Chapter 5.6 Posture and movement

1. Introduction⇫

The reconstructions presented in this chapter cover a wide semantic range of verbs concerned with posture and movement. The semantic domain is a complex one, and the reader will find more reference to the typological literature than in other chapters in these volumes, in an effort to make sense of the domain’s structure in Proto Oceanic. Languages structure the posture and movement domain in different ways, and Oceanic languages have their own shared peculiarities in this regard, described at appropriate points in the chapter.

The most important division is into posture verbs and movement verbs. Posture verbs in turn are divided into cardinal posture verbs—those denoting ‘sit’, ‘stand’ and ‘lie’, found in most languages around the world and often used as simple verbs of location (‘be at’)—and non-cardinal posture verbs, which are more culturally specific. The subdomain of movement verbs is much more complex. There are straightforward movement verbs like ‘go’, ‘fly’ and ‘descend’ (§§6.3–6.5), which have a single argument, the theme (the person or thing that is moved), and there are caused movement verbs like ‘raise’, ‘carry’, ‘put’ and ‘send’, which have two arguments: an agent that causes movement and a theme which moves. These complexities are discussed at greater length below.

2. Posture verbs⇫

Posture verbs are verbs meaning ‘sit’, ‘stand’, ‘lie’, ‘hang’, ‘lean’, ‘squat’, ‘kneel’ and the like. These can be divided into cardinal and non-cardinal posture verbs. Cardinal posture verbs have the meanings ‘sit’, ‘stand’ and ‘lie’ and have a broader range of use in many languages than non-cardinal posture verbs, i.e. verbs denoting other postures (‘hang’ etc) and verbs denoting a more specific posture (e.g. ‘sit crosslegged’) than the cardinal posture verbs.

The cardinal posture verbs ‘sit’, ‘stand’ and ‘lie’ have two kinds of use in many of the world’s languages, including Oceanic. In the first use they are simply intransitive verbs denoting the posture of their subject (‘Mary is sitting’). They are often accompanied by a location (‘Mary is sitting on a chair’). This gives rise to their second use, as default verbs in locative constructions (Ameka & Levinson 2007, Lichtenberk 2002). A locative construction is one that answers the question ‘Where is X?’. Across languages locative constructions come in three main kinds:¹

a construction with no copula (‘The cup — on the table’)
a construction with a copula verb corresponding to English be (‘The cup is on the table’),
a construction with a posture verb (‘The cup sits on the table’. ‘The vase stands on the table’).

Oceanic languages tend to have constructions of types 1 and 3, but not type 2. The Seimat (Adm) sentence below is of type 1:²

- Seimat (Adm): (Wozna & Wilson 2005:66)
  ‘The chicken [is] under the house.’
  Tok mom hahitak-e tehu iŋ.
  
  CLF chicken under-CSTR CLF house

‘The chicken [is] under the house.’
Tok	mom	hahitak-e	tehu	iŋ.
CLF	chicken	under-CSTR	CLF	house

The difference between a postural and a locative use is that in the latter, the verb is bleached of its postural meaning. Thus Manam (NNG) has a type 3 locative construction where the verb -eno ‘lie’ is used in locative and existential sentences³ with abstract subjects where there is no postural orientation at all.

- Manam (NNG): (Lichtenberk 2002:277)
  ‘If there is work, I’ll help you.’
  Malipi di-eno, masa n-duma-iʔo.
  
  work S:3P-lie INDEF.IRR s:1S.IRR-help-O:2S

‘If there is work, I’ll help you.’
Malipi	di-eno,	masa	n-duma-iʔo.
work	S:3P-lie	INDEF.IRR	s:1S.IRR-help-O:2S

- Manam (NNG): (Lichtenberk 2002:277)
  ‘I’m sick in the belly.’ (Lit. ‘The sickness is in my belly.’)
  More ilo-gu-lo i-eno.
  
  sickness inner.belly-P:1S-in s:3S.REAL-lie

‘I’m sick in the belly.’ (Lit. ‘The sickness is in my belly.’)
More	ilo-gu-lo	i-eno.
sickness	inner.belly-P:1S-in	s:3S.REAL-lie

With a human subject, on the other hand, -soaʔi ‘sit’ is used in a locative sentence, again with no postural implication:

- Manam (NNG): (Lichtenberk 2002:275)
  ‘Two years I was/stayed/lived in the town.’
  Barasi rua ŋau taun-lo u-soaʔi.
  
  year two I town-in s:1S.REAL-sit

‘Two years I was/stayed/lived in the town.’
Barasi	rua	ŋau	taun-lo	u-soaʔi.
year	two	I	town-in	s:1S.REAL-sit

The choice of -eno ‘lie’, rather than -soaʔi, with a human subject is clearly a postural usage.

Manam (NNG): (Lichtenberk 2002:274)

‘Kadiarang is lying in his house (and is) crying.’
Kadiaraŋ	maka	bale	ne-na-lo	i-eno	be	i-tan~taŋ.
K.	here	men’s.house	CLF-P:3S-in	S:3S.REAL-lie	and	s:3S.REAL-REDUP~cry

As these Manam examples show, the choice of posture verb depends on two factors. The first is whether the intended construction is locative or postural. If it is locative, then the second factor comes into play, namely the conventional collocation of particular nouns or semantic categories of nouns with a given posture verb. These collocations vary from language to language and almost always involve ‘sit’, ‘stand’ or ‘lie’ (Lichtenberk 2002:273–274, 305; see also Early 2000). Non-cardinal posture verbs almost never occur in a locative construction.

The three main posture verbs are also often used as aspectual verbs in Oceanic languages, usually as the second verb of a serial verb construction, but Lichtenberk (2002:270) thinks these meanings are impossible to reconstruct (and I agree with him) as different languages reflect different developments. He also thinks that reconstructing uses of posture verbs as default locative verbs is impossible. Here I am a little more sanguine (§6.2.1).

Typically in Oceanic languages intransitive verbs encode both posture and assumption of posture (inchoative posture). The latter use is sometimes distinguished by a path element meaning ‘upward’ or ‘downward’. Path was probably encoded in POc as the second verb of a serial verb construction, as in Mussau toka sio ‘sit down’ and tiŋina sae ‘stand up’, where sio ‘go down’ and sae ‘go up’ are verbs of vertical direction (vol.2:260-267). Such usages were certainly conventionalised in POc and were perhaps also grammaticised, like Manam go-soaʔi-ria [2SG. IRR-sit-go. down] ‘sit down’ (Lichtenberk 2002:269-270) or Carolinian sɔ̄-tiw [settle-go. down] ‘alight, land (of flying things)’.

Some Oceanic languages have transitives with a location object, e.g. Bugotu mono ‘abide, stay, dwell, be at’ vs monoŋ-i- ‘abide in’; To’aba’ita ʔono ‘sit’ vs ʔonof-i- ‘sit on’ (Lichtenberk 2002:269), Wayan toko ‘squat; sit on one’s heels or haunches’ vs tokoði- ‘squat or sit on’. Our sources are often silent on this matter, but there are sometimes enough reflexes to support the reconstruction of transitives.

In addition to cardinal posture verbs, verbs meaning ‘squat’, ‘kneel’ and ‘hang’ can be reconstructed, but they are not nearly as widely reflected as cardinal verbs. Many languages also cut the main domains into smaller, more specific ones. For example, as well as soaʔi ‘sit’ Manam has basaʔi ‘sit cross-legged’ (Newman 2002:3). However, the data do not permit reconstruction of such semantically narrow posture verbs.

2.1. Sit, be located⇫

Two POc forms meaning ‘sit, stay, dwell’ are reconstructable: *mono(ŋ)/*monoŋ-i and POc *nopo(q). A third form, *mia[n], is reconstructed subject to certain caveats mentioned below. Significantly, a purely postural meaning (‘sit’) is reconstructable for none of these verbs. In all three instances, reflexes tend to mean ‘sit’, ‘stay’ or ‘dwell’ or a combination of these, whereas reflexes of reconstructions meaning ‘stand’ (§6.2.2) and ‘lie’ (§6.2.3) tend to have glosses with only a postural sense. There is a reason for this. In traditional Oceanic societies there were no chairs or custom-made seats. People sat or squatted on the ground, so any ‘sit’ verb would not have had the postural sense of English sit but would have meant ‘be located’ and have been a candidate for service as the default POc locative verb.⁴

Reflexes of the widely reflected POc *toka ‘come to rest, settle (of vessel, on reef)’ also fall into the ‘sit, stay, dwell’ domain, but its POc meaning seems to have been narrower and indeed inchoative, a matter discussed below.

POc		*mono(ŋ)	‘sit, stay, dwell’
POc		*monoŋ-i	‘sit on’
MM	Tabar	mon	‘dwell’
MM	Lihir	mon	‘dwell’
MM	Sursurunga	mon-mon	‘remain, stay behind at home’
MM	Patpatar	mona	‘stay at home; lounge around; keep watch’
MM	Tolai	mono	‘remain and take care of the home, boat etc.’
SES	Gela	mono	‘stop a while, stay’
SES	Bugotu	mono	‘abide, stay, dwell, be (at)’
SES	Bugotu	monoŋ-i	[VT] ‘abide in’
SES	Talise	mono	‘lie down’
SES	Tolo	mono	‘lie down’
SES	Arosi	mono	‘live, dwell, reside’
SES	Arosi	monoŋ-aʔi	[VT] ‘reside at’
SES	Bauro	mono	‘stay in another village’
NCV	Lewo	mono	‘lie down, be horizontal; remain (at)’
NCal	Nengone	meneŋ	‘reside, be in a place’

POc *nopo(q) ‘sit, stay, dwell’ is reconstructed with possible final *-q on the basis of Polynesian reflexes, as Polynesian languages preserve root-final consonants in transitive verbs more faithfully than SE Solomonic languages like Lau below. Wayan Fijian shows two apparent reflexes of *nopo(q). Wayan novo is the regular reflex, whereas nō is irregular, as is Bauan no. The items listed under ‘cf. also’ may be cognate, but they display irregularities.

POc		*nopo(q)	‘sit, stay, dwell’ (Milke 1968) ; ‘sit’ (Blust 1993)
PT	Motu	noho	‘dwell, remain’ (persons, not things)
MM	Patpatar	noh	‘dwell, lie’
SES	Lau	nofo	‘dwell, stop, stay’
SES	Lau	nofo-s-i	[VT] ‘dwell, stop, stay (somewhere)’
SES	Arosi	nohu	‘sit quiet’
NCV	Mota	(va)nov	‘cause to dwell’
NCal	Xârâcùù	nöö	‘stay’ (Moyse-Faurie 2015)
Mic	Carolinian	no	‘be (at), stay, remain, live’
Fij	Bauan	nō	‘lie (of things)’
Fij	Wayan	nō	‘live, be (at), stay, dwell’
Fij	Wayan	novo	‘keep still, be motionless, stay without moving, keep quiet, be subdued, acquiescent, peaceful’
Fij	Rotuman	noho	‘dwell’
PPn		*nofo	‘sit, dwell’
PPn		*nofoq-i	[VT] ‘sit on, dwell in’ (POLLEX)
Pn	Tongan	nofo	‘sit, stay, dwell’
Pn	Tongan	nōfoʔ-ia	‘be constantly occupied’
Pn	Tongan	nōfoʔ-i	[VT] ‘sit on, dwell in, occupy’
Pn	Niuean	nofo	‘sit, dwell’
Pn	Samoan	nofo	‘sit, dwell, stay’
Pn	Samoan	nōfo-ia	‘be occupied’
Pn	Samoan	nōfo-aʔi	[VT] ‘sit on/at, occupy’
Pn	West Uvea	nō	‘sit, be (at)’ (Moyse-Faurie 2015)
Pn	East Futunan	nofo	‘sit, be (at) (of animates)’ (Moyse-Faurie 2015)
Pn	East Futunan	nofoʔ-i	‘be inhabited by’
Pn	East Futunan	nofoʔ-aki	‘stay together’
Pn	Tikopia	nofo	‘sit, stay, dwell, live’
Pn	Tokelauan	nofo	‘sit, reside, stay’
Pn	Pukapukan	nō	‘sit, stay, dwell’
Pn	West Futunan	nofo	‘sit’
Pn	Rapanui	noho	‘sit’
Pn	Marquesan	noho	‘sit, dwell, stay’
Pn	Tuamotuan	noho	‘sit, dwell’
Pn	Tahitian	noho	‘sit, dwell’
Pn	Māori	noho	‘stay, live’
Pn	Māori	noh-ia	‘be sat on, be inhabited’

cf. also:

NNG	Malai	(u)nep	‘dwell’
NNG	Lukep	nepa	‘sit down’
MM	Kia	nohe	‘sit’
SES	Bugotu	nohe	‘sit’

If Drehet (Adm) -miŋ ‘sit’ and/or Ponapean (Mic) mi ‘exist, be (at)’ reflect *mia[n], as they appear to, then POc *mia[n] ‘sit, stay, live’ can also be reconstructed. If not, then the form is of PWOc antiquity. Final -ŋ of Drehet -miŋ appears to reflect an added morpheme, as Admiralties languages lose POc final consonants. Evidence as to whether *mia[n] has a final *-n is ambiguous, and Bali reflects it both without (mia ‘dwell’) and with *-n (miaŋa, a regular reflex).

POc		*mia[n]	‘sit, stay, live’
Adm	Drehet	-mi(ŋ)	‘sit’
NNG	Mangseng	mimi(ŋa)	‘alive’
NNG	Numbami	-mi	‘dwell, live, stay, remain, be (at)’
NNG	Bam	(i)mi	‘dwell’
NNG	Wogeo	mi-mia	‘dwell’
NNG	Sissano (Arop)	-mɛn	‘stay, remain, wait, be (at)’
PT	Iduna	-mia	[V] ‘stay, live’; [N] ‘staying, existence, living’
PT	Dobu	mia(toa)	‘sit’
PT	Saliba	-mia	‘stay’
PT	Ubir	-mian	‘dwell’
PT	Tawala	-mi-mie(ya)	[V] ‘live permanently’; [N] ‘permanent residence’
PT	Dawawa	-mia	‘live, stay’
PT	Dawawa	-(tau)mia	‘stay for a while’
PT	Misima	mi-	‘still; staying’
PT	Misima	mina	‘stay’
PT	Motu	-mia	‘remain’ (things, not persons)
PT	Gabadi	-mia(do)	‘sit’
MM	Vitu	mia	‘sit, dwell’
MM	Bali	mia	‘dwell’
MM	Bali	miaŋa	‘sit’
MM	Tigak	min(aŋ)	‘dwell’
MM	West Kara	mi(ta)	‘dwell’
Mic	Ponapean	mi	‘exist, be (at)’

The primary sense of POc *toka was apparently one of settling in a position following movement, giving rise to glosses like ‘settle down’, ‘hit the bottom (of particles in suspension in a liquid and of canoes’ and ‘land (after flying)’. In many languages its reflex denotes the result of settling, ranging from ‘stuck on the reef’ (Roro [PT]) and ‘be aground’ (Tongan [Pn]) to the more general ‘sit’, ‘stay’ or ‘dwell’. In Poeng, Gela, Sa’a and PCP reflexes occur, sometimes with a qualifier, with the meaning ‘sit on one’s haunches, squat’, and this seems to have been a subsidiary meaning, overlapping with the domain of POc *tike (§6.2.4.1).

A number of reflexes point to a possible alternant *toko, and a few to a short alternant *tok. Thus Nehan toko reflects *tok with a regular echo vowel; the expected reflex of POc *toka is †toa.

PMP		*tekas	‘come to rest in a place’ (ACD)
POc		*toka	‘come to rest, settle (on bottom of vessel, on reef)’
Adm	Mussau	toka	‘sit, live, dwell’
Adm	Tenis	toka(sio)	‘sit’
Adm	Seimat	to	[VI] ‘sit, stay, remain, live, dwell’
Adm	Lou	tok	‘sit, stay, settle down’
Adm	Baluan	tok	‘sit’
Adm	Baluan	to[k]	‘stay, exist, be at’
Adm	Loniu	tɔ[w]	‘be at, live’
NNG	Mengen	toa(kuru)	‘squat’
PT	Dobu	(mia)toa	‘sit’
PT	Molima	toa	‘sit’
PT	Roro	-toʔo	‘be stuck fast on reef’
MM	Lamasong	lok	‘dwell’
MM	Madak	lo	‘dwell’
MM	Bilur	tok	‘(post) be erect’
MM	Halia	(ha)toka	[VT] ‘erect (e.g. a post)’
MM	Babatana	(toka)toka	‘ladder’
MM	Nehan	toka, toko	‘stay put, perch (as a bird)’
MM	Teop	toka	‘perch, stand on’
SES	Gela	toɣa	‘dwell’
SES	Gela	kabu toɣa-toɣa	‘squat on one’s heels; be on sentry duty’ (kabu ‘attentive’)
SES	Sa’a	oʔa	[VI] ‘settle, of birds; squat on haunches’
SES	Arosi	oʔa	‘stay, dwell, abide; to settle, of birds’
SES	Bauro	oɣa	‘sit’
TM	Asuboa	to	‘sit’
PNCV		toka, toko	‘sit, stay, be in a place’ (Clark 2009)
NCV	Mota	toɣa	‘abide, dwell, endure, live, behave, be’
NCV	Mwotlap	toɣ	‘live’
NCV	Raga	toɣo	‘stay, sit, dwell, be’
NCV	Tamambo	toɣo	‘remain’
NCV	Lewo	tō	‘sit, stay, be at’
NCV	Namakir	tok	‘stand, stay, live, be in a place’
NCV	Nakanamanga	toko	‘stay, live; be in, be at’
PSV		*a-toɣ	‘sit, stay, live at, be at’ (Lynch 2001c)
SV	Sye	e-te	‘sit, stay’
SV	North Tanna	a-təŋ	‘live, dwell’
SV	Lenakel	a-rək	‘live, dwell’
SV	Kwamera	a-ra	‘live, stay at, be at’
SV	Anejom̃	a-teɣ	‘sit’
SV	Anejom̃	-teɣ	‘exist, be, stay (of inanimates)’
NCal	Ajië	tō	‘be at’ (Moyse-Faurie 2015)
PMic		*Toka	‘settle, alight’ (Bender et al 2003)
Mic	Marshallese	cekʷ	‘settle (of liquids), alight, land’
Mic	Mokilese	cok	‘swoop’
Mic	Ponapean	sok	‘land (of s.t. that flies), touch bottom with one’s feet in water’
Mic	Woleaian	soxo	‘remain on the bottom (as dregs)’
Mic	Chuukese	so, sō(tiw)	‘precipitate out (as starch derived from washing grated arrowroot or manioc), come to rest, land (of flying things)’
Mic	Puluwatese	hō	‘land (as a plane)’
Mic	Carolinian	sɔ̄(tiw)	‘alight, land (of flying things)’
Mic	Satawalese	sɔ, sɔ̄(tiw)	‘land, alight’
PCP		*toka	‘sit, squat, live, stay, settle, coagulate’ (Hockett 1976)
Fij	Rotuman	foʔa	‘land, come ashore’
Fij	Bauan	toka	‘squat; be placed, situated (of people and small objects), be at’
Fij	Boumā	toʔa	‘be at’
Fij	Wayan	toko	‘squat; sit on one’s heels or haunches; crouch with back of thighs and buttocks resting on the heels’
Fij	Wayan	tokoði-	‘squat or sit on s.t.’
PPn		*toka	‘sit, settle, coagulate, run aground’
Pn	Tongan	toka	‘(boat) run aground, rest on the bottom; (water, particles in suspension, people) come to rest, settle down’
Pn	Samoan	toʔa	‘(liquid) be still, settle; (boat) run aground; (people) stop, settle’
Pn	Tikopia	toka	‘stop, cease (blood flow etc), quieten down (sea)’
Pn	Māori	to-toka	‘solidify, set, congeal’

2.2. Stand⇫

Two POc verbs for ‘stand’ are reconstructable: *tuqur ‘stand’ and *madriRi ‘be standing upright’. Both have PMP antecedents. Apparently no language in the data reflects both terms, with a reflex of *tuqur sometimes occurring in one language and a reflex of *madriRi in a quite closely related language. The competition between them seems to have been won in EOc languages by *tuqur, as there are no EOc reflexes of *madriRi.

PNCV appears to have had two reflexes of POc *tuqur, namely *tuqu and *tuquru (Clark (2009) reconstructs *tuqu-ru). PNCV *tuʔu is the expected reflex. PNCV *tuquru reflects the POc form plus an echo vowel, a minority pattern that occurs from time to time in PNCV.

The Choiseul forms shown under ‘cf. also’ have expected consonant reflexes but the ‘wrong’ vowels. It is tempting to attribute them to an ‘Old Oceanic’ substrate (Ross 2010:263).

A set of forms reflecting apparent POc *tutu follows below, and may or may not ultimately reflect POc *tuqur.

PMP		*tuqud	‘stand’ (ACD)
POc		*tuqur	‘stand’
Adm	Seimat	tua	[VI] ‘stand’ (expected reflex is tu)
Adm	Seimat	tu	[VI] ‘sit, stay, remain, live, dwell’
Adm	Baluan	tu	(renders preceding stative verb ongoing)
NNG	Ali	tou	‘stand’
NNG	Mindiri	tekur	‘stand’
NNG	Takia	-tur	[VI] ‘stand up, be in erect position’
NNG	Patep	laɣ	‘stand up’
PT	Iamalele	tovoi	‘stand’
PT	Dobu	tōlo	‘stand’
PT	Bunama	toholo	‘stand’
PT	Tawala	towolo	‘stand, wake; (become) leader’
MM	Lavongai	tuŋ	‘stand’
MM	Tigak	tuk	‘stand’
MM	Nalik	tur	‘stand’
MM	Tabar	turi	‘stand’
MM	Notsi	til	‘stand’
MM	Patpatar	tur	‘stand’
MM	Sursurunga	tur	‘stand’
MM	Siar	tur	‘stand’
MM	Ramoaaina	tur	‘stand’
MM	Tolai	tur	[VI] ‘be, exist, stand, stop’
MM	Nehan	turu	‘stand’
MM	Selau	tur	‘stand’
MM	Tinputz	sun	‘wake up, get up, stand up’
MM	Teop	sun	‘stand’
MM	Taiof	tu-tun	‘stand’
MM	Banoni	ciɣom	‘stand’
MM	Piva	cuɣonu	‘stand’
MM	Nduke	turu	‘stand’
MM	Roviana	turu	‘stand’
SES	Gela	tuɣuru	‘stand’
SES	Bugotu	tū	‘stand’
SES	Lengo	tu	‘stand’
SES	Kwara’ae	ū	‘stand’
TM	Engdewo	tu	‘stand’
TM	Tanibili	sū	‘stand’
PNCV		tuqu, tuquru	‘stand’ (Clark 2009)
NCV	Mota	tur	‘stand, be sufficient, prevail’
NCV	Tamambo	turu	‘stand, be at’
NCV	Lonwolwol	(rd)u	‘remain, rest, stop, stay, endure, last; be, continue (of process or action); exist; keep on’
NCV	Raga	tu	‘stand’
NCV	Raga	tu(a-mare)	‘stand (up), arise’
NCV	Kiai	turu	‘stand’
NCV	Sakao	tür	‘stand (in line)’
NCV	Uripiv	-tur	‘stand’
NCV	Port Sandwich	tül	‘stand’
NCV	Labo	tox	‘dwell’
NCV	Lewo	sū	‘be at, remain, stay’
NCV	Lewo	su([m,mʷ]alu)	‘stand, get up’
NCV	Namakir	tu(marak)	‘stand’ (marak ‘rise’)
NCV	Nguna	dū	‘stand, be standing’
NCV	Nguna	tu(leana)	‘get up, stand up, rise’ (leana ‘straight’)
NCV	Neve’ei	tur	‘stand’
NCV	Rerep	tu	‘stand’
NCV	North Ambrym	to-tor	‘stand’
SV	Sye	e-tur	‘stand’
SV	Whitesands	a-tul	‘stand’
NCal	Jawe	cūt	‘stand’
NCal	Caaàc	cōr	‘stand’
NCal	Caaàc	cu-	‘stand’ (in compounds; Cauchard 2014:97)
NCal	Iaai	teṭ, tooṭ	‘stand’
PMic		*tuu	‘to stand; stopped, halted’ (Bender et al 2003)
Mic	Kosraean	tu	‘stand up; stop’
Mic	Marshallese	ciw(tak)	‘stand up’
Mic	Mokilese	u	‘stand’
Mic	Mokilese	ū(ta)	‘stand up’
Mic	Ponapean	ū	‘stand’
Mic	Ponapean	ū(ti)	[VI] ‘stop (of moving objects)’
Mic	Pulo Annian	θɨ̄	‘stand; stop’
Mic	Woleaian	sɨ̄	‘stand, take an upright position’
Mic	Carolinian	ɨ̄(tæ)	‘stand up’
Mic	Carolinian	ɨ̄(lɔ)	‘stop’
Mic	Puluwatese	wɨ̄(tæ)	‘stand up, rise’
Mic	Chuukese	wɨ	‘stand erect, be upright; be stopped, halted (of s.t. in motion)’
PCP		*tuqu	‘stand, be somewhere’
PCP		*tuqur-a	[VT] ‘stand on/near s.t.’
PCP		*tuqur-aki-	‘stand up with’
Fij	Rotuman	fū	‘stand, be situated’
Fij	Bauan	tū	[VI] ‘stand, be in a place’
Fij	Bauan	tūr-a	[VT] ‘stand on/near s.t.’
Fij	Bauan	tur-ak-a	‘stand up with’
Pn	Tongan	tuʔu	[VI] ‘stand; stop; be situated, exist’
Pn	Tongan	tuʔul-aki	[PLURAL SUBJECT] ‘stand in position, form up’
Pn	Rennellese	tuʔu	‘stand, be upright’
Pn	Tikopia	tū	‘stand, remain’
Pn	Tikopia	tūr-ia	‘be stood by’
Pn	Tikopia	tūr-aki	[PLURAL SUBJECT] ‘stand (together)’
Pn	Samoan	tū	‘stand, stand up, stand erect; be stationary’
Pn	Samoan	tūl-aʔi	‘stand up, get up on one’s feet’
Pn	Samoan	tul-ia	‘be stood on, occupied, guarded’
Pn	East Futunan	tuʔu	‘stand, be at’ (Moyse-Faurie 2015)
Pn	East Futunan	tuʔul-aki	‘walk together, walk carrying a crying child’
Pn	Tahitian	tū	‘stand, be upright’
Pn	Māori	tū	‘stand, be upright’
Pn	Māori	tūr-ia	‘be arranged, entered upon’

cf. also:

MM	Vaghua	deɣere	‘stand’
MM	Varisi	deɣere	‘stand’
MM	Ririo	der	‘stand’
MM	Babatana	dere	‘stand’
MM	Sisiqa	dere	‘stand’

The set below appears to reflect something like POc *tutu. It is not clear whether this is an idiosyncratic alternant of *tuqur or not. What is obvious is that the first *-u- has a variety of reflexes, suggesting that its POc stress fell on the second syllable, leading to reduction or loss of the first *-u-.

POc		*tutu	‘stand’
Adm	Lou	a-tatu-t	‘stand firm’
Adm	Nyindrou	a-tutu-n-iy	[VT] ‘put (it) up, stand (it) up’ (a- CAUSATIVE)
Adm	Nyindrou	tutu-n-i	[VT] ‘put up, set up, stand up, erect’
NNG	Manam	tu(i-rake)	‘stand up’ (rake ‘go upward’)
NNG	Sissano	-to	‘standing’
PT	Kilivila	-totu	[VI] ‘stand’
PT	Kilivila	-tito-k-i	[VT] ‘stand on’
MM	Patpatar	tut	‘rise, stand up’
MM	Roviana	tutu	‘stand’
MM	Kia	tetu	‘stand’
MM	Kokota	tetu	‘stand’
MM	Blablanga	tʰetu	‘stand’

The set below is restricted to WOc languages but is clearly inherited from PAn. However, a PMP form with *ma-, i.e. *ma-diRi, has not been reconstructed, and would in any case have resulted in POc †*ma-riri. Instead the presence of *-dr- in POc *madriRi suggests earlier †*ma-n-diRi. CMP languages with a nasal prefix occur (the ACD lists Fordata n-diri and Kei en-dir), but not enough is known about pre-Oceanic diachronic morphology to allow a firm pre-Oceanic reconstruction.⁵

PAn		*diRi	‘stand’ (Blust 1999)
POc		*madriRi	‘be standing upright’
NNG	Bariai	-mad-madid	‘stand’
NNG	Kilenge	-mari	‘stand’
NNG	Tami	moji	‘stand’
NNG	Sio	-madi	‘stand’
NNG	Mangap	meⁿder	‘stand up, be in vertical position; rest from, stop, take a break from activity’
PT	Dawawa	-(wai)midiri	‘stand s.t. up’
PT	Misima	mílil	‘stand’
MM	Bali	madiri	‘stand’
MM	Vitu	madi	‘stand’
MM	Nakanai	magiri	‘stand’

cf. also:

Meramera

magili

‘stand’ (apparently a Nakanai loan)

The NNG and PT forms listed below may also reflect POc *madriRi. They allow the reconstruction of putative PNGOc *midi ‘stand’. If the latter reflects POc *madriRi, then two idiosyncratic changes must be inferred: (a) *ma- became *mi- by assimilation to the *-i- of *-driRi; (b) loss of the final syllable *-Ri. Change (a) is also reflected in Dawawa -(wai)midiri ‘stand s.t. up’ and Misima mílil above, and change (b) in Kilenge -mari, Tami moji, Sio -madi and Vitu madi. It is thus not unexpected that the two changes have sometimes occurred simultaneously, and the terms listed below may well reflect independent parallel innovations rather than a single PNGOc innovation.

PNGOc		*midi	‘stand’
NNG	Amara	mid	‘stand’
NNG	Apalik	mid	‘stand’
PT	Iduna	-midi	‘stand’
PT	Gapapaiwa	-mīri	‘stand; get up’
PT	Dawawa	-midi	‘stand’
PT	Nimoa	-midi	‘stand’

cf. also:

NNG	Numbami	-ⁿdi	‘stand, bask (in/at)’
NNG	Misim	(va)diŋ	‘stand’ (the presence of the causative prefix va- is anomalous)

2.3. Lie⇫

One POc term, *qenop, is reconstructed for ‘lie, rest horizontally’. A good many reflexes also mean ‘sleep’, but this is a natural development from ‘lie’. Distinct terms for ‘sleep’ are reconstructed in §4.6.1.

Blust (ACD) reconstructs the PAn term *qinep, of which the expected POc reflex is †*qinop. He also reconstructs the doublet PEMP *qenəp,⁶ reflected as POc *qenop. He apparently infers that both forms were inherited into POc because a few terms in -i- occur in NNG languages, listed below under ‘cf. also’. One might alternatively infer, however, that these reflect a NNG height assimilation. Also listed under ‘cf. also’ is the form inep, reflected in three South New Ireland languages. The presence of -e- for †-o- in the second syllable is unexplained.

The distribution of reflexes of *qenop shows some curious restrictions. Regular reflexes occur nowhere in MM, in apparently only a few northern NCV languages, and not at all in Micronesian or Central Pacific languages. In each of these areas other terms, some of them words for ‘sleep’, have replaced *qenop, but it is not clear why this replacement is so widespread. The only fairly widespread replacement is PCP *koto/PPn *ta-koto below. PPn ta- appears to reflect POc *ta-, which marks the event denoted by the verb as spontaneous. However, it is difficult to conceive of lying down as spontaneous.

PAn		*qinep	‘lie down to sleep’ (ACD)
PEMP		*qenəp	‘lie down to sleep’ (ACD)
POc		*qenop	‘lie, rest horizontally’ (ACD)
Adm	Seimat	eŋ	[VI] ‘stay in one place, lie down, sleep’
Adm	Lou	en	‘lie down’
Adm	Nauna	en	‘lie down’
NNG	Manam	eno(ria)	‘lie down’
NNG	Wogeo	-eno	‘sleep’
NNG	Gedaged	en	‘lie, rest, sleep’
NNG	Matukar	en	‘sleep’
NNG	Bariai	eno	‘lie, sleep; be located’
NNG	Tuam	-ɣēn	‘sleep’
NNG	Gitua	ɣeno	‘sleep’
NNG	Lukep	kien	‘sleep’
NNG	Mangap	-keene	‘sleep, lying down, be in horizontal position’
NNG	Sio	keno	‘lie (of inanimates)’
NNG	Numbami	-e	‘lie (of inanimates)’
NNG	Mengen	keno	‘lie down’
NNG	Avau	kene	‘sleep’
NNG	Kela	-en(si)	‘lie’
PT	Ubir	en(rir)	‘lie down’
PT	Molima	ʔeno	‘sleep, lie down, remain’
PT	Tawala	eno	‘sleep’
PT	Saliba	keno	‘lie’
PT	Dobu	ʔeno	‘sleep’
PT	Dobu	ʔeno-ʔeno	‘lie down’
PT	Motu	eno	[VI] ‘lie down’
PEOc		*qenop	‘lie, rest horizontally’
PEOc		*qenop-i-	[VT] ‘lie on, rest on s.t.’
SES	Gela	eno	‘lie down’
SES	Gela	enov-aɣi	‘lay s.t. down’
SES	Talise	eno	‘lie down’
SES	Birao	eno	‘lie down’
SES	Lengo	eno	‘lie down’
SES	Longgu	eno	‘lie down’
SES	Lau	eno	‘lie down’
SES	Kwaio	eno	‘lie, sleep’
SES	’Are’are	eno	‘lie down’
SES	Sa’a	eno-eno	[VI] ‘lie down’
SES	Sa’a	enoh-i	[VT] ‘lie in/on’
SES	Arosi	eno	‘lie down, rest’
SES	Arosi	enoh-i	[VT] ‘rest on’
NCV	Hiw	en	‘lie down’
NCV	Mwotlap	en(hiy)	‘lie down’
NCV	Tamambo	eno	‘lie down, stay, be at’
NCV	Tamambo	enov-i	‘lay s.t. down’
NCV	Ambae	eno	[VI] ‘lie’
NCV	Raga	eno	‘lie down, be’
NCV	Maewo	eno	‘lie down’
NCal	Caaàc	kẽ-	‘lie’ (in compounds; Cauchard 2014:97)

cf. also:

NNG	Roinji	kinu	‘sleep’
NNG	Tami	gin	‘sleep’
MM	Ramoaaina	inep	[VI] ‘lie down, sleep’
MM	Kandas	inep	‘sleep’
MM	Tolai	inep	‘lie down, sleep’

PCP		*koto	‘lie down’
Fij	Bauan	koto	‘lie in a place’
PPn		*takoto	‘lie down’ (POLLEX)
Pn	Tongan	tokoto	‘lie down (of persons and animals)’
Pn	Niuean	takoto	‘lie down’
Pn	Samoan	taʔoto	‘lie’
Pn	East Futunan	takoto	‘lie down’
Pn	Kapingamarangi	takoto	‘lie down’
Pn	Takuu	takoto	‘lie down, recline’
Pn	Tikopia	takoto	‘stretch out, lie prone’
Pn	Nukuoro	dagodo	‘lie down, be in place’
Pn	Luangiua	kaʔoko	‘remain (things), leave, lie down (?)’
Pn	Tongarevan	takoto	‘lie down’
Pn	Pileni	takoto	‘lie down; abide, dwell’
Pn	Pukapukan	takoto	‘remain, be left over, be established’
Pn	Rarotongan	takoto	‘lie or rest in a reclining position’
Pn	Rennellese	takoto	‘lie down’
Pn	Tahitian	taʔoto	‘sleep, lie, lie down’
Pn	Tuamotuan	takoto	‘recline, lie down, be in a reclining position’
Pn	Māori	takoto	‘lie down’

2.4. Non-cardinal posture verbs⇫

2.4.1. Squat, sit on haunches⇫

Sitting on one’s haunches is still a posture frequently assumed in traditional villages in Northwest Melanesia, so the reconstruction of POc *tike ‘squat, sit on haunches’ is no surprise. More of a surprise is that its reflexes are few, perhaps because its meaning is not collected by linguists who use a standard word list, but their distribution points clearly to its presence in POc.

POc		*tike	‘squat, sit on haunches’
Adm	Lou	tik-tik	‘squat, hunker down, sitting on one’s toes’
NNG	Kove	(po)tike	‘squat, sit on heels’
PT	Kilivila	siki	‘sit down’
PT	Kilivila	sigigi(na)	‘squat on haunches’ (-g- for †-k-)
MM	Bola	(pa)sike	‘crouch’
PPn		*tike	‘squat’
PPn		*tike-tike	‘keep squatting’ (POLLEX)
Pn	Tongan	sike	‘squat, sit on heels’
Pn	Tongan	sike-sike	‘keep squatting’
Pn	Niuean	tike-tike	‘squat on the heels’
Pn	Samoan	tiʔe-tiʔe	‘ride, sit astride; sit (on a chair); to be seated on s.t. above the ground’
Pn	East Futunan	tike-tike	‘squat on heels’
Pn	Rennellese	tike-tike	‘crouch or squat, as on the heels’

2.4.2. Kneel⇫

No POc term for ‘kneel’ has been reconstructed, but a small cognate set points to PWOc *tudruŋ ‘kneel’. This is evidently related to POc *turu- ‘knee, joint’ (§3.6.5.2). As dr is the prenasalised equivalent of *r, it is possible that *tudruŋ reflects infixation of PMP *⟨um⟩, which formed intransitive verbs, i.e. *tudruŋ reflects earlier *t⟨um⟩uruŋ (§1.3.5.5). Final *-ŋ is unaccounted for.

PWOc		*tudruŋ	‘kneel’ (< (? ) POc *t⟨um⟩uruŋ)
NNG	Sio	tudu	‘kneel down’
MM	East Kara	turuŋ	‘kneel’
MM	Patpatar	tu-tuduŋ	‘kneel down with head to ground; bowed down’
MM	Ramoaaina	tu-tuduŋ	‘kneel’

Nakanai in the west and Fijian and Polynesian in the east have compounds with two components, the first a verb meaning ‘stand (a pole or spear) up straight’ or a noun meaning ‘supporting pole, stake’ and the second the term for ‘knee’. They form a collocation which perhaps occurred in POc and which one might translate into English as ‘to knee-stand’. The data below do not allow a POc reconstruction, but the Polynesian terms point to a possible PPn *toko-turi (*toko (V) ‘support, prop up’, (N) ‘supporting pole, stake’; *turi ‘knee’).

MM	Nakanai	toto-tulu	‘kneel’ (toto ‘stand s.t. up’; tulu ‘knee’)
Fij	Bauan	teki-duru	‘kneel’ (teki ‘stick upright in ground’; duru ‘knee’)
Fij	Wayan	tiki-turu	‘kneel’ (tiki ‘stand s.t. right side up’; turu ‘knee’)
Pn	Niuean	faka-toko-tui	‘kneel’ (toko ‘supporting pole’; tui ‘knee’)
Pn	Samoan	toʔo-tuli	‘kneel’ (tuli ‘knee’)
Pn	East Futunan	toko-tuli, tuʔu-tuli	‘kneel’ (tuʔu ‘stand’, tuli ‘knee’)
Pn	Tongan	tū-ulu-tui	‘kneel’ (tui ‘knee’)

2.4.3. Hang⇫

Of the postures, ‘hang’ has proven the most difficult to reconstruct forms for, but the reason for this is clear. Oceanic languages have a variety of intransitive verbs of hanging (some of which seem to be used transitively without derivation), but they differ from each other semantically, and the same was presumably true of POc. Often dictionaries, let alone wordlists, do not provide distinctive glosses for ‘hang’ words. The following three lists are illustrative.

**Nehan** (MM) (Glennon & Glennon 2006)
abeke	‘hang above’
horohoro	‘hang down’
uakusu	‘snag on something, hang against something’ (< POc *paqus ‘bind, lash; construct by tying together’)
lolo	‘tie up, hang up’
lilioro	‘hang; drape especially around neck or shoulders’

**Gela** (SES) (Fox 1955)
soɣa-mate	‘hang’
tari	‘hang’
havi	‘hang in a cluster’
kiloro	‘hang down’
ki-kiloro	‘hang out of hole’
saūdola	‘hang down’
leleɣi	‘hang down (as fruit)’
loɣoti	‘hang by rope’
loiloi	‘hang down (of pigs’ genitals)’
loro	‘hang down to the ground (of loin cloth)’
loro-loro	‘hang very low when carried’
salala	‘hang on a line (as clothes)’
saupiri	‘hang down in front; hang down suspended from neck’
soro	‘hang up’
sasarau	‘hang on a peg; hang (as a bat)’

**Wayan Fijian** (SES) (Pawley & Sayaba 2003)
lili	‘hang, be suspended’
riri	‘hang, be suspended by a cord or by draping over a support’
robe	‘droop, hang down’
teteru	‘hang down, droop (of laden branches, fringe of a mat, hair)’
tiro	‘hang down, droop (a dress, line, banana leaves)’

The most widely reflected, and perhaps the most neutral, term for ‘hang’ is POc *tau(r).

POc		*tau(r)	‘hang, be suspended’
NNG	Labu	-towi	[VI] ‘be hanging’
PT	Motu	-tau-a(dae)	‘hang up, of s.t. with string’ (-dae ‘go down’)
PT	Dawawa	dawe	‘hang’ (d- for †t-)
PT	Sinaugoro	tou-kau	‘hang up’
MM	Bola	tau	‘hang up’
SES	Longgu	dau	[VI] ‘hang down; drop anchor’ (d- for †t-)
SES	Longgu	dau-dau(lele)	‘be hanging’
SES	Longgu	daur-aʔini-a	[VT] ‘hang s.t. up’
SES	Kwaio	tao	‘hang down (of leaves and branches); hanging down in front (of a person’s hair)’
Fij	Wayan	tau	[VI] ‘be located, situated, positioned, placed; be in, at or on a place’
Pn	Tongan	tau	[VT] ‘hang’
Pn	Niuean	tau	‘hang, suspend’
Pn	East Uvean	tau-tau	‘hang, suspend’
Pn	Rennellese	tau	‘hang (as clothes), wear (as necklace)’
Pn	Pukapukan	tau-tau	‘hang, suspend’
Pn	Samoan	tau	‘hang (clothes on line, basket on hook etc.)’
Pn	Samoan	tāu-taul-ia	‘be hung up, be opposed’
Pn	Tikopia	tau	‘hang, as necklace on neck’
Pn	Tahitian	tau-tau	‘hang down’
Pn	Hawaiian	kau	‘hang, perch, rest’
Pn	Māori	tau-tau	‘hang suspended’

There is evidence to allow a more specific gloss for POc *tuku below, ‘(person) hang by the arms, dangle’, but data on which POc *sorop ‘hang’ and POc *kuRu (VI) ‘hang’ are based do not allow a more specific gloss. In the *kuRu set the To’aba’ita gloss indicates the kind of hanging the verb denotes, but a single gloss is an insufficient basis for a POc gloss.

POc		*tuku	‘(person) hang by the arms, dangle’
NNG	Lukep	-tuk-a	‘hang (as when climbing a tree trunk)’
NNG	Mutu	-tūk	‘hang (of a broken branch)’
NNG	Patep	yuɣ	‘(be) hanging’
Fij	Wayan	tuku	‘hang, dangle (while holding on to a support with one’s arms)’
Fij	Wayan	tukuci-	‘hang onto s.t. by the arms’

cf. also:

NNG

Bariai

tututu

‘hang’

POc		*sorop	‘hang’
Adm	Lou	so	‘hang’
Adm	Lou	sor-ek	‘hang s.t. up’
NNG	Mapos Buang	sō	‘swing, hang’
MM	Nehan	horo-horo	‘hang down’ (h- and s- both reflect POc *s-)
MM	Nehan	sosoro	‘hang s.t. up’
SES	Gela	soro	‘hang s.t. up’
SES	Gela	sorov-aɣi	‘hang (s.t. ) on s.t.’

cf. also:

NNG

Lukep

-saur-ai

‘hang s.t. up; get snagged’ (-au- for †-o-)

POc		*kuRu	[VI] ‘hang’
MM	Banoni	va-ɣū	‘hanging (of fruit etc)’
MM	Banoni	va-ɣu-ɣuru	‘hang up’ (va- CAUSATIVE)
SES	To’aba’ita	kulu	[VI] ‘hang down, be suspended (of things too short to dangle)’
SES	To’aba’ita	kulu-fani-	[VT] ‘hang, suspend’
SES	Lau	(dau) kulu-kulu	[VI] ‘hang’
SES	Lau	ku-kulu-	[VT] ‘hang up’
SES	Lau	fā-kulu	[VT] ‘hang up’ (fā- CAUSATIVE)
SES	Lau	fā-ku-kulu	‘cause to hang down’
SES	’Are’are	kuru-a	‘hang up, suspend’

The cognate set below formally resembles the set above, but this is probably a matter of chance.

POc		*(u)kuku(t)	‘hang, suspend’
NNG	Kaulong	kuk	‘hang, suspend’
PT	Misima	kuki	‘hang, swing’
MM	Patpatar	kukūt	‘hang oneself’
SES	Lau	ukūku	‘hang down (as vines)’
SES	’Are’are	kuku	[VI] ‘hang down’
SES	’Are’are	kuku-a	[VT] ‘hang up, suspend’
SES	Sa’a	ʔuʔu	‘hang down, depend’
SES	Ulawa	kuku	[VI] ‘hang down, depend’
SES	Ulawa	haʔa-kuku	[VT] ‘hang up, to suspend’

2.4.4. Lean, slant⇫

POc *pʷaralat ‘be leaning, slanting’ is a posture verb, but the glosses suggest that its typical use was with inanimate rather than human subjects. The medial consonant correspondences are irregular, and the most natural inference is that Seimat reflects *-r- and *-l- regularly (as do the Polynesian reflexes), whilst the remaining languages reflect assimilations. The Malaita-Makira form hatara is probably cognate, but its history is less clear.

POc		*pʷaralat	‘be leaning, slanting’
Adm	Seimat	palaho	[VI] ‘lean, slant’ (-o for †-a)
MM	Tolai	balala-n	‘crooked, aslant, as a tree; to sag’ (for †varala)
MM	Teop	vananata	‘lie on a slant or angle’
SES	Gela	varava	‘lean on physically’ (for †varala)
SES	Gela	varavaɣ-i	‘rely, lean on’
SES	Tolo	varara	‘leaning’ (for †varala)
SES	Tolo	vararah-i	‘lean on, lean against’
SES	Longgu	varara	‘lean against, top part touching but bottom part separate (e.g. of two trees in an X shape)’ (for †varala)
PPn		*falala	‘lean, stoop, slant’ (POLLEX)
Pn	Tongan	falala	‘lean on or against; rely, trust, confide’
Pn	Samoan	falala	‘lean (as a coconut tree leaning over the water)’
Pn	Rennellese	hagaga	‘slanting, leaning, bent; lean against’
Pn	Kapingamarangi	halala	‘slant; oblique (angle)’
Pn	Māori	farara	‘lean, stoop, decline (as the setting sun)’

cf. also:

SES	’Are’are	hatara	‘leaning (on s.t.)’ (for †harara)
SES	Sa’a	hatara	‘lean, rest upon’ (for †harala)
SES	Arosi	hatara	‘rest upon, lean, press’ (for †harara)

3. Manner of movement verbs⇫

There is a copious literature on movement verbs and movement constructions in the world’s languages, much of it stemming from the work of Leonard Talmy (1972, 1974, 1985, 1991, 2000). Talmy observed that languages tend to have either movement constructions in which the main verb encodes the manner of motion, as in English as in He ran greinto the cave, or movement constructions in which it encodes the path of motion, as in Spanish Entró corriendo a la cueva [entered.3SG running to the cave]. Path is encoded in the English example by the preposition into, and manner in the Spanish example by the participle corriendo ‘running’.⁷

These English and Spanish movement constructions each have just one finite verb. Talmy’s classification also allows for serial verb constructions like those which occur in Oceanic languages, discussed in Ross (2004a) and briefly in vol.2(257–258), and illustrated in this Hoava (MM) example:⁸

- Hoava (MM): (Davis 2003:155)
  ‘The possum came running out.’
  Hagala vura mae sa manue.
  
  run go.out come ART:S possum
  
  MANNER PATH DEIXIS … …

‘The possum came running out.’
Hagala	vura	mae	sa	manue.
run	go.out	come	ART:S	possum
MANNER	PATH	DEIXIS	…	…

In this construction, the elements of manner, path and deixis are each encoded by a finite verb, respectively of locomotion, geographic direction (vol.2:259–267) and deictic direction (vol.2:269–282). This ordering also occurs in other language families, e.g. in Thai (Thepkanjana 1986:136). Omission of any one of the three elements is quite common, as here in Hoava.

Hoava (MM): (Davis 2003:155)

‘The child climbed up the coconut tree.’
Keba	saɣeli-a	sa	koburu	sa	ŋohara.
climb	go.up.TR-O:3S	ART:S	child	ART:S	coconut
MANNER	PATH	…	…	…	…

Elaboration of this manner–path–deixis schema, often by the presence of more than one path verb, is quite common across Oceanic languages, as in this Tuam (NNG) sentence:⁹

Tuam (NNG): (Bugenhagen 2007)

‘Our canoe/boat went coastwards and floated to that place there.’
Waax	tsiei	i-looɣ	i-pot	i-zi	nuɣ	tawe.
canoe	our.EXC	s:3S-enter	s:3S-float	s:3S-descend	place	that.over.there
…	…	PATH	MANNER	PATH	…	…

Talmy’s original classification of movement constructions and verbs did not include a deictic element, but Talmy (2008) treats deixis as a subpart of path. Path and deixis are typically encoded separately in Oceanic languages, but in some of them verbs of deixis have been grammaticised as a separate morpheme class (Ross 2004a).

The remainder of this section is concerned with verbs of locomotion, i.e. verbs that occur in the manner slot of a movement construction. Verbs of geographic and deictic direction, filling the path and deixis slots, are reconstructed in vol.2(259–272).

Verbs of locomotion can be conveniently divided according to the medium on/in which movement takes place: on land (§6.3.1), in the air (§3.2) or in water (§3.3).

3.1. Locomotion on land⇫

Verbs of land-based locomotion include meanings like ‘go’, ‘walk’, ‘run’, ‘crawl’, ‘creep’, ‘limp’, ‘hop’ and ‘roll’.

3.1.1. Go, move⇫

The broadest of these meanings is ‘go’, which in some Oceanic languages also denotes sea- based locomotion. Traditional travel on land was always on foot, and so the ‘go’ verb often also means ‘go on foot, walk’. The verbs of deictic direction away from the speaker or towards a location other than the speaker, described in vol.2(273-282), all seem to have been polysemous insofar as they also served as verbs of locomotion with the basic meaning ‘go’. These verbs were:

POc *lako/*la ‘go (away, to)’
POc *pano/*pa ‘go (away)’
POc *ua ‘go towards addressee’
POc *[y]aku ‘go, go away’

These are presented and their locomotion uses discussed below. The verbs *lako (vol.2:258, 269-270, 277-279) and *pano (vol.2:279–280), also had short forms, la and pa, and in a number of languages the short form displaced the long form altogether. A few languages—Gitua (NNG), Manam (NNG), Mota (NCV), Woleaian (Mic), Puluwatese (Mic)—have verbal reflexes of a long and a short form side by side.

Where a form reflecting *lako or *laka (§6.3.1.3 below) has lost its final vowel, it can be difficult to know which verb it reflects. Tuam -laɣ and Kaiep -lak are assumed to be cognate with Gitua -laɣo and Manam/Wogeo -lako respectively, and Sursurunga lək-lək is taken to reflect *lako because the root lək(ə) occurs in the compound ləkə(m) ‘come’, reflecting *lako + *mai, a combination occurring in other MM languages.

PMP		*lakaw	‘move, go, walk’ (ACD)
POc		*lako	‘go (away, to)’
Adm	Mussau	lao	‘go far’ (la in compounds)
NNG	Mengen	lao	‘go’
NNG	Kove	la-lao	‘walk’
NNG	Tuam	-laɣ	‘walk’
NNG	Gitua	-laɣo	‘go, walk’
NNG	Manam	-lako	‘go, move away’
NNG	Wogeo	-lako	‘go’
NNG	Kaiep	-lak	‘go’
PT	Kilivila	-la	‘go’ (Senft 2000)
PT	Molima	-nao	‘go’
PT	Dawawa	-naɣo	‘go’
PT	Saliba	-lao	‘go, go across’
PT	Magori	-yao	‘go’
PT	Sinaugoro	-iaɣo	‘go’
PT	Roro	-ao	‘go, walk’
MM	Bola	laɣo	‘walk’
MM	Sursurunga	lək-lək	‘walk’
MM	Sursurunga	ləkə(m)	‘come’ (< POc lako + mai ‘come’)
MM	Solos	nao(meh)	‘come’ (< POc lako + mai ‘come’)
MM	Hahon	nao	‘go’
MM	Tinputz	no(h)	‘go; walk’
MM	Tinputz	no(mah)	‘come’ (< POc lako + mai ‘come’)
MM	Teop	nao	‘go’
MM	Teop	no(mo)	‘come’ (< POc lako + mai ‘come’)
MM	Papapana	nau	‘go’
MM	Torau	lao	‘go’
MM	Torau	lao(ma)	‘come’ (< POc lako + mai ‘come’)
MM	Mono-Alu	lao(ma)	‘come’ (< POc lako + mai ‘come’)
MM	Nduke	laɣo	‘go’
MM	Kia	lao	‘go’
MM	Kokota	lao	‘go’
SES	Sa’a	laʔo	‘approach; go/steer toward (a place)’
SES	Bauro	raɣo	‘go’
SES	Fagani	raɣo	‘go’
NCV	Mota	laɣo	‘step, stretch the legs’
NCV	Raga	laɣo	‘walk, travel’
NCV	Nokuku	lako	‘step, walk, go’
Mic	Kiribati	nako	‘go’
Mic	Chuukese	ɾɔ	‘go, walk’
Mic	Woleaian	nako	‘go’
Mic	Puluwatese	lɔ	‘go, walk’
Mic	Pulo Annian	ɾaxo	‘go, walk’
Fij	Rotuman	laʔo	‘go’
Fij	Bauan	lako	‘go’

The long vowel in ’Are’are and Oroha rā below probably reflects a preference for bimoraic roots.¹⁰ However, Andrew Pawley (pers. comm.) suggests that perhaps all forms reflecting _ *la_ are bimoraic, but that length is not represented in their orthographies. If this is so, then the POc form should also be reconstructed as bimoraic, i.e. *lā.

The forms listed below under ‘cf. also’ appear to reflect a POc form *lae, but it is not clear how this might have differed from *la. It is possible that these forms represent independent developments (perhaps associated with bimoraicity preference), the more so as the sequence *-ae is not otherwise found in reconstructed POc.

In the short-form set below Loniu (Adm) -lɛ, Titan (Adm) -le, both ‘go to’, and Lonwolwol (NCV) -le ‘leave (a place)’, reflect coalescence either of *la + transitive suffix *-i or possibly of *la + locative preposition*i.

PMP		*lakaw	‘move, go, walk’ (ACD)
POc		*la	‘go (away, to)’
Adm	Loniu	-la	‘go’
Adm	Loniu	-lɛ	‘go to’
Adm	Kele	-la	‘go to’
Adm	Titan	-la	‘go’
Adm	Titan	-le	‘go to’
Adm	Lou	-la	‘go to, walk’
NNG	Numbami	la	‘go, get a move on’
NNG	Mapos Buang	ya	‘go’
NNG	Gitua	-la	‘go, walk’
NNG	Mangap	-la	‘go’
NNG	Sio	lɔ	‘go’
NNG	Gedaged	la	‘go (away, on foot, by boat, etc. ); walk, depart, get underway, set out; continue, keep on (with an action), maintain a course’
NNG	Takia	la	‘move away from speaker; go round the island’
NNG	Manam	-la	‘go, move away’
PT	Kilivila	-la-	‘go (to some place away from here)’
PT	Iduna	-na	‘go’
PT	Misima	na	‘go’
PT	Motu	la	‘go away’
MM	Notsi	la	‘go’
MM	Nehan	la	‘go’
MM	Selau	la	‘go’
MM	Vangunu	la	‘go’
MM	Hoava	la	‘go away from speaker and addressee’
SES	Marau Sound	ra	‘go’
SES	’Are’are	rā	‘go’
SES	Oroha	rā	‘go’
TM	Tanema	la	‘go, walk’
NCV	Sa	-la	‘walk’
NCV	Lonwolwol	la	‘take a step, go’
NCV	Lonwolwol	le	‘go, come, pass from, leave’
NCV	Sakao	la(m)	‘come’ (< POc la + mai ‘come’)
Mic	Woleaian	la	‘go, walk’
Mic	Puluwatese	la	‘go, walk’

cf. also:

NNG	Mangap	lae	‘move in horizontal direction’
PT	Gapapaiwa	-nae	‘go’
PT	Tawala	nae	‘go’
SES	Longgu	lae	‘go’
NCV	Lonwolwol	lae	‘walk, go’

There was evidently a difference in usage between POc *lako ‘go (to)’ and POc *pano ‘go (away)’. As noted in vol.2(279), many verbal reflexes of *lako have a valency which implies or requires a destination (expressed, for example, as a prepositional phrase), whilst those of *pano are intransitive.

A few reflexes of POc *pano indicate that it was also a geographic directional verb meaning ‘move in a transverse direction’, contrasting with ‘go up, go inland’ and ‘go down, go seawards’. This is discussed in vol.2(279).

PMP		*panaw	‘go away, depart, leave on a journey’ (ACD)
POc		*pano	‘go (away)’; ‘move in a transverse direction’
NNG	Kaulong	van	‘move’
MM	Vitu	vano	‘go (away)’
MM	Tiang	pən	‘go’
MM	Bola	vano	‘go’
MM	Bilur	van	‘go’
SES	Bugotu	vano	‘go, come’; [DIR] ‘thither’; ‘beyond, more’ (used in comparisons)
SES	Gela	vano	‘walk’; ‘away, further off; to go’
SES	Ghari	vano	‘walk’
SES	Gela	vano	‘walk’; ‘away, further off; to go’
SES	Bauro	hano	‘go’
SES	Arosi	hano	‘make a journey, set out; go’
NCV	Mota	van[o]	‘go, come’
NCV	Ambae	vano	‘move in transverse direction’
NCV	Raga	vano	‘go’
NCV	Nokuku	vano	‘go’
NCV	Kiai	vano	‘go, move away’
NCV	Sakao	yan	‘go’
NCV	Merei	van[a]	‘move in transverse direction’
NCV	Tamambo	vano	‘go away from speaker’
NCV	Uripiv	-van	‘go’
NCV	Port Sandwich	van	‘to go, to walk’
NCV	Lonwolwol	van	‘go, pass (and so also of time); continue (to do s.t.)’; [DIR] ‘away’
NCV	Lewo	vano	‘go’
NCV	Nguna	vano	‘go’
PSV		*van	‘go’
PSV		*a(v,p)an	‘go, walk’
SV	Sye	a-van	‘walk’
SV	Whitesands	vən	‘go’
SV	Lenakel	vən	‘go’
SV	Lenakel	a-vən	‘go, walk’
SV	Southwest Tanna	vən	‘go’
SV	Southwest Tanna	a-vən	‘go, walk’
SV	Kwamera	vən	‘go’
SV	Kwamera	a-vən, u-vən, e-vən	‘go, walk’
SV	Anejom̃	han	‘go’
SV	Anejom̃	a-pan	‘go, walk’
NCal	Nyelâyu	van	‘go’
NCal	Nemi	hen	‘go’
NCal	Tîrî	(ã)va	‘there, the other side of stream’
Pn	Niuean	fano	‘go’
Pn	Samoan	fano	‘(of time) be gone, past; perish’
Pn	Nanumea	fano	‘go’
Pn	Rennellese	hano	‘go; depending on, according to; on and on; little by little; one by one’
Pn	Māori	fano	‘go, proceed; lead, of a road; verge towards; be on the point of; act, behave’

PMP		*panaw	‘go away, depart, leave on a journey’ (ACD)
POc		*pa	‘go away; move in a transverse direction’
NNG	Lukep	pa	‘go, walk’
NNG	Mutu	-wa	‘go’
NNG	Mangap	-pa	‘walk’
NNG	Roinji	pa(la)	‘walk’ (< pa + la ‘go’)
NNG	Adzera	fa	‘go’
PT	Kilivila	-va-	‘go to’ (Senft 2000)
PT	Sinaugoro	va(riɣo)	‘go down’
PT	Motu	ha	[V AUX] ‘go and …’
SES	Gela	va	[V AUX] ‘be going to …’
NCV	Mota	va	‘go, come’; [V AUX] ‘go on …-ing’
NCV	Araki	ð̫a	‘go; go in a direction other than north or south’
NCV	Nokuku	va	‘go’
NCV	Merei	va	‘move in transverse direction’
NCV	Big Nambas	ð̫a	‘go’
NCV	Lonwolwol	va	‘go’
NCV	Southeast Ambrym	ha	‘go, leave, depart’
NCV	Paamese	vā	‘go’
NCV	Nguna	vā	‘go’ (short form of vano, Clark 1996)
SV	Sye	ve	‘go, arrive’
SV	Ura	va	‘go’
SV	North Tanna	va	‘come’
SV	Whitesands	va	‘come’
SV	Lenakel	va	‘come’
SV	Anejom̃	ha-m	‘come’ (< pa + mai ‘come’)

A small number of forms from Schouten (NNG) and Reefs/Santa Cruz (TM) languages and one from Sobei (SJ) are listed below. It would be easy to attribute them to POc *pa above, but in all instances except Medebur initial w- reflects *w- or *u-. They evidently reflect the POc deictic directional verb *ua ‘go towards addressee’ (vol.2:273–274). There is no other evidence to suggest that POc *ua functioned as a locomotion verb, and its extended use as a verb of locomotion may have developed separately on the north coast of New Guinea and in the Reefs/Santa Cruz languages.

NNG	Medebur	-wa	‘go’
NNG	Ali	-wa	‘go’
NNG	Sissano	-wa	‘walk’
NNG	Sera	-wa	‘walk’
SJ	Sobei	-wo	‘go’
TM	Äiwoo	wæ	‘go’

The medial consonant of POc *[y]aku ‘go, go away’ is reconstructed on the basis of Yabem low tone, which reflects the loss of a Proto Huon Gulf voiced obstruent, probably either *v or *ɣ, lenis reflexes of POc *p or *k. Of these, both are lost intervocalically in Takia, but only *k is lost in the Admiralties languages.

POc		*[y]aku	‘go, go away’
Adm	Nyindrou	au	‘leave, go away’
Adm	Loniu	yaw	[DIR] ‘away’
Adm	Kele	aw	‘go away’
Adm	Titan	aw	‘leave, go, disappear’
NNG	Yabem	-yà	‘go (to her/him/them)’ (?< *yaɣ)
NNG	Bilibil	au	‘walk’
NNG	Takia	-ao	‘go, go away, depart, flow’
NNG	Megiar	-au	‘go’
SJ	Sobei	awo	‘walk’
MM	Vaghua	zao	‘go’
MM	Maringe	zao-zaʔo	‘walk’
MM	Sisiqa	zo-zo	‘walk’
TM	Tanema	au	‘walk’

Only a few reflexes of POc *oRo ‘come, go’ have been found outside Polynesia. The glosses of its reflexes indicate that it was not a deictic directional verb, but it is otherwise difficult to further specify its meaning.

POc		*oRo	‘come, go’
NNG	Manam	oro	‘go landwards’
SES	To’aba’ita	olo	‘arrive’
SES	Lau	olo	‘come, go’
SES	Arosi	oro	‘come, go for a purpose’
Fij	Rotuman	ō	‘come, go’ (probable Pn loan)
Pn	Tongan	ō	‘go’ (non-singular form of both hau ‘come’ and alu ‘go’)
Pn	Niuean	ō	‘come, go’ (non-singular form of both hau ‘come’ and fano ‘go’)
Pn	Samoan	ō	‘go, get’ (plural of alu); ‘go side by side, go together with’
Pn	Anutan	ō	‘come, go’
Pn	East Uvean	ō	‘come, go’ (plural)
Pn	Takuu	ō	‘come, go’ (plural)
Pn	Tikopia	ō	‘proceed’ (plural)
Pn	Sikaiana	ō	‘come’ (plural)
Pn	Tokelauan	ō	‘go’ (plural)

Just as it is sometimes difficult to distinguish between reflexes of *lako (above) and *laka (§6.3.1.3), so it is sometimes difficult to distinguish between reflexes of *pano (above) and *pana (below). Languages as far apart phylogenetically as Vitu and Māori have reflexes of both forms, indicating that both occurred in POc, but there are few languages reflecting both and it is difficult to be sure what the POc meaning contrast was.

Blust (ACD) notes that, ‘In many languages reflexes of this form [*pana] are indistinguishable from reflexes of the far more common PAn *paNaw, POc *pano ‘go away, depart, leave on a journey’. Label han is assumed to reflect POc *pana rather than *pano since its closest relatives which have preserved the final vowel point to *-a.’ The same assumption is made here regarding other New Ireland reflexes, except for those with reduplicated versions that mean ‘walk’. These are attributed to *pano (see POc (?) *pa(no)-pano ‘walk’ in §6.3.1.3), but this decision may be mistaken.

POc		*pana	‘go, move, walk’ (ACD)
PT	Sudest	vana	‘go’
PT	Sinaugoro	fana-	‘step’ (in compounds: fana-beretoɣa ‘step to the side’, fana-kau ‘step on’, fana-tari ‘step down’)
MM	Vitu	vana	‘walk, travel’
MM	East Kara	pan	‘go’
MM	Patpatar	han	‘go, walk’
MM	Tolai	vana	‘go, move; come’
MM	Sursurunga	han	‘go’
MM	Konomala	fan	‘go’
MM	Label	han	‘go, go away; come’
Pn	Kapingamarangi	hana	‘go’
Pn	Rarotongan	ana	‘move (of a person)’
Pn	Rarotongan	ana mai	‘come/move here (toward speaker)’
Pn	Rarotongan	aná-atu	‘go/move away (from speaker)’
Pn	Māori	fana	‘travel, come, go’

3.1.2. Move from one location to another⇫

POc *(ali)ali ‘move from one location to another’ was inherited from PMP. In Vanuatu excluding the Banks and Torres Islands in the north, it has undergone a semantic shift to mean ‘walk’.

PMP		*aliq	‘move, change place’ (ACD)
POc		*(ali)ali	‘move from one location to another’
NNG	Kaulong	al	‘move, change location of, transplant’
NNG	Bing	al	‘go’
SES	Gela	ale-ale	‘move’
NCV	Mota	al	‘move’
NCV	Mota	ali-al	‘move from place to place’
NCV	Kiai	ali-ali	‘walk’
NCV	Sakao	yal-yal	‘walk’
NCV	Merei	ali-ali	‘walk’
NCV	Wusi	ali-ali	‘walk’
NCV	Maskelynes	-iar	‘walk’
NCV	Paamese	ali	‘walk, go to work in the garden’
NCV	Lamen	-yali	‘walk’
SV	Lenakel	-ali(uok)	‘walk’
SV	Whitesands	-ali(wak)	‘walk’

3.1.3. Walk, step, stride⇫

It is difficult to distinguish between the meanings of *lako/*la and *pano/*pa ‘go on foot’ (§6.3.1.1) and of the terms in the present section, but the terms below each perhaps denote the specific action of walking as opposed to generic going on foot.

Two POc terms meaning ‘go, walk’, *raka(t) and *laka, are formally similar to POc *lako ‘go (away, to), walk’ above. Where the nucleus of the second syllable reflects *-o or *-a, the distinction between *lako and *raka(t)/*laka is straightforward. The attribution to *lako of reflexes that have lost the final vowel is briefly explained in §6.3.1.1. Where a verb reflects the short form *la, it is taken to reflect *lako.

It is more difficult, however, to distinguish between reflexes of POc *raka(t) ‘go, walk’ and of POc *laka ‘go, walk’ in languages that have merged POc *r and *l, but terms that have the sense of stepping over something are attributed to *laka. This motivates the attribution of PPn *laka ‘step, march; pass, cross over’ and its many reflexes to *laka rather than to *raka(t), although phonologically they could reflect either. It seems likely that reflexes of the two terms were conflated in some languages. The PMP ancestors of POc *raka(t) and *laka are respectively *rakat and *la(ŋ)kaq, but transitive Wayan (Fij) lakat-i reflects *l- (not *r-) but *-t (not *-q), pointing to a hybrid *lakat.

Lihir merges *r and *l as l, but on semantic grounds Lihir laka(n), listed under ‘cf. also’, probably reflects *raka(t) rather than *laka. Other items listed under *raka(t) have an unambiguous initial consonant.

PAn		*rakat	‘walk’ (ACD)
PMP		*rakat	‘walk’ (ACD)
POc		*raka(t)	‘go, walk’
PT	Dobu	laga	‘go towards bush, southwards’ (-g- for †-ʔ-)
PT	Motu	raka	‘step, go, walk’
PT	Sinaugoro	raka	‘go, walk’
SES	Baelelea	rā	‘go up’

cf. also:

Lihir

laka(n)

‘go’

Motu has dozens of compounds denoting manners of walking, e.g. raka-edea ‘walk alone’, raka-hanai ‘walk past’, raka-hekapu ‘walk side by side’, raka-muri ‘walk backwards’, raka-roho ‘stride along’, raka-kahira ‘draw near’, and raka-kava ‘walk aimlessly’. Its Sinaugoro cognate behaves similarly: raka-rosi ‘go outside’, raka-vesiri ‘go away’, raka-ɣeɣeraɣi-ɣeɣeraɣi ‘go from place to place’.

PMP		*la(ŋ)kaq	‘step, stride; omit or skip over’ (ACD)
POc		*laka	‘go, walk; step over’¹¹
MM	Patpatar	laka	‘go inside, enter’
MM	Sursurunga	ləkə-i	‘step over, disregard’
NCV	West Ambae	laka	‘walk’
Fij	Rotuman	laʔa	‘go’
Fij	Wayan	laka	‘go, move along, proceed’
Fij	Wayan	lakat-i	‘go to or over (a place)’
Fij	Wayan	laka-laka	[VI] ‘go, keep going’; [N] ‘route, method, procedure; conduct, behaviour; style, characteristics; contributions to a feast or presentation, what one brings’
PPn		*laka	‘step, marcss, cross over’
Pn	Tongan	laka	‘go, walk (usually a short distance only); step, march; move forward, proceed, progress, develop, advance; pass, cross over; surpass, exceed, omit, skip over’
Pn	Niuean	laka	‘step; cross over’
Pn	Niuean	fe-laka	‘step over s.o./s.t. (formerly considered an insult or desecration)’
Pn	Niuean	laka-aŋa	‘a step, a pace’
Pn	Niuean	lakaf-ia	‘stepped over, exceeded’
Pn	Samoan	laʔa	‘step, march’
Pn	Samoan	laʔa-laʔa	‘go step by step’
Pn	Samoan	la-laʔa	‘step over; put someone above (in estimation or respect)’
Pn	Samoan	laʔas-ia	‘step over, go beyond’
Pn	Tuvalu	laka	‘step’
Pn	Tuvalu	laka-laka	‘take several steps’
Pn	Tuvalu	la-laka	‘press down with foot (as when firming soil)’
Pn	Anutan	raka	‘step over something’
Pn	Anutan	raka-raka	‘walk with large brisk steps; walk quickly’
Pn	East Futunan	faka-laka	‘pass over’
Pn	East Futunan	lakaf-ia	‘be passed over’
Pn	Nukuoro	laga-laga	‘put down one foot after the other (as in walking or marching in place)’
Pn	Māori	faka-raka	‘walk, step out’

A reduplicated reflex of POc *pano ‘go (away)’, (?) ‘move in a transverse direction’ (§6.3.1.1) means ‘walk’ in a number of languages with a discontinuous distribution including some MM languages of New Ireland, some SES languages, and a number of northern Vanuatu languages. The reduplicated form is tentatively attributed to POc.

POc		*pa(no)-pano	‘walk’
MM	Tiang	pan-pan	‘walk’
MM	Bilur	van-an	‘walk’
SES	Tolo	va-vano	‘walk’
SES	Ghari	va-vano	‘walk’
SES	Talise	va-vano	‘walk’
SES	Fagani	ha-hano	‘walk’
SES	Bauro	ha-hano	‘walk’
NCV	Mwotlap	van-van	‘walk’
NCV	Dorig	van-van	‘walk’
NCV	Araki	ð̫aano-vano	‘walk’
NCV	Aore	ð̫an-ð̫ano	‘walk’
NCV	Tambotalo	ðan-ðano	‘walk’
NCV	Tamambo	vʷano-vʷano	‘walk’
NCV	Tutuba	vano-vano	‘walk’
NCV	Mafea	ð̫an-ð̫anō	‘walk’
NCV	Narango	fan-fan	‘walk’

At first sight POc *pajale ‘walk about, take a walk’ looks suspiciously like the root *jalan ‘path’, preceded by the causative prefix *pa-. However, this would give rise to Lavongai †asalen (salen ‘path’) and Tigak †asalan (salan ‘path’) instead of pasal in both languages.

POc		*pajale	‘walk about, take a walk’
NNG	Takia	-padal	‘go astray, disappear’
NNG	Manam	alale	‘walk’ (for †adale)
NNG	Kairiru	-vyal	‘walk’
NNG	Kis	asali	‘walk’
MM	Lavongai	pasal	‘walk’
MM	Tigak	pasal	‘walk’
MM	Teop	pahana	‘cross (s.t.), go over (to s.t.)’
MM	Tinputz	pahan	‘go across’
PMic		*faSale	‘walk, move around’
Mic	Marshallese	yetal	‘walk’
Mic	Carolinian (Tanapag)	fatal	‘walk, journey, take a walk’
Mic	Mortlockese	fətal	‘walk about’
Mic	Satawalese	fetæl, fetæn	‘walk’
Mic	Woleaian	fetaɾe	‘move by its own power, go’
Mic	Pulo Annian	θataɾe	‘walk around’

3.1.4. Move quickly, hurry, run⇫

Terms in this area of meaning are difficult to reconstruct, as lexical replacement seems to have been frequent. For this probable reason reflexes of the reconstructed etyma have spotty distributions.

POc *(i,a)ropu ‘run’ bears a superficial resemblance to POc *Ropok ‘fly’ (§6.3.2.1), and indeed some terms meaning ‘run’ reflect the latter (e.g. Longgu [SES] lovo-lovo ‘run quickly’, a reduplicated form of lovo ‘fly’). The terms listed below, however, reflect POc *r-, as reflexes of *r and *R are distinct in Admiralties languages. Furthermore, terms for ‘fly’ reflecting *Ropok—Mussau loo, Titan yo, Misima you—are different from those reflecting *(i,a)ropu ‘run’.

POc		*(i,a)ropu	‘run’
Adm	Mussau	ilou	‘run’
Adm	Titan	ilow	‘run’
Adm	Nyindrou	arou	‘(people) rush in a disorganised manner’
PT	Misima	lou	‘run away’
PT	Tawala	lowo	‘run, run away’

The cognate set supporting POc *joŋas ‘move quickly’ entails two small phonological difficulties. First, the non-Micronesian reflexes reflect POc *j-, but Bender et al. (2003) reconstruct PMic *caŋa. PMic *c- regularly reflects POc *d or *dr. However, Kiribati r- reflects POc *d, *dr or *j-, and the most economic assumption is that Ponapean t- and Chuukese c̣- reflect an irregular intra-Micronesian sound change. Second, evidence as to the first vowel is conflicting. Admiralties and NNG languages reflect -o-, MM and Micronesian languages -a-. I have assumed here that the protovowel was *-o- and that -a- is the result of assimilation.

POc		*joŋas	‘move quickly’
Adm	Lou	sɔŋ	‘run, run away’
Adm	Baluan	soŋ	‘run’
Adm	Titan	coŋ	‘run away and hide’
NNG	Aria	suoŋ	‘run, quickly’
NNG	Apalik	soŋ	‘run, (wind) blow’
MM	West Kara	saŋas	‘quickly; walk’
MM	East Kara	saŋas	‘walk’
MM	Nalik	saŋas	‘walk’
MM	Sursurunga	saŋar	‘hurry, (be) quick; quickly’ (-r for †-s)
PMic		*Saŋa	‘move quickly’ (Bender et al. 2003: *caŋa)
Mic	Kiribati	raŋa	‘move fast’
Mic	Ponapean	taŋ	‘run, flee, swim (as fish)’; ‘run to (s.t.)’
Mic	Chuukese	c̣aŋ	‘fly, be flying (as a bird)’

3.1.5. Crawl, creep⇫

Two POc terms for ‘crawl, creep’ can be reconstructed. The term with the most widely distributed reflexes is POc *kaRaka ‘crawl on all fours’. Less widely distributed, with only one non-WOc reflex is POc *(d,dr)aRaC ‘crawl (along the ground)’ below.

POc		*kaRaka	‘crawl on all fours’
Adm	Mussau	kalaa	‘crawl’
NNG	Takia	karak	‘creep, crawl’
NNG	Mangap	kara-ra	‘crawl’
MM	Bola	karaka	‘crawl’
NCV	Nokuku	kerak	‘climb’
NCV	Tamambo	haraha	‘(child) crawl’
NCV	Araki	hadaho	‘crawl, walk on all fours’
NCV	Rerep	karah	‘creep on the knees’
NCV	Port Sandwich	kalax	‘crawl’
NCV	Paamese	keaa	‘crawl’
NCV	Nakanamanga	karaka	‘walk bent over, hands supporting body (indicates child will soon walk properly)’
SV	Anejom̃	a-ɣraɣ	‘creep, crawl’

POc		*(d,dr)aRaC	‘crawl (along the ground)’
NNG	Mutu	ⁿdarab	‘crawl’
NNG	Mengen	ralaŋ	‘crawl; walk on hands and feet/knees’
NNG	Mengen	ralasi	‘crawl; go underneath’
PT	Gumawana	dale	‘crawl on the ground; bow (to s.o.)’
PT	Iamalele	dale	‘crawl, walk, move across ground’
PT	Tawala	dala	‘crawl, swim’
PT	Saliba	dala	‘crawl’
PT	Wedau	dara	‘crawl’
PT	Dawawa	dara	‘crawl’
PT	Sinaugoro	dara	‘crawl’
MM	Nakanai	gala	‘crawl’
SES	’Are’are	da-dala	‘creep, crawl, glide as a snake’

PCP *dolo ‘crawl (along the ground)’ bears a vague formal similarity to POc *(d,dr)aRaC above, but in fact only their initial consonants correspond and they are not related.

PCP		*dolo	‘crawl (along the ground)’
Fij	Wayan	dolo	‘crawl, wriggle or slide along on one’s belly, like a snake’
Pn	Tongan	to-tolo	‘crawl’
Pn	Niuean	to-tolo	‘crawl’
Pn	Rennellese	togo	‘crawl’
Pn	Samoan	tolo	‘crawl, swarm (of creatures found in large numbers)’
Pn	Samoan	to-tolo	‘crawl, creep’
Pn	Tikopia	toro	‘crawl, creep’
Pn	Tahitian	toro	‘run or creep as vines or roots of plants’
Pn	Māori	to-toro	‘creep, crawl’
Pn	Hawaiian	ko-kolo	‘creep, crawl’

3.1.6. Limp, hop⇫

The senses of reflexes of POc *tige vary between ‘hop on one leg’, ‘limp’ and ‘walk on tiptoe’. Iduna sike and Longgu tike occur where respectively †ɣige and †ige would be the regular reflexes and are evidently loans. PPn *teki reflects vowel metathesis.

POc		*tige	‘hop on one leg, limp’
PT	Iduna	(-lufa)sike-sike	‘hop on one leg’ (s- for †ɣ-; -k- for †-g-)
PT	Gapapaiwa	sike	‘limp’
PT	Sudest	sike	‘hop’ (-k- for †-g-)
MM	Sursurunga	sik-si(kok)	‘limp, walk on tiptoe’ (-k- for †-g-)
MM	Teop	sige	‘lame’
MM	Teop	sige-sige	‘limp, hop’
SES	Longgu	ti-tike	‘hop’ (t- for †∅; -k- for †-g-)
NCV	Avava	si-sige(t)	‘hop’ (John Lynch, pers. comm.)
NCV	Labo	səge-səge	‘hop’ (John Lynch, pers. comm.)
NCV	South Efate	tig-tig	‘hop’
PPn		*teki	‘hop’ (POLLEX)
Pn	Emae	saere fakatuu-teki	‘walk or stand on tiptoe’ (saere ‘go’)
Pn	Marquesan	teki	‘limp, lame, hop’
Pn	Tongarevan	teki	‘hop’
Pn	Tongarevan	(pere)teki	‘hopscotch’
Pn	Pukapukan	teki	‘hop on one foot’
Pn	Tahitian	teʔi	‘hop, limp’
Pn	Tuamotuan	teki	‘hop on one leg’
Pn	Māori	(hii)teki	‘hop on one foot’

3.1.7. Roll⇫

This subsection concerns ‘roll’ as an intransitive locomotion verb expressing the manner in which its subject moves, e.g. ‘The stone rolled down the hill.’ At first sight, the items in the three sets below seem to form a single cognate set, but closer examination shows that they reflect three different POc etyma. The first, *buliŋ (VI) ‘roll’, was intransitive (or just possibly ambitransitive, i.e. used both transitively and intransitively). The other two, *(p,pʷ)uluk-i- (VT) ‘roll’ and *(p,pʷ)uri- (VT) ‘roll’, were transitive, and intransitive forms meaning ‘roll’ in its locomotion sense were formed with the spontaneity prefix *ta- (§1.3.5.4). Drehet ta-pulu-i and Samoan ta-fuli (VI) reflect POc *ta-, while Bariai ma-pul reflects POc *ma-. These prefixes are described by Evans (2003:268–271, 279–284), who notes that both were only semi-productive in POc, both were valency-decreasing, removing the agent, while *ta- also emphasised the spontaneity of the event (2003:300).

PMP		*puliŋ	[VI] ‘turn round, rotate’
POc		*buliŋ	[VI] ‘roll’
NNG	Takia	-bu-buli	[VT] ‘roll’
PT	Dobu	buni	[VI] ‘roll (as a canoe at sea)’
MM	Tolai	buli	[VT] ‘turn, roll, turn s.t. upside down’
MM	Ramoaaina	buli	[VI, VT] ‘roll’ (dialect variant bulu)
SES	Gela	pili	[VI] ‘roll over and over’
SES	To’aba’ita	(a)buli	[VI] ‘roll’
SES	Lau	(ā)buli	[VI] ‘roll; headlong’
SES	Arosi	buri	‘wallow, roll on the ground (as a dog)’
SES	Arosi	buriŋ-aʔi	‘roll on (s.t.)’

POc		*(p,pʷ)uluk-i-	[VT] ‘roll’
Adm	Drehet	ta-pulu-i	[VI] ‘roll, capsize’
MM	Patpatar	puluk-ane	[VT] ‘overturn; roll’ (-ane APPLICATIVE)
PNCV		*[v,b]uluk-i	‘fold, bend’ (Clark 2009)
NCV	Mota	wuluɣ	‘close over’
NCV	Uripiv	-ʙulk-i	‘bend, fold’
NCV	Port Sandwich	ᵐburuᵑg-i	‘bend knees; fold (material); roll’

POc		*(p,pʷ)uri-	[VT] ‘roll’
Adm	Mussau	puli	[VT] ‘roll’
NNG	Bariai	ma-pul	[VI] ‘roll spontaneously’
PT	Gumawana	si-pula	[VI] ‘roll (by accident)’
PT	Dobu	puli	[VI, VT] ‘roll (of a stone)’
PT	Molima	puli-	[VT] ‘push or roll along (as a log)’
PT	Tawala	wili	‘roll’
Fij	Wayan	vuli	‘(of a baby lying down) turn over, move position’
PPn		*fuli	‘turn round or over’
Pn	Tongan	fuli	‘turn (round or over)’
Pn	Samoan	fuli	[VT] ‘turn over, roll over (as a tree trunk)’
Pn	Samoan	ta-fuli	[VI] ‘roll’
Pn	Anutan	puri	‘turn’
Pn	Tahitian	huri	‘turn over, roll (as a cask)’
Pn	Tuamotuan	huri	‘turn (round or over)’

3.1.8. Climb⇫

‘Climb’ has two main senses in English: ‘go up’ and ‘propel oneself up or down (a tree, a cliff etc)’. The first of these senses is a verb of direction, and its POc equivalent is *sake, reconstructed in vol.2(263-264). The sense intended here is the second, although the glosses below suggest that its meaning was readily extended to include the first sense. The POc verb is *panaik, and one of the activities it denoted was one that is still readily seen in Melanesian villages, nicely defined in Ivens’ (1918) dictionary definition: ‘climb trees … with a rope round the feet, ascending by alternate jumps of hands and feet’.

Blust (ACD) reconstructs *panek, as supported by his data. POc *panaik is reconstructed here to take account of Lukep painak, which appears to reflect metathesis of the two syllable nuclei.

PMP		*panahik	‘climb’ (ACD)
POc		*panaik	‘climb (tree etc.)’ (ACD: *panek)
NNG	Lukep	-painak	‘climb up a steep hill’
PT	Molima	vane	‘climb’
PT	Bunama	hane	‘climb, go up’
PT	Sudest	vana	‘climb up’
PT	Misima	pani	‘climb (tree)’
SES	Gela	vane	‘grow abnormally’
SES	Longgu	vane	‘climb, rise’
SES	Longgu	vaneʔ-i-	‘climb s.t.’
SES	Lau	fane	‘rise up in sky, of sun, moon, or cloud’
SES	Kwaio	fane	‘climb, go up’
SES	’Are’are	hane	‘climb, ascend, rise (of sun and moon); rise, ferment’
SES	Sa’a	hane	[VI] ‘climb trees … with a rope round the feet, ascending by alternate jumps of hands and feet’
SES	Arosi	hane, haneʔ-i	[VI, VT] ‘climb (not a hill) a rope, tree, cliff, using hands’

3.2. Locomotion in the air⇫

There are far fewer terms for locomotion in the air than for on land or in/on water simply because human beings in traditional societies do not engage in aerial locomotion. The two main meanings associated with movement through the air are ‘fly’ and ‘fall’.

3.2.1. Fly⇫

The most widespread cognate set for ‘fly’ reflects POc *Ropok. It is not found, however, in Micronesian or in this meaning in Central Pacific languages. Reflexes exist in Polynesian but with changed meaning (PPn *ofo ‘be startled, wake up’; §4.6.4) In Micronesian languages *Ropok is replaced by various terms. In Eastern Fijian we find Bauan vuka, Boumaa vuʔa, in Western Fijian Wayan ðavu, whilst the PPn term was *lele (see below). The Fijian terms listed under ‘cf. also’ look at first sight like reflexes of *Ropok, but are not, as the expected reflex is †ovo.

It is interesting that some reflexes of POc *Ropok and PPn *rere taken from sources that provide more detailed definitions are glossed not only as ‘fly’ but also as ‘jump’, implying perhaps that the core meaning of the POc and PPn terms was something like ‘move through the air without touching the ground’.

PMP		*Rebek	‘to fly’ (ACD)
POc		*Ropok	‘to fly’
Adm	Mussau	lō	‘fly’
Adm	Loniu	wɔh	‘fly, jump’
NNG	Mengen	lo	‘fly, go’
NNG	Yabem	-lob	‘fly’
NNG	Takia	-rou	‘fly’
NNG	Manam	ro	‘fly’
NNG	Kairiru	-ruo	‘fly’
PT	Gapapaiwa	rovo	‘fly’
PT	Muyuw	yow	‘fly’
PT	Motu	roho	‘fly, leap, skip’
MM	Bali	rovoko	‘fly’
MM	Nakanai	lovo	‘fly’
MM	Lavongai	ŋoi	‘fly’
MM	Tiang	io-io	‘fly’
MM	Tabar	rovo	‘fly’
MM	Lihir	lah	‘fly’
MM	Sursurunga	roh	‘fly’
MM	Ramoaaina	rowo	‘fly (as birds)’
MM	Siar	rofoi	‘fly’
SES	Bugotu	ðovo	‘fly’
SES	Longgu	lovo	‘fly’
SES	Lau	lofo	‘jump, fly’
SES	Sa’a	loho	‘fly, swoop’
TM	Äiwoo	luo	‘fly’
PNCV		*rovo	‘run, flow, jump, fly’
NCV	Mota	rowo	‘spring, leap, move quickly up forward, rise, grow; fly (of birds and flying fish)’
NCV	Raga	rovo	‘run, sail, flow, fall’
NCV	Kiai	rovo	‘move, get out of the way, flee’
NCV	Uripiv	-row	‘fly, jump’
NCV	Lonwolwol	roo	‘run (also of liquids ); go well, prosper’
NCV	Paamese	loho	‘run’
NCV	Namakir	dow	‘go, go by, pass’
NCV	Nguna	dowo	‘fall’
SV	Lenakel	ivək	‘fly’

cf. also:

Fij	Wayan	rō	‘fly up, rise up through the air; land, alight, come down’ (for †ovo)
Fij	Wayan	rōv-i	‘land, settle, alight on s.t.’
Fij	Boumā	rō	‘alight (of housefly or bird)’ (for †ovo)
Fij	Boumā	rōv-a	‘alight on s.t.’

PPn		*lele	‘fly, run, leap’ (POLLEX)
Pn	Niuean	lele	‘fly’
Pn	Tongan	lele	‘run’
Pn	Rennellese	gege	‘jump, fly, leap, swing’
Pn	East Futunan	lele	‘fly, run quickly’
Pn	East Uvean	lele	‘fly, run’
Pn	Pukapukan	lele	‘run, swim quickly, of fish’
Pn	Samoan	lele	‘fly’
Pn	Tikopia	rere	‘run, fly, rush’
Pn	Māori	rere	‘fly, jump, run’
Pn	Tahitian	rere	‘fly, leap’
Pn	Hawaiian	lele	‘fly, jump, run’

3.2.2. Fall (from a height)⇫

The English term ‘fall’ is polysemous, and only the locomotion sense ‘fall from a height’ is considered here. As Zlatev & Yangklang (2004) point out in relation to serial verb constructions in Thai, ‘fall’ is not strictly a verb of locomotion, but rather a path verb, i.e. a verb of geographic direction. What is more, unlike most locomotion verbs it is not agentive. These facts perhaps provide a clue as to why no Oceanic-wide cognate set for ‘fall’ is found: a number of languages instead use their geographic direction verb ‘go down’ for ‘fall’, thus treating ‘go down’ as unmarked for agentivity.

The most widely reflected form in the data is POc *(p,pʷ)u(k,kʷ)a ‘fall’. Under ‘cf. also’ are listed a number of WOc items that reflect the appropriate consonants but lack the ‘right’ vowels. This is a conundrum that awaits explanation.

POc		*(p,pʷ)u(k,kʷ)a	‘fall’
MM	Vitu	poke	‘fall’
MM	Tigak	puka	‘fall’
MM	East Kara	poɣo	‘fall’
MM	West Kara	poko	‘fall’
MM	Tiang	pukə	‘fall’ (from tree +)
MM	Lihir	puok	‘fall’
MM	Konomala	puka	‘fall’
MM	Patpatar	puko	‘fall’
MM	Kandas	puku	‘fall’
MM	Ramoaaina	puka(pari)	‘fall’
MM	Uruava	uka	‘fall’
MM	Torau	uka(u)	‘fall’
MM	Hoava	uke	‘fall’
MM	Kokota	uka	‘fall’
MM	Maringe	uka	‘fall’
SES	West Guadalcanal	puka	‘fall’
SES	Talise	puka	‘fall’
SES	Birao	puka	‘fall’

cf. also:

NNG	Numbami	-peka	‘fall, fall down’
PT	Iduna	(-)beʔu	‘fall down’
PT	Gapapaiwa	peku	‘fall’
PT	Tawala	peu	‘fall’
PT	Dawawa	beku	‘fall’
MM	Halia	puku	‘fall’

In support of the PMP monosyllabic root *-buq ‘fall’, Blust (ACD) cites Yamdena (CMP) fu-fu ‘fall from a height’ and Arosi ahu below. With the NNG items these imply the possibility of POc *(a)pu.

POc		*(a)pu	‘fall’
NNG	Mangseng	pu	‘fall’
NNG	Mengen	pu	‘fall from standing position or from height’
SES	Arosi	ahu	‘fall, come down’

3.3. Locomotion in and on water⇫

In a study of verbs used in various languages for locomotion in and on water Lander, Maisak & Rahilina (2012) divide this semantic domain into three semantic subdomains: (a) swimming (self-propelled motion of an animate being); (b) floating (uncontrolled and non-agentive motion); and (c) sailing (motion of vessels and of the people aboard). Some languages (e.g. English) have verbs for each of these domains and little more. Others elaborate the subdomains. For example, Indonesian elaborates (c) with many verbs formed from nouns denoting vessels or areas of water. Others, on the other hand, conflate the three subdomains, like Russian, where plyt’/plavat’ is a verb denoting locomotion in and on water in general.

Given that the speakers of POc were skilful sailors, in many cases long-distance sailors, one might expect POc to have been like Indonesian. Indeed, perhaps it was, but the available data do not allow reconstruction of an elaborate set of terms for (c). Only POc *palau(r) ‘go to sea, make a sea voyage’ is reconstructable (§6.3.3.1). This may be because relevant terms in modern languages have been lost with the demise of the ocean-going canoe, or simply because researchers have not been conscious of the need to collect such terms. POc terms can be reconstructed for the three subdomains proposed by Lander et al., but not for much more. There was also a verb for ‘wade’ (§6.3.3.2), a concept not included in their analysis. There are two reconstructions for ‘swim’ (§6.3.3.3) and one for ‘go under water’ (§6.3.3.4). The ‘swim’ terms presumably differed in meaning in some now irretrievable way. Modern languages often distinguish between swimming on the surface and swimming under the surface. There were also terms for ‘float’ (§6.3.3.5). Thus if one is to believe the testimony of the reconstructions presented in this section, POc was like English in its lexicalisation of locomotion in and on water, but this may well be an artefact of data collection and/or modernisation.

3.3.1. Travel by sea⇫

No term for a locomotion verb ‘sail’ is reconstructable. POc *palau(r) meant ‘go to sea, make a sea voyage’, and is derived from PMP *lahud/POc *laur ‘sea, seawards’ (vol.2:91–92).

PMP		*pa-lahud	‘go down to the sea or coast’
POc		*palau(r)	‘go to sea, make a sea voyage’ (vol.1:206–207)
SES	Tolo	vola-volau	‘run, race’
NCV	Raga	walau	‘guide, steer, direct’
NCV	Mota	wala-walau	‘paddle all together’
NCV	Nguna	wo-wolau	‘steer canoe’
Mic	Kiribati	porau, po-porau	‘travel by sea’ (prob. borrowed from a Pn source)
Fij	Bauan	volau	[V] ‘make a sea voyage’; [N] ‘boat house’
Pn	Tongan	folau	‘voyage, travel by sea’
Pn	Tongan	folau(ʔaŋa)	‘boat in which one voyages’ (-ʔaŋa < NOM)
Pn	Tongan	folau(ʔia)	‘be constantly visited by ships’
Pn	Rennellese	hogau	(1) ‘ocean voyage’; (2) ‘canoe making an ocean voyage’
Pn	Samoan	folau	‘travel by sea, make a voyage; depart, sail’
Pn	Samoan	fōlau(ŋa)	‘voyage’ (-ŋa < NOM)
Pn	Samoan	folau(vaʔa)	‘sailor’
Pn	Tikopia	forau	‘voyage overseas, travel abroad’
Pn	Māori	farau	(1) ‘travel, particularly by water’; (2) ‘company of travellers’

cf. also:

Misima

alalau

‘(go on a) voyage (by boat)’

3.3.2. Wade⇫

POc had a term for wading, i.e. for walking through water deep enough to impede free movement, namely *tuRu(p). Blust (ACD) is uncertain whether the medial consonant is *R, but this is confirmed by the Baluan reflex. He suggests on the basis of SES transitive reflexes that the POc form may have ended in *-p. The latter is parenthesised here as we so far have evidence of it only in SES languages.

POc		*tuRu(p)	‘wade’ (ACD; Geraghty 1990: 76: PEOc)
Adm	Baluan	tui	‘wade through water’
PT	Motu	turu	‘wade’
MM	Vitu	turu	‘walk in shallow water’
MM	Nakanai	tulu	‘wade’
MM	Bola	tulu	‘walk’
SES	Gela	tulu	‘wade; float’
SES	Gela	tuluv-i-	‘wade to (s.t.)’
SES	Lau	ulu	‘wade’
SES	Lau	uluf-i-	‘wade to (s.t.)’
SES	To’aba’ita	ulu	‘cross a stream, river’
SES	’Are’are	uru	‘wade’
SES	Kwaio	ulu	‘wade’
SES	Sa’a	ulu	‘wade’
SES	Arosi	uru	‘wade’
SES	Arosi	uruh-aʔi	‘wade with, wade and carry’
Mic	Ponapean	sūr	‘sink or wade in mud’

3.3.3. Swim⇫

Four POc terms possibly meaning ‘swim’ are reconstructable, namely *kaRu and *qasa, both ‘swim’, *loso ‘bathe, swim’ and *kʷaya ‘immerse self, swim’. It is reasonable to infer that their meanings differed, but, because there is a tendency for reflexes to shift meaning, the protoglosses are inevitably tentative. A major distinction that was evidently made is that between swimming as a physical activity entailing movement of arms and legs, and immersing oneself in water, typically with the intention of washing one’s body. The first two verbs, *kaRu and *qasa appear to have denoted the physical activity of swimming by moving arms and legs, whereas *loso and *kʷaya apparently denoted immersion. Hence *kaRu and *qasa were locomotion verbs, and their cognate sets are listed below. POc *loso and *kʷaya, on the other hand, denoted a physical activity other than locomotion and appear in chapter 7.

POc		*kaRu	‘swim’ (Pawley 1976: Blust 1993: *kakaRu)
Adm	Seimat	ki-kau	[VI] ‘wash’
Adm	Titan	kaw	‘swim’
Adm	Nauna	kokaw	‘swim’
NNG	Sio	kɔlo	‘wade in water’
SES	To’aba’ita	ʔar(aŋa)	‘swim (of people and animals, but not fish)’
SES	Kwara’ae	ʔaru	‘swim’
NCV	Mota	ɣaru	‘wade/swim advancing by movement of arms and legs’
NCV	Kiai	karo	‘swim’
NCV	Mwotlap	ɣɛy	‘swim’
NCV	Vurës	ɣɛr	‘swim, wade’
NCV	Maewo	ɣa-ɣaru	‘swim’
NCV	Raga	ɣa-ɣaru	‘swim, float by moving arms and legs; take a bath’
NCV	Sakao	ɣɛr	‘swim’
NCV	Nguna	karu	‘swim side-stroke’
Fij	Wayan	gau	‘swim’ (g- for †k-)
Fij	Wayan	gau-vi-	[VT] ‘swim across, swim to’
PPn		kau, ka-kau	‘swim’ (POLLEX)
Pn	Tongan	ka-kau	‘swim (of man and ducks, but not fish)’
Pn	Niuean	ka-kau	‘swim’
Pn	Samoan	ʔa-ʔau	‘swim’
Pn	Samoan	ʔāus-ia	‘be swimming with, be full of’
Pn	Samoan	ʔaus-aʔi	‘swim with (s.o. or s.t.)’
Pn	Tuvalu	ka-kau	‘swim’
Pn	Pukapukan	ka-kau	‘swim’
Pn	Tikopia	ka-kau	‘swim (of man, but not fish)’
Pn	Rennellese	ka-kau	‘swim’
Pn	East Futunan	ka-kau	‘swim’
Pn	Kapingamarangi	kau	‘swim’
Pn	Rapanui	kau	‘swim’
Pn	Hawaiian	ʔau	‘swim’
Pn	Mangarevan	kau	‘swim’
Pn	Tahitian	ʔau	‘swim’
Pn	Rapanui	kau	‘swim’
Pn	Rarotongan	kau	‘swim’
Pn	Māori	kau(hoe)	‘swim’

POc		*qasa	‘swim’
Adm	Mussau	asa	‘swim on the surface’
NNG	Sissano	-yiəs	‘swim’
SJ	Sobei	-yas	‘swim’
MM	East Kara	ŋəs	‘swim under water’
MM	Patpatar	ias	‘swim’
MM	Kandas	was	‘swim’
MM	Laghu	aha	‘swim’

Also reconstructable is PEOc *olo ‘swim’. A possible WOc cognate is Teop (MM) oro ‘drift, float, be carried by the current’.

PEOc		*olo	‘swim’
SES	Gela	olo	‘swim’
SES	Bugotu	oðo	‘swim’
SES	West Guadalcanal	olo	‘swim’
SES	’Are’are	ōro	‘swim’
SES	Sa’a	olo	‘swim’
TM	Vano	wowo	‘swim’
NCV	Vunapu	olo	‘swim’
NCV	Paamese	olo	‘swim’

3.3.4. Dive, go under water⇫

Although most Oceanic languages have verbs that denote various kinds of diving as a human physical activity, the data do not allow POc reconstructions to be made for these meanings. The directional aspect of diving is often indicated by one of the geographic verbs for ‘go down’, POc *sipo and *sobu, reconstructed in vol.2:261-262. Nonetheless, one of the meanings of POc *ruku was ‘go under water’.

POc		*ruku	‘go under water, duck under (s.t.), bow the head’
NNG	Bam	-ruk	‘swim’
NNG	Roinji	luɣu	‘dive’
MM	Nakanai	lu-lū	‘duck under’
MM	Solos	ruku	‘dive’
SES	Bugotu	rugu	‘duck the head’
SES	Lau	rū	‘enter’
SES	Sa’a	ruʔu	‘draw back, retire’
SV	Sye	o-ruɣ	‘swim’
SV	Anejom̃	e-reɣ	‘swim, bathe, be in water’

PMic		*ruku	‘tip, go under water’ (Bender et al. 2003)
Mic	Carolinian	rux	‘go under water, of outrigger’
Mic	Chuukese	ruk	‘tip, be unsteady, tip with the outrigger going under water’
Mic	Woleaian	ruxu(tiw)	‘bend forward from waist, bow’ (tiw ‘go down’)
Mic	Mokilese	ruk	‘to bow, bend’
Fij	Wayan	ruku	‘bow down, stand or walk bent over or in a stooping posture’
Pn	Rarotongan	ruku	‘dive under water’
Pn	Hawaiian	luʔu	‘dive, plunge into water, immerse, dip (as shrimp net)’
Pn	Māori	ruku	‘dive feet first, sink, dive, submerge’

3.3.5. Float, drift⇫

Two POc terms meaning ‘float, drift’ can be reconstructed, but it is debatable whether either of them was a locomotion verb. POc *ma-qañur ‘floating, adrift’ is a promising candidate, as its Seimat and Arosi reflexes have glosses that imply movement. POc *saliR ‘flow, of water’ denoted movement of water, but the glosses of some of its reflexes also denote the movement of an object floating on flowing water. Whether the latter sense is reconstructable to POc is unclear. NCV reflexes do mean ‘float’, but the unexpected final vowel of PNCV *sale calls into question its descent from POc *saliR. PWOc *pʷati ‘float, drift, be carried on water’ does seem to have denoted movement, but no EOc reflexes have been found.

PAn		*qañud	‘drift on a current, carried away by flowing water’ (ACD)
POc		*maqañur	‘floating, adrift’ (ACD; vol.2:92–93)
Adm	Seimat	man	‘drift, float on a current’
SES	Lau	manu-manu	‘float’
SES	’Are’are	manu-manu	‘float’
SES	Sa’a	manu	‘float’
SES	Arosi	manu	‘float in water or air’
SV	Anejom̃	a-man-a-man	‘float (VI)’
PMic		*maanu	‘adrift, drift’ (Bender et al. 1983)
Mic	Chuukese	māɾ	‘be becalmed; drift’
Mic	Mortlockese	mān	‘float’
Mic	Puluwatese	mān	‘drift’
Mic	Carolinian	māl	‘drift’
Mic	Satawalese	mān	‘drift, float’
Mic	Woleaian	mālʉ	‘drift, be adrift’
Mic	Pulo Annian	māɾɨ	‘float, drift’
Fij	Rotuman	mɔnu	‘float’
PPn		*maqanu	‘be afloat’ (POLLEX)
Pn	Tongan	maʔanu	‘be afloat, not to be resting on or touching the bottom’
Pn	Samoan	mānu	‘come to the surface, emerge (as a turtle)’
Pn	East Futunan	maʔanu	‘bathe, wash oneself’
Pn	East Uvean	maʔanu	‘afloat, float’
Pn	Nukuoro	mānu	‘float in place’
Pn	Kapingamarangi	mānu	‘buoyant; float’
Pn	Rennellese	maʔanu	‘float, drift, soar; leap, as in a dance’
Pn	Tongarevan	mānu	‘afloat, float’
Pn	Tikopia	mānuu	‘floating on water’
Pn	Māori	mānu	‘float; be launched; overflow; be flooded’

PMP		*saliR	‘flow, of water’ (ACD)
POc		*saliR	‘flow, of water’ (vol.2:94)
NNG	Takia	-sarir	‘submerge, drown, sink, bury’
PT	Gumawana	salili	‘drown, sink’
MM	Nakanai	sali	‘flow’
MM	Tigak	salik	‘flow’
MM	Tolai	alir	‘swim, float, drift; float through the air, as a bird with motionless wings; adrift, drifting; run’
MM	Ramoaaina	alir	‘flow; float, drift’
MM	Bilur	alir	‘swim’
SES	Gela	hali-hali	‘flow swiftly’
PNCV		*sale	‘float, flow’ (-e for †-i)
NCV	Mota	sale	‘float, drift, soar with open wings; flow, run with water’
NCV	Nokuku	sale	‘float’
NCV	Kiai	sale	‘float, fly, sail’
NCV	Tamambo	sale	‘float, travel’
NCV	Sakao	hal	‘float’
NCV	Uripiv	-sal	‘float’
NCV	Lonwolwol	hal	‘gush out (of liquids); float, spread, flow; drift (of water or objects floating on water)’
NCV	Nguna	sale	‘float, drift’
Fij	Bauan	sali	‘flow’
Fij	Wayan	sali	‘flow’

PWOc		*pʷati	‘float, drift, be carried on water’
NNG	Bariai	pat	‘float, drift, float away’
NNG	Mangap	pot-pot	[VI] ‘float’
NNG	Sio	pɔti	‘float in one place’
NNG	Mangseng	(mon)pot	‘float downstream’
NNG	Mengen	potue	‘carry on water, float’
NNG	Takia	-fot	‘float, drift’
NNG	Bing	fat-at	‘float’
NNG	Manam	poati	‘float’
MM	Vitu	pati	‘float’
MM	Bulu	pasi	‘float’
MM	Bola	pasi	‘drift’
MM	Nakanai	pati	‘float’
MM	Nehan	pat-pat-aŋa	‘float’

4. Direction verbs⇫

A section with this title also occurs in vol.2(256-282). It presents reconstructions of the semantically most basic verbs of geographic direction, i.e. *sipo, *sobu and *[s,j]u[(a,u)] ‘go downward’, *sake ‘go upward’, *surup ‘enter, penetrate’, and of deictic direction, i.e. *ma[i] ‘come (towards speaker)’, *ua/*watu ‘towards addressee, *la[ko], *pa[no] and *[y]aku ‘away from speaker’. The purpose of the present section is to offer reconstructions of verbs that are semantically more complex but have a directional semantic component.

4.1. Return⇫

Verbs of returning arguably encode both an element of geographic direction, in that there is a reversal of path direction, and an element of deixis, as movement is towards a goal identical with an earlier starting point of movement.

Three verbs of returning reflect a single PMP root *uliq. Final *-iq is often raised to Oceanic -e. POc evidently had reflexes of both the plain root *uliq (POc *uli(q), *ule) and of *um-uliq (POc *mule), where PMP *um- is the marker of actor voice, present by default as the PMP verb was intransitive.

PMP		*uliq	‘return home; return something; restore, repair; repeat; motion to and from’ (ACD)
POc		uli(q), ule	‘come back’
MM	Bulu	ule(mai)	‘come back’
MM	East Kara	ula	‘come back’
MM	Nalik	uli	‘come back’
MM	Konomala	ule(i)	‘come back’
MM	Uruava	ure	[VI] ‘turn round’
Fij	Wayan	uli	[VI] ‘(of a visitor’s speech) be replied to’
Fij	Wayan	uli-	[VT] ‘reply to (of a visitor’s speech)’
Pn	Rarotongan	uri	‘turn round or turn about; to translate, as one language into another’

cf. also:

NNG

Mengen

ule

‘put behind one with a sense of completion’

In serial verb constructions reflexes of POc *mule are often treated as verbs of geographic direction, as in this Takia (NNG) example:

- Takia (NNG)
  ‘You’ve come back.’
  u-mul ∅-palu =wa
  
  2SG-return 2SG-come =PRF
  
  PATH DEIXIS …

PMP		*um-uliq	‘return home’ (ACD)
POc		*mule	‘return, restore’ (ACD)
Adm	Titan	muli-e	‘return’
NNG	Gitua	mule	‘return, back, again’
NNG	Lukep	-mulu	‘come back, turn around, go back’
NNG	Mangap	-mīli	[VI] ‘return, turn oneself around, go until, go as far as, stop at’
NNG	Kilenge	-mule	‘come back’
NNG	Tami	mul	‘backside; come back’
NNG	Mengen	mule	‘come back’
NNG	Kaiwa	(lu)mol	‘come back’ (lu- < POc *liu ‘turn aside, change direction’)
NNG	Bilibil	-mul	‘come back’
NNG	Takia	-mul	‘come back’
NNG	Manam	-mule	‘come back’
NNG	Manam	mule-aki	‘return (s.t.)’
NNG	Bam	-mul	‘come back’
MM	Vitu	(va)mule	‘come back’
MM	Bola	(ɣa)mule	‘come back’
MM	Notsi	mulə	‘come back’
MM	Tabar	mure(ŋai)	‘come back’
MM	Lihir	miel	‘come back’
MM	Madak	ml(oŋ)	‘return’
MM	Tolai	mule	[VI] ‘come back, return, of persons or things; again, afresh, more, yet’
MM	Simbo	mule	‘return; go back, come back; again, once again’
NCV	Mota	mule	‘come, go; refresh, restore in sickness’
NCV	Nokuku	mule	‘go back, go home’
NCV	Kiai	mule	‘go home, return’
NCV	Kiai	mule-i-	‘give back’
NCV	Ambae	mule	‘go home’
NCV	Raga	mule	‘reach a certain point’
NCV	Tamambo	mule	‘return home’
NCV	North Ambrym	mol	‘return, back’
SV	Anejom̃	(a-θu)mʷoc	‘return’

Blust (1983–84b:11–13) draws attention to an idiosyncratic innovation shared by CMP and Oceanic languages. This is the raising of PMP *u- to *o- giving a PCEMP variant *oliq ‘return’, reflected in Oceanic only in SES languages.

PMP		*uliq	‘return home; return something; restore, repair; repeat; motion to and from’ (ACD)
PCEMP		*oliq	‘return’ (ACD)
POc		*oli(q)	‘go back, come back’
SES	Bugotu	oli	‘change’
SES	Bugotu	olih-i	[VT] ‘exchange’
SES	Bugotu	va-oli	[VT] ‘exchange; in turn’ (va- CAUSATIVE)
SES	Gela	oli	‘come back, return; back, again’
SES	Gela	oli-oli	‘go and return’
SES	Gela	oliv-i	‘return to’
SES	Longgu	oli	[VI] ‘to return; go back’
SES	Kwaio	oli	‘return, come back to; commit incest’
SES	Kwaio	faʔa-oli-	‘cause to return, send back; return something to a person’ (faʔa- CAUSATIVE)
SES	Sa’a	oli-oli	‘return, relieve, replace’
SES	Sa’a	olis-i	‘change, alter’
SES	Arosi	ori	‘return; stroll about’
SES	’Are’are	ori	‘return; come back, go back (home)’

There are relatively few EOc reflexes of the terms reconstructed above, partly perhaps because their function had been taken over by reflexes of PEOc *poki ‘return’.

PEOc		*poki	‘return’
SES	Gela	voɣi	‘go back, turn over’
SES	Sa’a	hoʔi	‘return’
SES	Arosi	(a)hoʔi	‘return, go back; change, turn into, become’
NCV	Mota	woɣ	‘change, in character or colour’
Mic	Kiribati	oki	‘return’
Mic	Carolinian	woxo-wox	‘turn food when cooking’
Mic	Woleaian	wegi-ti	‘turn, change, transfer, convert (it)’
Fij	Bauan	vuki, vuki-ca	‘turn, change’
Fij	Bauan	(ta)vuki	‘turned over, capsized’
Fij	Bauan	voki	‘shift, of the wind’
Fij	Wayan	vuki	‘be turned around/over; turn oneself round; change direction’
Fij	Rotuman	hoʔi	‘go back, return’
Pn	Niuean	foki	‘turn, return’
Pn	Niuean	fo-foki	[VT] ‘turn inside out, turn over’
Pn	Tongan	foki	[VI] ‘return’
Pn	East Uvean	foki	‘return’
Pn	Rennellese	hoki	‘go or come back, return’
Pn	Samoan	foʔi	‘return’
Pn	Tikopia	foki	‘return, retreat’
Pn	Hawaiian	hoʔi	‘return’

4.2. Turn round⇫

POc apparently had two homophonous verbs with the root *liu, one glossed ‘turn aside, change direction’ (this section) and the other *liu ‘go beyond, pass, surpass’ (§6.4.3). It is possible that both reflect PAn *liuS, which Blust (ACD) glosses ‘circumambulate, circumvent’. However, its Formosan reflexes suggest that PAn *liuS formed both intransitive and transitive verbs meaning ‘turn around’,¹² the sense inherited by POc. Whether PAn *liuS was also ancestral to PMP *liu ‘surpass, exceed’ is not clear.

In addition to POc *liu ‘turn aside, change direction’, a reduplicated *li-liu ‘turn around, go back’ is also reconstructable (the different glosses of POc *liu and *li-liu are tentative at best). Many reflexes of these two forms are similar in meaning to reflexes of the items in §6.4.1, but the overall pattern of their glosses suggests that *liu and *li-liu profiled the semantic element of turning around rather than simply of going back.

The forms listed under ‘cf. also’ below appear to reflect a POc *[ri]riu ‘turn’, i.e. a form in which *l had been replaced by *r. This is not a regular sound change.

POc		*liu	‘turn aside, change direction’
POc		*li-liu	‘turn around, go back’ (ACD)
Adm	Baluan	li-liu	‘go back’
NNG	Mengen	(ga)liu	‘return (said to person nearby)’
NNG	Mengen	(kaga)liu	‘return (said to someone going far away)’
NNG	Mengen	(taoga)liu	‘return (by turning around and going back), chase (s.t.)’
NNG	Numbami	-leleu	‘return’
NNG	Kaiwa	lu(mol)	‘come back’ (mol < *mule ‘return’)
NNG	Manam	(-soa)li-li	[VI] ‘turn round’
MM	Minigir	(ta)li-li	‘come back’
MM	Tolai	li-li(kun)	[VI] ‘come back, turn back, return’
MM	Tolai	(ma)li-liu	[VI] ‘having gone before, as shown by flattened grass’
MM	Bilur	(va)li-li(kun)	‘come back’
MM	Roviana	li-liu	‘turn around’
SES	Bugotu	(pi)liu	‘turn’
SES	Gela	liu	‘turn aside, go in another direction or by another way’
SES	Gela	li-liu	‘become, change into’
SES	Lengo	(pi)li-liu	‘turn’
SES	Malango	lel-eo	‘turn’
SES	Longgu	liu-	[VT] ‘turn (s.t.) over, e.g. in the fire’
SES	Sa’a	(a)līu	‘turn; change one’s mind’
SES	Sa’a	li-li	[VI] ‘change, move about’; [VT] ‘change position of, push to one side’
SES	Kahua	ri-ri	‘turn’
PNCV		*li(u)-liu	[VI] ‘return; be backwards, be upside down’ (Clark 2009: *liu)
NCV	Raga	vʷa-liu	‘answer’
NCV	Uripiv	-luwi	‘return’
NCV	Uripiv	-li-li	‘turn (end to end), change positions’
NCV	Port Sandwich	li-l	‘backwards’
NCV	Port Sandwich	li-li	‘upside down’
NCV	Port Sandwich	li-li-in-i	‘back to front’
NCV	Lonwolwol	lu-lu (ne)	‘backwards, back to front, wrong way round, upside down’
NCV	Tolomako	li-liu	‘return’
NCV	Neve’ei	ma-li-li	‘return’
NCV	Naha’ai	ma-li-liu	‘return’
NCV	Lewo	li-liu	‘go upside down’
NCV	Nguna	li-liu	‘return’
NCV	Nguna	laki-pi-liu	‘sister exchange marriage’ (laki ‘marry’)
Fij	Rotuman	liu	‘jibe (of a boat)’
PPn		*liu	‘turn round’ (POLLEX)
Pn	Tongan	li-liu	[VT] ‘turn round, turn or change into s.t. else’
Pn	Niuean	liu	‘to turn, change’
Pn	Samoan	liu	‘alter, change; turn into’
Pn	Samoan	li-liu	[VT] ‘turn’
Pn	Samoan	liu-liu	‘turn over and over’
Pn	East Futunan	li-liu	‘turn back, return’
Pn	East Uvean	li-liu	‘return, come back’
Pn	Rennellese	giu	‘return, go back’
Pn	Kapingamarangi	riu	‘turn, reverse’
Pn	Emae	riu-a	[VT] ‘turn over’
Pn	Rarotongan	riu	‘turn round’
Pn	Tuamotuan	riu	‘turn round’
Pn	Tahitian	riu-riu	‘be moving around’
Pn	Hawaiian	liu-a	‘spinning, whirling, dizzy; shocked; fascinated’

cf. also:

MM	Sursurunga	-ri-riu	[VI] ‘turn round’
MM	Roviana	ta-liri	‘turn’ (consonant dissimilation?)
SES	Bugotu	riu	‘move position, change place of’
SES	Bugotu	riu-riu	‘wander about’
SES	Gela	riu	‘turn round’
SES	Lengo	ri-riu	‘turn’
SES	Longgu	ri-riu	‘turn’

PPn *ma-liu reflects the same root with the addition of the detransitivising prefix *ma- (see also the entries for Neve’ei and Naha’ai above).

PAn		*ma-liuS	‘turn round’¹³
POc		*ma-liu	‘change direction, turn’
PPn		*ma-liu	‘change direction, turn’ (POLLEX)
Pn	Niuean	maliu	‘turn towards the speaker; change one’s mind, attitude, etc.’
Pn	Tuvalu	maliu	‘turned’
Pn	East Futunan	maliu	‘return’
Pn	East Uvean	maliu	‘return, change direction’
Pn	Kapingamarangi	maliu	‘swerve’
Pn	Rarotongan	mariu	‘change direction, return, go, come’
Pn	Tuamotuan	mariu	‘turn to one side’
Pn	Hawaiian	maliu	‘turn toward’

Three further POc terms meaning ‘turn round’ or ‘turn back’ are reconstructable. The first, *likot, is reconstructed on the basis of evidence from western Malayo-Polynesian languages noted in the ACD and a single regular Oceanic reflex, Bali (MM) liɣota. The Vangunu reflex likoso is irregular, as the expected form is †liɣoto.

PMP		*liget	‘turn, rotate’ (ACD: PWMP)
POc		*likot	[VI] ‘turn round’
MM	Bali	liɣota	[VI] ‘turn round’

cf. also:

Vangunu

likoso

‘turn’

The other two reconstructions are formally similar, and it is hard to believe that they are not historically connected in some way, as both display the pattern *LABIAL-ulos. They are:

POc *[ta]bulo(s) (VI) ‘turn round, turn back’, *bulos-i- (VT) ‘turn round, turn back’
POc *pulo(s) (VI) ‘turn round’, *pulos-i- (VT) ‘turn (s.t.) round’

However, there is no straightforward way to unify them.

The first, *[ta]bulo(s), *bulos-i-, is internally consistent. Most of the intransitive forms reflect the semi-productive POc prefix *ta-, which decreased valency by removing the agent, and emphasised the spontaneity of the event (§1.3.5.4; Evans 2003:300). However, in this instance, the prefix simply functions to reduce valency.

POc		*[ta]bulo(s)	[VI] ‘turn round, turn back’
POc		*bulos-i-	[VT] ‘turn round, turn back’
NNG	Medebur	ta-wul	[VI] ‘turn round’
NNG	Wogeo	ta-bul	[VI] ‘turn round’
NNG	Kis	-tu-bu-bl-i	[VI] ‘turn round’
NNG	Kairiru	-ta-bul	[VI] ‘turn round’
MM	Blablanga	bulo	‘turn’
SES	Bugotu	ta-bulo	‘suddenly’
SES	Bugotu	ta-bulos-i	‘be sudden’
SES	Gela	ta-bulo	‘dash off suddenly’
SES	To’aba’ita	a-bula	[VI] ‘turn, veer to the side’
SES	Langalanga	bulus-i	‘turn’
SES	Lau	a-bulo	‘turn round, turn back, turn aside’
SES	Kwara’ae	a-bʷel	‘turn’
SES	Kwaio	a-bulo	‘turn’
SES	Kwaio	bulos-i-	[VT] ‘turn’
SES	Sa’a	pulo	[VI] ‘reach, turn back, return’
SES	Sa’a	ʔa-pulo	‘reversed, returned, turned back on a journey’
SES	Arosi	buro	[VI] ‘turn around, revolve’
SES	Arosi	a-buro	‘turn back, return, fail to reach home’
NCV	South Efate	plos	‘wring clothes’
SV	Ura	o-pre-i	‘turn’
Mic	Kosraean	tæ-pʌl	[VI] ‘turn’

POc *pulo(s) (VI) ‘turn round’, *pulos-i- (VT) ‘turn (s.t.) round’ entails some formal challenges:

The presence of fortis p- in Bulu, Harua, Roviana, W Guadalcanal, Talise and Birao, where a lenis reflex is expected, is probably explained by the fact that the sequence *pu- sometimes gives rise to [*pʷu], where [*pʷ] resists lenition.
The sequence [*pʷu] sometimes becomes [*pʷi] in Oceanic languages, i.e. the rounding feature is fully transferred from [*u] to preceding [*pʷ], accounting for -i-, instead of †-u-, in W Guadalcanal, Talise, Birao and all NCV reflexes except Kiai.
The second vowel, *-o-, is fronted to -e- (or -i- in Tabar and Neve’ei) in both intransitive and transitive forms. This is especially common in NCV reflexes, leading Clark (2009) to reconstruct two PNCV forms, *viles-i and *vilos-i. There is no obvious explanation for this alternation, but it is very unlikely that †*pules-i- occurred in POc, as *e is usually stem-final, reflecting PMP *-ay.

These considerations point to the reconstruction of POc *pulo(s) (VI) ‘turn round’, *pulos-i- (VT) ‘turn (s.t.) round’, and also suggest that early Vanuatu Oceanic must also have retained the form *vulos-i.

POc		*pulo(s)	[VI] ‘turn round’
POc		*pulos-i-	[VT] ‘turn (s.t.) round’
Adm	Seimat	hūlo	[VI] ‘turn about, as a man’
NNG	Gitua	pule-le	[VI] ‘turn round’
NNG	Kove	-pale-le	[VI] ‘turn round’
NNG	Bariai	-pale-le	[VI] ‘turn round’
NNG	Kilenge	-pulie	[VI] ‘turn round’
NNG	Amara	-pul	[VI] ‘turn round’
NNG	Maleu	-pule	[VI] ‘turn round’
MM	Bulu	pulo	[VI] ‘turn round’
MM	Harua	pule	[VI] ‘turn round’
MM	Nakanai	vulo	[VI] ‘turn round’
MM	Tabar	vurisi	‘turn’
MM	Babatana	vule	‘return, come/go back’
MM	Roviana	pule	‘again, once more; to return’
SES	West Guadalcanal	pilo	‘turn’
SES	Talise	pilo	‘turn’
SES	Birao	pilo	‘turn’
SES	Arosi	a-huro	[VI] ‘turn round, twist, slip around’ (a- SPONTANEOUS)
SES	Bauro	a-horo	‘turn’ (a- SPONTANEOUS)

PNCV		viles-i, vilos-i	‘turn’ (Clark 2009)
NCV	Mota	vile-ris	‘change, be changed’ (ris ‘change’)
NCV	Mwotlap	vɪy	‘reverse direction’
NCV	Kiai	ri-vules-i-	[VT] ‘turn’
NCV	Naman	vələs	‘turn around; change’
NCV	Port Sandwich	vløs-i	‘to roll on one’s thigh’
NCV	Neve’ei	vilih	‘turn around; put on back to front; stir; change’
NCV	Southeast Ambrym	-hilesi	‘turn’
NCV	Paamese	hiles	[VI] ‘turn around, turn over; turn corner (in vehicle); change, become different’
NCV	Paamese	hilesi	[VT] ‘turn around, turn over; sprain muscle’
NCV	Namakir	viloh	‘turn, turn back, come back’

The POc stems *bulos and *pulos which are central to the reconstructions above must have a common origin, but they are not related by any known morphological pattern.

4.3. Go beyond, pass by, surpass⇫

The formal identity of POc *liu ‘go beyond, pass, surpass’ and POc *liu ‘turn aside, change direction’ is discussed in the introduction to 6.4.2 above.

Grammaticisation of a verb meaning ‘surpass’—often a figurative extension of a verb meaning ‘go beyond, pass’—as a comparative marker is common in Oceanic languages, such that My house is bigger than yours is expressed by a clause containing a serial verb construction with the sense ‘My house big surpasses yours.’ Such a use is noted in the ’Are’are and Arosi dictionary entries from which items below are drawn.

PMP		*liu	‘surpass, exceed’ (ACD)
POc		*liu	[VT] ‘go beyond, pass, surpass’
MM	Tolai	liu	[VT] ‘pass’
SES	Bugotu	liu	[VT] ‘go beyond, pass’
SES	Lau	liu	‘go, pass by, move about’
SES	’Are’are	riu-	[VT] ‘surpass, win, excel, go beyond, pass over, of time’
SES	Sa’a	liu	‘pass by’
SES	Arosi	riu	‘go on, go past; beyond’
PNCV		*liu	‘go beyond, exceed’ (Clark 2009)
NCV	Raga	liu	‘surpass, more than’
NCV	Nokuku	liu	‘above, greater (used in comparative constructions)’
NCV	Kiai	lui-	‘lead, go first, be ahead; go beyond’
NCV	Tamambo	liu	‘win; be better’
NCV	Sakao	lö(d)	‘exceed, surpass, go beyond’
NCV	Paamese	lii (muli)	‘pass, go past; miss; more than’
NCV	Lewo	liu	‘pass, go past; miss; more than’
NCV	Namakir	liw	‘too much, too many, very’
NCV	Nguna	lius-	‘past, beyond; to surpass, pass’
Fij	Bauan	liu	‘precede, surpass, excel, go before, be the first to do a thing’

4.4. Arrive, appear⇫

Four POc verbs are reconstructed in this section: *potu ‘appear, come into view’, *pʷosa ‘appear’, *pura ‘arrive, appear’, *pʷaka ‘come into view’. They share similar clusters of meanings: ‘arrive’, ‘come into view’, ‘rise (especially of the moon)’ and ‘appear’. They are included here because reflexes of the most widely reflected of the four, *potu, serve as direction verbs in serial verb constructions in at least the languages of southern New Ireland, as in this Ramoaaina (MM) example.

- Ramoaaina (MM): (Fritzell & Davies n.d.)
  ‘He came.’
  i= wan pat
  
  3SG= go arrive
  
  … MANNER DEIXIS

Clark (2009) notes that in north and central Vanuatu languages reflexes of *potu are used of the emergence of the new moon: e.g. Mota o vula we wot ma ‘a new moon appears’, and occur as an attribute in the phrase ‘new moon’: e.g. Nokuku wul wotu ‘new moon’, Tamambo vitu-votu ‘new moon’, Rerep nebur mi uet ‘new moon’.

PMP		*betu	‘appear, come into view’ (ACD)
POc		*potu	‘appear, come into view’ (ACD)
Adm	Mussau	otu	[VI] ‘show up’
NNG	Amara	pod, uod	[VI] ‘appear, arrive’
NNG	Bariai	ot	‘arrive; emerge’
NNG	Kove	put	‘appear, enter a garden, come out’
NNG	Mangseng	=pot	‘out, off’ (‘=’ marks an enclitic boundary)
PT	Sinaugoro	votu	‘arrive’
MM	Madak	vot	‘come, arrive’
MM	Sursurunga	hut	‘come, arrive’
MM	Ramoaaina	pat	‘come, arrive’
MM	Tinputz	vōs	‘arrive at, reach home’
PNCV		*votu	‘emerge, appear’ (Clark 2009)
NCV	Mota	wot	‘rise up, stand up, appear, shoot up (as land coming into sight when at sea)’
NCV	Raga	votu	‘arrive, come, reach’
NCV	Merei	vu-vut	‘appear’
NCV	Nokuku	wot(ʔer)	‘appear, happen’
NCV	Uripiv	-wut	‘come out (guts from a wound; new moon)’
NCV	West Ambrym	fur	‘appear, come clear, come out’
NCV	Paamese	hotu	‘(of reef) come out and disappear under waves, come out at low tide; (of something floating) bob up and down in waves’
NCV	Lewo	woru	‘appear, arrive, reach’
NCV	South Efate	put(sak)	‘emerge, as from water’
Fij	Rotuman	hofu	‘rise, of moon; emerge, esp. from interior at coast’
Fij	Bauan	votu	‘appear, become visible (as a ghost)’
Fij	Wayan	votu	‘appear, come into view, become visible’
Pn	Tongan	fotu	‘emerge, come into view; appear (as fish in season), become manifest, stand out, or to be or become prominent’
Pn	Rennellese	hotu	‘come forth, appear’
Pn	Samoan	fotu	‘appear (as a boat coming around a point); (of trees) blossom out, come into blossom’
Pn	Samoan	fōtu-aʔi	‘emerge, heave in sight (as a boat)’
Pn	Tokelauan	fotu	‘appear, emerge’
Pn	Māori	hotu	‘dawn’

The other three verbs are less widely reflected and their usage is less well understood than that of *potu.

POc		*pʷosa	‘appear’
NNG	Lukep	-pa-pos-i	‘cause to appear, reveal’
NNG	Mengen	pota	‘appear’
PT	Dobu	(a)pʷesa	‘appear, arrive, arrive there, depart, happen’
PT	Bunama	(ha)pʷesa	‘appear’
MM	Patpatar	posa	‘appear, stick out’
MM	Patpatar	po-posa, pa-posa	[N] ‘appearing of sun, moon, stars’
SES	’Are’are	huta	‘be born, appear, rise up’
SES	Sa’a	uwota	‘appear on the horizon’

POc		*pura	‘arrive, appear’
NNG	Manam	pura	‘come, arrive’
NNG	Bam	-pur	‘come back’
NNG	Kairiru	-fur	‘appear, arrive on the scene’
MM	Roviana	vura	‘appear’
MM	Roviana	va-vura-ia	‘cause to appear, reveal’
NCV	Lonwolwol	fur	‘appear, come clear, (moon) come out’
NCV	Lewo	ula	‘arrive at; attain’
Fij	Bauan	vura	‘arrive, emerge, appear’
Fij	Wayan	vura	‘appear, come into view’

The presentation of POc *pʷaka ‘come into view’ below retains Bender et al.’s (2003) presentation of Micronesian data supporting PMic *pʷax(a,e) but separates off under ‘cf. also’ the transitive reflexes they list meaning ‘show’ or ‘announce’. This is done to highlight a difficulty in Bender et al.’s analysis. They reconstruct PMic medial *-x- (< POc *g) on the basis of the seemingly regular sound correspondence Kosraean -k, Ponapean, Mokilese and Pingelapese -r and Chuukese, Puluwatese, Carolinian and Woleaian zero (Bender et al. 2003:4). This is achieved by segmenting the transitives as Chuukese pʷǣ-ri etc. But if, as is usual, the consonant preceding the transitive suffix -i belongs to the PMic root (Chuukese pʷǣr-i etc.), then Chuukese, Puluwatese, Carolinian and Woleaian have -r in this position, and all items except Kosraean reflect PMic *pʷar(a,e), not PMic *pʷax(a,e). If this reinterpretation is correct, then it is possible that all Micronesian items other than Kosraean reflect POc *pura above, rather than POc *pʷaka.

POc		*pʷaka	‘come into view’
NNG	Sissano	pak	‘come up, appear, come into being, stay close’
NNG	Mangap	pok	‘burst forth into the open, appear, come into view, break’
SES	Arosi	pʷā	‘rise’
SES	Sa’a	pʷaʔa	‘rise (of heavenly bodies)’
PMic		*pʷax(a,e)	‘come into view, reveal’ (Bender et al. 2003)
Mic	Kosraean	fæk	‘say, tell, announce’
Mic	Ponapean	pʷar	‘appear’
Mic	Mokilese	pʷar	‘emerge, come to view’
Mic	Pingelapese	pʷar	‘appear, rise (of sun)’
Mic	Chuukese	pʷæ, pʷǣ-	‘come into view’
Mic	Puluwatese	pʷá, pʷǣ-	‘appear’
Mic	Carolinian	bʷæ	‘be(come) visible, appear’
Mic	Carolinian	pʷpʷæ	‘appear suddenly’
Mic	Woleaian	ɸʷā, pʷā	‘emerge into view, come from behind’

cf. also:

Mic	Chuukese	pʷǣr-i	‘show’
Mic	Puluwatese	pʷǣr-i-	‘show’

5. Accompanied movement verbs⇫

5.1. Accompany⇫

POc		*ud(r)u	‘accompany, go with (s.o.)’
NNG	Sio	wuru	‘accompany, go with (s.o.)’
MM	Ramoaaina	we-ur	‘accompany, go together’ (we- RECIP)
SES	Bugotu	udu	‘walk in file’
SES	Gela	udu	‘accompany’

5.2. Gather, congregate⇫

A POc verb of the form *so(k,g)o(n) is reconstructable with a meaning that has to do with gathering or assembly, but both its form and its semantics are problematic. The Mussau, Mota and Paamese reflexes point to medial *-k-, the Lonwolwol and Wayan reflexes to *-g- (and Nguna to *-ŋ-). The glosses make it difficult to determine which form is intransitive, which transitive. On formal grounds, *so(k,g)o(n) was presumably intransitive, *so(k,g)on-i transitive, but the Nguna and Wayan Fijian use of *so(k,g)on-i as intransitive render this a little uncertain.

POc		*so(k,g)o(n)	[VI] ‘gather, congregate’
POc		*so(k,g)on-i	‘gather, bring together’
Adm	Mussau	ai-soko-soko-iaa	[VT] ‘gather together’ (ai- RECIP)
PNCV		*soko	‘add, join’ (Clark 2009)
NCV	Mota	soɣo	‘give, bring, contribute, distribute; descriptive prefix to numerals, of things together in a bunch, such as coconuts; to measure money, a measure of money’
NCV	Mota	soɣon	‘bring together, pack, stow’
NCV	Lonwolwol	sogo	‘come together, to be close together’
NCV	Paamese	se-soon-i	[VT] ‘join’
NCV	Nguna	soŋon-i-a	[VI] ‘gather, congregate’
Fij	Wayan	sogo, sogon-i	[VI] ‘gather, assemble, come together in a crowd, congregate’
Fij	Wayan	sogon-i-ti-	[VT] ‘draw or attract a crowd, cause people to come together’
Pn	Tongan	hoko	‘join or unite (two things)’
Pn	Tongan	hoko-hoko	‘join or unite (many things)’
Pn	Samoan	soʔo	‘join, splice’
Pn	Samoan	soʔo-soʔo	‘join (many things)’

Just four reflexes of POc ‘gather, congregate’ have been found.¹⁴

POc		*lupun	‘gather, congregate’
NNG	Bariai	lup	‘meet, gather, collect together, join, put together’
NNG	Kaulong	lup	‘join, gather’
MM	Patpatar	luhu	‘gather about’
SV	Lenakel	ləpun	‘join (VI, VT)’

5.3. Precede and follow⇫

Two POc locative nouns *muri- ‘rear, back part’ and *muqa- ‘front’, along with PWOc *muga, a variant form of *muqa, are reconstructed in vol.2(247–249, 251-252). It is noted there that *muqa/*muga and *muri both also occurred as verbs, respectively meaning ‘be in front’ and ‘be behind, be after’. Only verbal reflexes are listed below.

POc		*muqa	‘be in front, precede’ (vol.2:247–248)
Adm	Mussau	mua	‘go first’
NNG	Manam	mua	‘go first, precede’
NNG	Labu	-mo	‘lead; go first’
NCV	Mota	mʷoa-i	‘first, foremost, principal; to be first’
NCV	Lewo	(va)mo	‘front’ (va ‘go’)
PMic		*mʷ(o,u)a	‘ahead, going before’ (Bender et al. 2003)
Mic	Chuukese	-mʷmʷe-ɾi	‘lead him, go before him’
Mic	Puluwatese	mʷmʷe-	‘lead’
Mic	Woleaian	mʷmʷa-	‘go ahead of it’
Fij	Wayan	mua	[VI] ‘head for a place, set course for a place’
PPn		*muqa	‘be first, precede’
PPn		*muqa-ki	‘before, first’
Pn	Tongan	muʔa	[VI] ‘precede, be/go in front’
Pn	Tongan	muʔa-ki	‘be the first or leading person’
Pn	Samoan	mua	‘be first, arrive first’
Pn	Rennellese	muʔaʔa-ki	‘to lead, direct’
Pn	Samoan	mua-ʔi	‘be first’
Pn	Tikopia	mua-ki	‘give/send/perform etc. first’

PWOc		*muga	‘be in front, precede’ (vol.2:249)
NNG	Gitua	muᵑga	‘precede, go ahead, future’
NNG	Lukep	-pa-mugu	‘cause to go first’ (pa- CAUSATIVE)
NNG	Mangap	-mūᵑgu	‘go ahead, go first, precede, go before’
NNG	Sio	muga	‘precede; before’
NNG	Gedaged	mug	‘precede’
NNG	Takia	mug	[VI] ‘go ahead, go first’
NNG	Mengen	muge	‘go ahead of, lead’
NNG	Yabem	muŋ	‘precede’
NNG	Numbami	-muᵑga	‘precede, go first’
NNG	Mapos Buang	muɢ-in	‘go first, go ahead’
NNG	Hote	-mɔŋ	‘pass, go first’
NNG	Adzera	moŋʔ(an)	‘precede’
PT	Muyuw	mug	‘go ahead, go first, precede’
PT	Suau	-muga-i	‘precede’
MM	Nakanai	muga	‘go ahead, forward, onward’
MM	Bola	muga	‘go first; lead, go first’
MM	Madak	mgo	‘go first’
MM	Ramoaaina	muga	‘go first’
MM	Ramoaaina	muga-in	‘lead’
MM	Siar	muga-i	‘lead the way’

PMP		*ma-udehi	‘be last; be after or behind; be late, be later; future’ (ACD)
POc		*muri	‘be behind, be after, follow’ (vol.2:251-252)
PT	Gapapaiwa	muri	‘follow’
PT	Dobu	muli-	‘follow’
PT	Bunama	muli-	‘follow’
MM	Nakanai	(ku)muli-	‘chase after (+s.t.)’
MM	Ramoaaina	muru	‘follow; behind, back; last’
MM	Teop	muri-na	‘follow (s.o., s.t.)’
MM	Banoni	mū-muri	‘last, follow last’
SES	Arosi	muri-	‘follow; behind, back; outside of s.t.; afterwards; left hand when facing an object’
Fij	Bauan	muri	‘following, after’
Pn	Samoan	mui	‘be/go behind, be late/last’
Pn	Samoan	mui-aki	‘be the last person(s), bring up the rear’
Pn	Samoan	muli	‘come last, be last; young, new’
Pn	Rennellese	mugi	‘follow, be or go behind or after; rear end, esp. lower or western end’
Pn	Hawaiian	muli	‘behind, afterwards; last, following behind; younger, youngest; (canoe) stern’

6. Caused movement verbs⇫

Caused movement refers to an event where an agent, usually a person, causes a theme (the person or thing that is moved) to move. The semantic frame of caused movement is rather complex. It involves

an agent (who/which moves the theme)
a theme (what is moved)
a source (the location the theme is moved from)
a goal (the location the theme is moved to)
the movement event itself

Thus in an English sentence like

She	moved	the pot	from the table	to the ground
AGENT	EVENT	THEME	SOURCE	GOAL

all five frame elements are profiled.¹⁵ Caused movement constructions differ as to whether these elements are all profiled. In this construction the source and the goal are each optional. In construction with, say, put (instead of moved), the goal is obligatory and for many English-speakers the source is obligatorily absent.

Crucially, however, the verb itself may also profile certain frame elements. Predictably some caused movement verbs profile elements associated with posture and movement verbs, discussed in earlier sections. These include:

the resulting posture of the theme (cf §6.2): transitive sit/seat/set, stand, lay, hang
locomotion (cf §6.3), i.e. the manner of movement of the theme: drop (‘let fall, cause to fall’), send (‘cause to go’), throw (‘cause to fly’), transitive roll
the direction of vertical movement (§6.4): lift, raise (‘cause to go up’), lower, drop (‘cause to go down’)
the path of movement (§6.4)): insert (‘cause to go into’), extract (‘cause to come out of’), immerse (‘cause to go into water’), box (‘cause to be in boxes’)
Other frame elements profiled by caused movement verbs relate to human causal agency:
whether or not the agent accompanies the theme in its movement:¹⁶
not specified, with take, push and pull
1. yes, with carrying verbs like carry, bring, drag
2. if no, then whether or not some part or tool of the agent touches the theme until it reaches its goal
  1. yes, with putting verbs like put, insert, immerse
  2. no, with sending verbs like send, drop, throw
  3. the kind of force that the agent applies in order to move the theme: push, pull, drag, throw
change of possession: give, present, distribute, receive, exchange, steal

The elements listed above are not mutually exclusive. Some of the English verbs given as examples occur more than once because their meaning potentially profiles more than one element. Because of this, verbs can be somewhat arbitrarily categorised in various ways. The organisation below is an attempt to categorise verbs by meaning in a way that is appropriate to Oceanic languages. A primary three-way division is based on item 9 which distinguishes among the following: a few verbs like take, push and pull where the agent may or may not accompany the theme in its movement (9a); carrying verbs, quite numerous in Oceanic languages (9b); and verbs of putting and sending (9c). Explanations of lower-order categories are distributed through the sections below, which are arranged as follows:

§6.6.1. Simple caused movement:
- §6.6.1.1. Taking (9a)
- §6.6.1.2. Verbs that profile vertical direction: raising and lowering (3)
- §6.6.1.3. Force-profiling verbs: pushing and pulling (9a, iii)
§6.6.2. Accompanied caused movement (9a):
- §6.6.2.1–12. Carrying verbs that profile the physical relationship between agent and theme
- §6.6.2.13. Dragging
§6.6.3. Unaccompanied caused movement:
- §6.6.3.1. Putting (9c-i)
  - §6.6.3.1.1. Simple putting
  - §6.6.3.1.2. Putting verbs that profile the resulting posture of the theme (1)
  - §6.6.3.1.3. Putting verbs that profile the path of movement (4)
  - §6.6.3.1.4. ‘Putting’ in reverse: taking
- §6.6.3.2. Sending (9c-ii)

The verbs reconstructed in the sections below consist of a root or, in the case of transitives formed from a root ending in -a or a consonant, of root + *-i. POc evidently had two morphological means of forming a caused movement verb from a locomotion (§6.3) or direction verb (§6.4). One was to prefix it with the causative derivational prefix *pa-, a process reflected in this Kwaio (SES) example:

Kwaio oli ‘return, come back to’ > faʔa-oli- ‘cause to return, send back’

The other was to add the applicative suffix *-akin[i]. Semantically this had two functions, as Evans (2003:203) notes. In the first it produced a verb like those formed with *pa-, in which the agent causes the theme to perform the action denoted by the root, as in

Manam	gege (VI) ‘roll’	>	gege-ak (VT) ‘roll’
	alale (VI) ‘walk’	>	alale-ak (VT) ‘help s.o. walk’
NE Ambae	saga ‘go on top’	>	sagat-agi[ni] ‘put s.t. on top’
Bauan	ðuru ‘enter’	>	ðurum-aki ‘insert s.t.’
	dromu ‘sink’	>	dromuð-aki ’push s.t. under (water),

In the typology above, these are simple or unaccompanied caused movement verbs.

In the second function *-akin[i] formed a verb in which the agent performs the action denoted by the root, and the theme is a concomitant:

NE Ambae	toa ‘run’	>	toa-gi[ni] ‘run off with s.t.’
	hivo ‘go down’	>	hivo-gi[ni] ‘go down with s.t., take s.t. down’
	vano ‘go’	>	vano-gi[ni] ‘go with s.t., take s.t.’
Bauan	ðiði ‘run’	>	ðiðiv-aki ‘run off with s.t.’
	galo ‘swim’	>	galov-aki ‘swim with s.t.’

Such verbs are verbs of accompanied caused movement, resembling verbs of carrying. However, in a carrying verb it is the physical relationship between the agent and the theme that is profiled. In an *-akin[i] verb it is the manner or direction of the agent’s movement that is profiled.

Verbs formed with *pa- and *-akin[i] are not considered further here, because we cannot be sure which derived verbs occurred in POc. But we can be sure that the morphosemantic processes reflected in these examples occurred in POc. The functions of *pa- have been widely discussed in the literature, and Evans (2003:195, 203) shows that the two functions of *-akin[i] mentioned here are reflected so widely that they must be reconstructed for POc.

6.1. Simple caused movement⇫

The word ‘simple’ is used here of verbs that are unspecified for a certain frame element. Verbs of taking (§6.6.1.1) and the force-profiling verbs of pushing and pulling (§6.1.2) are unspecified for item 9 in the list above. That is, they sometimes denote accompanied caused movement, sometimes unaccompanied. In the case of verbs of pushing and pulling, this is rather obvious: one can push a cart along a street (accompanied caused movement) or push a pot over (unaccompanied caused movement). In the case of ‘take’, the lack of specification is less obvious, and is explained in the following section.

6.1.1. Taking⇫

Simple verbs of caused movement are often glossed as ‘get’, ‘take’ or ‘bring’. They are used in constructions that also profile the source, e.g. ‘he took the knife from the table’, and in constructions that only profile the caused movement, e.g. ‘he took/brought my knife’.

English take occurs in dozens of idiomatic phrases (take heart, take hold, take a sip, take a break, take over, etc), but its core meaning involves accompanied movement by an agent of something (the theme) from one location to another. Either the source location or the goal location may be profiled, as for instance, in ‘take it off the ground’ and ‘take it to its mother’. In this respect Oceanic languages are similar to English. The Vitu (MM) verb pele- is used with a source in this example. The source is ‘her betelnut basket’.

Vitu (MM)

‘The spirit woman took a stamper from her betelnut basket.’ (129)
Na	tavine	hanitu	e	pele-a	na	tureturea	na	ka-na	kolopi	mamaha.
ART	woman	spirit	RLS:3	take-3SG	ART	stamper	PREP	PCL-3SG	basket	chew.betelnut

However, English take also profiles direction away from the deictic centre, contrasting with bring, which profiles direction towards it. Oceanic languages, however, encode this deictic contrast with a deictic direction verb—or a deictic morpheme derived from a verb (vol.2:273-282)—in the last slot of a serial verb construction, as in these Vitu (MM) examples.

- Vitu (MM): (Berg & Bachet 2006:180)
  ‘He took it away.’
  Ia pele-a vano.
  
  He take-3SG go
  
  … MANNER DEIXIS

- Vitu (MM): (Berg & Bachet 2006:181)
  ‘Bring the spear here!’
  Pele-a haine mai!
  
  take-3SG iron.spear come
  
  MANNER … DEIXIS

This leaves pele- profiling only the means of movement, namely that the agent moves the theme, i.e. simple caused movement. The verb pele- is the same in all three examples. This is a typical Oceanic configuration, such that one caused movement verb corresponds to both English take and bring. A practical result of this is that these Oceanic verbs are often glossed as one or more of ‘take’, ‘get’ and ‘bring’, but the verbs have the same functions whichever of these glosses is used. Thus the verbs reconstructed here overlap two semantic domains. The first is ‘reverse putting’, e.g. ‘taking off the table’ as opposed to ‘putting on the table’ (§6.6.3.1.4), a subdomain of unaccompanied caused movement. The other is carrying (§6.2), a subdomain of accompanied caused movement.

A second semantic complication, discussed by D’Jernes (2013) with regard to Lukep (Pono) -kaua/-kap ‘get, give’ (where the first form takes a singular object, the second a plural) is that a verb meaning ‘take, get’ in some languages also means ‘give’. In others, the ‘give’ verb is derived historically from a ‘take, get’ verb. This is less strange than it appears. If the basic meaning of these verbs is simply that the agent moves the theme, then the addition of a recipient is enough to narrow the movement to ‘giving’.¹⁷

The verbs reconstructed below also raise formal challenges. Broadly, these are due to the fact that words of high token frequency may fail to undergo changes that would otherwise affect them. For example, the Lukep (Pono) pair mentioned above, -kaua and -kap, are exceptional in encoding their object respectively as singular and plural. Lukep (Pono) has generally lost POc object indexing enclitics, but retains -a 3SG on a small number of now ‘irregular’ verbs (D’Jernes (2002).

However, at least one such irregularity originated long before the emergence of POc. Blust (ACD) reconstructs PAn/PMP *alap, PMP *alaq both ‘fetch, get, take’, and PAn *ala ‘take, get, fetch, obtain’. Their similarities of form and meaning indicate that they have a shared origin that cannot be reconstructed. PMP *alap and *alaq were both inherited into POc, but with a complication. The predicted POc transitive forms are *alap-i- and *alaq-i-. The former is reflected in Gela and Tongan (immediately below). But more widely reflected are forms in which initial *a- has vanished, i.e. *lap-i- and *la(q)-i-. The *a-less forms evidently already occurred in POc, and for clarity’s sake are reconstructed separately below

PAn/PMP		*alap	‘fetch, get, take’ (ACD)
POc		alap, alap-i-	‘get, take’
SES	Gela	alav-i	‘get, take’
Pn	Tongan	alaf-i	‘catch it’

PAn/PMP		*alap	‘fetch, get, take’ (ACD)
POc		*lapi	‘take, get, give’
Adm	Baluan	lɪp	‘get’
SES	Gela	lavi	‘bring, take, get’
PNCV		*lavi	‘carry, take’
NCV	Mota	lav	‘take, receive’
NCV	Nokuku	lap-i	‘give’
NCV	Valpei	lavi-	‘take’
NCV	Tangoa	lavi	‘take, receive’
NCV	Ambae	lawe	‘to’ (dative)
NCV	Big Nambas	-laði	‘take’
NCV	Port Sandwich	liv	‘carry’
NCV	Aulua	levi-	‘take, give, receive’
NCV	Maskelynes	lavi	‘take’
NCV	Maskelynes	lav-xin-i	‘give’
NCV	Naman	lev	‘give; take, get, fetch’
NCV	Apma	lev	‘take’
NCV	Paamese	lahi	‘carry; pick up (especially of many small things)’
NCV	Lewo	lavi-	‘take’
NCV	Nguna	lavi	‘bring, take’

PMP		*alaq	‘fetch, get, take’ (ACD)
POc		*ala(q)	‘take, get’
NCV	Loh	ɔlə	‘bring, take’
NCV	Hiw	ɔyə	‘bring, take’
PMic		*ala	‘take, get’ (Bender et al. 2003)
Mic	Kiribati	ana-	‘take, subtract, take away, remove, take off’
Mic	Ponapean	alɛ	‘take, get (s.t.)’
Mic	Mokilese	ɔlɔ	‘get, take, gather (s.t.)’
Mic	Mokilese	ɔlɔ-ɔl	‘get, take, gather’

PMP		*alaq	‘fetch, get, take’ (ACD)
POc		*la(q)-i-	‘take, get, bring’
Adm	Titan	la-i	‘take, get’
PNCV		*la-i	‘take, give’ (Clark 2009)
NCV	Mota	le, la	‘give, take’ (disambiguated by directional particles)
NCV	Raga	lai	‘give, take, bring, get’
NCV	Nokuku	lɔ	‘put, take, receive, have’
NCV	Nokuku	lei-a	‘take’
NCV	Nokuku	la-ma	‘bring’
NCV	Tamambo	lai	‘take, get, bring’
NCV	Uripiv	-la-i	‘take’
NCV	Uripiv	-la	‘give’
NCV	Lewo	la	‘bring, give’
NCV	Namakir	la	‘hold’
SV	Anejom̃	le	‘get, take, receive (SG OBJ)’ (John Lynch, pers. comm.)

Three further verbs which, like those above, denote simple caused movement, are reconstructed below. The first, POc *pa, is phonologically unusual in that its root is a monosyllabic CV form. Most POc roots are polysyllabic. The New Caledonian forms in this set are from Ozanne-Rivierre (2004b).

POc		pa, pa-i-	‘get, take, bring’
NNG	Amara	pei	‘get’
NNG	Kaulong	va	‘get, take, bring’
PT	Tawala	wa-i-	‘carry, get, select, take’
PT	Iduna	-va-i-	‘get, take, fetch’
SES	Kwaio	fee-	‘take, convey, take in marriage’
NCal	Nyelâyu	pʰa	‘take’
NCal	Nêlêmwa	fʰe	‘take, carry’
NCal	Nemi	fe	‘take’ (< *pa-pa)
NCal	Paicî	pá	‘take, bring’ (< *pa-pa)
NCal	Cèmuhî	pé	‘take’ (< *pa-pa)

POc		*ŋal(e,i)	‘get, take, carry, bring’
NNG	Gedaged	-ŋale	‘get, lay hold of, acquire, come to hand, obtain, procure, come by’
NNG	Takia	-ŋale, -ŋili-	‘get, obtain’
SES	To’aba’ita	ŋali-	‘take, carry’
SES	Lau	ŋali-	‘take, carry, bring, get’
SES	Kwaio	ŋali	‘hold, carry’

The exact form of the reconstruction below is not entirely clear. Evidence for the phoneme following initial *ka- is conflicting. Seimat and Kilivila kau, Lukep (Pono) -kau-a, Patep ko, Sinaugoro -ɣau, Wayan Fijian kau and Samoan ʔau-mai support the reconstruction of POc *kʷau. Other evidence points to a labial, but which labial is unclear. Tuam ɣam, Mangap kam, and Nehan kaba support *kʷabV, while a number of items indicate *kʷap. There are two possible explanations here. One is that the cognate set reflects more than one POc morphological variant, just as Lukep (Pono) preserves two variants (D’Jernes 2013). The other is that this is not one cognate set but two, reflecting similar but separate POc forms. The forms listed under ‘cf. also’ complicate the picture, but appear to reflect a separate form *ka[-i-].

POc		kʷau, kʷa(p,b)-i-	‘get, take’
Adm	Seimat	kau	‘bring, carry, take’
Adm	Seimat	kau-ma	‘bring, carry hither’ (ma ‘come’)
NNG	Tuam	-ɣam	‘take, bring, get, give’
NNG	Lukep	-kap	‘get, give’ (with plural object)
NNG	Lukep	-kau-a	‘get, give’ (with singular object)
NNG	Mangap	-kam	‘take, bring, get, give, put, do, cause’
NNG	Mangseng	ke(ne)	‘get’
NNG	Patep	ko	‘get, take’
PT	Kilivila	-kau-	‘take (away), carry’
PT	Dawawa	-kaua	‘do’
PT	Sinaugoro	-ɣʷa	‘carry (in general)’
MM	Sursurunga	kip-i	‘get, carry’
MM	Patpatar	kap	‘get, take’
MM	Tolai	kap	‘carry’
MM	Label	kep	‘carry’
MM	Siar	kep	‘get’
MM	Nehan	kaba	‘carry, lift’
SV	Anejom̃	a-kau	‘carry (wood etc) in the middle (twisting and turning to avoid obstacles)’
Fij	Wayan	kau	[VI, inanimate subject] ‘be carried, taken’; [VT, inanimate object] ‘carry, take s.t.’
Pn	Samoan	ʔau-mai	‘bring’
Pn	Samoan	ʔav-atu	‘take’ (< *kau-atu)

cf. also:

Adm	Mussau	ga	‘get’
NNG	Kove	kea	‘take’
NNG	Sio	kai	‘hold, get; carry’
NNG	Kaulong	kai	‘be stuck; grasp, hold’
PT	Dawawa	-kʷaya	‘grab s.t. from s.o.’
MM	Madak	ka-ka	‘get’
SES	Owa	ɣa-i-	‘remove s.t.’

Almost certainly related to the above is POc *(k,kʷ)awe ‘carry, carry away’, but the relationship does not reflect a known derivational process.

POc		*(k,kʷ)awe	‘carry, carry away’
NNG	Sio	kawe-	‘snatch, grab s.t. and flee with it’
PPn		*kawe	[VT] ‘carry, bear’
Pn	Tongan	kāve-	‘be carried off one’s feet, especially by love’
Pn	Samoan	ʔave	‘give (s.t.) to, hand (s.t.) to; carry, take s.t.; send s.o./s.t.’
Pn	Tuvalu	ka-kave	‘carry’
Pn	East Uvean	kave-kave	‘carry’
Pn	Sikaiana	k-kave	‘carry to’
Pn	Tikopia	kave	‘carry, bear off’
Pn	Tokelauan	kave	‘take, carry, give’
Pn	Tuamotuan	kave	‘carry’
Pn	Pukapukan	kave	‘take, remove, carry, give’
Pn	Rennellese	ka-kabe	‘escort, accompany, take, as in a canoe; be escorted, taken’
Pn	Marquesan	kave	‘carry’
Pn	Māori	kawe	[VT] ‘carry, convey, bring, go to fetch’

6.1.2. Raising and lowering⇫

A verb for a certain kind of raising, POc *laŋa(t), *laŋat-i- ‘raise, pull up, lever up’, is reconstructable.

POc		laŋa(t), laŋat-i-	‘raise, pull up, lever up’
Adm	Lou	laŋ	‘lift up’
PT	Molima	laɣas-i-	‘raise’
MM	Nakanai	laga	‘pull up on snare in which animal is caught; pull a rope; fish with a net, lift fish out in net’
SES	Arosi	raŋa-i-	‘raise, lift up’
PNCV		*laŋa	‘lift (flat object from surface)’ (Clark 2009)
NCV	Mota	laŋa	‘lift up, turn up, so as to show underside’
NCV	Raga	laŋa-i	‘raise as on hinge; raise thatch layer, mend roof’
Fij	Rotuman	laŋa	‘raise one side of’
PPn		*laŋa	‘raise up’ (POLLEX)
Pn	Niuean	laŋa-aki	‘raise up’
Pn	Tongan	laŋa	‘raise up’
Pn	East Futunan	laga	‘lever up, dig up yams etc’
Pn	Pukapukan	laŋa	‘pull up, raise, pull in a line in fishing’
Pn	Samoan	laŋa	‘raise up, as a heavy weight or a conquered party; rise from a sitting position’
Pn	Tuvalu	laŋa-laŋa	‘pull up’
Pn	Tikopia	raŋa	‘raise, rise, swell; arrange’
Pn	Kapingamarangi	laŋa	‘lift or move with a lever’
Pn	Māori	raŋa	‘raise up, cast up, pull up by roots’

No verb of lowering can be reconstructed, but a number of languages form a verb ‘lower’ by attaching a reflex of the POc causative prefix *pa- to a verb meaning ‘be deep’ or ‘go down’. Examples include Bariai (NNG) pa-sil from sil ‘deep’, Sursurunga (MM) a-sih-əi from sih ‘go down’, To’aba’ita (SES) faʔa-sifo-a from sifo ‘go, move down, descend, more or less vertically’.

6.1.3. Force-profiling verbs: pushing and pulling⇫

Verbs of pushing and pulling profile the force that the agent exerts on the theme in order to move it. Whether the agent accompanies the theme along the path of movement is not part of the verb’s lexical meaning, but is conveyed by the construction in which the verb is used, e.g. They pulled the canoe up the beach (accompanied caused movement) vs He pulled the bag open (unaccompanied caused movement).

6.1.3.1. Pushing

Oceanic languages tend to have a number of path-profiling verbs of pushing, e.g. ‘push open’, ‘push in’, ‘push through’, ‘push aside’, ‘push away’, ‘push under’, ‘push against’. None of these can be reconstructed from available data, but these ‘push’ terms from Lou (Adm), Lewo (NCV) and Sinaugoro (PT) give an indication of one reason for this.

Lou (Adm)
suek	‘push’
suer	‘push ground’
susuer	‘push ground with foot’
suput	‘push away, make go’
susuŋ	‘push clothes up’

Lewo (NCV)
suponia	‘push’
suponipu	‘push and turn s.t.’
supʷoni	‘push, shove’
susuni	‘move, push’
suwani	‘push into hole to block it’

Both the lists above, geographically separated as they are, consist of words with initial su-. The fact that so many ‘push’ verbs in each language begin with the same syllable cannot be coincidence, and probably tells us that these forms reflect earlier nuclear serial verb constructions, a productive construction in which two verbs formed a compound (Lou suek ‘push’ is an exception: -ek reflects the POc applicative suffix *-aki(n)). This hypothesis is supported by the fact that in Sinaugoro, such forms still are transparent compounds where the second element has its own lexical meaning (forms preceded by a hyphen in the third column always occur as the final part of a compound).

Sinaugoro (PT)
dori	‘push’
dori-faka	‘open by pushing’	faka	‘between’
dori-ɣau	‘close by pushing’	-ɣau	‘cover’
dori-maraɣe	‘push up’	raɣe	‘go up’
dori-riɣo	‘push down’	riɣo	‘go down’
dori-rosi	‘push out’	-rosi	‘out’
dori-tari	‘push down on the ground’	-tari	‘down’
dori-toɣa	‘push inside’	-toɣa	‘inside’

Sinaugoro dori ‘push’ is not cognate with Lou or Lewo su- but the latter are probably cognate with each other and probably reflect POc *qusur ‘push, shove’ with loss of the first syllable. Other Oceanic ‘push’ forms with initial su(r)- are listed under ‘cf. also’.

PMP		*qunzur	‘thrust out, extend forward’ (ACD)
POc		*qusur	‘push, shove’ (ACD)
SES	Talise	usu-	‘push’
SES	Tolo	usua-	‘push’
SES	To’aba’ita	ūsu(tani-)	‘push’
SES	Lau	usu	‘push’
SES	Lau	usu-i-	‘push, impale’
SES	Kwaio	usu	‘push’
SES	’Are’are	usu	‘push, shove’
SES	Sa’a	usu	‘push, shove, launch a boat’
SES	Sa’a	usu-i-	‘push, shove, launch a boat’
Fij	Bauan	usu	‘thread through, as a rope through a hole’
Fij	Wayan	usu	‘press against s.o., importune s.o.’
Pn	Samoan	usu-i-	‘thrust’
Pn	Tuvalu	uhu	‘push’
Pn	Nanumea	uhu	‘push’
Pn	Rennellese	usu	‘push, as a canoe into water’

cf. also:

NNG	Lukep	-surpak-i	‘push’
NNG	Bariai	su-sur-an	‘push’ (-an APPLICATIVE)
PT	Dawawa	suriɣa	‘push over violently’
MM	Nakanai	susu	‘push into, pierce’
MM	Kokota	huzu-i-	‘push’ (h- < POc *s-)

A second POc term for ‘push’, *juju(n), *juni- ‘push’, is reconstructed below. There are some unanswered marginal queries about this reconstruction. It is reflected to my knowledge in only one non-WOc language, Lau, but Lau does not usually contain WOc loans, and is a reliable witness. Ughele and Marovo juno- suggest a transitive form †*juno-, but all other reflexes point to instantiation of the paradigm identified by Blust (1977a) (vol.1:25), i.e. the forms as reconstructed.

POc		juju(n), juni-	‘push’
NNG	Sio	zuzu-	‘push forward; shove from behind with force’
NNG	Takia	-duduni-	‘push, push through, push inside with something, urge on’
NNG	Patep	du	‘move, push’
PT	Gapapaiwa	dudu	‘push’
PT	Tawala	dudu	‘push’
PT	Sinaugoro	dudu(gʷanu)	‘put the spear on an object and push’ (gʷanu ‘stab’)
MM	Kubokota	juju-	‘push’
MM	Lungga	juju	‘push’
MM	Simbo	juju	‘push’
MM	Nduke	zuzu-	‘push’
MM	Ughele	juno-	‘push’
MM	Marovo	juno-	‘push’
SES	Lau	dudu	[VI] ‘push, move’
SES	Lau	du	[VT] ‘push, move’

6.1.3.2. Pulling

There are a good many path-profiling ‘pull’ verbs in Oceanic languages. One, POc *pupu(t), *puti- ‘pick (fruit +), pluck (feathers +), pull out (weeds +)’, is widely reflected (vol.1:277–278). An apparently simple verb of pulling, POc *Rape ‘pull, drag’, has quite widely distributed reflexes. POc *dradra ‘pull’ has just three known reflexes, all NNG, but non-Oceanic cognates support the POc reconstruction.

POc		*Rape	‘pull, drag’
NNG	Takia	-rae	‘inhale, pull (e.g. on a rope), attract, extend (e.g. talk), drag, stretch out (by pulling)’
NNG	Sio	lae	‘pull behind, drag’
MM	Nakanai	lave	‘pull’
MM	Madak	rep	‘pick, pull’
MM	Patpatar	rahi	‘pull, drag’
MM	Tinputz	ræh	‘pull’
SES	Lau	lafi	[VI] ‘pull, draw, draw a tooth, pull a string’
SES	Lau	lafi-	[VT] ‘drag, pull up’
SES	Kwaio	la-lafi-	‘pull (tight), pull on, snatch’
PNCV		*rave	‘pull’ (Clark 2009)
NCV	Mota	rave	‘pull; draw out fish, catch fish with a line’
NCV	Mota	rave-g	‘drag s.o./s.t.’
NCV	Raga	rava	‘pull, draw, write’
NCV	Nokuku	rav-rav	‘catch fish’
NCV	Kiai	reve-a	‘pull, drag’
NCV	Tamambo	reve	‘pull’
NCV	Uripiv	-rev	‘pull’
NCV	Lonwolwol	re	‘hold, take, pull, etc.’
NCV	Lonwolwol	rvi	‘to pull along, tow, haul’
NCV	Paamese	lehe-lehe	[VI] ‘pull’
NCV	Paamese	lehe	[VT] ‘pull’
PSV		*a-yevi
SV	Sye	yevi	‘pull’
SV	Ura	yevi	‘pull’
SV	North Tanna	i	‘pull’
SV	Lenakel	vi	‘pull’
SV	Lenakel	evi	‘pull out’
SV	Lenakel	evi-evi	‘pull in jerks’
SV	Anejom̃	a-yihi-i	‘pull’

PCEMP		*dada	‘pull, haul, drag’ (ACD: PCMP)
POc		*dradra	‘pull’
NNG	Bariai	dada	‘pull’
NNG	Hote	-dadi	‘pull (on ground or from hole)’
NNG	Mapos Buang	dad	‘pull’

6.2. Accompanied caused movement⇫

Verbs of accompanied caused movement are verbs of carrying and dragging. Many Oceanic languages have no default verb of carrying, but employ a number of carrying verbs, depending on how something/someone is carried and to a lesser degree on what is carried.¹⁸ Meanings that crop up with considerable frequency include

carry on the head
carry hanging from the head
carry a child in a sling on the back
carry on the shoulder
carry hanging from shoulder
carry piggyback
carry with a shoulder pole
carry on a long shoulder pole between two people
carry under the arm (and on the hip)
carry in both arms
carry on a canoe

Of these, the first three modes listed are usually restricted to women, whilst carrying something on or hanging from the shoulder is typically a male mode. Hill (2015) remarks, ‘In the villages of Longgu district there is only one way to transport goods or children on land and that is for people to carry them. There are no animals used to carry loads, no bicycles, wheelbarrows or cars.’ This was true of all pre-contact Oceanic speaking communities and was true of communities speaking POc. Hill goes on to point out that while Longgu does not have a generic verb ‘carry’, it does have a verb zabe ‘not carry anything’, ‘behavior that is considered either unusual or unacceptable’ in a community where every able-bodied person must contribute to the daily labour of the village.

6.2.1. Carrying in general⇫

POc, however, appears to have used *puat as a generic verb of carrying and transporting, although when the agent was male it apparently also could have the specific meaning ‘carry on the shoulder’. In its generic sense it probably overlapped with POc *kʷau, *kʷap(i) ‘get, take’ (§6.6.1.1), but the glosses in the cognate set below suggest that generic *puat profiled carrying or transporting a load, whereas *kʷau, *kʷap(i) denoted carrying in a more general sense.

PMP		*buhat	‘lift, stand up, arise, emerge, begin, depart, carry; cargo; take something; take a wife’ (ACD)
POc		*puat	‘carry, transport from place to place; carry on shoulder’
NNG	Kaiwa	vuat	‘carry’
PT	Motu	hua-	‘carry on the shoulder’
SES	Longgu	hua-	‘carry something in a bag slung from the head’
SES	Lau	fūa	‘carry, as a haversack’
SES	Kwaio	fua	‘carry a load (of vegetables, coconuts, water); bring food’
SES	Kwaio	fua-geni	‘marry, take a woman in marriage’ (geni ‘woman’)
SES	’Are’are	hua	‘bring, carry’
PMic		wua, wuawua, wua-ti, wua-ta	‘carry, convey, transport’
Mic	Kiribati	uot-a	‘carry (s.t.), carry in arms, rock, lull to sleep’
Mic	Kiribati	uou-a	‘engage in carrying a child’
Mic	Chuukese	wuwa	‘convey, ship, carry from one place to another’
Mic	Chuukese	wue-yi, wue-e-	‘carry, convey, transport (s.t.)’
Mic	Mortlockese	ua	‘carry’
Mic	Mortlockese	uæ-æ-	[VT] ‘carry’
Mic	Puluwatese	wua-	‘carry, transport (s.t.)’
Mic	Puluwatese	wuɔ̄wo	‘carry’
Mic	Woleaian	uate	‘load, belongings’
Mic	Ponapean	wwa	‘carry’
Mic	Kosraean	wiwɛ	‘do carrying’
Mic	Kosraean	wʌ-n	‘carry, transport (s.t.)’
Fij	Bauan	vua	‘bear two burdens on the shoulder (one at each end of a pole’
Fij	Wayan	vua	‘be carried on a pole’
Fij	Wayan	vuat-i	‘carry (s.t.) on a pole resting on the shoulder’
Pn	Tongan	fua	‘lift, carry on shoulder with a stick’
Pn	Tongan	fue-si-a	‘carry, bear (a burden)’
Pn	Niuean	fua	‘carry on the shoulder; weigh’

cf. also:

SES	Sa’a	fuʔe	‘carry suspended from the head’
SES	Arosi	hua, huas-i	‘carry suspended from the head’

6.2.2. Carrying on the head⇫

The central meaning of POc *suqun was with reasonable certainty ‘carry on the head’, attested by non-Oceanic cognates from Taiwan to CMP (ACD). In Takia and Megiar the meaning has been extended to carrying in general. In NCV, as Clark (2009) recognises, the extension is to wearing on the head, then to wearing in general. That these terms are cognate is confirmed by the presence of the glottal stop in Namakir hiʔin, Namakir being the only NCV language regularly to reflect POc/PNCV *q.

PAn		*suquL	‘carry on the head’ (ACD)
POc		suqun, suqun-i-	‘carry on the head’
NNG	Megiar	-suni	‘carry’
NNG	Takia	-sini-, -suni-	‘carry or bear’
MM	Siar	sun, su-sun	‘carry on the head’
MM	Sursurunga	sun-sunun	[VI] ‘carry on the head’
PNCV		suqun, suqun-i-	‘carry on the head, wear on the head’ (Clark 2009: *suquni)
NCV	Ambae	huna	‘umbrella’
NCV	Tolomako	su-suni	‘wear on head’
NCV	Tolomako	sunu-n	‘hat, umbrella’
NCV	Raga	huni	‘carry or wear on head’
NCV	Nokuku	suni-m	‘hat’
NCV	Nokuku	sun	‘umbrella’
NCV	Uripiv	-sun-sun	‘carry or wear on head’
NCV	Port Sandwich	cün-i	‘carry on head, dress oneself, put on a hat, put flowers in the hair’
NCV	Lonwolwol	sun-e	‘put on, wear in belt at back (used of scented leaves worn by chiefs)’
NCV	Paamese	sinu	‘dressed; dress up; get dressed’
NCV	Namakir	hiʔin	‘wear’
NCV	Nguna	sūn-i	‘carry on head; put on, wear (on upper body)’

cf. also:

Fij	Wayan	sū	‘(inanimate, e.g. fruit) be carried in one’s skirts or shirt’
Fij	Wayan	sūn-i	‘cover or wrap s.t.; carry s.t. in one’s skirts’

PWOc *kud(r)u is also reconstructable as a term for carrying on the head. How this differed in meaning from POc *suqun is not clear, but the Mangap and Nakanai glosses below note that women carry burdens in this way, and it may be that *kud(r)u denoted a female carrying style.

PWOc		*kud(r)u	‘carry on the head’
NNG	Takia	-kud-i	‘wear on one’s head, carry on head’
NNG	Bing	kud	‘hold, carry on head, carry child on shoulder’
NNG	Sio	kudu	‘carry on head’
NNG	Mangap	-kūⁿdu	‘carry s.t. on head (women carry like this)’
NNG	Bariai	ud	‘carry on head’
NNG	Lukep	-kudu	‘carry on head’
NNG	Mangseng	ur	‘carry on head’
NNG	Mengen	kuru	‘carry on head’
NNG	Mengen	kur-e	‘carry s.t. on head’
NNG	Mapos Buang	kud	‘carry; carry on head or shoulders’
NNG	Mangga	kud	‘carry on head’
NNG	Vehes	kud	‘carry on head’
PT	Iduna	-kedu	‘carry on head’ (for †-kudu)
PT	Sinaugoro	ɣuru	‘carry on head’
MM	Nakanai	hugu	‘carry on head, of women’s loads or men’s masks’ (h- for †k-; h- < *q-)

cf. also:

NNG	Kaulong	un	‘carry on head’ (for †kuh)
NNG	Numbami	-kuku	‘carry on head, astride shoulders’ (for †kudu)

6.2.3. Carrying hanging from the head⇫

When a woman from New Guinea carries a loaded netbag (bilum in New Guinea Tok Pisin), she lets it hang down behind her resting on her back and she pulls its woven handle over her head to rest across her forehead so that the bag is suspended from it. Occasionally other loads are carried in this way too. Interestingly, no POc term for this mode of carrying can be reconstructed, and this matches the fact that no term for a netbag can be reconstructed either (vol.1:79), apparently because the bilum was a New Guinea highland artefact that spread to the lowlands after the dispersal of Oceanic languages from NW Melanesia. This carrying mode has spread into the Solomon Islands, but is attested only at odd locations in the north of Vanuatu.

Some Admiralties and WOc terms for this carrying mode, along with one NCV term, look as if they are reflexes or borrowings of reflexes of PWOc *kud(r)u ‘carry on the head’ (§6.6.2.2). Note that Tawala and Gapapaiwa gedu share the replacement of -u- by -e- also found in Iduna kedu above.

Adm	Loniu	kun	‘carry suspended from the head’
Adm	Lou	kun	‘carry on the back; basket worn on the back’
Adm	Baluan	kun	‘small basket made of tree bark fibre worn on the shoulder’
NNG	Lukep	-gud	‘carry on the head or by the second rope on the head’
NNG	Mengen	kul-e	‘carry hanging from the head’ (borrowed? -l- does not reflect *-d(r)-)
PT	Tawala	gedu	‘carry supported by cord around forehead’
PT	Gapapaiwa	gedu	‘carry suspended from the head’
NCV	Hiw	kʷɔt	‘carry on back (one burden) using headstrap’

The most likely borrowing above is Poeng kule as Poeng -l- reflects POc *-r-, -R- or -l- but not *-d(r)-. Significantly Madden (n.d.) annotates this entry with a comment that Poeng speakers do not normally make or use bilums.

6.2.4. Carrying piggyback⇫

The main POc term for carrying someone piggyback (or pick-a-back, the older English form) is *pʷa(p,pʷ)a, which is of PAn antiquity. Its exact POc form is uncertain, as the phonological evidence is conflicting. At the same time, the similarities in both form and meaning make it probable that these items form a cognate set. PMP *baba could regularly have become either POc *baba or *papa. Clark (2009) correctly observes that NCV items reflect either PNCV/ POc *baba or PNCV *bava (POc *bapa). One might reconstruct alternant POc forms, but there is a solution which unites the apparently conflicting lines of evidence. Lynch (2002e) shows that PMP labials surface sporadically as POc labiovelars, and that the reflex of POc *pʷ in WOc languages is often a fortis p. The evidence below suggests that the initial POc consonant was indeed *pʷ, an inference supported by the rounded vowels of Poeng pope and Arosi boha. It is less clear whether the medial consonant was *-pʷ- or *-p-, as some reflexes are fortis, others lenis.

There are also forms reflecting POc *papa, but I argue in §6.6.2.5 that these form a separate cognate set.

PAn/PMP		*baba	‘carry a person pick-a-back; ride pick-a-back’
POc		*pʷa(p,pʷ)a	‘carry pick-a-back’ (ACD)
Adm	Mussau	bao	‘carry pick-a-back’
Adm	Lou	pap	‘carry s.o. on the back’
NNG	Mengen	pop-e	‘carry on the back’
NNG	Vehes	(pir)pev	‘carry piggyback’
NNG	Mapos Buang	piv	‘carry piggyback’
NNG	Mangga	pēv	‘carry piggyback’
NNG	Patep	piv	‘carry piggyback’
NNG	Piu	pep	‘carry piggyback’
PT	Kilivila	-papi-	‘carry piggyback’
MM	Roviana	papa	‘carry a child on one’s back’
SES	Bugotu	papa	‘ride on back’
SES	Gela	papa	‘carry piggyback’
SES	Longgu	papā	‘carry s.o. on the back’
SES	Arosi	baha, boha	‘carry pick-a-back or in cloth on back’
PNCV		bava, baba	‘carry child; bear child’ (Clark 2009)
NCV	Hiw	pep	‘carry on back (child+)’
NCV	Mota	pepe	‘carry a child on the back or astride on the hip’
NCV	Raga	bava	‘carry (child or load) on back; carry on hip; bear (child)’
NCV	Ambae	baba	‘give birth’
NCV	Nokuku	papa	‘be born’
NCV	Nokuku	pap-i (jura)	‘bear (child)’
NCV	Kiai	pava	‘give birth’
NCV	Uripiv	-pepe	‘carry child on back’
NCV	Port Sandwich	ᵐbav-e	‘bear a child on the back; carry on one’s back’
NCV	Lonwolwol	baba	‘used of a baby, ride (on its mother’s back), or of the mother, to carry by slinging on her back’
NCV	Lewo	papa	‘carry (children’s talk)’

6.2.5. Carrying a child in a sling on the back⇫

During the research for this chapter the set below was included in the set in §6.6.2.4 above. Two facts led to its separation and to the reconstruction of POc *papa. First, Lonwolwol has two forms, baba and fefa (N Ambrym fafa), admittedly not very different in meaning. Second, Polynesian forms reflect PPn *fafa rather than *papa.

It may well be that there is a historical connection between POc *pʷa(p,pʷ)a ‘carry pick-a- back’ and POc *papa, tentatively ‘carry a child slung on the back’. It is also possible that reflexes of one have been contaminated by reflexes of the other.

POc		*papa	‘carry a child slung on the back’
Adm	Wuvulu	fafa	‘carry s.o. on one’s shoulders (legs straddling neck)’
SES	To’aba’ita	fafā	‘carry piggyback’
SES	Lau	fafa	‘carry on shoulders, pick-a-back; carry a bag round the neck’
SES	’Are’are	haha	‘carry one the back’
SES	Sa’a	haha	‘carry s.o. on one’s back’
NCV	Lonwolwol	fefa	‘carry baby or child slung on the back’
NCV	Paamese	hehe	‘carry child on back in cloth’
Fij	Bauan	vava	‘carry a child on the back’
Pn	Tongan	fafa	‘carry on the back; be carried, have a ride on someone’s back’
Pn	Niuean	fafa	‘carry on back’
Pn	Samoan	fafa	‘carry (child or other load) on one’s back’
Pn	Anutan	papa	‘carry s.o. on one’s back, piggyback’
Pn	Hawaiian	waha	‘carry on the back, as a child’
Pn	Māori	waha	‘carry on the back’

6.2.6. Carrying on the shoulder⇫

See also the generic carrying verb POc *puat (§6.6.2.1), which also seems to have had a specific use ‘carry on shoulder’.

The terms listed below reflect POc *[qa]paRa- ‘shoulder’, reconstructed in §3.5.3. It is possible that the term has been repurposed as a verb at various times and places, but two facts speak against this. First, all the forms listed reflect *qapaRa rather than simply *paRa. As noted in the reconstruction of *[qa]paRa-, there are a number of body-part and other nouns that are reflected with and without *qa-. If the repurposing had taken place on various occasions, we would expect some reflexes without *qa-, but none are found. Second, the repurposing must be quite old, as the Malalamai, Gapapaiwa and Longgu forms are no longer used for ‘shoulder’, and the Gumawana, and Misima forms are no longer identical to the noun ‘shoulder’.

PAn		*[qa]baRa	‘shoulder’ (ACD)
POc		*[qa]paRa-	‘shoulder; carry s.t. on the shoulder’
NNG	Malalamai	avala	‘carry, esp. on the shoulder’
PT	Gapapaiwa	kavara	‘carry’
PT	Gumawana	kavala	[VI] ‘carry on shoulder’ (vala-vala- ‘shoulder’)
PT	Gumawana	kavale	[VT] ‘carry s.t. on shoulder’
PT	Tawala	awala	‘carry on shoulder’ (awala- ‘shoulder’)
PT	Misima	havala	‘carry hanging from the shoulder’ (probable loan)
PT	Misima	kaval	‘carry, carry on shoulders, load carried’ (vevela- ‘shoulder’)
SES	Longgu	kaveria	‘carry s.t. by hanging it on shoulder’ (-r- for †-l-)
SES	Arosi	ʔabara	‘carry on the arms clasped on chest; shoulder’ (-b- for †-h-)
PMic		*afara	‘shoulder’ (Bender et al. 2003)
Mic	Puluwatese	yayefar	‘load carried on the shoulder; shoulder’
Mic	Pulo Annian	yaθaθala	‘carry on the shoulder’

6.2.7. Carrying hanging from shoulder⇫

POc *sape below is not widely attested, but non-Oceanic cognates support its reconstruction.

PMP		*sampay	‘drape over the shoulder or from a line, as a cloth’ (ACD)
POc		*sape	‘carry by a strap over the shoulder’
Fij	Wayan	ðove	‘be carried slung from the shoulder or slung from a hook, peg, or line’ (-o- for †-a-)
Pn	Tongan	hafe	‘carry by means of a strap or rope, etc. across the shoulder’
Pn	East Futunan	safe	‘be slung over the shoulder’

6.2.8. Carrying with a shoulder pole⇫

A carrying mode that is quite common among men in Oceanic societies is to attach a load to one end of a short pole and to balance the pole on the shoulder with the load hanging behind and the hand holding the pole’s front end. The POc term for this was *sola(t), *solat-i-, widely reflected in Oceanic languages.

POc		sola(t), solat-i-	[VT] ‘carry with a shoulder pole’
NNG	Sio	sola	‘carry object with stick slung over one’s shoulder’
NNG	Lukep	-solo, -sol-a	‘carry on the shoulder with a stick’
NNG	Bing	sōl	‘carry on a pole’
MM	Tabar	sorak	‘carry’
MM	Madak	solok	‘carry on shoulders’
MM	Sursurunga	sol-solat	‘carry s.o. on one shoulder with legs straddling one shoulder’
SES	Bugotu	hoða	‘carry on pole’
SES	Gela	hola, holat-i	[VI, VT] ‘take, carry, bring, fetch’
SES	Lengo	ðola	‘carry’
SES	Lau	tole	‘carry’
SES	Arosi	tora	‘carry’
SES	Bauro	tora	‘carry’
PNCV		solo, zolo	‘carry over shoulder on a stick’ (Clark 2009)
NCV	Raga	(ɣa)holo	‘carry on the shoulder; stick used for this’ (ɣai ‘stick’)
NCV	Nokuku	sol	‘carry’
NCV	Uripiv	-solo	‘carry with a pole on the shoulder; feel burdened; be pregnant’
NCV	Port Sandwich	co-co-ini	‘carry balanced on the shoulder; carry on the end of a stick’
NCV	Lonwolwol	hol	‘carry a load in a basket on a stick over shoulder; be pregnant’
NCV	Southwest Bay	(aj)hol	‘carry on the end of a stick’
NCV	South Efate	sol, slati	‘carry, bring’
PSV		*a-curia	‘carry on pole or shoulder’ (Lynch 2001c)
SV	Anejom̃	a-helui-i	‘carry on shoulder’
SV	Ura	e-surye	‘carry on shoulder’
SV	Lenakel	a-sulie	‘carry on stick over shoulder’
SV	Kwamera	a-soria	‘carry by hanging on an elongated object (pole or finger)’
NCal	Nyelâyu	cʰõlĩ̄n	‘carry on shoulder’
Fij	Bauan	colat-a	‘carry on shoulder’
Fij	Wayan	ðolat-i-	‘carry a burden of work for s.o.’

6.2.9. Carrying on a long shoulder pole between two people⇫

Once a pig has been captured for slaughter, the favourite way to carry it in Melanesian communities is to tie its legs to a pole, such that the pole can be placed across the shoulders of two men, standing one behind the other, one at each end of the pole, the pig hanging between them. Terms for this mode of carrying occur in numerous Oceanic languages, but no extensive cognate sets are found. The three items below are nonetheless attested.

POc		*sirip-i	‘carry on a long shoulder pole between two people’
Adm	Loniu	siʔihi	‘carry suspended from shoulder or from pole’
Adm	Lou	sirip	‘carry a load on a pole, of one or two men’
SV	Sye	surie	‘tie pig by legs to a pole so it can be carried by two people’

POc		*tibʷola	‘carry on a long shoulder pole between two people; long shoulder pole (?)’
MM	Banoni	ci-cibora	‘tie up (a pig)’
MM	Babatana	sigolo	‘poles or handles for carrying two poles between two people’
NCV	South Efate	sipʷol	‘carry balanced on the shoulder’

PNGOc		*pak(u,o)	‘carry on a long shoulder pole between two people’
NNG	Mengen	pau-e	‘carry, lift (by two people)’
NNG	Mapos Buang	vaqu	‘tie, fasten; carry on a pole between two people’
PT	Gapapaiwa	-pawo	‘carry on a pole over the shoulder or between two people’

6.2.10. Carrying under the arm (and on the hip)⇫

The stem of POc *qa(p,pʷ)i(s), *qa(p,pʷ)is-i- ‘carry (a child) on the hip or under the arm’ appears to reflect PAn *qapiC/PMP *qapit ‘tongs, anything used to hold things together by pinching’ (ACD). However, aside from the extension in meaning, this derivation is not straightforward. An irregular reflex, POc *kapit ‘tongs’; *kapi(t), *kapit-i- ‘grasp (with tongs)’ is reconstructed in vol.1(148) (with hindsight this should more accurately be *ka(p,pʷ)it ‘tongs’; *ka(p,pʷ)i(t), *ka(p,pʷ)it-i-). Its irregularity lies in the presence of *k-_ for expected †*q-. The terms in the set below, however, do reflect *q-.

POc *qa(p,pʷ)i(s), *qa(p,pʷ)is-i- is nonetheless irregular in a different respect. Instead of stem-final *-t it reflects *-s. The one exception is the Teop term under ‘cf. also’ below, which reflects *ka(p,pʷ)it-i- ‘grasp (with tongs)’. The apparent replacement of *-t by *-s in the POc form may reflect palatalisation or borrowing at a very early stage, perhaps pre-POc. Alternatively, the resemblance between *ka(p,pʷ)it and *qa(p,pʷ)i(s) may be a matter of chance.

POc		qa(p,pʷ)i(s), qa(p,pʷ)is-i-	‘carry (a child) on the hip or under the arm’
NNG	Lukep	-wis	‘carry under the armpit’
PT	Tawala	aweh-i	‘carry (under arm)’
PT	Kilivila	-pʷesi-	‘carry under one’s arm’
SES	Gela	avi	‘carry (a child) under the arm’
SES	Gela	aveh-i	‘carry under the arm’
PMic		afi, afis-i-	‘carry on the hip or under the arm’ (Bender et al. 2003: afi, afi-Si)
Mic	Marshallese	ab(cāce)	‘carry tucked under arm’ (jaja ‘carry on the hip’)
Mic	Ponapean	apit	‘carry (s.t.) on one’s side or under one’s arm’
Mic	Mokilese	apit	‘carry (s.t.) under arm’
Mic	Mortlockese	afiy-af	‘carry under the arm’
Mic	Mortlockese	afit-i	‘carry (a child) on the hip’
Mic	Puluwatese	yafit-i-	‘to carry (a child) on one’s side’
Mic	Carolinian	afit-i	‘carry (a child or object) on one’s side’
Mic	Satawalese	æfiy-æf	‘carry on the hip (as a baby)’
Mic	Satawalese	æfit-i	‘carry (a child) in the arms on the side’
Mic	Woleaian	yafiy-efi	‘carry under one arm’
Mic	Woleaian	yafit-i-	‘carry s.t. under the arm’
Mic	Pulo Annian	yaθit-i	‘carry s.t. under the arm’
PPn		qafi, qafis-i	‘hold or carry under the arm’
Pn	Tongan	ʔefi-ʔefi	‘hold or carry under the arm’
Pn	East Futunan	ʔefi	‘carry under the arm (as a letter, another’s hand)’
Pn	East Uvean	ʔefi-ʔefi	‘carry under the arms’
Pn	Samoan	ʔafis-i	‘carry under the arm or on the hip’
Pn	Tokelauan	afih-i	‘carry or hold s.t. under arm; carry (child) on hip’

cf. also:

Teop

kapis-i

‘carry s.t. under your arm’

Despite its formal similarity to the set above, the cognate set below evidently reflects PMP *qabin ‘hold or carry under the arm’, POc *qapi(n), *qapin-i- ‘hold or carry under the arm’. However, contamination from POc *qapiŋa ‘armpit’, itself a nominalisation formed from POc *qapi(n) (§3.5.4), has evidently occurred in Yalu, Mota, Raga and Rennellese and in the Dangal and Buang forms under ‘cf. also’, which have ŋ for †n. All the forms under ‘cf. also’ reflect *k- rather than †*q-, apparently reflecting contamination from POc *kapi(t), *kapit-i- ‘grasp (with tongs)’ (vol.1:148).

PMP		*qabin	‘hold or carry under the arm’ (ACD)
POc		qapi(n), qapin-i-	‘hold or carry under the arm’
NNG	Yalu	-apıŋ	‘carry’
NNG	Mapos Buang	kpiŋ	‘carry, hold; carry astraddle the hip; hold under the arm, against the side’
MM	East Kara	(ɣə)kəpin-e	‘carry under the arm’
MM	Halia	api-api(c)	‘armpit; carry under the arm’
SES	Bugotu	avin-i	‘carry in the arms’
PNCV		*qavin-i-	‘carry under arm’ (Clark 2009: *qavi-ŋa)
NCV	Mota	avŋa-g	‘carry in the arm, on or under’ (viŋa-i ‘armpit’)
NCV	Raga	viŋa-i	‘carry in arms, embrace’ (malaviŋa- ‘armpit’)
NCV	Paamese	ahin-i	‘carry under arm’ (hiŋo- ‘armpit’)
NCV	Nguna	avin-i-	‘hold under arm’
PPn		*qafin-i	‘hold or carry under the arm’
Pn	Ifira-Mele	avin-i-a	‘stand close to, hold by side’
Pn	Rennellese	ʔaviŋ-i	‘carry under the arm or on the hip (as a child)’

cf. also:

NNG	Dangal	kafiŋ	‘carry’
NNG	Mapos Buang	kpiŋ	‘carry, hold; carry astraddle the hip; hold under the arm, against the side’
NNG	Mangga	kāpin	‘carry on hip’
NCV	Port Sandwich	xavin-i	‘carry under arm’

6.2.11. Carrying in both arms⇫

The common factor among the glosses of items reflecting POc *tabe is one of holding something with both arms.

POc		*tabe	‘carry in both arms’ (ACD: ‘hold tightly or firmly’)
PT	Tawala	tape(uni)	‘grab, catch hold of’
MM	Tolai	tabe	‘hold s.t. so that s.o. else will not take it’
MM	Babatana	töbe	‘hold, carry in the arms’
SES	Gela	tabe	‘take, hold’
SES	Arosi	abe	‘carry against chest, as firewood’
NCV	Nguna	tape-	‘take, carry; to bear (a child)’
PMic		tap(a,e), tap(a,e)-ki	‘lift up, carry, bear in one’s hands’
Mic	Kosraean	taptap	‘support, uphold, bear’
Mic	Kosraean	tæpʌ-k	‘support, uphold, or bear (s.t.)’
Mic	Kiribati	tapetape	‘engage in carrying, carry repeatedly’
Mic	Kiribati	tape-ka	‘lift or take (s.t.) up in the hands’
Mic	Kiribati	tape-ki-	‘lift s.t. up’
Mic	Marshallese	cepak	‘support, hold up’
Mic	Chuukese	sap	‘be holding up with open palm (of the hand)’
Mic	Chuukese	sapa, sassapa	‘hold (s.t.) up in the open palm’
Mic	Mortlockese	sapa	‘carry (s.t.) with both hands, support (s.t. or s.o.)’
Mic	Ponapean	sapɛ	‘carry (s.t.) in one’s arms’
Mic	Mokilese	capa-k	‘lift (s.t.) up from the rear end’
Fij	Bauan	tabe	‘hold or carry with the hands under’
Fij	Wayan	tabe	‘hold or carry a burden in the extended arms, esp. resting on both palms extended horizontally in front’

6.2.12. Carrying on a canoe⇫

Blust (ACD) reconstructs three formally similar PMP terms for ‘load a canoe’: *Rujan, *lujan, and *ujan. He offers no explanation for the existence of three similar forms, and nor can I. In vol.1(198) it was inferred that Oceanic items reflected POc *lujan, and *ujan. With a better understanding of Oceanic sound correspondences, *Rujan is inferred here, rather that *lujan, and the two POc forms are reconstructed as *Rujan, *Rujan-i- and *ucan, *ucan-i-, both ‘load (s.t.) onto a canoe, transport by canoe’. The two reconstructions differ in two consonant correspondences. The key lies in the medial correspondence. SE Solomonic, Fijian and Polynesian terms unambiguously reflect POc *-j- and are assigned to a single cognate set, whose initial consonant correspondence—PSES *l-, PCP *∅- —regularly reflects POc *R-.

FIXME: update/replace/deprecate reconstruction in vol1?

The second set reflects POc *ucan, *ucan-i-. Several irregularities in the medial consonant reflex are noted below, but none is assignable to *-j-.

Finally, the Micronesian set assembled by Bender et al. (2003) may reflect either POc *Rujan or *ucan. POc *R is lost in many Micronesian etyma, while *c and *j are merged except in Kosraean, which does not have a known reflex of either *Rujan or *ucan.

PMP		*Rujan	‘load a canoe; cargo’ (ACD)
POc		Rujan, Rujan-i-	‘load (s.t.) onto a canoe, transport by canoe’ (Geraghty 1983: PEOc *Ruja)
MM	Teop	ruhana	‘transport s.t., ship s.t.’
PSES		*luda	‘load (s.t.) onto a canoe, transport by canoe’
SES	Gela	luda	‘load a canoe or ship with cargo; embark passengers; cargo’
SES	Longgu	ludā	‘load s.t. on a truck, boat or canoe’
SES	Lau	luda	‘carry in a canoe, carry as cargo; load a canoe or ship’
SES	Lau	luda-i	‘carry cargo’
SES	Kwaio	luda	‘load in a canoe’
SES	Sa’a	luda	‘carry cargo, load a canoe’
SES	Sa’a	lude	‘carry cargo, load a canoe, be heavily laden’
SES	Sa’a	ludeŋ-i	‘carry as cargo; recruit men’
SES	Ulawa	luda	‘carry cargo, load a canoe, be heavily laden’
SES	’Are’are	ruta	‘charge, load, carry a load’
SES	Arosi	ruta	‘load a canoe, carry to canoe and stow’
PCP		*uja	‘be loaded onto a canoe, be transported by canoe’
PCP		*ujan-i-	‘load (s.t.) onto a canoe, transport by canoe’
Fij	Bauan	usa	‘be carried by boat’
Fij	Bauan	usan-a	‘carry (s.t.) by boat’
Fij	Wayan	usa	‘be shipped, carried (as cargo), by boat or other vehicle’
Fij	Wayan	usan-i-	‘carry s.t. as cargo or passengers’
Pn	Tongan	uta	‘carry or convey by boat or vehicle; goods so carried’
Pn	Niuean	uta	‘load (as a canoe or truck); a load’
Pn	Kapingamarangi	uda	‘transport (cargo)’
Pn	Rennellese	uta	‘put aboard a canoe or ship; bear nuts, as a coco- nut palm’
Pn	Rarotongan	uta	‘convey from one place to another; load up, as a canoe’
Pn	Māori	uta	‘put persons or goods on board a canoe’

PMP		*ujan	‘load a canoe; cargo’
POc		ucan, ucan-i-	‘load (s.t.) onto a canoe, transport by canoe’
Adm	Seimat	uxan-i	‘load s.t. into a boat’
Adm	Nauna	us	‘load s.t. into a boat’
Adm	Titan	usun-i	‘carry from one place to another, usually by canoe’ (-s- for †-l-)
NNG	Mutu	ɣūza	‘carry, transport’
NNG	Mangap	-ūzu	‘transport, convey by canoe, vehicle’
PT	Gapapaiwa	uan	‘load s.t. into a container or boat’
PT	Misima	usan	‘put inside; pick up people or things (in car); (be) overloaded’ (-s- for †-h-)
PT	Sinaugoro	ɣura(udi)	‘load a canoe’
PT	Motu	uda-uda	‘load pots into a trading canoe’
MM	Meramera	uda	‘carry’ (-d- for †-s-)
MM	Nehan	ute	‘load up a container with contents’ (-t- for †-h-)

PMic		u[s,S]a, u[s,S]an-i	‘load (s.t.) onto a canoe, transport by canoe’
Mic	Puluwatese	wɨtæn	‘(cargo) be loaded’
Mic	Carolinian	-wut	‘be loaded, have s.t. inside’
Mic	Satawalese	-yɨt	‘load s.t.’
Mic	Satawalese	-yɨta	‘load it’
Mic	Ponapean	itan	‘carry (s.t.) in a vehicle’
Mic	Mokilese	itan	‘transport (s.t.)’

6.2.13. Dragging⇫

Dragging is a form of accompanied caused movement, in that the agent follows the same path as the theme (the thing dragged). In this respect it differs from pulling, where, as with putting, some part of the agent remains in contact with the theme but the agent doesn’t follow the theme’s path (§6.6.3). Only one dragging verb is reconstructable, and only to PROc.

PROc		*(q)ara	‘haul, drag’
Fij	Wayan	ara	‘be hauled, dragged, pulled’
Fij	Wayan	ara-ki-	‘drag, haul, pull s.t. along; trail s.t. behind one’
Fij	Bauan	yara	‘be hauled, be dragged’
Fij	Bauan	yara-ka	‘haul, drag’
PMic		are, arek-i	‘haul, pull, tow’
Mic	Kiribati	ā-i	‘tow’
Mic	Kiribati	aeae-	‘tow, drag (s.t.)’
Mic	Kiribati	aek-	‘take (a passenger) ashore in a canoe; take (food) from the fire’
Mic	Kosraean	ælʌk	‘lift, pull, scoop (s.t.)’
Mic	Marshallese	yar	‘haul a canoe or vessel up on shore’
Mic	Marshallese	yarek	‘haul (a boat) up on shore’

6.3. Unaccompanied caused movement⇫

Verbs of unaccompanied caused movement form two categories, based on whether or not some part or tool of the agent remains in contact with the theme until the theme reaches its goal. If it does, then we have a putting verb like put, insert or immerse (§6.6.3.1). If it doesn’t, the verb is a verb of sending, like send, drop, throw or pour (§6.6.3.2).

6.3.1. Putting⇫

‘Putting’ is used here in a specialised sense. The English verb put is ubiquitous. It occurs in dozens of idiomatic phrases like put out the light, put on clothes, put to death, put down (‘humiliate’), put out (‘annoy’), put off (‘postpone’). Its core meaning, however, has to do with moving something to a named location, as in put on the table, put into the bag and so on.

Some verbs of putting profile the resulting posture of the theme, e.g. He sat the child on the chair (result: the child was sitting on the chair). Others profile the path to the resulting location, e.g. He inserted the key into the keyhole (result: the key was in the keyhole).

Verbs of putting in the world’s languages vary as to whether they have a simple verb of putting like English put, whether they have verbs that specify the target posture of the theme (i.e. the thing that is put), like German setzen ‘sit, seat’, stellen ‘stand’ and legen ‘lay’, or whether they have a number of putting verbs that, like insert, specify the path of putting, e.g. ‘put into’, ‘put through’, ‘put on top of’, ‘put down’, ‘put up (e.g. onto a shelf)’, ‘put out (e.g. to dry)’, ‘put on a pile’ and ‘put in a line’ (Narasimhan et al. 2012).

These three possibilities are not mutually exclusive. Oceanic languages tend to be of the third type, and the list of path-of-putting categories in the previous sentence is based on Oceanic tendencies. At the same time, most Oceanic languages have a simple ‘put’ verb (§6.6.3.1.1), and many have verbs for ‘stand (s.t.) (on s.t)’ and ‘lay (s.t.) (on s.t)’ (§6.3.1.2). Few, however, have a verb for ‘sit (s.t.) (on s.t)’, presumably because, as noted in §2.1, sitting is typically not encoded by a dedicated verb.

Oceanic dictionaries sometimes appear to be incomplete when it comes to putting verbs. One which does contain a comprehensive collection is Pawley & Sayaba’s (2003) dictionary of Wayan Fijian, which includes the putting verbs tabulated below (two leftmost columns) together with the bases from which they are derived (third and fourth columns). All have the theme as their object except drosomi-/drosoti- ‘put s.t. inside’, which takes the location as its object.

Some interesting patterns emerge from the tabulation. There are far more path-profiling than posture-profiling verbs. All the verbs listed except the last are derived from a base listed in the dictionary, and a majority are derived from undergoer-subject verbs (U-verbs; vol.1:23). Four are derived from a noun denoting the path’s destination. Each of the derived verbs is simply a transitive form of its base, with the exception of vaka-dureni ‘stand s.t. up, make s.t. stand’, formed with the causative prefix vaka-. The latter is a posture-profiling putting verb, and supports the view that Wayan Fijian is indeed, in the preliminary typology of Narasimhan et al. (2012), a language that elaborates path-profiling rather than posture-profiling verbs of putting.

None of these observations is inconsistent with the data available for other Oceanic languages (other than perhaps Polynesian). Certainly, path verbs play a major role everywhere. Whether they are commonly derived from U-verbs it is impossible to say, as comprehensive dictionaries are not available for other languages that may share this derivation. Another extensive set of verbs of putting is from Mangap-Mbula, which, like all Oceanic languages on or near the New Guinea mainland, does not have U-verbs. and lacks transitive derivational morphology other than the causative prefix pV-, which occurs on three posture verbs (the language is exceptional in having a ‘sit’ verb here) and two path verbs. Mangap-Mbula putting verbs are shown in Table 17.

Table 16 Verbs of putting in Wayan Fijian
Simple	tauni-	put or place s.t., lay s.t. down, deposit s.t.	tau	be located, situated, positioned, placed; be in, at or on a place
Simple	nāki-	put or place s.t.	naki	be situated, put, placed
Posture-profiling	tavani-	put s.t. up lengthwise	tava	(of a long object) be placed lengthwise in an elevated position, as on a shelf or stack.
Posture-profiling	ruva-, ru-vani-	lay s.t. in place, fix or place s.t. in or on a place (object the thing fixed).	ruva	(of a flat object) be laid, put in place, fixed in position
Posture-profiling	vaka-dureni	stand s.t. up, make s.t. stand	dure	stand, stand up
Path-profiling	ðilini-	put or rest a thing on s.t.	ðili	be on top, rest on s.t.
Path-profiling	ðuruni-, ðuruti-	bag s.t., secure s.t. in a bag or basket.	ðuru	be bagged, put into a bag or basket with top fastened
Path-profiling	koroni-	gather s.t. in a heap, accumulate or pile up s.t.	koro	be gathered in a heap or mound; be heaped, piled, mounded
Path-profiling	livati-	take s.t. off, put down a burden, leave a position	liva	be taken off, removed
Path-profiling	duŋui-, duŋuni-	heap, pile, stack things	duŋu	be stacked in a heap; be heaped, piled, stacked
Path-profiling	tavo-, tavoni-	put s.t. in a container, bag s.t., etc.	tavo	be put in a container; thus, bagged, put in a sack, boxed, crated, put in a basket, etc.
Path-profiling	uðuni-	put s.t. in a container	uðu	go into a house or other living place, disappear in-side.
Path-profiling	drosomi-, drosoti-	put s.t. inside, insert into, enclose in, sheathe in, penetrate	droso	go inside, enter
Path-profiling	bolani-	put s.t. in a case, record s.t., commit s.t. to memory	bola (N)	basket woven from coconut leaves, large enough to carry several long yams or a small pig
Path-profiling	taŋani-	bag or pocket s.t., put s.t. in a bag or pocket	taŋa (N) any deep woven or netting receptacle with open top: bag, sack, pocket
Path-profiling	tāvata	be put on a platform, bed, etc; be shelved, tabled	tāvata (N)	any constructed surface acting as a platform; thus shelf, table
Path-profiling	atuni-	line (things) up	atu (N)	group of things standing in a row or line
Path-profiling	abani-	line (things) up, place (things) in a line or row	—

Table 17 Verbs of putting in Mangap-Mbula
Simple	-kam	do, give, take, get, put, cause	—
Simple	-ur	put, place	—
Posture-profiling	-pa-ᵐbutul	put into a sitting position; -	-ᵐbutul	sit
Posture-profiling	-pe-gēne	put to bed/ put into lying position;	-kēne	sleep, lie down, be in horizontal position
Posture-profiling	-pa-meⁿder	put into a standing position;	-meⁿder	stand
Posture-profiling	-suᵑgun	put into a standing position	-su	go down
			-ᵑgun	plant s.t. upright in the ground
Path-profiling	-dāba	put into, fill with	—
Path-profiling	-zēbe	put into	—
Path-profiling	-kuru	put through a hole, thread	—
Path-profiling	-ⁿdou	collect/ put together/ heap/ gather/ accumulate	—
Path-profiling	-pi-zil	put, slip in through a small hole/ hide inside	-zil	go through a hole
Path-profiling	-pa-māla	put out to be seen/ show/ advertise	māla	look at, watch, observe
Path-profiling	-salakāla	put on top of	-salakāla	be on top of
Path-profiling	–san	put out to dry	—
Path-profiling	-pa-rāza	put out to dry	rāza	dry out

Simple verbs of putting often occur in serial verb constructions. In terms of the division into manner (locomotion), path (geographic direction) and deictic direction verbs put forward in §6.3 on the basis of occurrence in these constructions, simple verbs of putting occupy the manner slot, i.e. they are verbs of (caused) locomotion. This follows from their encoding the act of putting and from their lack of directional meaning, as well as from their place in serial verb constructions.

- Mangap (NNG): (Bugenhagen 1995:166)
  ‘They put it on the ground.’
  Zin ti-ur i-su tōno
  
  they 3PL-put 3SG-go.down ground
  
  … MANNER PATH …

- Kusaghe (MM): (Davis 2003:166)
  ‘Put it in the basket.’
  Lohe lae-ni-a pa beha
  
  put go-TR-3SG PREP basket
  
  MANNER DEIXIS … …

- Ambae (NCV): (Hyslop 2001:302)
  ‘We put the pudding down.’
  …da-mo tau na loko mo hivo
  
  1INC-RLS put ACC pudding RLS go.down
  
  … MANNER … … … PATH

Posture-profiling verbs of putting occur so rarely in sentence examples that one cannot classify them in this way. Path-profiling verbs of putting, as their label suggests, are verbs of geographic direction.

6.3.1.1. Simple putting

The most widely reflected simple ‘put’ verb is POc *taRu(q) ‘put down, lay down’. The Mapos Buang and Meramera forms are shown under ‘cf. also’ as the expected forms are Mapos Buang †taɣu, Meramera †talu.

POc		*taRu(q)	‘put down, lay down’ (Geraghty 1990: PEOc)
Adm	Mussau	tau	‘give’
Adm	Wuvulu	ʔau	‘put’
Adm	Titan	to, taw-i	‘give, send, bring, put; make pregnant’
Adm	Lou	tu	‘give’
NNG	Mengen	talu-	‘hide’
PT	Motu	taru-	‘cover, as with a sheet; wrap oneself in’
PT	Motu	taru(adiho)	‘put down, of a loaded netbag’
MM	Vitu	taruɣ-i-	‘put, place’
MM	Nakanai	talu	‘put down on the ground, as a bundle carried on the head’
SES	Bugotu	talu	‘put, place, appoint’
SES	Gela	talu	‘put, place, set’
SES	Tolo	talu-	‘put, place’
SES	To’aba’ita	alu-	‘put, put down, place’
SES	Lau	alu	‘place, put, lay down’
SES	Kwaio	alu, alu-a	‘put, keep’
SES	Sa’a	ʔalu	‘put, place’
NCV	Mota	tau	‘set in place so as to catch or intercept’
NCV	Tamambo	tau	‘put’
NCV	Ambae	tau	‘put’
Fij	Bauan	tau	‘(of burden) be put down’
Fij	Bauan	tau-ði	‘put down (of burden), lay one thing on another’
Fij	Wayan	tau	‘be located, situated, positioned, placed; be in, at or on a place’
Fij	Wayan	tau-ni-	‘put or place s.t., lay s.t. down, deposit s.t.’

cf. also:

NNG	Mapos Buang	taɢu	‘put; hit against, knock over, fall on, smash’
MM	Meramera	tau	‘put’

Other apparently simple ‘put’ verbs have only two or three known reflexes. However, whether POc had more than one or two simple ‘put’ verbs is questionable, and it is possible that both *taRu and the two verbs below each had a more specific meaning.

POc		*aso	‘put’
NNG	Sio	o	‘put, place’
NNG	Mangseng	as[o]	‘put’
SES	Lau	ato	‘put, place’

PEOc		*naki-	‘put’
SES	Longgu	naʔi-	‘put, leave’
Fij	Wayan	naki	‘be put, be placed’
Fij	Wayan	nāki-	‘put or place s.t.’

Two other apparently simple ‘put’ etyma have reflexes that are not widespread enough to support POc reconstruction.

PNGOc		*ku(rR)a	‘put’
NNG	Mutu	ɣur	‘put, appoint, give, take’
NNG	Mangap	-ur	‘put, place’
PT	Sinaugoro	ɣura	‘put into’

PSOc		*liŋi	‘put, leave’ (Clark 2009)
NCV	Nokuku	lin	‘put, leave’
NCV	Kiai	lini-	‘deliver, bring, leave’
NCV	Uripiv	-liŋ-i	‘put, put on; select out; allow, let; leave’
NCV	Big Nambas	ln	‘leave’
NCV	Port Sandwich	riŋ-i	‘put’
NCV	Lonwolwol	liŋ-i	‘put, place; let, allow; let go, let down; leave, miss; leave out; to forget; bear (a child), beget; leave alone’
NCV	Paamese	liŋi	‘put; leave behind’
NCV	Lewo	liŋ-ani	‘let go, put, leave, place’
NCV	Araki	liŋ-i-	‘carry, take (s.o.) on a vehicle’
NCV	Naman	ləŋ	‘put’
PSV		*a-liŋi-i	‘put, leave’
SV	Anejom̃	i-cñi-i	‘put, leave (plural subject)’

6.3.1.2. Posture-profiling verbs of putting

Assuming that lexical sources are reliable, Oceanic languages have far more underived path- than posture-profiling verbs of putting. Where posture-profiling ‘put’ verbs occur, they tend to be transparently derived from posture verbs, as the Wayan and Mangap examples above show. Other instances include those in Table 18.

Table 18 Posture-profiling verbs of putting in Oceanic languages
Adm:	Titan	taRul-ani	stand s.t. up, erect; set up, establish	taRul	stand, stand up
NNG:	Bariai	pa-gun	build; put.upright	—
	Takia	-guni	plant (in ground) vertically	guni	stand up (e.g. posts in the ground)
	Mangseng	pa-ŋou	put; make lie	ŋou	lie down
NNG:	Poeng	pa-keno-e	lay s.o./s.t. down	keno	(of inanimates) lie horizontally
		pa-maisi-a	stand up (something)	meisi	stand
	Hote	-ɛk	put in a horizontal position	-ɛk	sleep
PT:	Misima	pa-kenu	lay s.t. on its side	kenu	lie down, sleep
	Gumawana	va-tao	stand s.t./s.o. up	taoya	stand (up)
	Dawawa	wai-midir-i	stand s.t. up	midi	stand
	Tawala	lu-towolo	stand s.t. up	towolo	stand, wake
MM:	Nakanai	hi-mavuta	lay (s.o. down to sleep)	mavuta	lie down, sleep
	Madak	vaxa-mdi	lay s.t. down	mdi	recline, lay down
	Patpatar	ha-tu-tur	cause to stand; raise up	tur	stand
SES:	Gela	enov-agi,	lay s.t. down	eno	lie down
		koliv-agi	lay s.t. down	koli	lie down

No underived POc posture-profiling verbs of putting can be reconstructed, but it is likely that there were verbs consisting of cardinal posture stems (§2.1–2.3) prefixed by causative *pa-. The data in §6.2.4.3 also suggest that transitive ‘hang’ was formed from an intransitive ‘hang’ verb.

6.3.1.3. Path-profiling verbs of putting

Despite the high numbers of verbs of putting in Oceanic languages that specify a path (§6.6.3.1), only one POc path-profiling verb of putting can be reconstructed, namely *soŋo ‘put into, insert’.

At first blush, the absence of POc reconstructions seems to contradict the statement that Oceanic languages have numerous path-profiling ‘put’ verbs. However, it is probably the very plethora of such verbs in Oceanic languages that leads ironically to the absence of cognate sets. The large number of ‘put’ verbs and their propensity to be derived apparently means that the lexical replacement rate is high in this semantic domain, so that cognates often vanish.

POc		*soŋo	‘put into, insert’
Adm	Lou	soŋ(pek)	‘put into (e.g. a bag)’
NNG	Patep	zoŋ	‘put into, force into; aim at’
MM	Madak	saŋa	‘put, place’ (for †soŋo)
MM	Patpatar	saŋ	‘put into’ (for †soŋo)
MM	Konomala	saŋa-i	‘give’ (for †soŋo)
MM	Nehan	oŋo	‘put into’ (for †hoŋo)
MM	Teop	vā-hogo	‘put s.t. (into s.t.)’ (hogo ‘be crowded, overfilled, packed with things or people’)
MM	Maringe	hono	‘load, put inside a bag, basket or case’
SES	Lau	toŋ-i	‘pack, put in a box, etc’
NCV	Ambae	hoŋo-ni	‘put inside’

6.3.1.4. Putting in reverse: removing

‘Removing’ in the title of this section refers to acts of unaccompanied caused movement that are opposite in direction to ‘putting’ (e.g. He took the knife from the table). In §6.6.1.1 it is argued that POc probably had no simple verb of removing that denoted the reversal of putting. Instead it had a simple verb of caused movement that covered both the reversal of putting and accompanied caused motion, i.e. ‘bring’ and ‘take’.

Across the world’s languages verbs of putting often profile both the path along which the object is moved and/or its resulting posture, but verbs of removing tend to profile only the source from which it is moved, not its former posture. Languages in which a verb of removing also encodes the object’s posture are rare (Narasimhan et al. 2012:10), but one of them occurs in the heart of Melanesia. In the Papuan language Yélî Dnye, located on the periphery of WOc, there are verbs meaning both ‘put into’ and ‘take out of’, ‘put onto’ and ‘take off of’, ‘stand/lay/hang at a location’ and ‘unstand/unlay/unhang from a location’ (Levinson & Brown 2012). Oceanic languages, however, do not resemble Yélî Dnye in this respect. This is no surprise, as even their verbs of putting tend not to profile posture (§6.6.3.1.2). Instead, we typically find a simple verb of removing (‘take out of/from/ away’) along with numerous specialised removal verbs, depending on what is being removed and from where. Many of these verbs concern common domestic processes: removing bark from wood, removing the husk from a coconut, removing fruit from a branch, removing flesh from a coconut or a mollusc, removing the stones from an earth oven, or removing the serrated edge from pandanus leaves. No attempt is made to reconstruct verbs for these meanings here, because they lie outside the semantic domain with which this chapter is concerned and are reconstructed elsewhere (vol.1:165–168, 277–279). However, Blust (ACD) has reconstructed a simple verb of removal, POc *unus, *unus-i- ‘withdraw, pull out, extract’ (I take ‘pull’ in the glosses below simply to be a consequence of caused movement towards the agent).

PAn		*SuLus	‘withdraw, pull out, extract’ (ACD)
PMP		*hunus	‘withdraw, pull out, extract’ (ACD)
POc		unus, unus-i-	‘withdraw, pull out, extract’ (ACD)
MM	Roviana	unusu	‘pull out, as a tooth, a nail, a post’
SES	Arosi	unu	‘take out of’
SES	Arosi	unus-i	‘take body out of shirt, i.e. take off shirt; fall out of’
Mic	Chuukese	wɨnɨ	‘drawn out, extracted’
Mic	Woleaian	ʉrʉru	‘pull, draw, pull something from a group’
Pn	Tongan	unu	‘pull out, draw out’
Pn	Tongan	uhuh-i	‘pull out, draw out, e.g. a sword from its sheath, or a tooth’
Pn	Rennellese	unu	‘take off, remove, pull out, peel’
Pn	Anutan	unu-i	‘pull out’
Pn	Nukuoro	unus-i	‘pull out (e.g. a sword from its scabbard)’
Pn	Hawaiian	unuh-i	‘take out, withdraw (as money from a bank, or a drawer from a desk); unsheath; take off (as a ring)’
Pn	Māori	unu	‘pull off, put off, doff; draw out, pull out, withdraw’

6.3.2. Sending, dropping, throwing and pouring⇫

‘Send’ verbs denote caused movement where the agent neither accompanies the theme in its movement nor, unlike ‘put’ verbs, remains in contact with the theme until it reaches its goal. In this semantic domain one might expect to be able to reconstruct a simple ‘send’ verb and one or two more verbs whose meaning satisfies these criteria, namely ‘drop’ (‘let fall, cause to fall’) and ‘throw’.

Curiously, neither a simple ‘send’ verb nor one meaning ‘drop’ is reconstructable on the basis of available data. There is no immediately obvious reason for this, as Oceanic languages often have verbs with each of these meanings, but they do not form widespread cognate sets. Many ‘send’ verbs appear to reflect the POc causative prefix *pa[ka]- but the roots to which it is attached do not occur independently in the data. Two that do are Teop (MM) va-nao ‘send’ (nao ‘go’) and Wayan Fijian vā-kauti- ‘send s.t., have s.t. carried or taken’ (kau ‘be carried, taken’). Takia uses a periphrastic causative -gane -ao ‘send’ (-gane ‘do’, -ao ‘go’). A number of ‘drop’ verbs similarly reflect the causative prefix: Poeng (NNG) pa-pu-e ‘drop, cause to fall’ (pu-pu ‘fall’), Iduna (PT) -ki-ve-beʔu- ‘drop from hand, make something fall’ (beʔu ‘fall’, ki- ‘do with hand’, ve- CAUSATIVE), Misima (PT) pa-bun ‘drop’ (bun ‘fall’), Wayan Fijian vaka-lutu-ni- ‘drop (s.t.)’ (lutu [VI] ‘drop, fall’). It thus seems possible that POc used causatives for ‘send’ and ‘go’, and that the plethora of ‘send’ forms in Oceanic languages reflects a re-purposing of other verbs.

The situation with ‘throw’ and ‘pour’, however, is quite different, perhaps because ‘throw’ also profiles the kind of force applied to the theme and ‘pour’ the action applied to the vessel containing the liquid.

6.3.2.1. Throwing

It seems fairly clear that throwing stones was a primary sense of POc *piri(ŋ).

PCEMP		*biriŋ	‘stone, throw a stone at’ (ACD)
POc		*piri(ŋ)	‘stone, throw a stone at’ (ACD)
NNG	Mangap	-piri	‘throw, cast away, toss, throw oneself into’
MM	Tolai	bir	[V] ‘stone; throw or fling a stone; chase away with stones’
SES	Gela	piri	‘throw with a twist, with finger and thumb’
SES	Bugotu	piri	[V] ‘stone with stones’
NCV	Mota	vi-vir	[V] ‘throw, giving a twist, twirling motion, as to a stone’
NCV	Vurës	vi-virr	‘throw’
NCV	Nume	vi-vir	‘throw’
Fij	Bauan	viri	[VI] ‘throw at , pelt’
Fij	Bauan	viri-ka	[VT] ‘pelt s.t.’
Fij	Bauan	viri-tak-a	[VT] ‘throw s.t.’
Fij	Wayan	viri	[VI] ‘throw or hurl overarm’
Fij	Wayan	viri-ki-	[VT] ‘pelt s.t., hit s.t. with a thing thrown’

6.3.2.2. Pouring

Several verbs of pouring can be reconstructed. The most widely reflected is POc *liŋi(s), *liŋis-i- ‘pour out, spill (liquid)’. POc *puRi ‘pour water on’ took the goal as its object. PWOc *jiwaR, *jiwaR-i- ‘pour out (liquid)’ is reflected only in WOc languages, and seems to have replaced *liŋi(s)/liŋis-i-. It is not clear whether its object was the (liquid) theme or the goal of pouring. As noted below, it may have an earlier shared history with PCP *suqi ‘pour water on/into’, which took the goal as its object.

PMP		*iliŋ	‘pour’ (ACD)
PCEMP		*liŋi	‘pour’ (ACD)
POc		liŋi(s), liŋis-i-	‘pour out, spill (liquid)’
NNG	Tuam	-liŋ	‘pour’
NNG	Mangap	-liŋ	‘pour out, shed’
NNG	Sio	liŋi	‘pour out, spill; throw away or discard’
NNG	Mengen	liŋ-a	‘pour out’
PT	Sudest	liᵑgi	‘pour’
PT	Misima	ligí-n	‘pour out (onto the ground), spill’
SES	Sa’a	liŋi	[VI] ‘pour’
SES	Sa’a	liŋis-i-	[VT] ‘pour’
SES	Arosi	riŋi	[VI] ‘pour, incline a vessel’
SES	Arosi	riŋis-i-	[VT] ‘pour, incline a vessel’
SES	Owa	riŋi-	‘pour s.t.’
PNCV		*liŋi	‘pour’
NCV	Mota	liŋ	‘pour gently’
NCV	Raga	liŋ	‘pour out’
NCV	Sakao	lüg	‘pour’
NCV	Nguna	liŋi	[VI] ‘spill over’
NCV	Nguna	liŋis-i	[VT] ‘spill’
NCV	South Efate	liŋ	[VI, VT] ‘pour out’
Fij	Wayan	liŋi	[VI] ‘be poured, pour’
Fij	Wayan	liŋi-	[VT] ‘pour (liquid)’
Pn	Tongan	li-liŋi	[VT] ‘pour s.t. out’
Pn	Tongan	liŋi-ʔi	[VT] ‘pour s.t. out’
Pn	Samoan	liŋi, li-liŋi	[VT] ‘pour s.t.’
Pn	Samoan	ma-liŋi	[VI] ‘(rain, tears) pour, run’

POc		*puRi	‘pour water on’ (Geraghty 1990: PEOc *vuRi ‘pour water on, rinse’)
NNG	Kairiru	-pul	‘pour’
NNG	Mangap	-puri	‘rinse, clean something small’
PT	Motu	huri	‘wash, scrub’
SES	Gela	vuli	‘pour water’
SES	Gela	vu-vuli	‘pour, sprinkle’
PNCV		*vui	‘pour water on’ (Clark 2009)
NCV	Mota	vu-vui	‘pour water upon, cool with water’
NCV	Nguna	vue	‘pour water on, water, sprinkle’
PSV		*a-vwi(i)	‘to water, pour water on’
SV	Sye	avwi	‘to water, pour water on’
SV	Lenakel	vi	‘to water, pour water on’
SV	Anejom̃	ahwi-i	‘to water, pour water on’

The relationship, if any, between the two reconstructions below, PNGOc *jiwaR, *jiwaR-i- ‘pour out (liquid)’ and PCP *suqi ‘pour water on/into’, is somewhat puzzling. Each is based on a regionally restricted cognate set, and the formal similarities between them suggest that they may share an earlier history. Indeed, the Fjian (Bauan and Wayan) terms under PCP *suqi could also reflect *jiwaR-i-, if one assumes that the sequence *-iwa- is readily reduced to *-ua- (as in Takia and Sio), then *-u- (as in Kela and Numbami). This would give PCP *juRi, hence, e.g. Wayan sui. Concomitantly, *suqi would then be demoted to PPn status. Either way, though, PPn *-q- cannot be reconciled with PWOc *-R- without resorting to ad hoc speculation.

PWOc		jiwaR, jiwaR-i-	‘pour out (liquid)’
NNG	Takia	-suari	[VT] ‘pour out, empty’
NNG	Sio	zuala	‘rinse out with water; slosh back and forth’
NNG	Kela	zui	‘pour out’
NNG	Numbami	-zu-zula	‘pour, flow’
PT	Gumawana	siwo-i	‘pour’
PT	Dobu	siwa	‘pour (water)’
PT	Dobu	siwala(ga)	‘pour more water in’
PT	Dobu	siwal(olo)	‘pour water on’
PT	Iamalele	iwa(ga)	‘pour (water)’
PT	Iduna	-ɣiwa-	‘pour, fill’
PT	Are	sewa-i-	‘pour’
PT	Tawala	hiwo(ga)	‘pour out, tip out, unload’
PT	Misima	hol	‘pour’
MM	Babatana	jili-ni	‘pour out, spill’

PCP		*suqi	‘pour water on/into’
Fij	Bauan	sū	‘pour water on, irrigate, quench (fire)’
Fij	Bauan	sūy-a	[VT] ‘pour water on, irrigate’
Fij	Bauan	sui-bokoð-a	[VT] ‘quench (fire +)’ (boko ‘extinguished, erased’)
Fij	Wayan	sui	‘be watered, have water poured or sprayed on’
Fij	Wayan	sui-	‘water s.t., quench a fire’
PPn		*suqi	‘dilute, mix with liquid’ (POLLEX)
Pn	Tongan	huʔi	‘thin down, dilute’
Pn	Samoan	sui	‘add water to, dilute’
Pn	East Futunan	suʔi	‘dilute’
Pn	Tuvalu	hui	‘mix with water’
Pn	East Futunan	sui	‘dilute dry or thick substances with water’
Pn	Nukuoro	s-sui	‘wet’
Pn	Rennellese	suʔi	‘mix’
Pn	Tokelauan	hui	‘water down, thin down (soup +)’

Notes⇫

This is a simplified version of Ameka & Levinson’s (2007) categorisation, which see for more detail.↩︎
Abbreviations in these examples are as follows: ASP aspect marker, CLF classifier, INDEF indefinite, IRR irrealis, REAL realis; ~ connects reduplicates.↩︎
Oceanic existential constructions tend to be similar to locative constructions, with the one difference that their subject is indefinite (e.g. ‘A cat is sitting under me’ = ’There is a cat sitting under me). For present purposes, existential constructions are treated as a subtype of locative construction, as in the first Manam example below.↩︎
A number of sources gloss a verb as ‘be’, but ‘be (at)’ is used here to indicate locative usage,↩︎
See also the discussion of terms for ‘cold’ in §4.8.1.↩︎
Blust labels this form PCEMP, implying that there are also CMP reflexes, but none are listed.↩︎
More recent work (Aske 1989, Huang & Tanangkingsing 2005, Croft et al. 2010) emphasises that many languages have constructions of both types, depending on the semantics of the event.↩︎
There has been debate (Slobin 2004, Zlatev & Yangklang 2004, Talmy 2008, Ameka & Essegbey 2013) as to whether a serial verb construction contains a main verb. One criterion offered by Talmy is that the verb belonging to the set with the largest number of members is the main verb. In Oceanic languages this would be the manner verb if there is one, otherwise the path verb.↩︎
Most descriptions are silent as to how such constructions should be analysed.↩︎
The expected reflex of *laka (§6.3.1.3) in both languages is †laʔa.↩︎
John Lynch points out that Clark (2004) reconstructs PNCV *lakau~*lakawa ‘cross over’. I suggest this may be a lexicalised serial verb construction made up of *laka and POc *ua ‘go towards addressee’ (vol.2:284).↩︎
Budai Rukai ma-liu-lius (VI) ‘turn round’, Tamalakaw Puyuma mu-liu-lius (VT) ‘turn (s.t.) round’, Amis liyuc (VT) ‘turn (s.t.) round’.↩︎
Reconstructed on the basis of Formosan data: Budai Rukai ma-liu-lius (VI) ‘turn round’ (Ferrell 1969, cited ACD).↩︎
We thank John Lynch for drawing our attention to the Lenakel reflex.↩︎
Semantic frames, frame elements and profiling are terms from Frame Semantics, a theory of meaning deriving from the work of Charles J. Fillmore (see especially Fillmore 1982, 1985). Croft & Cruse (2004:8–22 and passim.) provide an overview. A short set of definitions is given at https://framenet.icsi.berkeley.edu/fndrupal/about .↩︎
This distinction is adopted from Levin’s (1993) analysis of English verbs. Note, though, that by no means all distinctions that work for English verbs work for Oceanic verbs. English take and bring contrast in deixis (‘away from deictic centre’ vs ‘towards deictic centre’), but deixis is not usually profiled by caused movement verbs in Oceanic languages, as the discussion in §6.6.1.1 indicates.↩︎
Narasimhan et al. (2012:10) note that Tzeltal, a Mayan language of Mexico, has a semantically general ‘put’ verb that encompasses both ‘put’ (inanimate goal) and ‘give’ (animate recipient) meanings.↩︎
I am indebted to Andrew Pawley for a copy of Pawley (1997), to Deborah Hill for allowing me to read a pre-publication draft of Hill (2016), and to people who provided Hill with lists of carrying verbs: Wolfgang Sperlich for Nakanai (drawing on Chowning & Goodenough 2014), Ralph Lawton for Kilivila and Alexandre François for Hiw.↩︎

Zin	ti-ur	i-su	tōno
they	3PL-put	3SG-go.down	ground
…	MANNER	PATH	…

Lohe	lae-ni-a	pa	beha
put	go-TR-3SG	PREP	basket
MANNER	DEIXIS	…	…

…da-mo	tau	na	loko	mo	hivo
1INC-RLS	put	ACC	pudding	RLS	go.down
…	MANNER	…	…	…	PATH