG	Manam	sua(pu)	[N] ‘fish spear’; [V] ‘spear’
NNG	Mengen	sue	‘bone, needle’
SES	Gela	sua	‘spear without barbs’
SES	Longgu	sua	[N] ‘spear’
SES	Longgu	sua-	[VT] ‘pull out (spear, stick +)’
SES	Lau	sua	‘spear’
SES	Kwaio	sua	‘spear’
Fij	Bauan	sua(k)	‘stab; pierce with a spear (retaining hold of the weapon); husk a coconut’
Pn	Samoan	sua	‘thrust’

Other terms used to describe the action of spearing are derived from POc *soka, soka -i- (V) ‘pierce; stab, poke hole in (s.t.)’ (Ch. 9, §4.1).

The following is a lower-level reconstruction for spear:

PEOc		*saRi	‘k.o. spear’
NCV	Tamambo	sari	‘spear’
Fij	Wayan	sai-sai	‘arrow or spear with three or more prongs, used for fish, bats +’

The bow, although evidently not as widespread as the spear, is used both for fighting and, in a smaller version, for shooting birds and other small game. It seems that *pusuR was the POc term for the bow and arrow, *(p,pʷ)anaq more probably for the act of shooting.

PAn		*busuR	‘hunting bow’ (ACD; Dempwolff 1938)
POc		*pusuR	‘bow and arrow’
NNG	Kis	us	‘arrow’
PT	Kuni	budu	‘arrow’
MM	Petats	husul	‘bow’
NCV	Southeast Ambrym	his	‘bow and arrow’
Fij	Wayan	vuðu	‘bow’

PMP		*panaq	‘shoot’ (Dempwolff 1938)
POc		*(p,pʷ)anaq	‘bow’
POc		*(p,pʷ)anaq, *(p,pʷ)anaq-i-	‘shoot’5
NNG	Tuam	paneg	‘bow’
NNG	Mapos Buang	vaneh	‘shoot’
NNG	Medebur	pan	‘bow’
MM	Tolai	panak	‘bow for shooting’
SES	Gela	vanahi-	‘shoot at’
SES	Kwaio	fana	‘hunt’
NCV	Mota	vene	‘shoot with a pointed arrow’
NCV	Tamambo	(i)vine	‘arrow’
NCV	Kiai	vinai-	‘shoot (with arrow)’
Mic	Kiribati	pana	‘shoot at fish with band of rubber and long arrow’
Fij	Bauan	vana	‘shoot with arrow or gun’
Pn	Tongan	(kau)fana	‘bow’

With regard to the reconstruction of *(p,pʷ), see Chapter 2, §2.1.

POc		*tibʷa(ŋ)	‘dart, arrow (not a fighting weapon)’ (Ross 1994a: 464)
POc		*tibʷaŋ-i-	‘shoot with dart, arrow’
PT	Dobu	tupu	‘bow, arrow’
PT	Tawala	diba	‘small pretend spear’
PT	Taboro	diba	‘spear, fish spear’
PT	Motu	diba	‘arrow’
MM	Notsi	tipi	‘shoot’
MM	Roviana	tupi	‘arrow or dart’ (vowel metathesis)
NCV	Mota	tigʷa	‘blunt arrow for birds; shoot (not in fighting)’
NCV	Mota	tigʷaŋ	‘shoot and hit’
NCV	Tamambo	tibua	‘shoot an arrow’
NCV	Raga	tibʷa	‘shoot’
NCV	Atchin	cip	‘blunt arrow’
NCV	Paamese	a-tuvo	‘arrow’
NCV	Paamese	tuvo	‘shoot’
NCV	Nguna	na-tipʷa	‘arrow, (fish) spine, (pine) needle’
NCV	Nguna	tipe-	‘shoot with bow’
Fij	Bauan	tiga	‘reed dart, used in game of veitiqa’
Fij	Wayan	tigʷe	‘throw reed or dart horizontally controlled by end of forefinger, with aim of making dart skip up when it hits ground’
Pn	Tikopia	tika	‘k.o. arrow thrown in game’
Pn	Māori	teka	‘dart used in a traditional game’
Pn	Hawaiian	keʔa	‘dart, bow; shoot with bow and arrow’

The glosses of the reflexes of POc *tibʷa(ŋ) suggest strongly that it referred prototypically to small, arrow-like objects or darts that did not have a killing function. The initial-consonant voicing of the PT forms (other than Dobu tupu, which may not be cognate) is unexplained.

PAn		*deles	‘bowstring’ (ACD)
POc		*lolo(s)	‘bowstring’
SES	’Are’are	i-roro	‘bowstring’
SES	Sa’a	i-lolo	‘bowstring’
Fij	Wayan	lolo	‘bend, flex; bent, flexed’
Fij	Bauan	(ka)lolo	‘bent to one side by excessive burden on shoulder’
Pn	Tikopia	(ka)roro	‘bowstring’
Pn	Tahitian	(a)roro	‘stay to mast of canoe’

The Fijian and the Tahitian cognates have strayed semantically, yet still retain the idea of something held so tautly that it distorts.

10. Club⇫

Across the Pacific, the kind of weapon that is labelled ‘club’ varies from the great knobbed, obsidian-studded roots that can be seen in the Bishop Museum in Honolulu, to the flat round stone clubs of the Motu (gahi), to the smaller, lighter throwing clubs of Fiji (i ula: Clunie 1977), and to the Kiribatese kati-popuki—about four feet long, pointed at both ends, used for warding off a spear, making a thrust, or wielding as a club (Hudson 1841, quoted in Koch 1986:249), which might equally be classified as a shield or lance. I consider these to be more or less marginal examples of what is basically a blunt, heavy instrument, wielded in the hand for bludgeoning.

In the Southeast Solomons, Fiji and Polynesia at least (I have little information from Western Oceania), clubs have an importance beyond that of simply weapons. They often carry ceremonial weight. Many are heirlooms with names and magical powers. Special reverence is accorded the war club. To the extent that it is possible to distinguish war clubs from hunting clubs, I do not intend to delve further into the culture and terminology of the former. It is my guess that hunting clubs would have carried less cultural significance, and as a result, been labelled more broadly. Two terms have been reconstructed.⁶ In each case I have included all known cognates. It will be apparent that neither reconstruction is particularly soundly based. Nor do their cognates provide much information as to the shape, material or specific function of each club:

POc		*paru	[N, v] ‘club’
PT	Molima	pulu(mai)	‘war club, wooden and flat-sided, used to kill by hitting throat or back of neck’
MM	Teop	varu	‘club’
NCV	Paamese	vau(lev)	‘club’
SV	Lenakel	nɛ-pɛrāu	‘club wielded by the leader of various dances in a nɛkoviār (dance cycle)’
Fij	Bauan	ravu(ta)	‘club s.o. or s.t.’ (metathesis)

and an even more questionable reconstruction, in view of irregular final vowels:

POc		*gapi	‘(stone?) club’
PT	Motu	gahi	‘flat stone club’
SES	Fagani	ɣafe	‘club’
SES	Lau	afui	‘k.o. club’

11. Slingshot⇫

POc		*maga	‘stone; slingshot’
MM	Vitu	(ta)maga	‘slingshot’ (maga-maga ‘sand’)
MM	Bulu	(ta)maga	‘slingshot’
MM	Nakanai	(tu)maga	‘sling made of bark-cloth, used for hunting birds and bats’
SES	Talise	maka(ra)	‘stone’
Pn	Tongan	maka(tā)	‘sling’ (maka ‘stone, rock’, tā ‘hit’)
Pn	Tikopia	maka	‘stone for sling; sling for hurling stones’
Pn	Rarotongan	maka	[v] ‘throw; hurl; sling’; [N] ‘sling, stone’
Pn	Tahitian	maʔa	[N, V] ‘sling’

POc		*kaio	‘sling; to turn round and round’
MM	Bilur	aio	‘slingshot’
SES	Arosi	ʔaro(rabu)	[N] ‘sling’ (ʔaro ‘turn round and round’, rabu (V) ‘strike or knock s.t.’)

Slingshots would have been used for hunting birds, flying foxes and perhaps the cuscus.⁷ Although it is assumed that ordinary well-shaped stones were used, Green (1979: 39) has documented a pointed-end Tridacna shell slingstone from Main Reef Islands circa 1000 BC.

12. Trail and pitfall spikes⇫

PMP		*suja	‘bamboo trail or pitfall spike’ (Blust 1976b)
POc		*suja	‘sharpened stake set in ground to stop or wound animals or enemies’
MM	Sursurunga	sus	‘sharp-pointed stick, used in traps or for breaking coconuts open; make holes in’
SES	Arosi	suda	‘stake set slanting and sharpened in a pit for the enemy’

The setting of sharp spikes or slivers of wood or bamboo, hardened in the fire, either directly into a path, or at the bottom of a pit, is also known in Fiji (Bauan soki, ‘spike’; lovosa ‘pitfall man trap’), but I cannot locate any other Oceanic reflexes of PMP *suja. Nonetheless, given the external evidence, the reconstruction must stand. It may be that this was primarily a warfare technique, and that a different term was used for the pit-trapping without spikes of pigs and smaller ground animals like lizards.

13. Snare trap⇫

Birds are caught for food throughout the region, often by netting or by ground snares, fashioned from a spring-laden twig and a noose. Although I can locate many terms which refer to snare, noose and bird trap, the cognate sets collected are few. The following term was both a noun and a verb. Terms reflecting **piti in MM and SES languages seem to be local inventions formed from a noun **pit (after loss of POc *-a) with the addition of the transitive suffix *-i-.

POc		*pʷita, *pʷita-i-	[N] ‘snare’; [V] ‘tie by encircling, ensnare’
NNG	Lukep	wit	‘tie by encircling’
NNG	Lukep	wit(kala)	‘tie together’
NNG	Lukep	pit	‘snare, trap; to trap’
NNG	Mangap	mbit	‘tie up with rope, fasten’
NNG	Mangap	(na)pit-pit	‘snare, trap for rats, pigs, bandicoot’ (na- indicates a borrowing from Kilenge)
NNG	Mengen	(sam)pite	‘tie securely, tighten’
NNG	Mengen	(bago)pita	‘fasten (to help it to float)’
NNG	Takia	pite(lak)	‘tie on (as grass-skirt)’
MM	Sursurunga	puti	‘tie together’
MM	Ramoaaina	pita	‘hang up in the house; tie to the end of anything’
MM	Ramoaaina	pit	‘tie’
MM	Ramoaaina	pit-pit	‘line, snare’
MM	Ramoaaina	piti	[v] ‘snare’
SES	Gela	piti-	‘tie’
SES	West Guadalcanal	piti	‘tie’
SES	Longgu	piti-	‘trap an animal’s leg; tie s.t. around ankle or wrist’
Pn	Tokelauan	fi-fita	‘(garments) be too tight’
Pn	Māori	ɸita	‘firm, secure, fast’

With regard to the reconstruction of *(p,pʷ), see Chapter 2, §2.1.

PEOc		*taRi	‘noose, snare’
NCV	Mota	tar	[V] ‘lay (net +)’
Fij	Bauan	dai	‘snare or trap for catching animals’
Fij	Wayan	tai	‘snare or trap for catching animals’

14. Birdlime⇫

Another widespread method of catching birds involves the use of birdlime. Hooley describes the technique used by Mapos Buang speakers in the Huon Gulf region of New Guinea:

The sap of this tree (dagΣem) is used as birdlime for catching birds. The sap is collected in a length of bamboo and is then heated over a fire. When it is boiling a stick is dipped in and twisted and the sap adhering to it is chewed to make it soft and then wrapped in cordyline leaves. The sap is then spread on a suitable tree branch so that when birds come to eat the fruit of that tree they are caught. (from word list held on computer file at ANU)

Blust (1983–84a) has reconstructed Proto WMP *pikat or *piket ‘birdlime’ and *mamikat or *makiket ‘snare birds with birdlime’, but I have not been able to locate any Oceanic reflexes. The reconstruction given below evidently refers generally to gum or resin, although in one instance, in Arosi, it is the first element in a compound, buruhasi, meaning ‘bird trap’. The second element, hasi, means ‘adhere’.

PMP		*bulit, *pulit	‘caulk, fill up a hole or crack with viscous material’ (ACD)
POc		*bulit	‘gum; resin’
MM	Tolai	bulit	‘gum of any tree, especially that used for putty or gum’
MM	Nakanai	bulubulu(aga)	‘plant with sticky fruit used for catching birds’
SES	Arosi	buru(hasi)	‘bird trap; gum of banyan or other tree put at intervals along a pole’
SV	Lenakel	(noua-neha)pʷit	‘sap, glue’
Fij	Bauan	bulu-bulut(i)	‘(soil) sticky, doggy’
Fij	Rotuman	pulu	‘gum; sap’
Pn	East Uvean	pulu	‘resin’

15. Archaeological evidence⇫

Evidence comes both directly from the existence of artefacts themselves in archaeological sites—the fishhooks, trolling lures, stone fish traps, spear tips and so on—and indirectly through analysis of remains in Lapita middens, which provide evidence of Lapita people’s diet and (indirectly) for food procurement techniques.

Spriggs (1997a: 114) summarises non-vegetable food traces from an Eloaue site on Mussau, east of the Admiralties, believed to have been occupied between about 3,500 and 2,500 years ago. (Eloaue is a low, flat, coral island south-east of the high island of Mussau.)

Large quantities of shell midden occurred in the site … The three Oceanic domesticates, pig, dog and chicken, were present but only in small numbers. 95 per cent of the bone was of fish, mainly inshore species, and turtle. Present too in small quantities were porpoise, reptiles, birds, rats and also the phalanger Spilocuscus maculatus (p. 117).

A similar range of fauna to Mussau’s occurs in the Lapita sites in the Arawe Islands, off the south-west of New Britain, with the addition of a wallaby, Thylogale browni and the flightless cassowary as introduced species (p. 120). A site on Watom, off the north-east tip of New Britain, adds bandicoot to the faunal repertoire (p. 124). Lapita sites from the Solomons provide little evidence, while the Reef-Santa Cruz Islands, next in line as one moves east, show a decline in the range and diversity of animal life available for food, a decline that increases as one moves progressively further out into the Pacific. Spriggs records that “hunting took place of what local species of birds, bats and sea creatures such as turtles and dugongs there were, as well as collecting of reef and lagoon fish and shellfish”. (p. 135)

Artefacts are considered alongside the linguistic evidence. With regard to angling, we have the following POc reconstructions:

*apon	‘fishing line’
*kawil	‘fishhook’
*ta(g,k)o	‘barbless (?) fishhook’
*qunap	‘turtle shell; fishhook’
*bayan	‘trolling lure’
*bani, *baŋi	‘bait; fodder’

Of these items, fishhooks have proved the most useful in developing local culture sequences. One-piece fishhooks made from several kinds of shell are now well documented from a number of Lapita sites, ranging from Talepakemalai (Mussau) to Lolokoka (Niuatoputapu, Tonga), one site at the latter dating back to about 2800 BP. “Most of the Mussau hooks are fairly large (about 5 cm in shank length) and were probably designed for hand-lining from canoes, in order to catch larger benthic species such as groupers. A few hooks are smaller, however, and could have been used for taking smaller species on reefs and along coastlines.” (Kirch 1997: 200). Most hooks so far found are of Turbo, Trochus and pearlshell. Linguistic evidence supports clamshell and tortoiseshell. Presumably all of these were used, as well as the less durable coconut shell, bone and Pemphis acidula wood, which are in use today. Also from Mussau sites come examples of carefully crafted lures made from Trochus shell. These trolling lures “are streamlined for hydrodynamic lift in the water, and have finely carved grooves for attaching both the line and hackles (probably feathers or pig bristles) near the recurved point”. (pp.200-201)

More elaborate two-piece fishhooks with separate shank and point were apparently a later development, appearing in Eastern Polynesia and possibly originating in the Marquesas Islands (Bellwood 1987:58).

A pertinent observation on the connection between present-day fishing techniques and those of the early Lapita settlers is made by Kirch and Yen (1982) in their study of the prehistory of Tikopia (initial settlement dates to around 2875-2750 BP). They have assembled a collection of one-piece fishhooks and three distinctive early trolling hooks made from trochus shell, which are associated with Lapita sites. They note (p.243) that:

While we must be cautious in interpreting archaeological fishhooks on the basis of ethnographic observations of the use of modern metal hooks, the data are suggestive that Tikopia utilised a similarly broad range of angling strategies in prehistoric times. Group I hooks would have been best suited to catching small fry on the fringing reefs. Group II hooks, the most varied, were likely used for more than one strategy, and towline and bottom-fishing techniques, among others, seem likely. The largest specimens, Group III, would seem to have been intended for towline capture of carniverous pelagic fish or bottom fishing for large groupers, Ruvettus, and the like. Such an interpretation of the angling gear is consistent with the range of fish taxa identified in the Tikopia faunal material.

We have eight POc terms that relate to netting:

*pukot	‘fishing net, seine’
*reke	‘fishing net, seine net’
*lawa((n,q))	‘k.o. fishnet’
*ku(p,pʷ)ena	‘net, generic term (?)’
*utoŋ	‘net float’
*buli(q)	‘cowrie shell used as net sinker’
*sika	‘netting needle’
*mata-	‘mesh of net’

Although netting itself will not endure, we have limited physical evidence of dragnetting from Tikopia in the form of cowrie shell net weights. Kirch and Yen (p.245) have been able to identify shells and stones from a Lapita site as net weights and line sinkers:

Tikopia women frequently scour the reef flats with small two-handled dip nets, te kuti, the edge of which is weighted with small Cypraea shells. Such shells have the dorsum removed to facilitate lashing … Two line sinkers were [also] found, each consisting of a natural cobble … with an artificially pecked groove running laterally round the stone. Such grooved cobbles are still used on occasion by Tikopia fishermen.

It is unlikely that the remainder of our fishing terminology, with the possible exception of the stone fish weirs, can ever be supported by archaeological evidence:

*pupu	‘basketry fish trap’
*baRa	‘fence’ (‘fish weir’ only in PCP)
*tupa	‘derris fish poison’
*puna	‘vine used for fish poison’
*(d)rama(R)	(N) ‘torch’; (V) ‘fish at night with torch’

The pointed weapon terminology consists of:

*sao(t)	(N, V) ‘spear’
*sua	(N, V) ‘spear, weapon retained in the hand’
*soka, *soka-i-	‘pierce; stab’
*qio(r,R)	‘spear’
*bako	‘spear’
*kuj(u,i)r	‘fish spear’
*tara	‘fish spear’
*koto	‘obsidian head of spear’
*pusuR	‘bow and arrow’
*(p,pʷ)anaq, (p,pʷ)anaq-i-	‘bow’; (V) ‘shoot’
*tibʷa(ŋ), *tibʷaŋ-i-	‘dart, arrow; shoot with dart, arrow’
*lolo(s)	‘bowstring’

The only archaeological evidence we have for spears, bows and arrows, darts, etc. would seem to be obsidian and bone spear points. Triangular-sectioned obsidian blades are an artefact form produced only in the Manus Group and were almost certainly spearpoints (Spriggs 1997a:163). They were manufactured in the vicinity of the main obsidian sources in the Lou Island area by 2100 BP (pp. 162-164). Green (1979:39) documents a bone spear-point from Main Reef Islands circa 3100 BP.

I can find no archaeological evidence for any kind of club, our two (weakly supported) terms notwithstanding:

*paru	(N, V) ‘club’
*gapi	‘stone? club’

However, for the slingshot we have pointed-end tridachna shell slingstones from Main Reef Islands documented by Green (1979:39):

*maga	‘slingshot; stone’
*kalo	‘sling; turn round and round’

The trail and pitfall spikes, the snare trap and the birdlime lack any supporting evidence from archaeology.

*suja	‘sharpened stake set in ground’
*pʷita, *pʷita-i-	(N, V) ‘snare’
*bulit	‘gum, resin (birdlime?)’

Although we can look to archaeology to confirm the hypothesis that durable artefacts will be found in sites associated with Oceanic-speaking communities, there is little hope of archaeological recovery of the perishable artefacts in question. Here, linguistics adds an extra dimension to research on the prehistory of Oceania.

Notes⇫