Chapter 4.5 Mammals, reptiles, amphibians

Meredith Osmond and Andrew Pawley

1. Introduction⇫

This chapter is concerned with the Proto Oceanic lexicon for three groups of vertebrates: mammals, reptiles and amphibians. Birds and fish each have a chapter to themselves.¹

Attempts to reconstruct POc names for land vertebrates face a number of difficulties. The most fundamental of these relate to discontinuities in the fauna that reflect long-established ocean gaps. Most mammal and reptile species, of course, live on land and cannot cross large ocean gaps. As a consequence cognate sets for certain animal taxa have highly restricted distributions. This is especially true of mammals.

A major zoogeographical boundary divides the mammals native to the Austronesian-speaking world. The Wallace Line, which runs north-south through the Indo-Malaysian Archipelago, separating Sumatra, Borneo, Java and Bali from islands to the east, marks (among other things) the westernmost limits of marsupial mammals. West of this boundary only placental mammals occur. The native land mammals that would have been familiar to speakers of Proto Oceanic, living in the Bismarck Archipelago, were almost entirely different from those known to their ancestors in the Proto Malayo-Polynesian homeland, and from those found in Remote Oceania, where most of the best-described Oceanic languages are now spoken.

East of the Wallace Line there is a region of the Indo-Malaysian Archipelago, Wallacea, where relatively few land mammals occur but there is some overlap between placentals and marsupials. Lydekker’s Line forms the eastern boundary of Wallacea. It separates the Australia-New Guinea faunal zone, home to a rich marsupial fauna, from the islands of eastern Indonesia: Sulawesi, Lombok, Flores, Timor, and the Moluccas. Only bats and rodents are common to both the Asian and the Australia-New Guinea regions.

These faunal regions reflect geomorphological boundaries. The islands west of Wallace’s Line lie on the Sunda Shelf, a shallow continental shelf that was dry land at various times during the ice ages. As a consequence, the western islands of the Indo-Malaysian Archipelago share a similar mammal fauna with the Southeast Asian mainland and, to a lesser extent, with the Philippines and Taiwan. However, a deep trench separates Bali from Lombok, and few Asian mammal species were able to cross this. Similarly, during parts of the ice ages Australia and New Guinea formed a single land mass, Sahul (also known as Greater Australia or Meganesia), but were always separated from the Moluccas and Timor by deep water. Also part of Sahul were the Aru Islands southwest of New Guinea, but not the Kei Islands. Off the Bird’s Head, the West Papuan islands of Batanta, Gebe, Waigeo, Salawati and Misool (Mysol) were included, as were the islands of Cenderawasih Bay (Biak, Yapen, Numfoor) and the islands of the north coast shelf (Kairiru, Manam, Karkar, Bagabag). At the south end of New Guinea the D’Entrecasteaux Group, Louisiade Archipelago, Woodlark Island and the Trobriands were all part of the mainland. However, New Britain, New Ireland, the Solomons and the Admiralties were still separated from the New Guinea mainland and each other.

Robert Blust (1982, 2002, 2009a) has demonstrated the relevance of the Wallace Line to the subgrouping of Austronesian languages. In the most generally accepted subgrouping of Austronesian languages, all the Austronesian languages of New Guinea fall into a lower order group, namely Eastern Malayo-Polynesian. Blust writes

Since placental mammals are not native to areas east of the Wallace Line, PAN and PMP must have been spoken west of that boundary. In crossing the Wallace Line [Austronesian] speakers would have encountered marsupials for the first time. If this happened through separate migrations into eastern Indonesia and the western Pacific there would be no basis for expecting that the terms for marsupial mammals would be cognate, since they would have been independently invented (or borrowed) at different times and in different places. As it happens, however, many languages of eastern Indonesia and the Pacific reflect *kandoRa ‘cuscus’ and *mansar/manser ‘bandicoot/marsupial rat’ which could not have been present in PAN or PMP.

… it is simplest to attribute these innovations to a single speech community that was ancestral to the [Central Malayo-Polynesian, South Halmahera-West New Guinea and Oceanic] languages (2009a: 715-716).

Further significant zoogeographical boundaries occur within Oceania. As one travels from the New Guinea mainland to the Bismarck Archipelago, then east down the main Solomons chain and on into Remote Oceania, the land masses become smaller and the island groups further apart, and the faunal range declines markedly. Relatively few species of terrestrial mammals are widely distributed across Oceania. A good many taxa are restricted to New Guinea or to New Guinea, New Britain and New Ireland.

Reptiles and amphibians are not subject to the same geographical division. Terms for both crocodile and turtle occur on both sides of the Wallace Line and are reconstructable to PAn. Crocodiles were formerly present in Taiwan (see footnote on p.251), and their present breeding range extends from south-west India and the Philippines to the Santa Cruz group. Lizards and snakes are capable of swimming short distances, and are readily transportable on floating logs or rafts of vegetation, but the number of species shrinks as land masses decline in size and island groups become more widely scattered.

In the most generally accepted subgrouping of Austronesian languages, all the Austronesian languages of New Guinea fall into a lower order group, namely Eastern Malayo-Polynesian. It follows that for kinds of mammals found only in New Guinea we will not find a Proto Austronesian (PAn) or Proto Malayo-Polynesian (PMP) reconstruction.

The mammals known to POc speakers may be divided into those indigenous to the Bismarck Archipelago² and perhaps the north coast of New Guinea and those that humans brought to this region from island southeast Asia, namely the pig and dog and certain rat species.

2. Indigenous terrestrial mammals⇫

2.1. Introduction⇫

The indigenous mammal groups of Melanesia recorded in historic times are listed by Flannery (1995a: 13-15). The main island of New Guinea has one tree kangaroo genus and four genera of wallabies. It has several species of bandicoot and a large variety of possums, including cuscuses, pygmy possums, ringtail possums and gliders. It has two genera of spiny ant-eaters (monotremes). A less well known family is the Dasyuridae (formerly known as Phascogale), a family to which Tasmanian devils belong. The New Guinea dasyurids consist of quolls and antechinuses. The Bismarcks have only one bandicoot, one wallaby, and two species of phalanger. The Solomons have one phalanger. Murids and bats are represented by many different genera throughout New Guinea, the Bismarcks and the Solomons. In Vanuatu, New Caledonia, Fiji, Polynesia and Micronesia, the only native terrestrial mammals are bats. The range of a few indigenous mammals has been extended by human introductions.

2.2. Macropodidae⇫

2.2.1. Wallabies (Genera Dorcopsis, Dorcopsulus, Macropus, Thylogale)⇫

Four genera of wallabies occur in the New Guinea area. Dorcopsis is a genus of small rainforest wallabies with an extensive range round the perimeter of New Guinea. One species, the black dorcopsis, is common on mountainous Goodenough Island in the D’Entrecasteaux group. A second genus, Dorcopsulus, the smallest of all the Macropodids, lives in mountain forests. A third genus, Macropus, the large sandy or agile wallaby, is primarily a savanna woodland animal of the south-eastern New Guinea mainland, but is also found on many of the Trobriand and D’Entrecasteaux islands and around Merauke in West Papua. It is hunted by many coastal people, an activity no doubt contributing to its somewhat erratic distribution. Flannery (1995a:82) mentions several reports of a large wallaby, possibly Macropus agilis, from southern New Ireland, which he has so far been unable to verify. The fourth genus, Thylogale, is the small scrub wallaby or pademelon. Its range includes the lowlands of northern New Guinea as far west as Jayapura, the hills round Port Moresby, the Huon Peninsula, New Britain and New Ireland. Bali-Vitu, spoken off the north-west coast of New Britain, has a term for wallaby. Lapita sites in the Arawe Islands, off the south-west coast of New Britain, have yielded bones of Thylogale browni as well as the cuscus Phalanger orientalis and the flightless cassowary, all introduced species.

One wallaby species is found on the islands west of the Bird’s Head, and others on Aru and on Kei. No wallabies are present in the Solomons or the Admiralties group. Although wallabies are now absent from the St Matthias group, bones of wallabies and a cuscus (Spilocuscus maculatus) have been recovered from Lapita middens excavated from the low flat coral island of Eloaue south-east of the high island of Mussau (Green 1979:37). Flannery and White state that the northern pademelon, Thylogale browni, was introduced into New Ireland some 7000 years ago, and that animals were possibly carried from there to the St Matthias group, although the species never became established there (Flannery 1995a:83).

The general pattern seems to be that different species adapt to particular areas, and are restricted to those areas. Although it may happen that two or three species are found within the territory familiar to a single language community, sources for each language usually give only one term for wallaby and it is likely that this is the generic.

A PWOc term for wallaby, *wakin, is well supported, with reflexes in North New Guinea, Papuan Tip and Meso-Melanesian languages. If terms in two non-Oceanic languages spoken on the West Papuan coast are cognate, then *wakin is reconstructable to PEMP and POc. The non-Oceanic terms, from Smits and Voorhoeve (1992:241), are glossed ‘kangaroo’, but the reference must be to small macropod species, popularly known as wallabies, as the larger species known as kangaroos are absent from New Guinea.

PEMP		*wakin	‘wallaby’
SHWNG	Dusner	waen	‘kangaroo’ (wallaby)
SHWNG	Mor	uʔana	‘kangaroo’ (wallaby)
POc		*wakin	‘wallaby’
NNG	Lukep	ok	‘wallaby’
NNG	Mangap	wok	‘wallaby’
NNG	Bing	woy	‘wallaby’
NNG	Bilibil	wei	‘wallaby’
NNG	Gedaged	woi	‘kangaroo (generic), incl. Macropus, Dorcopsis’
PT	Kilivila	waila	‘wallaby’
PT	Iduna	waɣi(tuli)	‘large grey wallaby’
PT	Gabadi	vai(aru)	‘wallaby, Macropus agilis’
PT	Hula	waɣi	‘wallaby, Macropus agilis’
MM	Sursurunga	wak	‘wallaby, includes Agile Wallaby, Forest Wallaby’
MM	Konomala	uakin	‘wallaby’
MM	Minigir	okin	‘wallaby’
MM	Tolai	okin	‘wallaby’
MM	Tolai (Nodup)	okini	‘wallaby’
MM	Label	uakin	‘wallaby’
MM	Siar	uakin	‘wallaby’

Several additional Papuan Tip terms support a PPT reconstruction *wagita ‘wallaby’ These appear to be compounds derived from *wakin + *ta(u), the second element possibly from POc/PPT *tau ‘person, body’.

PPT		*wagita	‘wallaby’
PT	Diodio	vayita	‘wallaby’
PT	Bwaidoga	vaɣita	‘wallaby, Agile wallaby’
PT	Molima	wagilita	‘bandicoot-like animal that eats coconuts’
PT	Iamalele	vagitau	‘opossum’

Agreements between NNG and PT languages point to a PNGOc term, *makan. Reflexes in PT languages predictably add a final -i. Retention of final *n plus the added vowel is unexpected in Motu and Magori, and suggests borrowing from a SE Papuan source, perhaps a dialect of Suau. At least two Central Papuan languages use reflexes of *makan ‘wallaby’ as a generic for ‘fish’. Ross (1994a: 422) uses the example of the two Sinaugoro dialects, one inland (Taboro), the other coastal (Balawaia), to suggest that the term was in some places generalised to cover protein food, and then restricted in some coastal languages to the main protein food available, namely fish.

PNGOc		*makan	‘wallaby, Macropus agilis’
NNG	Dami	maga	‘wallaby’
NNG	Tumleo	magə	‘wallaby’
PT	Magori	magani	‘wallaby, Macropus agilis’
PT	Hula	mani	‘wallaby’
PT	Tawala	makani	‘wallaby type’
PT	Taboro	maɣani	‘wallaby’
PT	Balawaia	maɣani	‘fish’
PT	Motu	maɣani	‘wallaby, Macropus agilis’
PT	West Mekeo	maka	‘fish’
PT	Mekeo	maʔi	‘wallaby, Macropus agilis’

cf. also:

NNG	Matukar	maku(sik)	‘wallaby’ (-sik perhaps from POc *siki ‘small’)
NNG	Kairiru	meik	‘wallaby’
NNG	Ali	meik	‘wallaby’
PT	Kuni	mau	‘wallaby, Macropus agilis’

2.2.2. Tree kangaroos (Genus Dendrolagus)⇫

Within the Oceanic region, tree kangaroos (genus Dendrolagus) are restricted to mainland New Guinea and some close offshore islands. There are about eight species in the region. Dendrolagus inustus, the grizzled tree kangaroo, is found in a wide band across northern New Guinea from the Bird’s Head Peninsula to the Torricelli Mountains. Dendrolagus scottae, Scott’s tree kangaroo, occurs in lowlands forests in Sandaun Province, Papua New Guinea. The other species have distributions that make them less likely to be known to speakers of Oceanic languages. Dendrolagus matschiei, the Huon tree kangaroo, with chestnut-red pelage and long yellow tail, is restricted to the heavily forested mountains of the Huon Peninsula. Dendrolagus dorianus. Doria’s tree kangaroo, is a large brownish-black animal that lives in the cloud and moss forests of montane New Guinea. Dendrolagus goodfellowi and the closely related Dendrolagus pulcherrimus, the smallest species, are found in mid montane zones, D. stellarum in the highlands of west-central New Guinea, Dendrolagus spadix, in the south-western lowlands of Papua New Guinea and D. mbaiso in the highlands of west-central New Guinea.

To date, tree kangaroos are known outside the New Guinea mainland only on Umboi (presumably carried by humans from the nearby Huon Gulf area), Yapen and Salawati and Cape York, Australia (Van Deusen 1972:713, Flannery 1995a: 126-143). There is no record of their existence in the Admiralties or the Solomons, although we have a term for tree kangaroo from Sursurunga, New Ireland, and another from Kaulong, New Britain, presumably because people there are aware that they exist elsewhere.

Fewer than twenty terms for Dendrolagus have been collected in total, all bar one from North New Guinea, the odd one out being Meso-Melanesian: Sursurunga atal ‘wallaby, includes Matschie’s tree kangaroo’. Of the NNG terms, one is from New Britain (Kaulong elan ‘possum, also tree wallaby’), two are from a Schouten Chain language on the north coast (Sissano nep and ʔapar, both apparently meaning ‘tree kangaroo’) and one is from Bing, a Vitiaz Strait language north of the Huon Peninsula (dagyiebiy ‘tree kangaroo’). The remainder are from Huon Peninsula languages. No reconstructions are possible from these data other than a single low level term, Proto Buang *kekuŋ.

2.3. Cuscuses (Phalangeridae)⇫

The phalangerids include the cuscuses of the New Guinea region and the brushtail possums of Australia. They are large arboreal herbivores. Several species of cuscus are present in mainland New Guinea and in the D’Entrecasteaux, Louisiade and Trobriand Islands. Phalangers are common in the islands immediately to the west and southwest of New Guinea, including Sulawesi, Timor, Ambon, Seram, Burn and Kei. Groves (1984) notes that Ambon and Seram have a distinct variety of Phalanger orientalis, and concludes that it got there without human intervention.

In prehistoric times certain species were translocated by humans to Island Melanesia and the Moluccas. New Britain, New Ireland and the Solomons each have an introduced species. Phalanger orientalis, the northern common cuscus (aka grey cuscus) is an abundant lowland species native to New Guinea that about 20,000 years ago was taken to New Ireland, where it became an important prey species (Leavesley 2005). It was introduced to the Solomon Islands and also to Timor at some time in the last 7,000 years (Flannery and Wickler 1990, Glover 1986). A single species, Spilocuscus maculatus, is found in the Admiralties, including Manus, Wuvulu and Ninigo. It is commonly hunted on Manus.

There is a well supported POc reconstruction for ‘cuscus’, *(k,kʷ)adroRa. For a number of languages, sources give the gloss ‘possum’ to its reflexes (‘possum’ is often used loosely in Australian English to refer to any member of the superfamily Phalangeroidea), but this can only refer to a phalanger.

PCEMP		*kandoRa	‘cuscus’ (Blust 2002)
CMP	Watubela	kadola	‘cuscus’
POc		*(k,kʷ)adroRa	‘phalanger, cuscus’ (Ross 1988: *kadroRa)
Adm	Mussau	aroa	‘phalanger; cuscus’ (Smythe)
Adm	Bipi	koxa	‘cuscus’ (metathesis)
Adm	Sori-Harengan	ohay	‘cuscus’ (metathesis)
Adm	Penchal	kotay	‘cuscus’ (metathesis)
Adm	Nauna	kocay	‘cuscus’ (metathesis)
NNG	Mangap	kendēre	‘opossum’
NNG	Gedaged	kodoɬ	‘k.o. phalanger with tan or greyish fur’
NNG	Manam	ʔodora	‘cuscus’
PT	Dobu	kʷadoiya	‘possum’
PT	Kilivila	kʷadoya	‘cuscus’
PT	Muyuw	kʷadoy	‘cuscus, possum’
MM	Tabar	adora	‘cuscus’
MM	Lungga	ɣadora	‘possum’
MM	Simbo	ɣadora	‘possum’
MM	Maringe	kadara	‘possum’

The following Southeast Solomonic terms reflect *kadora, with -r- for expected -l-. This is consistent with borrowing from Meso-Melanesian neighbours in the western Solomons.

SES	Gela	kadora	‘cuscus’
SES	West Guadalcanal	kadora	‘cuscus’
SES	Talise	kadora	‘cuscus’
SES	Malango	kadora	‘cuscus’

A second reconstruction is more weakly supported. It is well represented in the Cristobal-Malaitan branch of SE Solomonic but otherwise is reflected only in Gela, where the reflex refers to a juvenile cuscus, and in one Papuan Tip language.

POc		*pu(dr,d,j)(o,u)	‘species of possum or cuscus’
PT	Suau	hudu	‘striped possum’ (Flannery 1995: 203)
SES	Gela	vuto	‘a young cuscus’
SES	Owa	futu	‘possum’
SES	’Are’are	huto	‘possum’
SES	Kahua	huto	‘cuscus’
SES	Longgu	vuto	‘cuscus’
SES	Lau	futo	‘cuscus’
SES	To’aba’ita	futa	‘cuscus’
SES	Sa’a	huto	‘cuscus, phalanger’
SES	Arosi	huto	‘cuscus’

2.4. Other possum families (Acrobatidae, Burramyidae, Petauridae, Pseudocheiridae)⇫

Besides the cuscuses, four other families within the super-family Phalangeroidea are represented in New Guinea. They include little feather-tailed possums (Acrobatidae), pygmy possums (Burramyidae), sugar gliders, striped possums and long-fingered striped trioks (all Petauridae) and ringtail possums (Pseudocheiridae). The Acrobatidae and Burramyidae are restricted to the Australian-New Guinea land mass. Sugar gliders are also found in the D’Entrecasteaux, Louisiade and Trobriand Islands and New Britain, and outside the Oceanic region on Halmahera, Kei and Aru. They are absent from New Ireland, the Solomons and the Admiralties. The Pseudocheiridae are poor dispersers who prefer montane forests, and are confined to New Guinea and a few of the larger islands off the coast of western New Guinea. A PNGOc term for ‘sugar glider’ is supported by probable cognates shared by several Papuan Tip languages and at least one North New Guinea language. However, reflexes show a number of irregularities, and it is necessary to reconstruct doublets.

PNGOc		*si(r,R)(a,e)si(r,R)(a,e), *di(r,R)(a,e)di(r,R)(a,e)	‘Petaurus breviceps, sugar glider’
NNG	Gedaged	saɬesaɬe	‘sugar glider, Petaurus breviceps’ (vowel metathesis)
NNG	Riwo	siɬesiɬe	‘k.o. marsupial’
PT	Molima	selasela	‘sugar glider’
PT	Iduna	selasela	‘flying squirrel, sugar glider’
PT	Tawala	selasela	‘sugar glider’
PT	Wedau	sira-sira	‘squirrel’ (sugar glider)
PT	Lala	dilodilo	‘glider phalanger’ (-o unexpected)
PT	Motu	diledile	‘sugar glider, Petaurus breviceps’

cf. also:

PT	Gapapaiwa	surasura	‘flying squirrel’
MM	Tolai	diradira	‘flying squirrel’ (d- for †∅)

2.5. Bandicoots (Peroryctidae)⇫

New Guinea is home to several genera of the family Peroryctidae, the bandicoots. They include Echymipera (common or spiny bandicoot, long-nosed bandicoot), Peroryctes (giant bandicoot, Raffray’s bandicoot), and Microperoryctes (mouse bandicoot, striped bandicoot). The mouse bandicoot is found only in the west of West Papua but the other genera are distributed widely on mainland New Guinea. Bandicoots are not present in the Solomons. Spriggs (1997a: 54) writes that at Pamwak, in Manus, bandicoots appear suddenly in the archaeological record at about 13,000 BP and quickly become dominant in the faunal record. This suggests human introduction. To the west of New Guinea, outside the region occupied by speakers of Oceanic languages, bandicoot species are present on Seram, Yapen, Waigeo, Kei and Aru. Compilers of wordlists are often unfamiliar with bandicoots, describing them as ‘large bush rats’, ‘marsupial rats’ or similar.

A POc term for ‘bandicoot’, *mʷajar, is well supported. There are reflexes in the Admiralties, North New Guinea, Papuan Tip and Meso-Melanesian subgroups referring to the bandicoot, and probable reflexes in Ponapean and certain Fijian dialects with change of referent. External cognates are found in several CEMP languages of the Moluccas and islands to the south of the Bird’s Head of New Guinea, where the reference is often to cuscus rather than bandicoot.

PCEMP		*mansar	‘bandicoot, marsupial rat’ (Blust 1982, 2002)
CMP	Ujir (Aru Is.)	meday	‘cuscus’ (Antoinette Schapper, pers. comm.)
CMP	Ambelau	mate	‘bandicoot, marsupial rat’
CMP	Kei	medar	‘cuscus’
CMP	Yamdena	mande	‘cuscus’
POc		*mʷajar	‘bandicoot’
Adm	Lou	mʷas	‘bandicoot, marsupial rat’
Adm	Nauna	mʷac	‘bandicoot, marsupial rat’
NNG	Mangap	mōzo	‘bandicoot’
NNG	Wogeo	mʷaja	‘cuscus’
NNG	Ali	maic	‘bandicoot’
PT	Sudest	mʷaja-mʷaja	‘bandicoot’
PT	Taboro	moda	‘bandicoot’
PT	Motu	mada	‘k.o. bandicoot’
MM	Meramera	maso	‘bandicoot’
MM	Tolai	mā(ŋ)	‘bandicoot’
MM	Ramoaaina	ma(n)	‘bandicoot’
Mic	Ponapean	(keli)mʷet	‘k.o. small bat’ (keli- ‘prefix signifying danger, connection with spirit world’; Blust 2001b)
Fij	Namosi	ŋʷaða	‘k.o. large rat’ (Blust 1993)

The following are probably borrowed from a pre-Oceanic or early Oceanic language (Ross 1988:21):

NNG	Megiar	madar	‘bandicoot’
NNG	Takia	madar	‘bandicoot’
NNG	Yabem	mojaŋ	‘marsupial rat, cuscus, possum, tree kangaroo’
NNG	Mapos Buang	masaŋ	‘possum, Pseudocheirus corinnae, yellow-green ring-tailed possum’

2.6. Marsupial carnivores (Dasyuridae)⇫

The Dasyuridae are a family of carnivorous and insectivorous marsupial rodents, including various antechinus species and quolls and, most famously, the Tasmanian devil. Five genera of dasyurids are unique to New Guinea and three genera are common to New Guinea and Australia. Most dasyurids are small, weighing between 100-500 grams, although quolls of the New Guinea rainforest may exceed half a kilogram in weight (Flannery 1995: 83). Dasyurids may be either diurnal or nocturnal, and some antechinus species are partly arboreal. Many species are found in rainforest, often at middle to high altitudes. Few members of the family have been recorded from coastal Austronesian-speaking areas, although one antechinus is found on Normanby in the D’Entrecasteaux Group, and other dasyurids have been recorded from the Aru Islands. Flannery (1995:84) comments that quolls will enter villages to catch rats, and are respected as fierce predators by most New Guineans familiar with them, who show great caution when handling them.

No Oceanic terms for dasyurids have been recorded. It is likely that compilers of wordlists have generally not been familiar with dasyurids as such, and in some cases have referred to them by a term such as ‘small marsupial rodent’.

2.7. Echidnas (Tachyglossidae)⇫

Monotremes are represented in New Guinea by two genera of echidna: Tachyglossus (short-beaked echidna) and Zaglossus (long-beaked echidna). Both are regarded as rare. Flannery (1995: 67) suggests that hunting pressure has severely reduced their numbers. Very little is known about the distribution and natural history of Tachyglossus, but it has been recorded in the Markham valley and around Port Moresby, and from the island of Salawati off the Bird’s Head. Zaglossus has also been reported from both the western and eastern extremities of New Guinea, but is found only where population densities are low. A term for the echidna is present in only two languages in our sample (Sinaugoro modagini ‘anteater’ (cf. moda ‘bandicoot’) and Motu miɣu ‘short-beaked echidna’) and no reconstructions are possible.

2.8. Rats (Muridae)⇫

Rats were a significant source of protein for many Oceanic communities in prehistoric times. The diversity of rat species diminishes sharply as one moves from New Guinea to the Bismarcks and Solomons. In New Guinea the murid fauna is ‘abundant and diverse, and for the size of the land mass, one of the most species-rich found anywhere on earth’ (Flannery 1995: 231). There are some 27 genera there, including Hydromys, Leptomys, Mallomys, Melomys, Pogonomys, Uromys, Rattus and Stenomys, although only a subset of these are present in any one region. Most New Guinea area species are endemic.

Nine murid species are endemic to the Solomons, including two that are extinct. Three species of Solomys and one of Uromys have evolved into large or very large arboreal rats, which are probably the evolutionary counterparts of the possums of Sahul. However, single islands in the Solomons seldom have more than two or three species. No rat species were present in Remote Oceania before the advent of humans.

Five commensal rat species have been introduced to Oceania: Rattus rattus, Rattus exulans, Rattus tanezumi, Rattus norvegicus and Mus musculus. Rattus rattus, the black rat, and Mus musculus, the mouse were introduced into the Pacific within the past few centuries by European voyagers. The other three species were carried to Oceania in prehistoric times. Rattus exulans, the Pacific rat, is a commensal animal whose original range was in Asia, probably from Bangladesh to Vietnam (Groves 1984). It was carried to Oceania by humans several millenia ago and today is found widely in Melanesia, Polynesia and Micronesia. Rattus exulans has in recent years come to play an important role in tracing the directions of early human settlement of Oceania, especially Remote Oceania. Because it is an introduced animal and because it is common in archaeological sites, it provides sufficient prehistoric and contemporary populations for reliable genetic analysis (Matisoo-Smith and Robins 2004, Matisoo-Smith 2007). Rattus rattus occurs in Lapita sites from the Bismarcks to western Polynesia and was clearly associated with the initial Lapita dispersal into Remote Oceania. It was the only one of the commensal rats to reach Polynesia in ancient times.

Rattus praetor, the spiny rat, considerably larger than exulans, is native to New Guinea, where it is widespread north and south of the central ranges as far east as the Sepik-Ramu basin. It also occurs on offshore islands in the Bismarck Archipelago where it appears in archaeological assemblages dated to at least 13,000 years ago (White, Clark and Bedford 2000) and in the Solomon Islands, where its introduction is probably of similar antiquity. Rattus praetor is now extinct in Remote Oceania but occurs in archaeological sites in Vanuatu and Fiji, though not in Polynesia or Micronesia.

A third commensal species, Rattus tanezumi, the Asian house rat, has a narrow distribution in Oceania, being securely attested only in archaeological sites in the Caroline Islands and the northern Marianas, dated to within the last 1,000 years. Rattus tanezumi is native to South Asia and Southeast Asia and its Oceanic distribution is consistent with transport direct from Island Southeast Asia to the Carolines, rather than via Melanesia.

Sources for Oceanic language seldom give precise identifications for rat terms. Glosses tend to be of the order of ‘bush rat’ or simply ‘kind of rat’, glosses which fail to capture genuine taxa in the language. One investigator (Futscher 1959) recorded at least 10 terms for kinds of rat in Tolai but the glosses simply say ‘mouse’.

POc *kasupe, continuing PEMP *kanzupay, has reflexes applied to one or more species of the genus Rattus, rats living in and around human settlements. Reflexes in Cristobal-Malaitan and Vanuatu languages preserve initial syllable *-a-. Those in Western Oceanic and Guadalcanal languages and in Western Fijian show initial u-. These may represent several independent cases where the first vowel assimilated to the following u. However, the assimilation is so widespread it is reasonable to reconstruct a POc doublet, *kusupe.

PCEMP		*kanzupay	‘rat’ (Blust 1993)
POc		*kasupe, *kusupe	‘rat taxon, probably Rattus exulans and possibly other small commensal sp. or spp.’
Adm	Drehet	(n)osop	‘marmot’ (sic; cf. pʷek ‘rat’)
Adm	Seimat	usuh	‘k.o. rat’
NNG	Malai	uzuv	‘rat’
NNG	Mangap	kuzi	‘rat/mouse’
NNG	Malasanga	kui	‘rat’
NNG	Mindiri	kusue	‘rat’
NNG	Wogeo	kusive	‘rat’
NNG	Ulau-Suain	isup	‘bandicoot’
PT	Lala	uduve	‘rat’
PT	Ouma	ureve	‘rat’
MM	Tolai	kaupa	‘rat, mouse, Mus browni’
MM	Vitu	kuðuve, kuvuðe	‘rat’ (metathesis)
MM	Bali	kuzuveke	‘rat’ (final -ke unexpected)3
MM	Bola	kuruve	‘rat’
MM	Madak	kisap	‘rat’
MM	Siar	kusup	‘rat’
MM	Kandas	kusupu	‘rat’
MM	Teop	kuho	‘rat’
MM	Varisi	kuzu	‘rat’
MM	Ririo	kuj	‘rat’
SES	West Guadalcanal	ɣusuve	‘rat’
SES	Talise	ɣusuve	‘rat’
SES	Malango	husuve	‘rat’
SES	Lau	ʔasufe	‘rat’
SES	Kwaio	ʔasufe	‘rat’
SES	’Are’are	ʔasuhe	‘rat’
SES	Sa’a	ʔasuhe	‘rat’
SES	Arosi	kasuhe	‘rat’
NCV	Mota	ɣasuwe	‘rat’
NCV	Mwotlap	(na)ɣhōw	‘rat, Rattus rattus’
NCV	Paamese	asue	‘rat’
SV	North Tanna	kahap	‘mouse, rat’
SV	Lenakel	kahau	‘rat’
SV	Kwamera	(i)esukʷ	‘mouse, rat’
Fij	Wayan	kuðuve	‘generic for rats and mice’

Reflexes of the root PAn *-labaw ‘rat’ occur alone or in combination with a preposed CV-formative in various daughter languages, yielding a number of related terms including PAn *kulabaw and PMP *balabaw (Blust 2002:107).

As an independent form *labaw is traceable back to PMP, with Oceanic reflexes from the Te Motu group and, less certainly, the Malakula languages of Vanuatu.

PMP		*labaw	‘rat, mouse’
WMP	Kagayanen	ambaw	‘mouse, rat’
CMP	Manggarai	lavo	‘rat’
CMP	Rotinese	lafo	‘mouse, rat’
POc		*la(b,p)(o,u)	‘rat taxon’
TM	Äiwoo	lapū	‘rat’ (Tryon and Hackman 1983: 190; Wurm 1969: 69)
TM	Natügu	lopu	‘rat’ (Wurm 1969: 69)
TM	Tanibili	no-uvo	(Tryon & Hackman)
Pn	Pileni	lepū	(probable borrowing from TM)

cf. also: the following terms for ‘rat’ from Malakula languages:

NCV	Neve’ei	ne-labut	‘rat taxon’
NCV	Nāti	ne-lemput	‘rat taxon’
NCV	Naman	labët	‘rat taxon’
NCV	Southwest Bay	lambut	‘rat taxon’

The latter forms exhibit a final -t which is unexpected in comparison with PMP *labaw. Tagbanawa (Kalamian), a Philippines language, has a similar word laput ‘rat’ but this is likely to be an accidental resemblance.

PMP *bualabaw has probable Oceanic reflexes in Admiralty and NNG languages. The expected POc form is *bʷalapo or *pʷalapo, and Bing palāf supports the latter. However, the final vowels of the Admiralties forms are unexplained. The Titan and Aua forms also show irregular loss of *-p-.

PMP		*bualabaw	‘rat, mouse’ (ACD: *balabaw)
WMP	Malagasy	vualavu	‘rat’ (Blust 1972a)
WMP	Ngaju Dayak	belawaw	‘rat’ (Blust 1972a)
WMP	Uma	flulehuʔ	‘mouse, rat’ (Tryon 1995: 03.630)
WMP	Buginese	balawo	‘mouse, rat’
CMP	Konjo	balaho	‘mouse, rat’ (Tryon 1995: 03.630)
CMP	Buru	blafe(n)	‘marsupial rat’ (Grimes, pers. comm)
POc		*pʷalapo	‘rat taxon’
Adm	Titan	buléy	‘rat’ (Blust 1996:43)
Adm	Wuvulu	balafa(i)	‘rat’ (Blust 1993, 1996:43)
Adm	Aua	balā	‘rat’ (Blust 1993)
NNG	Bing	palāf	‘bush rat with very strong teeth’

PAn *kulabaw ‘rat, mouse’, with reflexes in Taiwan (Paiwan kulavaw ‘mouse, rat’, Rukai kolabaw ‘rat in the fields’) as well as in MP languages is reconstructed by Blust (2002). A probable Oceanic reflex has been recorded in Bauan Fijian: kalavo ‘rat’. Amber (SHWNG) agrees with Bauan in having initial ka rather than ku, pointing to PEMP *kalabaw, POc *kalapo.

PAn		*kulabaw	‘rat, mouse’ (Blust 2002)
PEMP		*kalabaw	‘rat, mouse’
SHWNG	Amber	kalubu	‘mouse’ (Smits)
POc		*kalapo	‘rat taxon’
Fij	Bauan	kalavo	‘rat (Rattus spp.)’

Two distinct cognate sets are widespread in Polynesia. The terms are simply glossed ‘rat’ in the sources, unless otherwise noted, but the proto-forms must all have referred to Rattus exulans.

PPn		*kumā	‘rat’ (pollex)
Pn	Tongan	kumā	‘rat’
Pn	Niuean	kumā	‘rat’
Pn	East Uvean	kumā	‘rat’
Pn	Anutan	kuma	‘rat’
Pn	Tikopia	kuma(ri)	(final -ri unexpected)
Pn	Marquesan	kuma-kuma	‘rat’

The second set is confined to Nuclear Polynesian languages.

PNPn		*kimoa	‘rat, Rattus exulans’
Pn	Samoan	ʔimoa	‘rat, Rattus exulans’
Pn	Tuvalu	kimoa	‘rat, Rattus exulans’
Pn	Nukuoro	kimoo	‘rat, Rattus exulans’
Pn	Ifira-Mele	kimoa	‘rat, Rattus exulans’
Pn	Rennellese	kimoa	‘rat, Rattus exulans’
Pn	West Uvea	kimoa	‘rat, Rattus exulans’

Lincoln (2007) points to terms for ‘rat’ in various PT languages from Milne Bay Province that do not correspond regularly to *kimoa but show more than a passing resemblance. The Milne Bay terms are from unpublished wordlists recorded by Pawley (1969).

PT	Oya’oya	gimau	‘rat’
PT	Logea	gimʷau	‘rat’
PT	Tubetube	gimʷau	‘rat’
PT	Saliba	ɣimʷau	‘rat’
PT	Gilagila	gibʷau	‘rat’

The following reconstruction is attributable to Proto Western Oceanic on the basis of cognates in North New Guinea, New Ireland and NW Solomonic.

PWOc		*kotu	‘rat’
NNG	Sio	koto	‘rat’
MM	Tabar	kotu	‘rat’
MM	Tangga	kut	‘rat’
MM	Hoava	kutu	‘rat’
MM	Marovo	kutu	‘rat, common type’ (kiitu vaka ‘big rat that eats coconuts’; Hviding)
MM	Vangunu	kutu	‘rat’

2.9. Bats (Order: Chiroptera, Suborders Megachiroptera, Microchiroptera)⇫

Bats form a very large order consisting of two suborders and at least six families (Flannery 1995). The Megachiroptera include the fruit-sucking bats (flying foxes and fruit bats) and bats which feed on nectar and pollen. Figure 5.6 shows the greater bare-backed fruit bat, Dobsonia moluccensis (aka D. mapna). This species and others of the genus Dobsonia are widespread in New Guinea and Island Melanesia, as far east as the Solomons, as are other members of the family Pteropodidae (fruit-bats). The Microchiroptera include tube-nosed bats, sheathtail bats, horseshoe bats and leaf-nosed bats, typically insectivorous, although at least one genus preys on small fish (Flannery 1995: 455). Several species of Microchiroptera are recently described, and revision of the genera is ongoing.

Reflexes of PMP *paniki ‘fruit bat, flying fox, Pteropns spp.’ occur in the Philippines and Indonesia, but have not been located in Oceanic languages. POc *bʷeka is a well supported reconstruction. It has reflexes denoting ‘fruit bat’ in most major subgroups of Oceanic, an exception being Papuan Tip. In NCV languages *bʷeka is generally lost but in a few languages the form is continued as the name for the barn owl, Tyto alba. For North New Guinea reflexes see under *mali-bʷeka below.

POc		*bʷeka	‘flying fox, large fruit bat’
Adm	Drehet	pʷek	‘rat’
Adm	Titan	pʷe	‘bat, flying fox’
MM	Bali	beɣa	‘flying fox’
MM	Vitu	beɣa	‘flying fox’
MM	Nakanai	beka	‘flying fox’
MM	Meramera	bea	‘flying fox’
MM	Sursurunga	bek	‘generic, fruit bat, flying fox’
MM	Patpatar	beka	‘flying fox’
NCV	Araki	peka	‘Tyto alba, barn owl’
NCV	Tangoa	beka	‘Tyto alba, barn owl’
NCal	Nemi	bʷek	‘flying fox’
NCal	Jawe	bʷek	‘bat, flying fox’
Mic	Kosraean	fak	‘bat’
Mic	Mokilese	pʷēk	‘bat’
Mic	Ponapean	pʷēk	‘bat’
Fij	Wayan	bekʷa	‘bat (generic)’
Fij	Bauan	beka	‘fruit bat’
Pn	Niuean	peka	‘flying fox, fruit bat’
Pn	Tongan	peka	‘bat, flying fox’
Pn	Rennellese	peka	‘flying fox: formerly important for teeth and food, and for the sport of snaring’
Pn	Samoan	peʔa	‘flying fox’
Pn	Tikopia	peka	‘fruit bat, both large and small’
Pn	Hawaiian	peʔa	‘bat’

A few Admiralties and North New Guinea languages reflect a complex form *mali-bʷeka. The etymology of *mali- is uncertain but it is most likely a modification of POc *manuk ‘bird, bat’.

POc		*mali-bʷeka	‘flying fox, large fruit bat’
Adm	Nyindrou	balimek	‘large fruit bat’ (metathesis of two foot-initial consonants)
Adm	Drehet	pelemik	‘bat/flying fox’
NNG	Vehes	mɛbɛk	‘fruit bat’
NNG	Mapos Buang	malibək	‘large fruit bat, flying fox’

cf. also:

Adm

Titan

pálimet

‘flying fox, fruit-eating bat’

The next reconstruction is also plainly a compound of *mali- and *boŋi ‘night’, reflecting the fact that bats are nocturnal. In one of the Motu forms and in Mwotlap a different qualifier has been substituted for the first morpheme, and in the Tarpia reflex that morpheme is missing.

POc		*mali-boŋi	‘flying fox’
SJ	Tarpia	boni	‘bat’ (Smits and Voorhoeve 1992: 244)
NNG	Tami	malimboŋ	‘flying fox’
NNG	Wab	malboŋ	‘flying fox’
NNG	Bing	malbaŋ-boŋ	‘flying fox (mid size)’
NNG	Bing	malboŋ	‘flying fox, fruit bat’
NNG	Manam	malaboŋ	‘flying fox’
PT	Suau	malibon	‘flying fox’
PT	Molima	niboi	(unexpected loss of first syllable)
PT	Motu	mariboi	‘large fruit bat, flying fox’ (for †maiboi) (Oram n.d.)
PT	Motu	(sisi)boi	‘small bat’ (sisi ‘small’)
PT	Gabadi	(manu)boʔi	‘flying fox’ (for †maiboni, possibly a folk etymology: manu < *manuk ‘bird’)
MM	Tiang	mələbiŋ	‘flying fox’
MM	Nalik	məlbuŋ	‘flying fox’
NCV	Mwotlap	ilbʷōŋ	(1) ‘a small dark bat’; (2) ‘Vanikoro swift’ (unexpected loss of initial nasal)

cf. also: the following, which are glossed’flying fox’ and appear to reflect *mali + *eba or *oba.

Adm	Mussau	marieba	‘flying fox’ (-r- for †-l-)
PT	Ouma	maiba	‘flying fox’
PT	Balawaia	maioba	‘flying fox’
PT	Maopa	maopa	‘flying fox’
PT	Hula	maopa	‘flying fox’

In many Oceanic languages small insectivorous bats (Microchiroptera) have a separate name from fruit bats. It is uncertain whether this was also the case in POc. In parts of Polynesia reflexes of *peka-peka are given both to a small bat and to a swift whose behaviour, form and habitat resemble superficially that of insectivorous, cave-dwelling bats. The Fijian languages have cognate reduplicated forms referring only to small cave bats. Reduplication is commonly used to show diminutives in some Oceanic languages (see also vol.3,50-51).

PCP		*beka-beka	‘small insectivorous, cave-dwelling bat’
Fij	Wayan	bekʷa-bekʷa	‘a small insectivorous cave-dwelling bat’
Fij	Bauan	beka-beka	‘insect-eating bat’
PPn		*peka-peka	(1) ‘small bat’; (2) ‘white-rumped swiftlet’
Pn	Niuean	peka-peka	‘bird, white-rumped swiftlet, nests in caves’
Pn	Tongan	peke-peka	‘swallow’
Pn	East Futunan	peka-peka	‘small bat’
Pn	East Uvean	peka-peka	‘k.o. small swallow’
Pn	Rennellese	peka-peka	‘glossy swiftlet’
Pn	Samoan	peʔa-peʔa	‘white-rumped swiftlet’
Pn	Māori	peka-peka	‘bat’
Pn	Hawaiian	(ʔō)peʔa-peʔa	‘bat’

An unrelated Mwotlap term, ilbʷōŋ (bʷōŋ from POc *boŋi ‘night’) refers to the same two creatures, a small dark bat and a Vanikoro swift, a bird with wholly dark plumage (A. François, pers. comm.), while Kwamera (SV) has kiri ‘flying fox’ and kiri-kapou ‘glossy swiftlet’.

3. Introduced terrestrial mammals⇫

3.1. Pigs (genus Sus, family Suidae)⇫

Island Southeast Asia is the centre of genetic diversity of the genus Sus, containing six of the 10 generally recognised species: S. verrucosus (the Javan warty pig), S. barbatus (the bearded pig of Malaysia, Sumatra and Borneo with a subspecies in Palawan), S. celebensis (the Sulawesi warty pig), S. cebifrons (the Visayan warty pig), S. philippensis (the Philippines warty pig) and Sus scrofa (the common wild pig, endemic to Europe, North Africa, parts of Asia and parts of the Indo-Malaysian archipelago). No pig species is native to New Guinea or any part of Oceania.

Most pig species have never been domesticated. The most important exception is Sus scrofa, which was domesticated in the Middle East and in China around 10,000 years ago and subsequently introduced to many regions. A wild variety of this species, Sus scrofa vittatus, the banded pig, inhabits Sundaland from the Malay Peninsula to Sumatra, Java and Bali. It is also present east of Wallace’s Line on Sumbawa and Lombok, possibly due to human transportation.

Sus celebensis occurs domesticated only in Timor and Roti. Groves (1984) argues that its natural range is Sulawesi and offshore islands, and that its presence on Timor, Roti, Flores, Halmahera (and even more surprisingly, on Simalue, west of Sumatra) is the result of human agency. A similar conclusion, based on genetic evidence, is drawn by Larson et al. (2007). Groves (1984:3) suggests that the wild and domestic pigs of New Guinea are a cross between the banded pig and the Sulawesi warty pig.

Pigs, domesticated and wild, appear to have been part of the Austronesian world since PAn times. A number of terms associated with pigs and pig husbandry can be reconstructed for PAn and PMP. In his Austronesian Comparative Dictionary (1995), Blust writes that

Although PAn *beRek evidently meant ‘domesticated pig’ PAn *babuy referred to pigs in general, with qualifications where needed.⁴ In other words, the glosses of *babuy and *beRek were not perfectly complementary, but partially overlapping …. Reflexes of both *babuy and *beRek are found in Formosan and WMP, but only *babuy is reflected in CMP and only *beRek in Oceanic.

The history of the introduction of pigs into Oceania has been a matter of debate. Pig bones occur in early Lapita sites from the Bismarck Archipelago to the Reef Is. in the period 3300-3100 BP. A recent genetic study supports the view that although Lapita pigs reached Melanesia via Indonesia they ultimately stem from Vietnam or nearby parts of mainland Southeast Asia, rather than from Taiwan (Larson et al. 2007), but domestic pig bones dated to 4000 BP have been found in cultural deposits in the Philippines (Piper et al. 2009). Some archaeologists argue that the wild pigs of New Guinea are feral descendants of Lapita pigs. Others hold that pig teeth are present in New Guinea highland assemblages predating Lapita, although it has proved difficult to securely date these finds.

POc *boRok is continued in all the major subgroups of Near Oceania and in parts of Remote Oceania as far east as western Fiji. However, no reflexes are found in New Caledonia and the Loyalty Is., where pigs have been absent since initial settlement by Lapita peoples, or in eastern Fiji and Polynesia.

PAn		*beRek	‘domesticated pig’ (ACD)
POc		*boRok	‘pig, Sus serofa’
Adm	Seimat	pou	‘pig, Sus serofa’
Adm	Lou	puo	‘pig, Sus serofa’
Adm	Nyindrou	bou	‘pig, Sus serofa’
NNG	Takia	bor	‘pig, Sus serofa’
NNG	Manam	boro	‘pig, Sus serofa’
NNG	Yabem	boʔ	‘pig, Sus serofa’
NNG	Mumeng	bʷok	‘pig, Sus serofa’
PT	Tubetube	buluka	‘pig, Sus serofa’
PT	Kilivila	bulukʷa	‘pig, Sus serofa’
PT	Bwaidoga	bulukʷa	‘pig, Sus serofa’
PT	Gapapaiwa	poro	‘pig, Sus serofa’
PT	Wedau	poro	‘pig, Sus serofa’
MM	Bali	boroko	‘pig, Sus serofa’
MM	Tabar	boro	‘pig, Sus serofa’
MM	Nakanai	bolo	‘pig, Sus serofa’
MM	Tolai	boro(i)	‘pig, Sus serofa’
MM	Simbo	boroɣo	‘pig, Sus serofa’
SES	Gela	bolo	‘pig, Sus serofa’
SES	Tolo	bolo	‘pig, Sus serofa’
PNCV		*boe	‘pig, male pig’ (final -e unexpected)
NCV	Mota	pʷoe	‘pig, male pig, barrow pig’
NCV	Kiai	poe	‘pig, Sus serofa’
NCV	Raga	boe	‘pig, Sus serofa’
NCV	Nokuku	poe	‘pig, Sus serofa’
Fij	Wayan	gō	‘pig, Sus serofa’

Reflexes of *boRok in many Solomons languages are irregular. Irregular loss of *R is exhibited in this term in most Cristobal-Malaitan languages, e.g. Longgu bo, Sa’a pō, Santa Ana bʷō, some Guadalcanal languages, e.g. Ghari bo, and many NW Solomonic languages, e.g. Roviana boko. North Malaitan languages show unexpected s as the medial consonant, as in Fataleka boso, suggesting borrowing from Santa Isabel languages, which typically show the same replacement.

POc *boRok is not retained in Polynesian languages, being replaced by PPn *puaka. This appears to continue POc *b(o,u)kas(i), which has reflexes in PT, NCV and SV. Reflexes of both *boRok and *b(o,u)kas(i) persist in PT and NCV. It is not clear whether there was a meaning difference between the two terms.

POc		*b(o,u)kas(i)	‘pig’
PT	Sudest	voɣadi	‘wild pig that dogs can’t kill’
PT	Dobu	boʔas(a)	‘village pig’
PSOc		*bukasi	‘pig (generic), and probably by default male pig, boar’ (John Lynch, pers. comm.)
NCV	Mota	puɣas	‘to rear with food, tame by feeding’
NCV	Paamese	vuasi	‘pig (generic)’
NCV	Nguna	(na-)pʷokasi	‘meat, flesh’
SV	Lenakel	pukas	‘pig’
SV	Kwamera	pukah	‘pig’
SV	Anejom̃	pikaθ	‘pig’
PPn		*puaka	‘pig’
Pn	Tongan	puaka	‘pig’
Pn	Samoan	puaʔa	‘pig’
Pn	Tikopia	puaka	‘pig’ (archaic)
Pn	Hawaiian	puaʔa	‘pig’

cf. also: The following are all Polynesian loans.

PT	Saliba	puaka	‘pig’
PT	Magori	uaka	‘pig’
PT	Yoba	uaʔa	‘pig’
NCal	Nixumwak	puaka	‘pig’
NCal	Iaai	buaka	‘pig’
Fij	Bauan	vuaka	‘pig’

Another term, *bʷasa, may have referred specifically to sows. This term has reflexes in NNG, PT, MM and SES.

POc		*bʷasa	‘pig, possibly sow’
NNG	Bilibil	bas	‘pig’
PT	Ouma	bura	‘pig’
MM	Patpatar	pes	‘female pig’
SES	Gela	baha	‘sow’
SES	Lau	gʷata	‘sow’
SES	To’aba’ita	gʷata(kʷala)	‘sow that has had young more than once’ (kʷala ‘bear young’)
SES	’Are’are	ata	‘sow’ (for †pata)
SES	Kwara’ae	gʷat	‘pig’ (Tryon and Hackman 1983: 185)

cf. also:

SES	Arosi	bʷa-bʷa	‘female (only of pigs)’ (zero for †-t)
SES	Sa’a	pʷa-pʷa	‘female of an animal; lay eggs’ (zero for †-t)

In NCV there is a widespread cognate set pointing to PNCV *bʷara-bʷara ‘female pig’.

PNCV		*bʷarabʷara	‘female pig’ (Clark 2009)
NCV	Uripiv	bʷer-pʷer	‘pig (generic)’
NCV	Lonwolwol	ba-bar	‘female pig, sow’
NCV	Aore	mbara-mbara	‘sow’
NCV	Naman	ba-bar	‘pig, esp. sow’
NCV	Neve’ei	(ne-)bʷebar	‘sow’

3.2. Dogs (Canis sp.)⇫

Dogs were brought by humans to Island Southeast Asia, Oceania and Australia from mainland SE Asia within the last few millennia. The pre-European dogs of Oceania and Australia did not bark, but rather howled.

PAn and PMP *asu ‘dog’ is reflected in many languages of Taiwan, the Philippines, western Indonesia, the Lesser Sundas and much of the southern and central Moluccas, but to date is unattested in any Eastern Malayo-Polynesian language. As others (Hudson 1989, Lynch 1991, Donohue 1995) have discovered, names for ‘dog’ are notoriously variable in Oceanic languages. This variability is consonant with a hypothesis, first suggested by Andrew Sharp (1956), that on small remote islands in Oceania the dog lost much of the economic value it had in insular Southeast Asia, became a competitor for food, and was itself eaten to extinction in many locations and then reacquired from areas where it had managed to survive. Lynch (1991:427) suggests that the Polynesians reintroduced the dog to various parts of Island Melanesia in their back-migrations from the Polynesian Triangle, with variants of Polynesian terms for ‘dog’ appearing in Vanuatu, the Reefs and parts of the Solomons. Various Micronesian languages also appear to have borrowed from Polynesia.

It is unclear whether or not POc speakers had dogs. No dog remains have been found in early Western Lapita sites in the Bismarck archipelago. The earliest dog remains associated with the Oceanic dispersal are those found in late or post-Lapita assemblages in the Reef Is., dated to around 2600 BP (R. Green pers. comm.). No POc term for dog can be reconstructed with any confidence. A number of languages of Bougainville reflect a term reconstructable as *kasi, but we hesitate to associate this term with PAn *asu because the initial *k- is unexplained.⁵

Proto North Bougainville		*kasi	‘dog’
MM	Selau	kas	‘dog’
MM	Taiof	kas
MM	Hahon	kas	‘dog’
MM	Tinputz	keh	‘dog’
MM	Teop	kahi	‘dog’

cf. also:

MM	Solos	kus	‘dog’
MM	Petats	kus	‘dog’
SES	Longgu	ʔusu	‘dog’
SES	Sa’a	ʔusu	‘dog’

Donohue (1995:226) has identified forms resembling *kapuna that are present in Western Oceanic languages and in various languages of eastern Indonesia, both Austronesian and non-Austronesian. The Austronesian languages include Sangir kapuna (Sangiric, WMP), Geser kafuna (Seram, CMP) and Mor auna (WNG). He argues that ‘this (the presence of this term in eastern Indonesia), and the patterned distribution of the reflexes of *kapuna in Oceania point to an origin of the form in eastern Indonesia with spread to Oceania, and not the other way round.’

PWOc		*kapuna	‘dog’
NNG	Sissano (Arop)	aun	‘dog’
NNG	Uvol	kaune	‘dog’
NNG	Kaiwa	uvun	‘dog’
MM	Notsi	kapuna	‘dog’
MM	Tabar	kapune	‘dog’

Compare also the following which point to a very similar form in PNNG *gavun:⁶

PNNG		*gavun	‘dog’
NNG	Tuam	gavun	‘dog’
NNG	Malai	gavun	‘dog’
NNG	Kilenge	(na)ɣaune	‘dog’
NNG	Lukep	gaunu	‘dog’
NNG	Gedaged	gaun	‘dog’

The following series offers some support for reconstructing PWOc *ka(u,w)ak, but there is probably onomatopoeic influence.

PWOc		*ka(u,w)ak	‘dog’
NNG	Kove	kawa, kauwa	‘dog’
NNG	Bariai	kaua	‘dog’
PT	Kilivila	kaukʷa	‘dog’
PT	Muyuw	awukʷ	‘dog’
MM	Bali	kauaka	‘dog’
MM	Harua	kaua	‘dog’
MM	Lavongai	kauek	‘dog’

cf. also:

MM	Torau	kau-kau	‘dog’
MM	Mono-Alu	ʔau-ʔau	‘dog’
MM	Roviana	ka-kau(la)	‘indigenous dog, gen. reddish brown. Always howls, never barks’
SES	Gela	kau	‘dog’
SES	Birao	kau	‘dog’
SES	Tolo	kau	‘dog’
SES	Lengo	kau	‘dog’

The two reflexes in the following set are too close geographically to exclude borrowing, possibly at post-POc dates.

PWOc		*paia	‘dog’
NNG	Mengen	paia	‘dog’
MM	Nakanai	paia	‘dog’

cf. also:

Adm	Mussau	paua	‘dog’
Adm	Seimat	paua	‘dog’
SES	West Guadalcanal	pai	‘dog’
SES	Ghari	pai	‘dog’
SES	Malango	pai	‘dog’

PPn *kuli ‘dog’ is well-attested but has no secure non-Polynesian cognates. All apparent cognates in languages of Melanesia and Micronesia are almost certainly borrowings from Polynesian sources. A sample of such borrowed forms, all glossed ‘dog’, is listed below. Although a term from Mota is included, Codrington and Palmer note in their 1896 dictionary that there were no dogs on Banks Islands at first contact.

NCV	Mota	kurut	‘dog’
NCV	Paamese	hul	‘dog’
NCV	Lewo	kuli	‘dog’
NCV	Raga	vʷiriu	‘dog’
SV	Anejom̃	kuri	‘dog’
SV	Lenakel	kuri	‘dog’
Mic	Kiribati	kiri	‘dog’
Mic	Marshallese	kiṛiw	‘dog’
Mic	Ponapean	kiti	‘dog’

4. Aquatic mammals⇫

The aquatic mammals of tropical Oceania consist of various cetaceans (whales and dolphins) and the dugong. In Oceanic languages these animals are often classified as kinds of fish (POc *ikan, cf ch.8, §2).

4.1. Cetaceans⇫

4.1.1. Whales⇫

POc speakers probably occasionally sighted certain species of whales. Few surveys appear to have been done of whale populations in Melanesia but sperm whales, Physeter macrocephalus, and a number of smaller whales, in particular pilot whales, Globicephala melaena, and orcas or killer whales, Orcinus orca, have been reported in New Guinea waters. Pilot whales are more visible, being highly gregarious and occurring in large schools. Large numbers of pilot whales have been reported in Astrolabe Bay, while orcas have been sighted in Kimbe Bay off the north coast of New Britain (website reportings, 2003). Whales were not hunted by Oceanic communities in pre-contact times but when strandings occurred communities made use of the meat and teeth.

Reflexes of *rato ‘whale’ are well represented in Micronesian languages. Elsewhere probable reflexes are known only in Nyindrou and Wuvulu in the Admiralties. Borrowing cannot be ruled out.

POc		*rato	‘whale’
Adm	Nyindrou	rak	‘whale’
Adm	Wuvulu	xaʔo	‘whale’
Mic	Kiribati	ato
Mic	Marshallese	ṛac	‘whale’
Mic	Woleaian	zaso	‘whale’
Mic	Mokilese	rɔc	‘whale, porpoise, any sea mammal’
Mic	Ponapean	rac̣	‘whale’

There is a widespread tendency in Oceanic languages to refer to whales by terms that otherwise describe the action of blowing something out or spraying or spurting, e.g. Teop kavuhu ‘to spray’, kavu-kavuhu ‘whale’. POc *puRa-puRas ‘whale’, which is attested in NNG, PT and SES, is evidently an extension of *puRas ‘spray water from the mouth’. It has as antecedents PAn *buReS, PMP *buRah ‘spray water from the mouth’ (ACD).

POc		*puRa-puRas	‘whale’
NNG	Lukep	puru	(-u by regular assimilation)
NNG	Bing	fur-fūr	‘whale’
PT	Molima	wula-wula	‘whale’
PT	Tawala	ula-ulasi	‘whale’
PT	Diodio	ula-ula	‘dugong’
PT	Sudest	ura-urasi	‘whale’
PT	Misima	ula-ulas	‘whale (generic)’
SES	Bugotu	vua-vula	‘whale’

cf. also:

NCV

Ambae

vure(gi)

‘whale’

PPn *taf(o,u)raqa ‘whale’ is well attested. Possible cognates occur in several non-Polynesian languages of central and southern Vanuatu, but these forms are probably borrowings from the Polynesian Outlier speech communities in these regions.

PPn		*taf(o,u)raqa	‘whale’ (pollex)
Pn	Tongan	tofuaʔa	‘whale’
Pn	Niuean	tafuā	‘whale’
Pn	East Uvean	tafolaʔa	‘whale’
Pn	Pukapukan	tawolā	‘whale’
Pn	Rennellese	tahogaʔa	‘whale’
Pn	Samoan	tafolā	‘whale’
Pn	Tuvalu	tafolā	‘whale’
Pn	Kapingamarangi	toholā	‘whale’
Pn	Tikopia	taforā	‘cetacean, including whale, porpoise, dolphin’
Pn	West Futunan	tafora	‘whale’
Pn	Hawaiian	koholā	‘whale’

cf. also:

NCV	Namakir	taviraʔ	‘whale’
NCV	South Efate	tafra	‘whale’
SV	Sye	tovura	‘whale’
SV	Kwamera	təfra	‘whale’
Fij	Wayan	tavuto	‘generic for all large whales’
Fij	Bauan	tavuto	‘sperm whale, whose teeth furnish the tabua’

4.1.2. Dolphins (Delphinidae)⇫

Several species of dolphins, including Tursiops truncatus, the bottlenose dolphin, Stenella longirostris, the spinner dolphin and Grampus griseus, Risso’s dolphin, are found in Oceania (Dawbin 1972: 270-276). Wordlists and dictionaries of Oceanic languages give only generic terms for ‘dolphin’. The synonym ‘porpoise’ is commonly used in glosses.

POc *ku[r,R]iap, with final *-ap, is indicated by reflexes in North New Guinea, Papuan Tip and Meso-Melanesian languages and by kulef ‘dolphin’ in an external witness, Sawei, an Eastern Malayo-Polynesian language of South Halmahera. Southeast Solomonic and Vanuatu reflexes point to PEOc *giRio(p). Admiralties and SES languages reflect medial *-r-, while reflexes from NCV indicate *-R-. Other reflexes are compatible with either.

PEMP		*kuriap	‘dolphin’
POc		*ku(r,R)iap	‘dolphin’
Adm	Nyindrou	wiri	‘dolphin’
Adm	Lou	wiri	‘dolphin’
Adm	Lenkau	uhi	‘dolphin’
Adm	Loniu	wiʔɛ	‘dolphin’
NNG	Mangap	kuri	‘dolphin’
NNG	Lukep	kuri	‘dolphin’
NNG	Mengen	kule	‘dugong’
NNG	Uvol	ulia	‘dugong’
NNG	Numbami	kuliawa	‘dolphin’
NNG	Mengen	kulia	‘dolphin’
PT	Dobu	kuli-kuliya	‘dolphin’
PT	Tawala	uli-uliyawa	‘dolphin’
PT	Kilivila	kʷalivia	‘dolphin’
MM	Tolai	kuriap	‘dolphin’
MM	Ramoaaina	kuriap	‘dolphin’
MM	Siar	ku-kuriap	‘dolphin’
SES	Bugotu	kirio	‘dolphin’
SES	Sa’a	ʔirio	‘dolphin’
SES	Kwaio	kilio	‘dolphin’
SES	Lau	kirio	‘dolphin’
SES	Bauro	kirio	‘dolphin’
SES	Arosi	girio	‘dolphin’
PNCV		*guRio	‘dolphin (generic)’7
NCV	Hiw	kʷʟi	‘dolphin’
NCV	Loh	kʷuriə	‘dolphin’
NCV	Mota	kio	‘whale’
NCV	Ambae	kio	‘dolphin’
NCV	Raga	gio	‘dolphin’
NCV	Lewo	(le)kio-kio	‘dolphin’
NCV	Shark Bay	kuo	‘dolphin’
NCV	Namakir	qio-qi	‘dolphin’
NCV	Nguna	giio	‘dolphin’
PMic		*kua	‘dolphin’8
Mic	Kiribati	te-kua	‘dolphin, whale’
Mic	Marshallese	ke	‘dolphin’
Mic	Woleaian	xüw	‘dolphin’
Mic	Puluwatese	kūw	‘porpoises, believed to be whales when fully grown’

cf. also:

SV	Sye	(so)furwav(oh)	‘dolphin’9
SV	Ura	(so)furwav(os)	‘dolphin’
Fij	Bauan	giō	‘shark (generic)’
Fij	Kadavu	giō	‘shark (generic)’

4.2. Dugong (Dugong dugon)⇫

Dugongs are large aquatic mammals, with paddle-like forelimbs, no hindlimbs and a fluked, dolphin-like tail. They graze on beds of seagrass and are found in shallow coastal waters throughout the Indo-Malaysian archipelago, New Guinea and nearby islands including the Admiralties and Bismarcks, and in the Solomons, Vanuatu and New Caledonia, but not in Fiji or places further east (Chambers et al. 1989:1). Dugongs are prized by Pacific Islanders for their flesh, oil and small ivory tusks. Spears and nets are the main methods of capture. In the Trobriands and Manus very strong nets made from pandanus leaves are used while in the Siassi islands nets are made from the bark of a tree (Pyne 1972:276).

There is a well attested PMP etymon, *duyuŋ ‘dugong’. While most Oceanic reflexes point to *rui, evidence that POc continued PMP *duyuŋ as *ruyuŋ survives in the Tiang (New Ireland) reflex, which preserves the final velar nasal, and in Marovo (New Georgia), which retains a nasal, though this is m for expected *-ŋ.

PMP		*duyuŋ	‘dugong’ (Blust 2002)
POc		*ruyuŋ	‘dugong’
Adm	Pak	hū	‘dugong’
Adm	Bipi	xū	‘dugong’
Adm	Nauna	tuy	‘dugong’
Adm	Drehet	kxu	‘dugong’
NNG	Kove	lui	‘dugong’
NNG	Bariai	rui	‘dugong’
NNG	Gitua	rui	‘dugong’
NNG	Bing	rui	‘dugong’
NNG	Malasanga	rui-rui	‘dugong’
NNG	Sio	(gu)rui	‘dugong’
NNG	Lamogai	dū	‘dugong’
NNG	Takia	rui	‘dugong’
PT	Iduna	lui	‘dugong’
PT	Motu	rui	‘dugong’
PT	Magori	rui	‘dugong’
MM	Bulu	rui	‘dugong’
MM	Tiang	iuŋ	‘dugong’
MM	Marovo	rumu	(-m- for †-ŋ-)

cf. also:

NNG	Kaiep	kuoi	‘dugong’
NNG	Kairiru	qui	‘dugong’
MM	Babatana	jui	‘dugong’

No reflexes of POc *ruyuŋ have been found in Eastern Oceanic languages. For Proto Southeast Solomonic, *paipu ‘dugong’ can be reconstructed. For Proto North Central Vanuatu *boe-ni-tasi ‘dolphin’ (literally ‘pig of the sea’) has been reconstructed by Clark (2009).

5. Reptiles⇫

5.1. Marine turtles (Chelonidae)⇫

At least four large species of marine turtles are common to the Oceanic region: the hawksbill (Eretmochelys imbricatd), with its beak-like mouth, the green turtle (Chelonia mydas), with brown or black carapace, the loggerhead (Caretta caretta), with its very large head, and the leatherback (Dermochelys coriacea), distinguished from other turtles by its lack of a bony shell, its carapace being covered by thick leathery skin. The much smaller olive fidley (Lep- idochelys olivacea) spends almost all of its time in the open ocean and is less familiar to Pacific Islanders. Turtles are prized by Oceanic communities for their flesh and fat and (in the case of the hawksbill) for their shells.

As well as having a generic for ‘turtle’, a number of languages have distinct names for all four species and some distinguish male and female turtles, as the following examples show.

NNG	Lukep	pon	‘turtle’
NNG	Lukep	olol	‘hawksbill turtle’
NNG	Lukep	padodo	‘green turtle’
NNG	Lukep	kawadui	‘leatherback turtle’
MM	Kia	tege	‘turtle generic’
MM	Kia	tege ŋapo	‘hawksbill turtle’
MM	Kia	tege mogaha	‘green turtle’
MM	Kia	tege babaru	‘leatherback turtle’
MM	Kia	tege kakapodoko	‘loggerhead turtle’
MM	Marovo	vonu	‘turtle (generic)’
MM	Marovo	vonu pede	‘hawksbill turtle’
MM	Marovo	vonu ihana	‘green turtle’
MM	Marovo	kariatolu	‘leatherback turtle’
MM	Marovo	kakapodoko	‘loggerhead turtle’
SES	Sa’a	honu	‘turtle (generic)’
SES	Sa’a	honu hapa	‘hawksbill turtle’ (hapa ‘turtleshell’)
SES	Sa’a	honu iʔe	‘green turtle’ (iʔe ‘fish’)
SES	Sa’a	kule oŋo	‘leatherback turtle’ (kule ‘shore’, oŋo ‘mangrove’)
SES	Arosi	honu	‘turtle (generic)’
SES	Arosi	honu hapa	‘hawksbill turtle’
SES	Arosi	honu hau	‘soft-shelled turtle’ (?)
SES	Arosi	honu marabona	‘large sp. of turtle’
SES	Arosi	honu orobiu	‘leatherback turtle’
SES	Arosi	honu ʔeʔero	‘female turtle’
SES	Arosi	honu maraw	‘male turtle’
Mic	Chuukese	wīn	‘generic for sea turtles’
Mic	Chuukese	māw	‘hawksbill turtle’
Mic	Chuukese	winic̣c̣en	‘hawksbill turtle’
Mic	Chuukese	winimʷen	‘green turtle’
Mic	Chuukese	woŋ	‘leatherback turtle’
Fij	Wayan	ika-bula	‘turtle generic’
Fij	Wayan	taku	‘hawksbill turtle’
Fij	Wayan	ika-dū	‘green turtle’
Fij	Wayan	tosiviwalu	‘leatherback turtle’
Fij	Wayan	tū-vonu	‘loggerhead turtle’
Fij	Wayan	bala	‘male turtle’
Pn	Tikopia	fonu	‘turtle (generic)’
Pn	Tikopia	fonu koroa	‘hawksbill turtle’ (koroa ‘goods, treasure’)
Pn	Tikopia	fonu tea	‘green turtle’ (tea ‘white, light-coloured’)

Three terms for turtle taxa are attributable to POc. POc *poñu, continuing a PAn etymon, was clearly the generic term, and is very well supported.

PAn		*peñu	‘turtle’ (ACD)
POc		*poñu	‘turtle (generic)’
Adm	Mussau	onu	‘turtle (generic)’
Adm	Wuvulu	fonu	‘large hard-shelled sea turtle’
Adm	Aua	fonu	‘large hard-shelled sea turtle’
Adm	Lou	puon	‘sea turtle’
Adm	Loniu	poñ	‘sea turtle’
NNG	Manam	poŋ	‘turtle’
NNG	Lukep	pon	‘turtle, Dermochelidae’
NNG	Gedaged	fon	‘leatherback turtle’
NNG	Kove	ponu	‘turtle’
PT	Tubetube	wonu	‘turtle’
PT	Kilivila	wonu	‘turtle’
MM	Tigak	un	‘turtle’
MM	East Kara	fun	‘turtle’
MM	Nalik	fun	‘turtle’
MM	Tabar	vo	‘turtle’
MM	Lihir	on	‘turtle’
MM	Sursurunga	hun	‘turtle’
MM	Tangga	fon	‘turtle’
MM	Patpatar	hun	‘turtle’
MM	Nehan	won	‘turtle’
MM	Marovo	vonu	‘turtle (generic)’
MM	Taiof	fon	‘turtle’
MM	Banoni	ɣom	‘turtle’
MM	Babatana	vunu	‘turtle’
MM	Lungga	voñu	‘turtle’
SES	Bugotu	voñu	‘turtle’
SES	Gela	vonu	‘turtle (generic) for hawksbill and green turtle’ (Foale)
SES	Lau	fonu	‘turtle’
SES	Kwaio	fonu	‘turtle’
SES	Sa’a	honu	‘turtle’
SES	Arosi	honu	‘turtle’
NCal	Jawe	pʷen	‘turtle’
NCal	Nemi	pʷen	‘turtle’
NCal	Xârâcùù	pʷê	‘turtle’
Mic	Kiribati	on	‘turtle’
Mic	Chuukese	woŋ	‘leatherback turtle’
Mic	Carolinian	wōŋ	‘turtle’
Mic	Woleaian	woŋ	‘turtle’
Mic	Ulithian	wōl	‘turtle’
Fij	Bauan	vonu	‘turtle (generic)’
Fij	Wayan	(tu)vonu	‘loggerhead turtle’
Pn	Tongan	fonu	‘turtle (generic)’
Pn	Niuean	fonu	‘turtle’
Pn	Rennellese	honu	‘green turtle’
Pn	East Futunan	fonu	‘turtle’
Pn	Tikopia	fonu	‘turtle (generic)’
Pn	Hawaiian	honu	‘turtle’

The shell of the hawksbill turtle is prized, being used to make ornaments like earrings and bracelets as well as bonito hook tips and cutting tools. In Tonga and E. Futuna this turtle is called fonu koloa, in Tikopia fonu koroa, the term for ‘property, treasure’ being used to identify the turtle with valuable shell.

A term for the hawksbill turtle, *keRaŋ is well supported. In Oceanic languages, the beaked profile of the hawksbill turtle is identified with that of a parrot rather than a hawk. In Lou (Admiralties), the same term refers to both the turtle and a red parrot. In Maori the term survives only for the parrot. (For discussion of the atypical /e/ vowel, see Blust 1983-84:8-10). All the Admiralties terms have problematic reflexes of *R and all show vowel metathesis.

PCEMP		*keRaŋ	‘hawksbill turtle’ (Blust 2002)
POc		*keRaŋ	‘hawksbill turtle, Eretmochelys imbricata’
Adm	Lou	kareŋ	‘hawksbill turtle; red parrot’
Adm	Lenkau	kareŋ	‘hawksbill turtle’
Adm	Pak	kareŋ	‘hawksbill turtle’
Adm	Titan	karéŋ	‘parrot turtle’
Adm	Andra	kareŋ	‘hawksbill turtle’
PT	Motu	era	‘turtle, shell turtle’
Mic	Kiribati	(tabʷa)kea	‘parrotbill turtle, brilliant shell’
Mic	Ponapean	(sapʷa)ke	‘hawksbill turtle’
PPn		*kea	‘hawksbill turtle’
Pn	Pukapukan	kea	‘k.o. turtle’
Pn	Nanumea	kea	‘a small turtle’
Pn	Rennellese	kea	‘qualifier for small turtles’
Pn	Māori	kea	‘parrot sp.’
Pn	Hawaiian	ʔea	‘hawksbill turtle; also term for its shell’

The green turtle is the most abundant and widespread turtle of tropical Oceania and the one most valued for food. It may weigh up to 200 kg, with carapace length around 150 cm. A POc term is reconstructable but reflexes have been found only in a few North New Guinea and Micronesian languages.

POc		*manak(V)	‘green turtle, Chelonia mydas’
NNG	Takia	manak	‘green turtle’
NNG	Gedaged	manak	‘green turtle’
Mic	Carolinian	mʷēl	‘turtle with thin shell, good for eating’
Mic	Chuukese	(wini)mʷən	‘green turtle, eaten’
Mic	Puluwatese	mʷol	‘green turtle, commonly caught and distributed to entire island’

The leatherback turtle, Dermochelys coriacea, is the largest turtle in the world with an average carapace length of two metres and average weight of around 360 kg. It nests widely along the north coast of New Guinea, Manus, New Britain and Bougainville, but always in very low densities. It is generally not favoured for eating because of its very oily and smelly meat (Mueller 1985:195). No widespread cognate set has been found.

5.2. Crocodiles (Crocodylidae)⇫

The archaeologist Simon Best has published a comprehensive review of the distribution of crocodiles in the Pacific (Best 1988). Two species occur in Melanesia: Crocodylus porosus, the estuarine or saltwater crocodile and Crocodylus novaeguineae. The latter is found in freshwater swamps and lakes in lowland New Guinea. The normal (breeding) range of Crocodylus porosus extends from south-west India to the Solomons, and from the Tropic of Capricorn in Australia to the Philippines. The easternmost limit for breeding populations is the Santa Cruz group. Individual animals are capable of very long journeys, over 1000 km, and have been reliably recorded in northern Vanuatu, Fiji and Kiribati, but these occasional occurrences do not represent an extension of their breeding range.

POc *puqaya ‘crocodile’ continues PMP *buqaya and, probably, PAn *buqaya.¹⁰ Reflexes of *puqaya are found widely in languages of the Bismarck Archipelago, New Guinea and the main Solomons group, i.e. where crocodiles are common, and are also present in a few languages of Vanuatu. No reflexes have been recorded in Micronesian or Central Pacific languages.

PAn		*buqaya	‘crocodile’ (ACD)
POc		*puqaya	‘crocodile’
Adm	Mussau	ua	‘crocodile’
Adm	Nyindrou	buwa	‘crocodile’
Adm	Lou	pua	‘crocodile’
Adm	Aua	fuara	‘crocodile’
NNG	Kilenge	na-pua	‘crocodile’
NNG	Malasanga	pua	‘crocodile’
NNG	Roinji	puɣaia	‘crocodile’
NNG	Adzera	fugai	‘crocodile’
PT	Arifama	uaya	‘crocodile’
PT	Dobu	uaya	‘crocodile’
PT	Hula	vuɣa	‘crocodile’
PT	Motu	huala	‘crocodile’
MM	Vitu	vuɣa	‘crocodile’
MM	Lavongai	uaia	‘crocodile’
MM	Patpatar	huaia	‘crocodile’
MM	Petats	hua	‘crocodile’
MM	Marovo	vua	‘crocodile’
SES	Gela	vua	‘crocodile’
SES	Lau	fuasa	‘crocodile’
SES	’Are’are	huara	‘crocodile’
SES	Sa’a	huasa	‘crocodile’
NCV	Mota	via
NCV	Mwotlap	ne-ve	‘crocodile’

5.3. Lizards⇫

5.3.1. Introduction⇫

Lizards are discussed here under the following headings: monitors (§5.3.2), iguanas (§5.3.3), geckos (§5.3.4), skinks (§5.3.5) and indeterminate (§5.3.6), but there is frequent transfer of name within cognate sets from one family to another. Monitor lizards and iguanas belong to different families, although these names are used interchangeably in some wordlists. The Australian goanna belongs with the monitor lizards of Indonesia, New Guinea and the western Pacific. The iguana is of South American origin, and is found in the Pacific today only in the Galapagos Islands, Fiji, Tonga, Wallis and Futuna. In many sources the gloss ‘iguana’ is incorrectly applied to monitor lizards.

5.3.2. Monitors (Varanidae)⇫

Nearly a score of different species of monitors are to be found in the Indo-Pacific region. The best known, because of its great size, is the komodo monitor, a species confined to the Lesser Sundas. The most widespread species in Oceania is Varanus indicus, a water monitor, valued as an important food source. Its range extends from Sulawesi through to the Aru and Kei Islands, the New Guinea land mass including the Schouten Islands, the D’Entrecasteaux and Trobriand Islands, New Britain, New Ireland, the Admiralties, the Solomons and northwards to the Carolines, Marianas and Marshalls (Loveridge 1946:60-64, de Rooij 1970:149). Water monitors can grow to over two metres in length, with the tail about one and a half times as long as the head and body. They are excellent swimmers, and have been seen swimming far out to sea.

De Rooij lists five other Varanus species in New Guinea (V. salvadorii, V. kalabeck, V. gouldi, V. prasinus, V. kordensis), but none are found outside the main island except for the last-named which is found also on Aru. The fact that Arosi has two terms for monitor lizard (incorrectly glossed ‘iguana’ by Fox), haha, ‘the larger’, ʔaraʔu, ‘the smaller’, may indicate sub-species or a distinction based on sex or age. The skin of the larger monitors is used in New Guinea to cover the end of hourglass drums. The POc term *ka(r,l)(i,u)pa(s) probably referred to large monitor lizards.

POc		*ka(r,l)(i,u)pa(s)	‘large monitor lizard, Varanus indicus and possibly other species’
PT	Muyuw	kalevas	‘tree goanna, big lizard’
PT	Motu	ariha	‘large lizard, iguana’ (= monitor; r for *l unexpected)
SES	Tolo	kolahe	‘k.o. lizard’ (vowel metathesis)
SES	Lau	kurafe	‘k.o. large lizard with smooth shining body’ (vowel metathesis)
PMic		*kalufa	‘monitor lizard’
Mic	Chuukese	konuf	‘Varanus indicus, monitor lizard’
Mic	Mortlockese	kaluf	‘lizard’
Mic	Puluwatese	kólūf	‘giant lizard’
Mic	Satawalese	kaluf	‘lizard’
Mic	Carolinian	xalūf	‘black and green forest lizard: gecko’
Mic	Woleaian	xarüfa	‘k.o. big lizard’

cf. also:

SES

To’aba’ita

gʷalu(aʔa)

‘green-blooded skink, Prasinohaema virens’

There is a second, more weakly supported POc reconstruction, *wak(i,e) ‘k.o. lizard, possibly monitor lizard’. It may be a contraction of PMP *bayawak ‘monitor lizard’ (Blust 2002).

POc		*wak(i,e)	‘k.o. lizard, possibly a monitor’
Adm	Lou	wak	‘monitor lizard, Varanus sp.’
PT	Minaveha	wake(neta)	‘lizard, rough-skinned’
PT	Dobu	waki-waki	‘small gecko’
PT	Dobu	waki(balu)	‘large gecko’
SES	’Are’are	weʔe	‘k.o. big lizard that eats the inside of the nari nut’

Reflexes of the next reconstruction, POc *(k,g)uma, form a compound in many PT languages with reflexes of POc *ka(r,R)ak ‘creep, crawl’.

POc		*(k,g)uma	‘monitor lizard’
PPT		*(k,g)uma-kara	‘monitor lizard’
PT	Ubir	umar	‘lizard, general term’
PT	Gapapaiwa	kuma-kara	‘lizard (monitor type?); the skin of this lizard is used for drum heads’
PT	Saliba	kuma-kala	‘k.o. lizard’
PT	Minaveha	kuma-kana	‘lizard which lives near ocean’
PT	Minaveha	um-ana	‘larger lizards, speckled or striped, with smooth skin’
PT	Dobu	goma-ʔara	‘monitor lizard’
PT	Wedau	umara	‘k.o. lizard’
PT	Tawala	umʷala	‘iguana, large black lizard’
PT	Iduna	uma-ʔala	‘goanna; goanna skin of drum’
PT	Galea	guma-ʔala	‘monitor lizard’
PT	Molima	goma-ʔala	‘monitor lizard’
MM	Tolai	gom	‘k.o. small lizard’
Mic	Woleaian	xumʷa(ṣere)	‘k.o. small lizard’ (ṣere ‘gruel’)

PWOc		*paRia	‘k.o. big lizard’
NNG	Manam	paria	‘k.o. big lizard’
NNG	Bariai	paria	‘lizard (generic)’
NNG	Kove	pahia	‘monitor lizard’
NNG	Sengseng	a-pahiya	‘monitor lizard’
NNG	Lukep	pari	‘large ground lizard’
NNG	Gedaged	paɬiu	‘k.o. lizard, edible, skin used as drum head. Gets to be seven feet long’
PT	Motu	aria	‘lizard (generic); monitor lizard’ (for †haria)
MM	Roviana	vari(lazu)	‘k.o. lizard’

cf. also:

MM

Nakanai

vareki

‘monitor lizard, skin used for drum-head’

There is a well supported Proto Admiralties reconstruction for monitor lizard.

PAdm		*watiRi	‘monitor lizard, Varanus spp.’
Adm	Wuvulu	waʔi	‘monitor lizard, Varanus spp.’
Adm	Seimat	wat	‘monitor lizard, Varanus spp.’
Adm	Penchal	wati	‘monitor lizard, Varanus spp.’
Adm	Nali	wariy	‘monitor lizard, Varanus spp.’
Adm	Lele	watiy	‘monitor lizard, Varanus spp.’
Adm	Titan	wati	‘monitor lizard, Varanus spp.’
Adm	Loniu	wati	‘lizard’
Adm	Bipi	wati	‘monitor lizard, Varanus spp.’

5.3.3. Iguanas (Iguanidae)⇫

True iguanas, in contrast to monitor lizards, would not have been known to speakers of POc. Iguanas (Brachylophus spp.) are present in the Oceanic region only in Fiji, Tonga, Wallis and Futuna. They are of South American origin and probably arrived in the central Pacific on floating vegetation. They are diurnal and strongly arboreal in behaviour. A Proto Central Pacific name for ‘iguana’, *vokai, is reconstructable from cognates in Fijian and Polynesian languages. No non-CP cognates have been recorded. In Rennellese the reflex of *vokai is applied to monitor lizards. The Tongan term identified in the dictionary as ‘chameleon’ probably refers to a lizard with chameleon-like ability to change colour; true chameleons are natives of Africa. Paddy Ryan reports that Brachylophus vitiensis, the crested iguana, has the ability to change colour when frightened or angry and will turn a very dark green, almost black, within five minutes (Ryan 1988:106).

PCP		*vokai	‘k.o. lizard’
Fij	Bauan	vokai	‘banded iguana (Brachylophus fascialus) and crested iguana (Brachylophus vitiensis)’
Fij	Wayan	vokai	‘crested iguana’
Pn	Tongan	fokai	‘chameleon’
Pn	Rennellese	hokai	‘general name for monitor lizards, Varanus indicus’

cf. also:

NCV

Tamambo

vohare

‘k.o. lizard’

5.3.4. Geckos (Gekkonidae)⇫

Geckos are widely distributed across mainland Southeast Asia, Indonesia and the tropical Pacific. They are one of the few reptiles with voices; they make small chirruping noises. They are mostly nocturnal, and many species frequent buildings, although others are forest dwellers. Most are less than 15cm long, and, in common with skinks, have the ability to discard their tails when threatened. They are generally welcomed by households, both for their efforts in keeping down moths, mosquitos, cockroaches and other household pests, and because they are widely regarded as good omens (Loveridge 1946:73).

A number of Western Malayo-Polynesian terms for gecko are onomatopoeic, typically reflexes of PMP *tektek. Although there are numerous gecko species throughout the Oceanic region, only one term which might be considered a reflex of PMP *tektek has been noted.

PMP		*tektek	‘gecko, house lizard (onom.)’ (doublet *cekcek) (ACD)
POc		*tok	‘gecko’
NNG	Hote	(kamu)tok	‘small gecko, found on the wall’

We have reconstructed POc *bili ‘k.o. lizard’, which may be an irregular reflex of a PAn term, *bulilik ‘lizard sp.’ (Blust 2002), but on present evidence the connection is uncertain. Its reflexes in Polynesian languages carry two meanings, ‘gecko’ (or sometimes ‘lizard (generic)’) and ‘adhere to’, pointing to a lizard name that reflects its character.

Some of the Polynesian reflexes may be compounded with the generic term for lizard, PPn *moko, from POc *mo(g,k)o ‘k.o. snake’ (see §5.4.1).

POc		*bili	‘k.o. lizard’
MM	Tolai	bil-bil	‘k.o. lizard’
PPn		*pili	‘gecko; adhere to’ (pollex)
Pn	Tongan	pili	‘k.o. bluish lizard’
Pn	East Futunan	pili(moko)	‘gecko’
Pn	Pukapukan	(ŋō)pili	‘gecko’
Pn	Samoan	pili	‘lizard’
Pn	Takuu	(moko)pili	‘small house lizard’

cf. also:

Pn

Tikopia

(moko)pili

‘gecko’ (for †piri)

5.3.5. Skinks (Scincidae)⇫

Most skinks are slim lizards capable of moving at high speed, although the family includes the more heavily built blue-tongue lizards. Skinks usually emerge during the day to feed and bask in the sun. The New Guinea blue-tongue skink, Tiliqua gigas, is widely distributed across eastern Indonesia, the New Guinea mainland, the Admiralty islands and Bismarck Archipelago (http://www.anapsid.org/bluetong.html). Various Emoia species are widely distributed through New Guinea and the Pacific. Contemporary languages often distinguish a number of skink taxa but no POc reconstructions that refer specifically to skinks are supported. However, John Lynch (pers. comm.) has proposed a reconstruction for Proto Southern Oceanic, based on cognates in NCV and New Caledonia, and there is evidence for a second term attributable to Proto Eastern Oceanic.

PSOc		*gala	‘green lizard, Emoia sp.’ (John Lynch, pers. comm.)
NCV	Mota	kala	‘small blue lizard’
NCV	Dorig	kal	‘green lizard’
NCV	Raga	ɣala	‘lizard, probably green skink, Emoia sanfordi’
NCV	Tamambo	kala	‘small green lizard, about 30 cm long, generally reviled.’
NCV	Ambae	kala	‘shiny green lizard w. long tail, grows to about 20cm, lives on coconut palms’
NCV	Lonwolwol	gal	‘green lizard’
NCV	Paamese (South)	a-kal	‘green lizard, Emoia sanfordi’
NCal	Nemi	gēna	‘lizard (generic)’

The presence of *kali in the following reconstruction raises the question as to whether this is an echo of the old qali-/kali- prefix that in pre-Oceanic times evidently meant something like ‘Warning. Take care!’ (See ch. 7 for detailed discussion.) Although the evidence is that the morpheme was no longer productive in Oceanic languages, an association of danger with a kind of iridescent lizard persists, at least in Tamambo (NCV) and Tikopia (Pn). Jauncey (pers. comm.) notes that in Tamambo they are generally reviled, people shuddering at the sight of them. Firth (1967:233) observes that Tikopia atua, which applies primarily to a supernatural entity, is extended as a category name to any natural species which should not be eaten. Among the latter he includes iridescent lizards.

PEOc		*kalis(i,u)	‘k.o. skink’
SES	Owa	ga-garisu	‘a brown skink’
PPn		*kalisi	‘skink, k.o. lizard’ (pollex)
Pn	Tongan	kalihe-lihe	‘k.o. lizard’
Pn	Rennellese	kagisi	‘skink (Emoia cyanura)’
Pn	Nanumea	kalihi	‘skink’
Pn	Kapingamarangi	galiha	‘skink’
Pn	Tikopia	kali-lisi	‘green lizard, small, iridescent’

5.3.6. Indeterminate kinds of lizards⇫

There is a weakly supported reconstruction based on an imperfect agreement between Southeast Solomonic languages and Tolai.

POc		*kuk(o,u)	‘k.o. lizard’
MM	Tolai	kuku	‘lizard (generic) name’ (kuku alira ‘small lizard w bronze tail’; also ‘green’)
SES	’Are’are	ʔuʔo	‘small lizard, green and gold’
SES	Sa’a	uʔo	‘house lizard, green and gold; first element in several lizard names’

A second weakly supported reconstruction rests on putative cognates in Mussau, Kokota and N Tanna.

POc		*kapoRu	‘k.o. lizard’
Adm	Mussau	kapou	‘monitor lizard, Varanus sp.’
MM	Kokota	kopro	‘gecko’
SV	North Tanna	kapʷer	‘lizard’ (-r for †∅) (Tryon 1995)

cf. also:

CB	Serui-Laut	kafei	‘house lizard’ (Smits 1992:239)
PT	Molima	kekepa	‘green tree skink’
MM	Nakanai	kavega	‘lizard’
MM	Tolai	keveu	‘k.o. iguana’ (monitor)

5.4. Snakes (sub-order Serpentes)⇫

5.4.1. Land snakes⇫

Around 100 species of land snakes occur in New Guinea, the Bismarck Archipelago and the Solomon Islands. They include pythons and boas (family Boidae), both of which constrict their prey, and a range of tree snakes and ground-dwelling snakes, some of which are highly venomous. The best known include the Papuan taipan, Oxyuranus scutellatus, the Papuan black snake, Pseudechis papuanus, and the death adder, Acanthophis antarcticus.

Only two species of snake have been reported for Vanuatu: the Flowerpot Snake, Typhlops braminus of Efate and the Pacific Boa, Candoia bibroni, also with restricted distribution (Medway and Marshall, quoted in Walsh et al. 2000:248). Fiji has two terrestrial snakes, the Pacific boa and the rare Ogmodon vitianiis, the latter found only in Fiji (Ryan 1988:111-114, Morrison 2003:90). In Polynesia land snakes are found only in Samoa, while the only land snake reported for Micronesia is the miniature blind snake found on Chuuk (Davis 1999:188). For PMP *nipay ‘snake’ only a single Oceanic reflex has been located.

PMP		*nipay	‘snake’ (ACD)
POc		*nipe	‘snake’
MM	Mono-Alu	nihe	‘snake’

POc *mʷata is the most widely reflected generic for ‘snake’ in Oceanic languages. It has no known non-Oceanic cognates, raising the possibility that this term entered Proto Oceanic from a non-Austronesian source.

POc		*mʷata	‘snake (generic)’ (Ross 1988)
Adm	Wuvulu	waʔa	‘snake (generic)’
Adm	Aua	waʔa	‘snake (generic)’
Adm	Loniu	mʷat
Adm	Nauna	mʷat	‘snake (generic)’
Adm	Bipi	mʷak	‘snake (generic)’
SJ	Ormu	mata-mata	‘snake (generic)’
SJ	Sobei	matu	‘snake (generic)’
NNG	Yabem	mɔaʔ	‘snake (generic)’
NNG	Manam	moata	‘snake (generic)’
NNG	Gedaged	mot	‘collective term for snakes and certain kinds of worms’
NNG	Kove	mota	‘snake (generic)’
NNG	Malasanga	mota	‘snake (generic)’
NNG	South Watut	mʷar	‘snake (generic)’
PT	Are	mota	‘snake (generic)’
PT	Tawala	mota	‘snake (generic)’
PT	Molima	mʷata	‘snake (generic)’
PT	Tubetube	mʷata	‘snake (generic)’
PT	Sudest	mʷata	‘snake (generic)’
MM	Bola	mata	‘snake (generic)’
MM	Bali	(gu)mata	‘snake (generic)’
MM	Tigak	mʷata	‘snake (generic)’
SES	Bugotu	umata	‘snake (generic)’
SES	Gela	umata	‘snake (generic)’
SES	Longgu	mʷa	‘snake (generic)’
SES	Lau	wa	‘snake (generic)’
SES	Kwaio	wā	‘snake; various kinds of worms; other snake-like creatures’
SES	Sa’a	mʷā	‘snake (generic)’
SES	Arosi	mʷā	‘snake (generic)’
NCV	Mota	mʷata	‘snake (generic)’
NCV	Mwotlap	na-mʷat	‘snake (generic)’
NCV	South Efate	mʷat	‘snake (generic)’
SV	Lenakel	mʷati(kalo)	‘worm’
Mic	Kiribati	mʷata	‘worm’
Mic	Woleaian	mʷata	‘worm’
Fij	Wayan	ŋʷata	‘snake (generic)’
Fij	Bauan	ŋata	‘snake (generic)’
Pn	Tongan	ŋata	‘snake’ (There are no snakes in Tonga, but Tongans know snakes exist in Fiji and Samoa.)
Pn	Niuean	ŋata	‘snake’ (Refers only to land snakes, even though there are none on Niue.)
Pn	Samoan	ŋata	‘snake (generic)’
Pn	Tokelauan	ŋata	‘snake (generic)’

The gloss for a third POc reconstruction, *mo(g,k)o ‘k.o. snake’, is supported by reflexes in three Meso-Melanesian languages and one Southeast Solomonic language. In Central Pacific languages the cognate form is applied to lizards in general.

POc		*mo(g,k)o	‘k.o. snake’
MM	Laghu	mogo	‘snake’
MM	Blablanga	mogo	‘snake’
MM	Maringe	mogo	‘snake (generic)’
SES	Gela	mogo	‘k.o. inert black snake’
Fij	Wayan	moko	‘lizard (generic)’
Fij	Bauan	moko	‘lizard’ (Capell); ‘gecko’ (Ryan)
PPn		*moko	‘lizard’ (pollex)
Pn	Tongan	moko	‘k.o. lizard, light brown’
Pn	Niuean	moko	‘generic for most creeping and crawling creatures’
Pn	Samoan	moʔo	‘gecko’
Pn	Rennellese	moko	‘geckos’
Pn	Tikopia	moko	‘lizard (generic); incl. black lizard (moko-uri), small gecko (moko-pili), large dark coloured gecko (moko-fata), crocodile (moko-toro)’
Pn	Hawaiian	moʔo	‘lizard; reptile of any kind’

cf. also:

PT

Saliba

(kenokeno)mugu

‘lazy snake’ (keno ‘lie down’)

The Ubir gloss of ‘cobra’ in the next set is wrong, as Asian cobras are not present in the New Guinea region (Cogger 1972a: 1013).

PWOc		*sibi	‘k.o. snake’
PT	Ubir	sib	‘cobra’
MM	Nakanai	sibi	‘small snake’
MM	Nakanai	si-sibi	‘red snake which turns black after death’

5.4.2. Sea snakes (Elapidae)⇫

Sea snakes are venomous elapid snakes that have adapted to living in the sea. Many are found in shallow coastal waters and estuaries. They have an eel-like appearance, with laterally compressed bodies and (in many species) paddle-like tails that aid swimming. Sea snakes are widespread in the Pacific, with one species, Pelamis platurus, the yellow-bellied sea snake, found across the entire Pacific region. Four species are known in Fiji, two of which, Laticauda colubrina and L. laticaudata, are semi-terrestrial (Ryan 1988:111-114).

Agreement between Central Pacific and Vanuatu languages allows reconstruction of Proto Remote Oceanic *takulasi or *takusali ‘sea snake’. Metathesis of consonants has occurred more than once in the final two syllables, making the original form uncertain.

PROc		*takulasi	‘sea snake’ (PSOc *ta(k,g)ulasi or *ta(k,g)usali ‘sea snake’; John Lynch, pers. comm.)
NCV	Loh	təɣlɛh	‘sea snake’
NCV	Lonwolwol	telsil	‘snake’
NCV	Paamese	tousali	‘sea snake’
SV	Sye	tuŋklah	‘sea snake’
SV	Ura	duŋlas	‘sea snake’
Fij	Wayan	da-dakulaði	‘Laticauda colubrina, black and white banded sea snake’
Fij	Bauan	dakulaði, da-dakulaði	‘black and white banded sea snake’
PPn		*takuhali	‘black and white banded sea snake’ (pollex)
Pn	Tongan	tukuhalo	‘sea snake with black and white bands’
Pn	Niuean	katuali	‘sea snake, banded black and white’ (metathesis; loss of *h unexpected)
Pn	East Futunan	takuali	‘sea snake, blue and white bands’
Pn	Tokelauan	takuali	‘k.o. very rare and small but most dangerous sea eel which attacks’

In the Banks languages of northern Vanuatu the sea snake, Laticauda semifasciata, is referred to by reflexes of *mʷai.

NCV	Mota	mʷai	‘sea snake’
NCV	Mwotlap	mʷe	‘sea snake characterised by black and white stripes; succubus, female demon, harmful’

6. Amphibians⇫

The only amphibians present in Oceania are frogs. Menzies (2006:3) writes that ‘at the latest count, New Guinea [including the Admiralties and Bismarck Archipelago] and the Solomon Islands have a total of 35 genera and 322 recognised species … and, despite lacking several frog families present in southeast Asia, the region remains one of the most species-rich in the world’. East of the Solomons the native frog population is almost negligible. Codrington (1891:17) notes that frogs are absent in Santa Cruz. An account by Medway and Marshall (quoted by Walsh et al. 2000:235) contains the following observation: ‘There are no amphibians native to the New Hebrides [Vanuatu], and none was taken by Dr Felix Speiser who collected herpetological material in these islands in 1910-12.’ Morrison (2003:11) reports that there are two endemic species of frogs in Fiji, the Fiji tree frog, Platymantis vitiensis, and the Fiji ground frog, P. vitianns. There are no endemic frogs in Micronesia, nor, with the exception of New Zealand, in Polynesia. New Zealand has four native species, all belonging to the genus Leiopelma, an ancient and primitive group of frogs.

Four families occur without human intervention in New Guinea, Australia and islands of the western Pacific (Menzies 2006:6): the Hylidae (mostly climbing frogs, found throughout New Guinea and the Solomons, and on Timor, Seram and Halmahera), Myobatrachidae (mainly ground-dwelling frogs, with distribution confined to Australia and New Guinea), Ranidae (pond frogs and some others, a vast family of almost world-wide occurrence), and Microhylidae, a wide variety of morphologically distinct types. Menzies (2006:172) writes that Microhylidae ‘is almost worldwide in distribution but reaches its greatest diversity in the tropics, and the origin of the New Guinean genera almost certainly lies with some currently unknown southeast Asian ancestor’. The families are unevenly distributed, with 89 percent of the New Guinea mainland population from just two families, Hylidae and Microhylidae, while in the Solomon Islands 92 percent are Ranidae. The occurrence of Bufonidae (toads) in Oceania is due to deliberate or accidental introduction in recent times. The native species occupy a wide range of habitats, many found well away from water provided humidity is adequate. A substantial number lack a tadpole stage, the young hatching on land. Complex topography and close juxtaposition of different habitats have facilitated species differentiation so that few species are widespread over the whole region. Most species for which adequate data exists have relatively restricted distribution.

Certain species of frog are eaten in parts of Melanesia, but we have very limited information as to which Oceanic communities eat frogs.

Although onomatopoeia plays a large part in naming kinds of frogs, the picture is complicated by the fact that different species may have very distinctive calls. Mussau ropa-ropa, Lou rek-rek, Dobu kʷe-kʷe, Yabem kʷeŋ-kʷeŋ, Takia waŋ-waŋ, Molima bʷe-bʷe, Tolai rok-rok are all frog names presumably reflecting particular calls. Only the males utter the characteristic species-specific calls.

None of the glosses for the several reconstructions proposed here are at all specific. It seems that wordlist compilers are generally content to collect one or two terms, probably generic or perhaps identifying a well-known frog of the area, but few sources give more than minimal details. However, Kalam speakers of the Central Highlands (a non-Austronesian language of the Trans New Guinea family) distinguish 35 different taxa (Bulmer & Tyler 1968), and it is probable that many Oceanic communities name more than have been recorded. In any event, as most species have limited distribution, species names are likely to be local innovations.

PMP *palakaq ‘frog’ (ACD: PWMP) has possible cognates in a few Oceanic languages, although correspondences are irregular.

POc		*pa(R,r)a(C)	‘frog’
NNG	Bing	far	‘frog (generic)’
PT	Ouma	(a)para-para	‘frog’
PT	Hula	pala(keo)	‘frog’
PT	Balawaia	fara(keo)	‘big frog’
PT	Motu	para-para	‘frog’
MM	Vangunu	para(goa)	‘large frog’
MM	Marovo	para(goa)	‘frogs (generic)’

cf. also:

MM	Nakanai	pere	‘small frog’
MM	Tolai	per	‘frog’
SES	Arosi	haro	‘tadpole’
SES	Arosi	pʷaro(ʔama-ʔama)	‘k.o. large frog’

The term *kuRau, for a kind of frog, is attested by reflexes in a single Sarmi Coast language and in Southeast Solomonic, with possible cognates in a number of other groups.

POc		*kuRau	‘frog taxon’
SJ	Ormu	kurau	(Smits 1992:896)
SES	Bugotu	kulau	‘tree frog’
SES	Gela	kulau	‘a small frog’
SES	Lau	ʔulao, ʔulou	‘k.o. very large frog, Rana guppyi’
SES	Bauro	urau	‘frog’ (initial k- or ʔ- expected)

cf. also:

NNG	Manam	kuru(basi)	‘k.o. green tree frog’
PT	Mekeo	kulua	‘k.o. tree frog’
PT	Sudest	kula	‘small frog’
PT	Balawaia	kura(fege)	(fege ‘small frog’)
MM	Halia	kura	‘frog, lives in dry places’
MM	Tinputz	kura	‘frog, Ranid’
MM	Kia	kule	‘frog’
MM	Nduke	(pa)karau	‘frog’
MM	Lungga	(ba)karau	‘frog’
MM	Roviana	(ba)karao	‘frog’
SES	Bugotu	kui-kuli	‘frog’
SES	Arosi	wairau	‘frog’

The following set is almost certainly onomatopoeic.

PNNG		*karok	‘frog’
NNG	Gedaged	gazok	‘various kinds of frog’
NNG	Kaulong	kahok	‘frog’
NNG	Sengseng	kahok	‘frog’
NNG	Kove	kalo	‘frog’
NNG	Lusi	kalo	‘frog’

cf. also:

PT	Saliba	kʷalu-kʷalu	‘frog’
PT	Wedau	qeru	‘frog’
MM	Nakanai	karu-karu	‘frog, not eaten’
MM	Halia	karou	‘frog, stays in fresh water’

An additional POc reconstruction, *kʷekʷe, is indicated by agreements between Admiralties, North New Guinea, Papuan Tip and Southeast Solomonic witnesses. However, the nature of the form suggests onomatopoeia, and the possibility of parallel development cannot be ruled out.

POc		*[kʷe]kʷekʷe	‘k.o. frog’
Adm	Seimat	keke(u)	‘k.o. frog’
NNG	Mangseng	keki	‘frog’
NNG	Hote	kokʷek	‘frog’
PT	Saliba	kʷekʷe	‘tree frog’
PT	Tawala	kʷekʷe	‘k.o. frog’
PT	Tubetube	kʷekʷe	‘frog’
PT	Dobu	kʷekʷe	‘frog’
PT	Dobu	(sikʷesi) kʷekʷe	‘small frog, undeveloped’
PT	Gumawana	kʷekʷe	‘frog’
SES	Gela	kekeke	‘k.o. frog’
SES	Lau	ʔiki-ʔiki	‘k.o. frog’

cf. also:

PT	Gapapaiwa	kʷere	‘frog’
SES	’Are’are	koʔe	‘frog’
SES	Gela	kere(rao)	‘k.o. frog’
SES	Lau	kʷere	‘k.o. frog’

The following cognate set is confined to Vitu, the New Ireland group, Nehan and Buka. The fact that rokrok is also Tok Pisin for ‘frog’, from Tolai, makes this set problematic. However, a portion of Tok Pisin vocabulary has its origins in southern New Ireland languages.

PMM		*rokrok	‘bullfrog’
MM	Vitu	rok-rok	‘frog’
MM	Sursurunga	rok-rok	‘frog, incl. marine toad’
MM	Patpatar	rok-rok	‘k.o. frog that stays on ground’
MM	Tolai	rok-rok	‘k.o. frog’
MM	Ramoaaina	rok-rok	‘frog’
MM	Nehan	rok-rok	‘frog’
MM	Tinputz	rok-rok	‘bullfrog’

cf. also:

Adm

Lou

rek-rek

‘large green frog, bullfrog’

PSES		*bʷari	‘large frog’
SES	Gela	bari	‘k.o. large frog’
SES	Tolo	bari	‘large water frog’
SES	’Are’are	pari	‘small frog’
SES	Longgu	bʷari	‘big frog, lives near river; bush frog’

A POc term for ‘tadpole’ has proved elusive. The etymon kʷasi ‘tadpole’ is found in two closely related Papuan Tip languages, Molima and Bwaidoga. Elsewhere we find kʷasi ‘k.o. frog’ in Lau, a SE Solomonic language. However, Lau kʷ reflects POc *w, so the term is evidently a borrowing. No other cognates have been located.

From the following cognate set, it appears that *kuni may be reconstructed as a PWOc term for ‘frog’, and with variable modifier as a term for ‘tadpole’ in PT languages.

PWOc		*kuni	‘frog, tadpole’
PT	Gapapaiwa	kuni(goagoa), kunu(goagoa)	‘tadpole’
PT	Molima	(bʷali)kunu-kunu	‘tadpoles’
PT	Dobu	(wari)kunu-kunu	‘tadpoles’11
MM	Simbo	kuni	‘frog (generic)’
MM	Roviana	kuni	‘small noisy frog’

7. Body parts⇫

7.1. Mammal body parts⇫

Most mammal body-part terms are also terms for human body-parts. For instance, ‘fur’ equates with ‘body hair’ (typically from POc *pulu), both ‘hind leg’ and ‘fore leg’ with ‘leg’ (POc *qaqe ‘leg, foot’ or POc *paqal ‘thigh’), ‘paw’ with ‘hand’ (POc *lima), and ‘claw’ with ‘nail’ (POc *kuku). The generic term for teeth, animal or human, was POc *[l,n]ipon, a term which could also include ‘tusk’ (see §7.1.2 below).

7.1.1. Canine tooth⇫

POc *bati evidently referred specifically to canine teeth. These teeth may have been singled out for special naming in POc times because of their distinctive shape, which made them suitable for decoration, but cognates occur in Vanuatu with particular reference to pigs. There, boars sometimes had their upper canine teeth removed to enable the lower tusks to grow in a circle and re-enter the jawbone. These pigs played a significant role in status rituals, and languages had numerous terms for boars at various stages of tusk growth.

POc		*bati	‘tusk, canine tooth’
NNG	Gedaged	bali	‘canine tooth of dog; ornamental band for forehead made of dog’s teeth’
PSOc		*bati	‘tusk, upper canine tooth’ (John Lynch, pers. comm.)
NCV	Mota	pati(u)	‘upper tusk in a boar, knocked out to let the lower tusk grow long; the corresponding teeth in man’
NCV	Uripiv	bati-n	‘upper canine tooth, of pig, porpoise or person’
NCV	Avava	bas-n	‘upper tusk of a pig, incisor’
SV	Sye	ne-pati	‘pig tusk’
SV	Kwamera	nə-pati-	‘tusk, horn’
SV	Anejom̃	ni-pat	‘tusk, horn, tusked pig’
Fij	Bauan	bati-na	‘tooth (generic); tusk’

cf. also:

MM

Nduke

patu

‘tooth’ (polite; reflects *p)

7.1.2. Pig tusk⇫

POc *[l,n]ipon ‘tooth’ has reflexes across subgroups which indicate that pig tusks were included within its meaning.

POc		*[l,n]ipon	‘tooth, tusk’
Adm	Lou	lia-n puo	‘tusk of pig’
NNG	Tuam	livo	‘tooth’
SES	Lau	lifo	‘tusk, tooth, porpoise teeth used as money’
SES	Sa’a	niho	‘tusk, tooth’
SES	Arosi	riho-na	‘tusk, tooth’
NCV	Mwotlap	nı-lıw	‘pig tusk’
NCV	Sa	lio bʷê	‘pig tusk’
Pn	Tongan	nifo	‘tooth, tusk, nippers’
Pn	Samoan	nifo	‘tooth, tusk, horn’

A NGOc term, *joqa, which apparently referred to ‘pig’s tusk’, is also reconstructable.

PNGOc		*joŋa	‘(boar’s) tusk’
NNG	Tami	joŋ	‘boar’s tusk’
NNG	Manam	zoŋo	‘tusk’
PT	Kilivila	doga	‘pig’s tusk’
PT	Wedau	tona	‘tusk’
PT	Saliba	dona	‘horn’
PT	Motu	doa	‘tusk’
PT	Muyuw	dog	‘tusk, horn’

cf. also:

MM	Nakanai	togo	‘tooth, tusk, claw’ (t- for †s-)
MM	Tolai	lokono	‘tusk, pig with tusks’

7.1.3. Tail⇫

POc *iku(R) ‘tail’ continues a PAn etymon. Although many Oceanic wordlists simply give ‘tail’ as the gloss, a substantial number indicate that reflexes of this term relate primarily to tails of quadrupeds. It is likely that more spectacular tail feathers of birds would have been separately named (Clark reconstructs POc *lawe, tail feathers, plume’ in ch. 6) but more commonplace bird tails were probably referred to by *ikuR. Tails of fish are possibly also included.

PAn		*ikuR	‘tail’ (ACD)
POc		*iku(R)	‘tail, of quadruped, some birds and possibly fish’ (cf ch. 6, p.277)
Adm	Nyindrou	kiwi-n	‘tail’
NNG	Wogeo	iku	‘tail’
NNG	Gitua	igu	‘tail’
PT	Motu	iu	‘tail (of animals)’
PT	Tawala	giu-na	‘tail’
MM	Nakanai	kiu	‘tail of dog +’ (metathesis)
MM	Bulu	iku	‘tail of dog’
MM	Tolai	uku-na	‘tail’
SES	Gela	iɣu	‘tail’
SES	Bugotu	iu-iɣu	‘tail, of dog, bird, fish’
SES	Longgu	iku-iku	‘tail of dog, pig, etc.’
SES	Lau	ki-kiu	‘tail of animal, fish (not bird)’
SV	Anejom̃	n-iɣe	‘tail of fish’
NCal	Xârâcùù	kʷi	‘tail, general term, can include fish’
Mic	Woleaian	iux	‘tail of bird’
Mic	Puluwatese	wik	‘tail, as of birds, chickens, fish, hindquarters as of dog’
Pn	Tongan	iku	‘tail (esp. of a quadruped)’ (hiku ‘tail of fish or reptile’)
Pn	Samoan	iʔu	‘tail (as of a fish); tail of a shark (as a portion) when it is divided according to custom’

A number of Polynesian languages use reflexes of PPn *siku to refer to fish and animal tails. These may show contamination from POc *siku(n) ‘elbow, corner’, PPn *siku ‘extremity; end; tail (esp. of fish)’

7.2. Reptile body parts⇫

7.2.1. Turtle shell⇫

A POc term for the shell or carapace of turtles rests on agreement between Proto Micronesian *pʷōca and a cognate in an Admiralties language, Wuvulu. The Wuvulu word may be a borrowing from Micronesian.

POc		*bo(dr,d)a	‘turtle shell’
Adm	Wuvulu	poxa	‘turtle shell’
PMic		*pʷōca	‘turtle shell’ (Bender et al. 2003)
Mic	Chuukese	pʷōc̣	‘armour-plating of turtle’s or crocodile’s back’
Mic	Mortlockese	pʷōṣ	‘turtle shell’
Mic	Puluwatese	pʷōṛ	‘outer layer of turtle shell, used for fishhooks, earrings etc.’
Mic	Carolinian	bʷōṣ	‘k.o. shell from turtle or coconut, used to make jewellery, bowls etc.’
Mic	Woleaian	ɸʷōṣe	‘turtle shell’
Mic	Kiribati	te-bʷora	‘turtle shell’

In many Polynesian languages and at least one Southeast Solomonic language, reflexes of POc *qunap ‘fish scales’ also refer to the shell of a turtle. (See also chapter 2, §66.3)

PEOc		*qunap	‘turtle shell, fish scales’
SES	Arosi	una-na	‘turtle shell, tortoiseshell’ (unahi ‘to remove shell from turtle, scale a fish’)
Pn	Niuean	una	‘turtle shell’ (hinafi ‘fish scales’)
Pn	Tongan	ʔuno	‘scale of fish, shell of turtle’
Pn	Rennellese	ʔuna	‘outer shell, as of turtle’ (ʔunahi ‘fish scales’)
Pn	Samoan	una	‘shell of hawksbill turtle, fish scale’ (unafi ‘to scale’)
Pn	Tikopia	una	‘carapace of marine turtle’ (unafi ‘fish scale’)
Pn	Hawaiian	una	‘turtle shell’ (unahi ‘fish scales’)

cf. also:

MM

Kia

ŋapo

‘turtle shell’

7.2.2. Gaps in reconstruction⇫

POc speakers almost certainly had names for several different parts of a turtle in addition to the shell. Arosi, for instance, has terms for ‘flat back plates of shell’, ‘curved side plates of shell’ and ‘turtle fat’. However, most dictionaries fail to record such details and we are unable to make secure reconstructions for other turtle parts.

From the few items located, it seems that in Polynesia flippers of turtle, stingray and whale may have been denoted by a derivative of the verb POc *kaba ‘wing’, PEOc *kaba-kaba ‘to flap wings’, reflexes of which can also refer to pectoral fins of fish (see ch.2, §66.4 and ch.6, §2). Terms for ‘turtle egg’ are commonly reflexes of the generic term for egg, POc *qatoluR. However, at least one language, To’aba’ita (SES) has a term, kʷalaruru applying specifically to ‘a collection of turtle or spider eggs’. Sa’a (SES) has apota ‘eggs of turtle or crocodile’.

Notes⇫