*R   

*Rabak tear, in tatters

Note:   Kapampangan normally shows *R > y, but has g from *R in some other forms (e.g. *Rapus ‘bind’ > gapús ‘binder, fetters, chain’, ga-gapus ‘be bound, tied’).

*Rabat driftwood; flotsam

*Ra(m)bej wind around

Note:   With root *-bej ‘wind around repeatedly’.

*Rab(e)qun cover with earth

Note:   With root *-bun or *-bequn ‘heap, cover with earth; collect, assemble’.

*Rabiqi late afternoon, evening; evening meal

Note:   Also Puyuma ɭabi ‘dinner’, Ida'an Begak gabpi ‘night’. The most serious problem with this comparison is that some languages point unambiguously to *Rabiqi, while others just as clearly indicate *Rabi. Both variants appear in Amis, where Fey (1986) gives lafi ‘dinner, the evening meal’, but lafii ‘midnight’, and several languages that should have a non-zero reflex of *q instead show no intervocalic consonant, as with Atayal gbi-an ‘late afternoon, evening’, Agutaynen labi ‘night’, Maranao gabi ‘meal at evening’, Tongan efi-afi and Rennellese ahi-ahi ‘afternoon, early evening’. One option would be to reconstruct doublets *Rabi, *Rabiqi, but for now I prefer to posit only PAn *Rabiqi, and assume that *q tended to be lost between identical vowels in words of three or more syllables.

*RabuC uproot, pull out by the roots

Note:   Also Bontok gábot ‘pull out by the roots, as weeds, grass or hair’, Kankanaey gábut ‘eradicate, uproot’, Ifugaw gábut ‘to weed, pull out weeds and grass’, Tagalog lábot ‘drawing out of something rooted in soft matter, pulling away of teats of cattle (or such animals) from the suckling mouth of the young’, Manggarai rambut ‘pull up grass, weed a garden.’ With root *-buC.

*Rabuk decaying wood; dust, powder

Note:   Also Manggarai rawok ‘dusty’. With root *-buk₁ ‘decay, crumble; powder’.

*Rabun drizzling rain, mist; fog

12612

POC     *ma-Rapun misty, hazy, dim

*RaCus hundred

Note:   Also Isneg ma-gasút ‘one hundred’, Casiguran Dumagat dátos ‘unit of a hundred, a hundred’, Agta (Dupaningan) ma-gátu ‘one hundred’, Ibaloy dasos ‘hundred’, san-dasos ‘one hundred’, d<in>asos ‘by hundreds, multiples of one hundred’, Pangasinan lasós ‘hundred’, Ida'an Begak ratu ‘hundred’, mə-ratu ‘one hundred’, Kadazan Dusun atus ‘hundred’, iŋ-gatus ‘a hundred times’, Bintulu ratus (< Malay), Melanau (Mukah) ratuih (< Malay), Simalur ratui, Sasak atus ‘hundred’.

The PAn status of this reconstruction depends crucially on Hoanya matala-gasut, recorded by Naoyoshi Ogawa from a 75-year old male in 1909 (Tsuchida 1982:40). However, this form is problem-free, since *R > g is independently attested in *um-aRi > magi ‘come’, *C > s in *Caliŋa > saŋila ‘ear’ or *CebuS > sibus ‘sugarcane’, and *s > t in *susu > tutu ‘female breast’. Unlike such Northern Luzon forms as Ilokano gasút or north Sulawesi forms as Sangir se-hasuʔ, then, the Hoanya form does not show metathesis in relation to PMP *Ratus. In fact, if Hoanya reflected *RasuC its expected reflex for ‘one hundred’ would be **(matala-)gatus, and so would appear to agree with unmetathesized reflexes of the PMP form. Both PAn and PMP reconstructions for this form thus appear to be firmly established, with consonants in the order reflected by the great majority of the languages.

*Radut pluck, pull out

Note:   With root *-dut ‘uproot, pluck, pull out’.

*Ragem clenched in the fist

Note:   Old Javanese agem may also reflect *agem. With root *-gem ‘grasp in the fist’.

*Rakéd to hitch, tie two things together

Note:   Also Tausug mag-gakut ‘to tie or fasten something securely’.

*Rakep seize, grasp, wrap around

Note:   With root *-kep₂ ‘seize, grasp, embrace’.

*Ra(ŋ)k(e)qaŋ spread the legs

Note:   With root *-kaŋ₁ (or *keqaŋ?) ‘spread apart, as the legs’.

*Rakit lay long objects side by side; raft

Note:   Also Agta (Dupaningan) rákit ‘raft’, mag-i-rakit ‘send something by raft’ (< Ilokano), Mapun gakit ‘raft (usually made from bamboo)’ (< Central Philippine source), Iban akit ‘raft; make or use a raft’, Javanese rakit ‘pair (as of carabaos), team, span; layout, arrangement; (forming) a set or pair’, Manggarai raŋki ‘raft’.

The *Rakit probably was made of bamboo, which is highly buoyant, and difficult to sink or become waterlogged. Although cognates are unknown in Taiwan, it is likely that Austronesian speakers reached Taiwan from the mainland of southern China (Fujian) by means of bamboo rafts, since there is no evidence for the outrigger canoe prior to Proto-Malayo-Polynesian. Rafts continued to be useful for purposes of short water crossings, long after outrigger canoes had become the primary means of long-distance oversea exploration. Their function probably was primarily to ferry people and goods across large rivers or short water crossings. For this reason rafts were retained in places like Mindanao or Borneo, where there were many rivers to cross, but lost in places like the Marianas, Micronesia or Polynesia, where open water crossings required more sophisticated watercraft. With root *-kit₂ ‘join along the length’, a raft being the major cultural production made by joining objects (bamboos or logs) along the length.

*Rakut tie together

Note:   Zorc (1971) gives PPh *Rakut ‘rope; bind, bundle’.

*Rambaŋ casting net; spread out, as a casting net

*Rambat large casting net

Note:   Possibly a Malay loan distribution.

*Rambia sago palm

Note:   Also Wetan lpia, Asilulu lapia ‘sago tree, Metroxylon spp.’, lapia mutin ‘cold sago porridge’, Alune lapia, Kayeli mpia ‘sago palm’, Likum lepi, Molima labiʔa ‘sago palm’. Dempwolff (1938) reconstructed *rumbi(y)a ‘sago palm’, but reflexes of *Rambia are more widely distributed, and less vulnerable to the interpretation that they are loanwords from Malay. Nonetheless, both *Rambia and *Rumbia appear to be justified by comparative evidence. In addition, nearly all known reflexes of this form in CMP languages show p for an expected voiced bilabial stop, and this may be evidence for a third variant, but one that cannot be attributed to a proto-languages higher than Proto-Central Malayo-Polynesian.

*Rames squeeze, knead, to mix with the hands

Note:   Also Gayō ramas ‘to massage, knead flour’.

*Rámit to use

*Rampaŋ passionate, quick-tempered

*Rampas rob, plunder

Note:   Dempwolff (1934-38) proposed *Ra(m)pas ‘tear loose’.

*Rampis thin, thin layer

Note:   With root *-pis₂ ‘thin (of texture)’.

*Ramut₁ fibrous roots, as of grass

Note:   Also Agta (Dupaningan) ramót ‘the root of a plant or tree’ (< Ilokano), Maranao gamot ‘poison’, Dobuan ramuna ‘root’. Surprisingly, this widespread form is missing from Dempwolff (1938) because no two of the critical witnesses on which he relied have a homosemantic reflex.

*Ramut₂ a charm used to control someone magically

Note:   Also Tboli lamat ‘a kind of charm; for someone to have this charm means to have a little knowledge about spirit-beings which can help other people by knowing the cause of their illness.’

*Ranebut pluck out

Note:   Also Tagalog gábot ‘uprooting’. With root *-buC ‘to weed, pluck, pull out’.

*Rantaŋ kind of large carrying basket

Note:   Also Proto-Minahasan *bantaŋ ‘large basket’.

*Raŋa₁ open the mouth wide

Note:   With root *-ŋa(q) ‘gape, open the mouth wide’.

*Raŋa₂ the spider conch: Lambis lambis

Note:   Most of this comparison (excluding the Malay cognates) was first noted in print by Geraghty (n.d.) who, however, proposed no reconstruction above the POc level.

*Raŋak the spider conch: Lambis lambis

*RaŋaR the spider conch: Lambis lambis

*Raŋaw₁ new shoot of vines or creepers

*Raŋaw₂ to dry up, wither

*Raŋkay dry

*Raŋkep to crawl, creep, as vines

*Raŋkub to cover

Note:   With root *-kub₁ ‘cover’.

*Raŋu dry

Note:   Also Manobo (Western Bukidnon) gaŋuʔ 'of plants, to dry up, Manggarai daŋo ‘dry’. The agreement of the Kalamian Tagbanwa and Kambera glosses suggests that this form may have referred to a contrastive degree of incomplete dryness.

*Rapiq crumble

*Rapuq₁ rotten, crumbling

Note:   With root *-puq ‘brittle’.

*Rapuq₂ branches carried along and deposited by water

Note:   Also Tboli lafuɁ ‘dry leaves or twigs’.

*Rapus to bind, tie the hands or feet, as a person in detention, or an animal to be slaughtered; restraining rope