There may not have been a domain of the Proto Oceanic lexicon
corresponding precisely to the Class Aves. The creatures referred to by
reflexes of POc *manuk (below) almost always include bats and
sometimes other flying animals. For example, Kwaio laŋasi (a
taboo replacement for manu) ‘in the broadest categorical sense,
includes birds, butterflies, moths, bats, dragonflies and one species of
frogs (Pseudophryne sp.)’ (Keesing 1975:119). At its widest
*manuk may have encompassed the entire non-aquatic animal
kingdom (for discussion see chapter 8, §§5 and 6). More restrictive
expressions are often coined by adding a modifier to *manuk,
typically the verb ‘to fly’, for example Chuukese māɾ ‘living
creature of land or air (other than human)’, mac̣c̣aŋ ‘bird’
(cf. c̣aŋ ‘fly’). The organisation of this book, however,
follows the Linnaean classification, and bats will be found in chapter
5, with only a couple of cross-references below to connect the two.
Birds played varied roles in the life of Oceanic speakers. Besides
the domesticated fowl (Gallus gallus), some types of birds
(including frigate birds and parrots) were caught and kept as pets.
Others were prized as food (pigeons and doves, some seabirds). The large
eggs of the megapode were gathered for food. The feathers of many
species had decorative uses, and their bones were used for a range of
artefacts from needles to flutes. Birds could be caught with snares,
nets or traps of various kinds, or shot with bow and arrow. (Special
blunt-tipped arrows avoided damage to valued feathers.) Apart from
direct exploitation, some birds could be sources of useful information,
such as flocks of terns indicating the location of schools of fish, or
landbirds far out at sea, indicating the presence of land nearby. Some,
on the other hand, were harmful, such as the raptors that preyed on
domestic fowl, or the swamp-hen with its destructive effects on taro
gardens.
A widespread belief was that the voices of certain species could
foretell such events as a death or the arrival of visitors. In eastern
Polynesia, the visitors themselves were referred to as ‘birds’ (PCEPn
*manu-firi ‘chosen (?) birds’). And in myth, folk tale and
religious belief, birds play almost as varied a range of roles as do
humans. Among the Dobuans, ‘the native term for totem is the term for
bird … [and] one stranger may ask another, “what is your bird?”’ (Fortune 1963:32).
A few aspects of avian anatomy and behaviour are different enough
from their human (or mammalian) analogues that they might be expected to
be separately lexicalised. In most cases, however, POc appears to have
found commonalities with more general anatomical features of the animal
kingdom.
Perhaps the most emblematic features of birds are feathers and wings.
But the most general term for feathers is the same as that for human
body hair and animal fur:
Both of the above refer to the overall covering of birds that is
analogous to the hair covering of mammals. Feathers particularly
conspicuous by length or colour have probably always been of special
interest and valued as decoration. A term for such feathers is
reconstructed for PMP, but its unambiguous Oceanic reflexes are
restricted in scope:
A distinct lexical item POc *banic ‘wing, arm, hand, fin
(probably pectoral)’ is widely attested. Its PMP ancestor
*panij denoted only wings, but its meaning had clearly been
extended in POc (see chapter 2 for a fuller cognate set and
discussion).
Dempwolff (1938) reconstructs PMP
*kapak ‘wings; flutter’ on the basis of Oceanic data and Malay
kepak, Ngaju Dayak kapak. This etymon appears to be
reflected by the widespread etymon POc *kaba (N) ‘wing’, (v)
‘flap wings’.
Blust (ACD) also reconstructs PWMP *kepek ‘flap wings’
(expected POc form †*kopo(k)) and PWMP *kepay-kepay
‘flap (wings, fins of fish, etc.)’ (expected POc form †*kope)
with supporting data, and crossreferences the doublets *kapay
and *kipay (without supporting data) (expected POc forms
†*kape and †*kipe). PMP *kapay is perhaps
reflected by Tawala apape ‘wing’. I find no straightforward
reflexes of Blust’s other reconstructions, but Iduna (PT)
ofa-na ‘(its) wing’ and Gela (SES) gapo-gapo ‘spread
wings, flap, flutter’ suggest that there were perhaps several
phonologically similar forms in POc cognate with the forms found in
Western Malayo-Polynesian languages.
A bird’s beak was most probably referred to as POc *ŋuju,
denoting the external or protruding aspect of the mouth. There is also
some evidence for POc *muju, in the form of non-Oceanic terms
from which Blust (ACD) reconstructs PWMP *mu(n)cuŋ ‘mouth (of
an animal), snout’,2 together with the Iduna, Gapapaiwa
and Balawaia terms listed below, reflecting PPT *mu(d,j)u.
However, it may be that this is a chance similarity and that PPT
*muju was an irregular reflex of POc *ŋuju.3
Many, perhaps most Oceanic languages make some lexical distinctions
among the tails of birds, fish, reptiles and mammals. The following very
common term for tail appears applicable to birds in many languages:
PAn
*ikuR
‘tail’ (Blust 1995)
POc
*iku(R)
‘tail, of quadruped, some birds and possibly fish’
(cf ch. 5, §7.1.3)
Other anatomical features such as the cock’s comb and spurs are
lexicalised diversely and there is no clear POc reconstruction.
The laying of relatively large (and often edible) eggs is distinctive
of birds (along with turtles and other reptiles). A well established
reconstruction is:
A word of more general reference, POc *puaq, originally
‘fruit’, but generalised as a classifier for a wide range of more or
less spherical objects, is sometimes used for eggs, but this may well be
the result of repeated local extensions.
Despite the great variety in the form and construction of birds’
nests (note PMic *fata ‘platform, nest’ from POc
*patar ‘platform’, vol.1,57 and p. 190), a common term
can be reconstructed (sometimes stated as also applying to nests or dens
of animals).
The actions surrounding eggs and nests are less distinctively
lexicalised. The verb corresponding to ‘lay’ is etymologically various
(e.g. Lau kʷala ‘give birth, beget, lay an egg’, Bauan
vaka-lutu ‘drop, lay (egg)’, Paamese mūmoni ‘make’
mūmon orelīte ‘lay an egg’), but note the following:
The action of incubating eggs by sitting on them is compared with
that of covering and protecting the young in the following verb
attributable to PEOc:
PEOc
*ovi(s), *ovis-i-
(1) ‘brood, sit on eggs’; (2) ‘cover chicks with
wings’
‘to brood, of hens; to cover up the chickens under her
wings’
When the egg hatches, the words used are generally those for the
breaking open of rigid containers (the coconut probably being the most
familiar comparison), but no consistent etymon emerges.
The action of flying is as emblematic of birds as their possession of
feathers and wings, and several verbs are associated with it (see also
POc *kaba ‘(n) wing, (v) flap wings’ above):
‘fly up, rise up through the air’ (expected
†rovo; cf. no ‘sit, stay’ < *nopo)
POc *tap(p,pʷ)a may perhaps have referred to gliding or
soaring; the apparently contradictory meaning of ‘flap wings’ in some
languages may represent contamination from POc *kaba ‘flap
wings’ (above).
There is a strong consensus among recent researchers locating the
Proto Oceanic homeland in Northwest Melanesia, with the island of New
Britain a focus of probabilities. I will follow this view insofar as it
helps give a perspective on the dimensions of the problem of
reconstructing this particular lexical field. In what follows I will use
the term ‘homeland’ to refer to New Britain and its neighbouring
islands. The birds among which referents for POc reconstructions will be
sought in this study are those which are residents of or regular
visitors to the Bismarck Archipelago, as described in recent handbooks
(Coates 1985, 1990; Coates & Peckover 2001).
This gives a total of slightly more than 200 species belonging to 126
genera, 77 of them monotypic.
The reconstructions presented below do not comprise all or even most
of a plausible avian terminology of Proto Oceanic. Entire families of
birds which were almost certainly known to the POc speakers are
unrepresented here, because the small number of terms so far recorded
for living Oceanic languages do not yield up any useful cognate sets.
Examples are the grebes (Podicipedidae), pelicans (Pelecanidae),
cormorants (Phalacrocoracidae), ibises (Threskiornithidae), nightjars
(Caprimulgidae), flowerpeckers (Dicaeidae), waxbills (Estrildidae) and
drongos (Dicruridae). This lack, in turn, is primarily a result of the
insufficiency of the data, particularly for languages of Northwest
Melanesia, where the avifauna is the richest. The situation has
improved, but only slightly, since my previous study (Clark 1994b). There is still an almost
complete lack of really comprehensive studies.
It may be useful, nevertheless, to have some idea of what such a
taxonomy would have looked like. In my earlier paper I reasoned by
analogy from two detailed studies from societies close to the Oceanic
region, though both non-Austronesian in language: Majnep & Bulmer (1977) on Kalam of the New
Guinea highlands, and Taylor (1990) on Tobelo of north Halmahera.
Both studies found a total of about 200 lexemes for avian taxa. The
typical taxon corresponds to a Linnaean genus, though as a substantial
proportion of genera were monotypic (represented by a single species)
within the area studied, there is a certain ambiguity between genus and
species. Genus-level taxon labels in Kalam, for instance, would include
ccp covering the Australian Goshawk (Accipiter
fasciatus) and Black-mantled Goshawk (Accipiter
melanochlamys), kwwt for both the Amboina Cuckoo dove
(Macropygia amboinensis) and the Black-billed Cuckoo dove
(Macropygia nigrirostris), mmañp labelling both the
Glossy Swiftlet (Collocalia esculenta) and the Mountain
Swiftlet (Collocalia hirundinacea), plolom both the
Mountain Yellow-billed Kingfisher (Halcyon megarhyncha) and the
Sacred Kingfisher (H. sanclay and walkobneŋ both the
Black-fronted White-eye (Zosterops minor) and the Mountain
Whiteeye (Zosterops novaeguineae). In some cases, conventional
distinguishing expressions exist to discriminate between the two
species, for example, ccp mseŋ-ket ‘Australian Goshawk’
(mseŋ-ket ‘open country’) versus ccp kamay-ket
‘Black-mantled Goshawk’ (kamay-ket ‘beech tree sp.’).
On the other hand, some Kalam taxa correspond to Linnaean groups
larger than the genus, and some cut across the terminology altogether,
recognizing convergent similarities: spsep refers both to the
Mountain Pygmy Parrot (Micropsitta bruijni) and the Pink-faced Nuthatch
(Daphoenositta miranda). sskl refers to both
Whitehead’s Swiftlet (Collocalia whiteheadi) and the Pacific
Swallow (Hirundo tahitica). (Cf. POc *kabakabal in §5.12.)
Most birds produce some sound audible to humans, whether territorial
song, flocking calls or alarm cries, and often this is distinctive
enough to be recognised as emblematic of the bird itself. The
transcoding of bird vocalizations into human speech sounds is a
universal phenomenon, and birds are probably onomatopoetically named
more than any other class of living things. In theory this can pose an
obstacle to the comparative method, if the independent application of an
onomatopoetic name to the same bird in different languages produces
pseudo-cognates. However, the more remotely related the two languages,
the less likely such pairs of names are to follow regular sound
correspondences. The approach taken here will be to accept exact cognate
sets as genuine, but cases where there are repeated irregularities
despite similarity of phonetic form will be taken as, at least, suspect
for reconstruction purposes.
In theory we would like to be able to say how POc speakers would have
named each species of bird they were familiar with. Some reasons why
this is not possible at the present state of knowledge have already been
reviewed. There is also the fact that, even if we take the location and
date of the POc-speaking community as fairly precisely known, the
composition of the avifauna at that place and time is by no means so
certain. The distribution of bird species is a changing set of
circumstances, affected both by expansion of successful species and by
extinction of others. In practice all we have to go on (except for some
scattered archaeological evidence) is the present situation. But even in
a situation where several reasonably good cognate sets are available,
which could in theory be assigned to several species present in the
Oceanic homeland, the assignment may be difficult since our data is
weakest precisely in the area where the fauna is richest.
Below I present nearly 200 cognate sets, of which about 80 are
attributed to Proto Oceanic and the remainder to various interstages.
Following a discussion of the domestic fowl, the families are arranged
into three large groups: the non-passerine land and fresh-water birds;
the passerines; and the sea and shore birds. A typical local inventory
contains roughly equal numbers of species in these three groups. (Bregulla 1992 on Vanuatu, for
example, has 36 passerine species, 45 sea birds and 40 others.) All
scientific and ‘common’ English names have been normalised to follow the
checklist of Howard &
Moore (1991).
The Red Jungle Fowl, (Gallus gallus), spread as a
domesticate along with Oceanic speakers throughout almost all the
Pacific islands. There does not seem to be any clear lexical distinction
between domestic and feral birds, though descriptors may be added, as in
Samoan moa ʔāivao ‘wild fowl’ (cf vao ‘bush, jungle’).
The glosses ‘chicken’ and ‘fowl’ have been taken as equivalent.
A number of Oceanic languages use a word for fowl which is most
commonly the generic term for ‘bird’.
Note that Bender et al. (2003:327) consider the Micronesian
forms to be loans from Chamorro mannok ‘chicken’.
A more widely distributed set of terms for the fowl is clearly an
imitative term for the distinctive call of the male bird. A pattern of
consonants occurs which could represent POc *k-k-r-k, though
the vowels are not consistent and one or other of the consonants may not
appear. Such distant semblances as Japanese kokekokko and
French cocorico ‘cock-a-doodle- doo’ suggest that this sequence
may have, contrary to appearances, something natural about it. And yet,
the shift of these words from sound-imitative to referential (including
the female) cannot have taken place many times independently. I
therefore hesitantly offer a reconstruction. However, it is noteworthy
that *k in this term is never lenited (to ɣ,
ʔ etc) in the many languages where this is a regular change.
This presumably reflects its onomatopaeic origin.
The common Polynesian word for the fowl has no certain external
cognates, though it has apparently been borrowed rather far afield.
However, it is possible that Wayan and Bauan mō are cognate
with PPn *moa, as PCP *oa often becomes ō, at
least in Western Fijian. If they are, then PCP *moa is
reconstructable, but with uncertain denotation.
A common observation about animal taxonomy is that terminology for
domesticated species will be more detailed than for most wild
species—specific names for sexes, ages and varieties are to be expected.
This does not seem to be the case for the fowl in Proto Oceanic.
One term, reconstructed as PMP _*qup_a ‘hen, egg-laying chicken’,
appears to have Oceanic reflexes only in Polynesian and Rotuman (the
latter a borrowing):
The male of the species is most commonly simply denoted by ‘male
fowl’ (Manam maŋ moane, Motu kokoruku
maruanena,’Are’are kua mane), and chicks as ‘child of
fowl’ (Motu kokoroku natuna, Cemuhi nahi-ja, Bauan
luve ni toa).
Though in species numbers they are roughly a third of the total, this
group of families accounts for well over half the cognate sets and
reconstructions presented here. Land birds are likely to be seen by
everyone in the community, and not just those who go to sea. The
non-passerines are on the whole larger and more conspicuous than the
passerines. They are therefore likely to have names known by everyone.
They are also objectively more diverse as a group, and hence more likely
to be recognised and lexicalised (Boster, Berlin & O’Neill 1986:
577-578).
A single species of this family of huge flightless birds, the Dwarf
Cassowary (Casuarius bennetti), lives in the Oceanic homeland.
Although cassowaries do not extend even into the Solomons, enough
reflexes exist in languages of the NNG, PT and MM groups to support the
reconstruction below. Attribution of such a form to POc is further
supported by numerous cognates in languages to the west, e.g. Biak
(SHWNG) (man)swar, Kaiwai (CMP) asawar (Anceaux 1961).
The herons of the homeland include three species of large herons or
egrets of the genus Egretta as well as the Green-backed Heron
(Butorides striatus), the Nankeen Night Heron (Nycticorax
caledonicus) and two bitterns (Ixobrychus spp.). Most
cognate sets refer to the large and conspicuous Egretta herons,
which may have been the basis for a generic term.
It appears that *kao may have been a generic term for herons
in PCP. Both Fijian and Polynesian introduce new terms (see below) for
Egretta, and *kao mainly survives in reference to less
conspicuous species. Among Central Pacific reflexes only the Takuu and
Maori reflexes may be considered unproblematic. The Nukuoro word is
probably borrowed from a Micronesian language.
The relation among the three cognate sets above is not entirely
clear. The ordering is based on their successively narrower
distribution, from which one might conjecture that *kao was the
original POc term, *kaopa an extension of this, and
*qopa a truncation of *kaopa restricted to
North-Central Vanuatu and its immediate neighbour Temotu area. The above
cognate complex is almost unrepresented in the Papuan Tip group, which
seems to replace it with the following:
Two of the identifications below are problematic: the Pied Heron,
Egretta picata, is a mainland species not known to occur in the
Bismarcks, and the Grey Heron, Ardea cinerea, does not occur in
the New Guinea area at all.
Additional clear but highly localised names in the Solomons,
Micronesia and Polynesia also centre around Egretta. POc
*sou below is suspect as a POc reconstruction as it only occurs
in the Solomons. However, Numbami (NNG) saole ‘egret’ may be
cognate. If it is, then the POc form was perhaps *saol or
*saul, but this in turn leaves Numbami final -e for
expected †-a unaccounted for, as well as the lack of the final
consonant and echo vowel from expected Marovo †coulu Nduke
†houlu.
The Rennellese term seems to agree with the Central Micronesian
reconstruction, but borrowing from this direction would be unexpected.
No heron terms like this appear to exist in the Southeast Solomons.
The East Polynesian forms prefixed with *kō- fall into a
large class of nouns so reshaped in this subgroup. However, the Hawaiian
and Maori terms with *kau- suggest the possibility of influence
of the *kao term reconstructed above.
Compared with extensive cognate sets for the Egretta herons, only two
small sets appear to denote the Night Herons:
The Pacific Black Duck (Anas superciliosa) is the most
common duck of the Oceanic region, but the Grey Teal (Anas
gibberifrons) is also present in the homeland, as are two species
of Whistling Ducks (Dendrocygna spp.), and various other
species of this family are known as migrants. There is no evidence of
domestication of any duck species in pre-European Oceania. Many
languages use a separate lexical item for the introduced domesticates,
as Samoan pato ‘domestic duck’ (from Spanish), toloa
‘wild duck’. Very few descriptions offer glosses more precise than
‘duck’ or ‘wild duck’. However, Iduna geluluva is explicitly
stated to cover both species of Anas mentioned above, and Nelemwa
kerorōp includes both these as well as the Mallard, Anas
platyrhynchos). There may be a terminological agreement between
Caaac and Fijian (Lomaivitilevu) in which the Whistling Duck is
described as ‘red duck’ (Caaac nīaŋ mīa, Fijian
gā-damu).
The following possible Maori reflex refers to a different waterfowl,
but the formal development would be regular. (The grebe is presently
found only in the South Island, where Maori *ŋ >
k.
A number of other Polynesian and Fijian languages have bird names
formally cognate with the above, but referring to shore and sea birds.
(A number of these innovations are in the languages of atolls, where
ducks are rare or non-occurrent owing to the absence of surface fresh
water.):
A strikingly resemblant form occurs in Motu. The only known source of
Polynesian loanwords here would be 19th century missionaries from the
Cook Islands, but it seems odd to have a wild species named in this
way.
These three families are treated together because of obvious
similarities in appearance and habits, as well as overlap in naming
practices. They are well represented in the homeland, with at least ten
species. Roughly from largest to smallest, they include the
White-bellied Sea Eagle (Haliaeetus leucogaster), the Osprey
(Pandion haliaetus), several goshawks of genus Accipiter, the
Brahminy Kite (Haliastur indus), the Black Honey Buzzard
(Henicopernis infuscata), the Crested Baza (Aviceda
subcristata), and probably both the Peregrine Falcon (Falco
peregrinus) and the Oriental Hobby (Falco severus).
A very large number of terms for birds of prey are recorded, from
which a number of cognate sets emerge; however, associating particular
names with particular species is much more difficult, owing in large
part to vague descriptions which use ‘hawk’ or ‘eagle’ to refer to any
bird of prey.
The fish-eagle (Haliaeetus) was probably the largest flying creature
known to the Proto Oceanic speakers. (Haliaeetus leucogaster of
the homeland is replaced in the Solomons by the very similar
Haliaeetus sanfordi.) So it is not surprising to find terms for
this species literally meaning ‘big bird’ in a number of languages. In
all but one of the following, the first element reflects POc
*manuk ‘bird’, and the second is the local word for ‘big’.
These second elements are etymologically diverse, but Patpatar, Nehan
and Puluwatese probably retain the POc form *lapuat ‘large’ (vol.2:190-191).
The following three sets of terms have fairly wide geographical
spread but they are not numerous or semantically consistent enough to
justify a confident attribution to Proto Oceanic.
In Southern Oceania, the sea eagles are no longer present, and the
familiar raptors are the goshawks (Accipiter spp.) and the
Pacific Marsh Harrier (Circus approximans)
The following could be an irregular development of *mala,
the languages shown are all from Malakula, Vanuatu; however some of the
Vanuatu reflexes of POc *pʷa(r,R)a ‘hawk’ above may belong
here, being from languages in which *l and *R would
both give zero in this environment.
Megapodes or scrub fowl (Megapodius spp.) are found from the
homeland south to Vanuatu and as far east as the island of Niuafo’ou in
Tonga. The representative in the homeland is the Bismarck Scrub Fowl
(Megapodius eremita). All three of the name-sets below may be
of imitative origin, given the bird’s two-syllable calls ending in
[au]-like sounds. (Hadden (2004a:78) gives ‘paa-au’,
‘keeawau keeawuh’, ‘keereow keerow’ and ‘kiau
kiau’ as representations of its calls.)
Figure 6.8:Megapodius eremita, Bismarck Scrub Fowl
POc *mʷalau is reflected from Sulawesi (Sangirese
maleo) to the easternmost bird of this family in Tonga (Wallace 1869: 202-4, Christian 1926, but see also
comments in Blust 2002).
Apart from the domestic fowl, the only representative of this family
in the homeland is the Indian Blue Quail (Coturnix chinensis).
In the following comparison, a POc reconstruction is justified if the
Mangap and Tolai forms are regarded as continuing the PAn etymon.
The numerous rails of Oceania include the very widespread Purple
Swamphen (Porphyrio porphyrio), easily recognised by its large
size and red forehead. Among the smaller species are the Buff-banded
Rail (Rallus philippensis) and White-browed Rail (Porzana
cinerea), which occur throughout much of the Oceanic region, and in
the homeland three others of the genus Rallus, as well as the
Bare-eyed Rail (Eulabeornis plumbeiventris), Red-necked Crake
(Rallina tricolor), and Rufous-tailed Moorhen (Amaurornis
olivaceus).
The only other rail forwhich widespread cognate sets exist is the
Buff-banded Rail (Rallus philippensis). The following possible
POc reconstruction has only slender support:
The similarity is enhanced by two SE Solomonic languages in which the
word, while still denoting the Purple Swamphen, show unexpected
k for *R in the final syllable:
The hypothetical Tahiti Rail (Rallus pacificus) is known
only from a painting made on Cook’s second voyage (duPont 1976:41), but
the name recorded agrees with the evidence of its appearance that it was
a species of Rallus.
The only reconstruction for the Sooty Crake (Porzana
tabuensis) is restricted to Polynesian.
The Nukuoro name looks like a regular development of PPn
*moso, but the Kapingamarangi suggests possible borrowing from
Micronesian (cf. Mokilese mʷiɔk).
This family is among the most richly elaborated in the Oceanic realm.
Almost every island has at least one or two representatives, and the
Proto Oceanic speakers may have been acquainted with twenty or more
species. The most widely distributed Columbidae belong to two genera:
the Imperial Pigeons (Ducula spp.), and the much smaller Fruit
Doves (Ptilinopus spp.). Other genera represented include
Gymnophaps, Columba, Macropygia,
Reinwardtoena, Chalcophaps, Gallicolumba,
Henicophaps, and the shaggy-plumaged Nicobar Pigeon,
Caloenas nicobarica.
The most likely candidate for a generic is the following, which is
identified as generic in some languages (Gedaged, Tawala, Muyuw), and
elsewhere refers to a wide range of columbid genera. In some local
areas, however, it appears to have become specialised.
Although the following is in near-complementary distribution with POc
*bune, and could almost be a metathetic development of it,
Bauan at least contrasts the two terms.
The following appears to be a local survival of a term
reconstructible for PMP, but the glosses, even within the SE Solomonic
subgroup, are confusingly diverse.
An isolated Micronesian form appears as a possible cognate of this
set, Kiribati pitin ‘ground dove, Gallicolumba sp.’.
The exact species reference of the Kiribati term is unclear. However, Pratt et al. (1987:194-5) indicate that this bird
was introduced to Kiribati (Abemama) about 1940 from some other Pacific
island. Since none of the possible source islands have words resembling
bitin, it would seem much more likely that this name
(phonetically [pisin]) is a borrowing from English pigeon.
The absence of glottal stop in the Tongan and E Futunan forms shows
that this name is likely imitative of the species’ voice, rather than
derived from tuʔu ‘stand’ (referring to the species’
ground-feeding habits). However, in languages where the glottal stop has
been lost the two would be homophonous, and tū-kere may be
understood by present day speakers as ‘stands on the ground’.
Imitative naming is a high probability here; cf. PAn *qekuŋ
‘owl’, PWMP *bekur ‘coo, turtle dove’, PMP *bukaw ‘owl
sp.’ (Blust 2002). The voice of
Macropygia mackinlayi is represented as ‘kor-wu’ or ‘vo-ku’ (Hadden 2004a: 114). A metathetic
variant is represented by Teop (MM) uvo ‘Macropygia
mackinlayi’, cf. Koiari (Papuan) kuvo ‘Lesser Bar-tailed
Cuckoo Dove, Macropygia nigrirostris’.
The following set undoubtedly owes something to imitation of typical
pigeon ‘coo’ vocalizations—low pitched but with strong formant
structure. Compare the Wayan Fijian name for a species introduced from
Asia in modern times: kukurū ’Spotted-necked Dove, Streptopelia
chinensis‘.
The three families united here are represented by nine diverse genera
in the homeland. By far the largest bird, not confusible with any other,
is the Blue-eyed Cockatoo (Cacatua ophthalmica). The large
Eclectus Parrot (Eclectus roratus) shows striking sexual
dimorphism, with males (green) and females (red) often lexically
distinguished. The Rainbow Lory (Trichoglossus haematodus) is
probably the most conspicuous and widespread parrot in western Oceania.
Among other middle-sized parrots are the Cardinal Lory (Chalcopsitta
cardinalis), the Purple-bellied Lory (Lorius
hypoinochrous) and the Singing Parrot (Geoffroyus
heteroclitus). Smaller parrots include two Lorikeets of the genus
Charmosyna, which, along with the closely related genus Vini, are found
as far as eastern Polynesia. Smallest of all are the Pygmy Parrots of
genus Micropsitta and the Orange-fronted Hanging Parrot
(Loriculus aurantiifrons). For the last (which is found no
further east than New Guinea) not a single name was found in any of the
sources.
Only Papuan Tip languages supply sufficient cognates to support
reconstructions for these birds. That the first term may be traceable to
POc, however, is suggested by possible cognates in SHWNG languages such
as Ansus kara, Serui-Laut karai, Ambai kara
‘cockatoo’ (Anceaux 1961).
This large parrot (Eclectus roratus) is most notable for its
sexual dimorphism, with the predominantly red-plumaged female and
green-plumaged male sometimes being separately lexicalised. The first
two reconstructions are similar in form, and appear to refer to the
male. Unfortunately, sources for the few languages that reflect both the
shorter and longer term are unhelpful as to the distinction of meaning.
Apparent reflexes of the first reconstruction extend well beyond the
range of the Eclectus, and are applied to various other species,
eventually reaching as far as New Zealand.
The following term may also be associated with Eclectus. Cf. POc
*keRaŋ ‘hawksbill turtle’ (ch. 5, 250), a name which is
eventually re-applied to a parrot in New Zealand (Maori kea
‘Nestor nolabilis’).
The following four reconstructions appear to be cut from the same
basic form: the first with a three-consonant sequence, and each of the
others missing one of the three. It would be natural to treat the others
as simply variously reduced forms of the longest form, but there are
apparent contrasts in several languages. The references are to
medium-sized parrots (Chalcopsitta, Lorius,
Trichoglossus), but it is difficult to see any consistent
association between a particular species and any of these forms. Beyond
the Solomons, the first two species disappear, and these terms are
applied to Trichoglossus, or in New Caledonia to the parakeets of genus
Cyanoramphus.
POc
*sipi(r,R)i
‘Rainbow Lory, Trichoglossus haematodus, or Cardinal
Lory, Chalcopsitta cardinalis’
Eight species of this family are recorded from the homeland, two as
seasonal visitors. They include parasitic cuckoos of the genera
Cuculus, Cacomantis, Chalcites,
Urodynamis and Scythrops, and two Coucals of the genus
Centropus. Names recorded, however, are far fewer than for the
parrots with a comparable number of species. Only one small cognate set
of Proto Oceanic scope has been found.
The Barn Owls (Tytonidae) and Typical Owls (Strigidae) are each
represented in Oceania by a single genus. The Barn Owl (Tyto
alba) is probably the most widespread owl of the Oceanic region. In
the homeland it coexists with the endemic New Britain Barn Owl (Tyto
aurantia). There are also two Hawk Owls of the genus Ninox. While
the voice of the Barn Owl is described as a screech, Typical Owls (like
pigeons) have calls whose acoustic profile suggests the low second
formant of [u]-type vowels. The Solomon Islands Hawk Owl (Ninox
jacquinoti), for example, has a call described as ‘’Kuurrroo’ with
rolled ’r” (Hadden 2004a: 155). The
prevalence of such vowels in the first three cognate sets places them
under a certain degree of uncertainty. Note, however, that reflexes of
*lulu are found mostly in areas where the Barn Owl is the only
owl present, and are therefore unlikely to be imitative.
There is at present an owl-free area in Eastern Polynesia, which may
account for the maritime application of some of these words. However,
owls and petrels have in common that they are seldom seen, and mainly
known by their calls at night. Spenneman (2004:149) notes that
licemᵚao, the present Marshallese term for the female of the
Short-eared Owl (Asio flammeus) was given by an earlier source
for a type of petrel. (The name is probably analysable as li-
‘feminine prefix’ + ? + mᵚa ‘cry’ + o ‘oh!’.)
5.12.
Swifts and swallows (Apodidae, Hemiprocnidae, Hirundinidae)⇫
These three groups are closely allied in general appearance and
habits and rarely distinguished terminologically. The swiftlets are
familiar birds of garden areas throughout much of Oceania. The most
common species noted are the White-rumped Swiftlet (Aerodramus
spodiopygius), Uniform Swiftlet (Aerodramus vanikorensis)
and White-bellied Swiftlet (Collocalia esculenta). These are
not in general lexically distinguished. The few terms recorded for the
Pacific Swallow (Hirundo tahitica) and the even fewer for the
Whiskered Tree-swift (Hemiprocne mystacea) are often lexically
associated with those for the Swiftlets, and where they are not, they do
not support any independently reconstructible term. These birds are also
terminologically linked with small insectivorous bats: SE Ambrym (NCV)
avœp ‘any small bat; swiftlet’; E Futunan pekapeka
‘White-rumped Swiftlet’, pekapeka saʔi ‘small bat sp.’ (see ch.5, §2.9).
POc
*kabakabal
‘swiftlet, Aerodramus or Collocalia spp.’
(cf. *kaba etc. ‘wing, flap’)
Ten species of Kingfisher are recorded for the homeland, half of them
in the genus Halcyon, which extends widely through Oceania.
Particularly widespread are the White-collared Kingfisher (Halcyon
chloris), the Sacred Kingfisher (Halcyon sancta) and the
White-headed Kingfisher (Halcyon saurophaga). Other kingfishers
of the homeland include the very large White-tailed Kingfisher
(Tanysiptera sylvia), and the small birds of genera Ceyx and
Alcedo. A number of good cognate sets exist, but assignment to species
is very problematic. The relatively few terms for non-Halcyon
species do not form any useful cognate sets. Blust (2002:113) compares POc *(sj)iko
with PMP *cikep ‘catch with hands’.
This word appears to have been reshaped in Polynesian to PPn
*tiko-tara, a compound analysable as ‘defecate’ + ‘end of
house’. Some outliers have returned to a historically earlier form
through non-Pn influence. The Bauan Fijian word for kingfisher has a
general resemblance to the Polynesian *tikotara forms, but
cannot be formally reconciled with them; more exact agreement is found
in the first part of the word sikorere, though with slightly
deviant reference:
The following two sets are likely to be at least influenced by
imitation of the voice of Halcyon kingfishers, represented as
ki-ki-ki-ki-ki-i-i-ee (Watling
2004:137).
POc
*ki(o)kio
‘kingfisher’ (cf. PAn *kiaw ‘puling sound of
a bird’; Blust 1995)
The single species in this family in the Oceanic homeland, the Dollar
Bird (Eurystomus orientalis) can be associated with two small
cognate sets, though there are problems of identification.
‘bird sp., (Graucalus sublineatus) with a rather shrill
note’
The obsolete generic in Roviana most probably refers to the Lineated
Cuckoo Shrike (Coracina lineata). Since the terms below are
distributed only around the Western Oceanic/SE Solomonic boundary, it is
possible that terms have been borrowed across the boundary. If so, there
is no support for a POc reconstruction.
There are very similar names for this bird in some Papuan languages
of Bougainville, such as Buin kikitou, Koromira
mekikiro, which may be borrowings from Oceanic languages.
The hornbill’s voice is described as ‘various deep grunts and honking
notes’ (Coates & Peckover
2001:113). Given the lack of exact agreement among the following
names, it is likely that at least some represent repeated imitative
coinages.
The very large bill of the hornbill would account for the
mis-identification as ‘toucan’ in the Bugotu and Gela sources. The
toucan is a tropical American bird also celebrated for its extremely
large bill.
Although the Passerines belong to a single order (Passeriformes)
among more than 20 within the class Aves, they account for more than
half of all bird species. Yet in the present study there are far fewer
cognate sets for passerines than for the rest. On the basis of the Kalam
and Tobelo studies, there seems no reason to think that the Proto
Oceanic speakers did not name the passerines in as much detail as other
types of bird. Two factors may have contributed to the relative paucity
of passerine cognate sets. First, many species are highly localised,
which may have produced a high rate of turnover (abandonment and
recreation of terms) as the Oceanic speakers migrated away from their
homeland. Second, since these birds are, on average, smaller and less
conspicuous (and of less economic importance) than the non-passerines,
an accurate identification would require both a researcher with the
means and inclination to go where the birds are, and a speaker as
thoroughly informed about the avifauna as were the Oceanic ancestors. It
is just this type of information that is lacking for most Oceanic
languages; passerines are undoubtedly over-represented among the
‘unidentified bird species’ entries in the dictionaries.
At least some of the following are likely to be independent
imitations of vocalizations variously represented as whee-see,
twill you twill you, keweeo, and squeeii,
produced by birds of this genus (Hadden
2004a: 185-7).
The highly distinctive appearance and behaviour of fantails, and
their tolerance of human proximity, make them familiar birds, much
better represented by widespread cognate sets than the other
flycatchers.
The first part of this compound clearly relates to riu
‘turn’, referring to the bird’s characteristic turns with spread tail
feathers, both in flight and perching. In two languages the following
element is a body part: Sa’a pote ‘buttocks’, Gela
pege ‘forepart of thigh’. But the more widespread second
element seems more likely to be related to POc *kapak, compare
Arosi kape ‘flutter’.
A single species of this family, the White-throated White-eye,
Zosterops meeki, occurs in the Oceanic homeland, but others of
the same genus are very widely found throughout Oceania, with at least
one local cognate set.
The honeyeaters of the homeland region include several belonging to
the genus Myzomela, as well as the New Britain Friarbird (Philemon
cockerelli) and Gilliard’s Honeyeater (Melidectes
whitemanensis). Other birds of this family can be found as far as
Fiji and western Polynesia.
‘red-fronted bird used for feather money’ (expected
†*miti)
The friarbirds (Philemon spp.) are a large and distinctive
type of honeyeater represented by three species in the homeland region.
The following terms present various formal problems, but could be
derived from some form such as POc *(sa)quka, with metathesis
to *(sa)-kuqa.9
‘Foulehaio carunculata, also Fiji Shrikebill,
Clytorhynchus vitiensis’10
This set may be related to an innovative generic for ‘bird’ in a
number of Central Vanuatu languages: Apma bʷihil, N Ambrym
pʷehel, Burmbar mbasir, etc.
Starlings of the genus Aplonis, and the Yellow-faced Mynah, Mino
dumontii, extend throughout much of Oceania. The first two cognate
sets may be related.
The Shrikebill has in common with the starlings only a similar size
and rather drab colouring. The Huia of New Zealand, a much larger bird,
presents even less obvious similarity, and may have been named in
imitation of its whistling call.
The definition of Drehet pʷisi is a good description of
Aplonis metallica. Yapese gœʔpluw ‘small black bird
sp.’ bears a striking resemblance to the Rennellese word, but given the
lack of precise identification and the absence of known borrowing in
this direction, this is probably coincidental.
The following appears to be a continuation of POc *midi,
reconstructed above (§6.6) as
originally applying to honeyeaters, but extended in Polynesia to the
starlings and toilers, on the basis of similarities which remain
unclear.
Again a single species, the Australian Crow (Corvus orru) is
present in the homeland, with congenerics elsewhere in Melanesia. Both
*(kao)kao and *kaka are likely to show some influence
of widespread representations of crow vocalizations such as English
caw.
The following reconstruction seems very likely to have denoted some
(probably black) passerine species, but the diversity of actual
referents makes it hard to be more precise.
The sea and shore birds have in common that the species tend to be
wide ranging rather than localised, many of them migrants or wanderers.
The larger families contain numerous species whose field discrimination
can be difficult. Observation of these birds becomes more common as sea
orientation becomes more dominant, so that the most precise and detailed
taxonomies are found in Micronesian and Polynesian languages.
Petrels come to land only to breed, and may be known primarily from
the strange noises they make at their burrows at night. Of several
species recorded for the homeland, only the Wedge-tailed Shearwater
(Puffinus pacificus) breeds in the vicinity and is present year
round.
There is a certain degree of crossover apparent between the families
of petrels, albatrosses and boobies.
The two species of tropic-bird (Phaethon) are rare in Papua
New Guinea waters but more frequently encountered in Remote Oceania. The
Red-tailed Tropic Bird (Phaethon rubricauda) and the
White-tailed Tropic Bird (Phaethon lepturus) differ, as their
names suggest, in the colour of the two extremely elongated tail
feathers by which this type of bird is readily recognised. They are
generally covered by a single lexical item, though there may be a
conventional specifier for one or the other, as in Tongan tavake
toto ‘Phaethon rubricauda’ (toto ‘blood’).
Of the three Boobies (genus Sula) found in Oceania, the Brown Booby
(Sula leucogaster) is present year round in New Guinea waters.
The Red-footed Booby (Sula sula), however, is also widely
reported, while the Blue-faced Booby (Sula dactylatra) appears
to be the least common. Only in Polynesia and Micronesia are the three
species lexically distinguished.
The following also appears to be basically a term for a booby sp.,
though there is some spread to other sea birds. (The Maori word refers
to a gannet, the temperate-zone counterpart of the booby, very similar
in appearance and habits.)
The following three forms suggest POc *kalau ‘booby’, but
may be scattered cognates of the previous set. The Nukuoro word is
probably borrowed from Micronesia.
Two species of frigate bird are common in New Guinea waters, the
Great Frigate Bird (Fregata minor) and the Lesser Frigate Bird
(Fregata ariel). (The identifications of other species for
Puluwat and Roviana below are likely to be erroneous.) Two clear POc
reconstructions can be justified, though there seems to be no consistent
lexical differentiation between the two species in the sources for
contemporary languages.
(The glosses for Puluwat yahaf and Roviana belama
are probably errors at the species level, as these frigate birds are not
known in the Oceanic region.)
More than twenty species of these shore birds frequent the coasts of
the Oceanic homeland, but most are migrants seen only seasonally. Among
the most commonly identified are the Bristle-thighed Curlew
(Numenius lahiliensis), Whimbrel (Numenius phaeopus),
Bartailed Godwit (Limosa lapponica), Wandering Tattler
(Heteroscelus incanus), Pacific Golden Plover (Pluvialis
fulva), and Ruddy Turnstone (Arenaria interpres). Field
identification of these birds can be difficult, and even in Polynesia
and Micronesia there are few cognate sets with completely consistent
reference.
The set above is perhaps cognate with the set from which PWOc
*kiwiwi ‘sandpiper sp.’ (p.360) is reconstructed. There are
also probable Western Oceanic cognates, namely Tawala (PT), Gapapaiwa
(PT) kiu ‘bird (generic)’ and Nakanai (MM) e-kivu
‘bird sp.’. Together with the set above these imply the reconstruction
of POc *kiu, but its gloss cannot be reconstructed, as the
Tawala and Gapapaiwa generic presumably reflects the promotion of a bird
taxon term to generic status, and the Nakanai gloss is not specific.
Hote (NNG) kakapʸak, glossed as a ‘small black bird like a
swallow; doesn’t sit in a tree but on rock; morning and afternoon it
catches insects in sky’ is a possible Western Oceanic cognate, but its
similarity to the items above may be due to chance.
The family Laridae includes both gulls and terns. Gulls are mainly
birds of the temperate zones, and are only rare visitors to tropical
Oceania. The common gloss ‘seagull’ in dictionaries and vocabularies of
languages in this area almost certainly indicates a tern. Several
species of tern (Sterna, Thalasseus) are common in the waters of the
homeland, along with noddies (Anous) and the very distinctive White Tern
(Gygis alba).
The following two terms from Polynesian Outliers show a closer
resemblance to the Halia form above than to the rest of Polynesian, and
are probably borrowed from some North Solomons language:
PAdm *baraŋ is reconstructed on the basis of the Loniu and
Nyindrou terms. Mussau rabaŋana points to earlier
*rabaŋan (or *baraŋan), but the relationship of Mussau
to the Admiralties languages is not well enough understood to know what
protolanguage this earlier term occurred in.
