p   

pain

Dempwolff (1938) posited *kedut ‘pain’, but the Tagalog form is irregular (expected **kurót), and is a not a convincing semantic match for Javanese kedut.

palate

palm tree sp.

palm:   sago palm

The resemblance of Balinese ental to the other forms cited here is attributed to chance. The Lau and Arosi forms must reflect earlier *osa, and Rotuman ota must be a loan, although its source remains problematic. This leaves Mota ota, Anuta ota 'sago palm' as the only clearly cognate forms, but given the evidence for borrowing in Rotuman this distribution is perhaps best regarded as a product of diffusion.

Palolo worm

pant, be out of breath

Tempting as it is to accept this as a valid comparison, the sound correspondences simply do not work. The Puyuma vowels disagree with those of the Paiwan and Ibaloy forms, and although the latter two languages appear to show a perfect agreement, Paiwan /s/ reflects *S and Ibaloy /s/ reflects *s, making reconstruction impossible. Like many other bases with initial velar nasal, this one probably is a product of convergent innovation involving the phonestheme ŋ-, widely associated in Austronesian languages with the nasal-oral area (Blust 2003a).

pardon, forgiveness

The forms in Javanese, Sundanese and Balinese are cognate, but the similarity of these to Tagalog apulaʔ is due to chance.

pare:   to peel, pare

parent-in-law

part fraction

particle:   affirmative particle

particle:   clause-final particle

particle

The Tagalog linker and the Hanunóo conjunctive particle are assumed to be cognate. The resemblance of the other forms to these is attributed to chance.

passive affix

The suffixed *-a in Oceanic languages probably is identical to *-a '3sg. object'. The prefixed a- in Malagasy and Fijian is assumed to have no historical connection, either between these two languages, or with any of the suffixes cited here.

past tense marker

Apparently reconstructible for PCEP. The Paiwan and Amis forms disagree in pointing to *Si and *i respectively. I take the resemblance of PCEP *i to all other forms cited here as a product of chance.

paste

Most likely a chance resemblance. Dempwolff (1938) reconstructed Uraustronesisch *lepa ‘paste’.

pattern:   textile pattern

paunch

pearl

peel:   to peel, pare

peel:   to peel

Reconstructible as PCP *isi. The resemblance of Rotinese isi to PCP *isi is attributed to chance.

peel off, separate

Chance, although both forms appear to contain the root *-bak₃ 'split off, separate'.

peeling, rind

Chance. For an alternative etymological connection for Cebuano paflut cf. POc *paro. For an alternative view of the Cebuano:Malay comparison which is now rejected cf. Blust (1970).

penetrate:   pierce, penetrate

Dempwolff (1938) posited *serep ‘penetrate’, a form that was accepted in Blust (1972), but for which the supporting evidence is now considered unconvincing.

perform:   do, perform

permission

Based on the Tagalog and Malagasy forms Dempwolff (1938) proposed ‘Uraustronesisch’ *tulut ‘permission’, despite a lack of comparative support for his proposed gloss.

persuade:   entice, persuade

The resemblance of Kankanaey imók to the Central Philippine forms is attributed to chance.

pick up

Mills (1981:75), who cites both the Buginese and Makassarese words with a simplex k, proposed PAn *si(ŋ)kit ‘to pick up’, but the known evidence does not appear strong enough to rule out chance convergence.

piece broken off:   fragment, piece broken off

pierce, skewer

Dempwolff (1938) included these forms under ‘Uraustronesisch’ *zuluk ‘pierce, penetrate’ (Stechen). However, there is little to recommend this comparison, which appears to be a collection of formally compatible words that have almost nothing else in common. It is true that the meaning ‘penetrate, bore through’ in Toba Batak man-julluk is fairly similar to ‘put in, insert, sheathe’ in Samoan sulu, but the medial geminate is unexplained (Dempwolff wrote ju<l>luk), and it is further noteworthy that Tongan hulu ‘to fix one’s loincloth by tucking in one end of it at the waist; to tidy up the edges of a mat by tucking in the protruding ends’, was omitted despite its obvious cognation, presumably because it would weaken the argument for comparing the Samoan form with those in insular Southeast Asia

pierce, penetrate

Dempwolff (1938) posited *serep ‘penetrate’, a form that was accepted in Blust (1972), but for which the supporting evidence is now considered unconvincing.

pig

pigpen

pinch

pith

pity, compassion

The similarity of Makassarese ballasaʔ to the other forms cited here probably is due to chance. The remainder of this comparison is best regarded as a late innovation in western Indonesia, with likely borrowing either from Malay or from Javanese.

place:   put in place

place of burning

Dempwolff (1938) proposed *(s)eRa(q) ‘place of burning’, but this comparison is both formally and semantically problematic, and is best treated as a product of chance convergence.

place of residence

Despite the clear phonological irregularity in the initial consonant, and the divergent semantics of the Malay and Polynesian forms Dempwolff (1938) posited PAn *tempet ‘place of residence’.

plan, design, project

plank:   board, plank

Dempwolff reconstructed *bani(r), and Tsuchida (1976) posits *baNiR/. I regard Saaroa, Tsou, Paiwan and Ngaju Dayak, Malay, Acehnese, Sundanese, Sasak as forming two geographically restricted cognate sets. The resemblance of these sets to one another and to Selaru and Yapese is attributed to chance.

plant:   food plant

plant sp.

plant sp.

plant sp.

plant sp.

plant sp.

plant sp.

plant sp.

plant sp.

plant sp.

Chance, or if related, Aklanon, Cebuano, Maranao, Mongondow reflect an innovation in PCPh.

plant sp.

Chance. Cebuano pálaw 'large vigorously growing aroid: Cyrtosperma/ merkushii', Maranao palao 'cultivated taro: Xanthosoma violaceum Schott. suggest a PCPh *pálaw 'large aroid: Cyrtosperma spp.'. The Tagalog word shows secondary suffixation.

plant sp.

plant sp.

The Old Javanese and Balinese words are cognate. The similarity of the Isneg and Kayan forms to these and to each other is due to chance.

plant sp.

plant sp.

plant shoot:   sprout, plant, shoot

Dempwolff (1938) cited Fijian i-so-somi ‘seedling’, but this is not the sense of this term in Bauan Fijian as given by Capell (1968), or in Wayan as given by Pawley and Sayaba (2003). In the absence of further evidence this comparison is best treated as a product of chance.

plants:   covering membrane of plants

pledge:   vow, pledge

This comparison proposed by Dempwolff (1938) appears to bring together completely unrelated forms. I am unable to find Tagalog sahot in either Panganiban (1966) or English (1986), and Samoan sau does not occur in this meaning in any dictionary I have been able to consult, including Pratt (1893), which was Dempwolff’s source.

plot evil

pluck

pluck (as fruit)

plural marker

point:   sharpen to a point

Chance?

point, tip

Dempwolff proposed *zuru ‘point, tip’ (dbl. *duRu ‘corner’), and included such semantically divergent forms as Malay juru ‘trained worker (at some occupation other than a handicraft)’ and Javanese juru ‘person who performs a certain [skilled] job’, under the general gloss ‘master’ (also cf. Old Javanese juru ‘head, leader, chief (of a division, military or administrative; of a group or trade); tradesman, trained worker’). The entire collection seems extremely forced, and is best abandoned as ill-conceived.

point out directions

Dempwolff (1938) reconstructed ‘Uraustronesisch *taki ‘point out directions’.

pole, mast

The Ngaju Dayak, Malay and Javanese forms are related, but probably a product of borrowing from Malay (with Ngaju Dayak tihaŋ from Banjarese). The similarity of the Oceanic forms to these is best treated as a product of chance. Dempwolff (1938) reconstructed ‘Uraustronesisch’ *tiaŋ ‘pole, rod’ (Pfahl, Stange) and dismissed the Ngaju Dayak form as irregular, but Old Javanese confirms the medial consonant (< *q).

poles:   suspend from poles

poor

pouch:   basket pouch

Dempwolff (1938) posited *kanTuŋ ‘Korb, Tasche am Kleid’ (basket, pouch on the clothing). However, as he noted, the last vowel of the Oceanic forms fails to agree with that of his reconstruction. The forms in western Indonesia appear to be loans from Malay which the Fijian, Tongan and Samoan forms resemble by chance.

pound:   beat, pound

pound:   beat, pound

Dempwolff (1938) used this comparison with Javanese kəwut ‘beat out’ instead of the Sundanese form cited here. However, I am unable to find a Javanese form meeting this description in either Horne (1974) or Pigeaud (1938), and the Tagalog form is both phonologically irregular and semantically deviant, leaving little comparative basis for a reconstruction.

pour:   stream, pour

poverty

Dempwolff (1934-1938) proposed ‘Uraustronesisch’ *dahup ‘weak’.

prayer

The Malay and Javanese words given here are clearly Arabic loans. The Tagalog word looks like it could be a borrowing from Malay, but if so it would be the only known case in which a Malay r has been borrowed as a Tagalog g, mimicking the correspondence reflecting PAn *R. For now it seems best to consider the resemblance of the Tagalog word to those in Malay and Javanese a product of chance.

precipitous:   aslant, steep, precipitous

Dempwolff (1938) proposed *si(dD)aŋ ‘steep, precipitous’, but this comparison appears to be confined to Ngaju Dayak and Malagasy. It is thus best treated as ‘noise’ not on grounds of false cognation, but rather on the grounds that it cannot be reconstructed for a proto-language of any considerable time-depth.

prefix

Despite a problem with the medial consonant correspondence, the Bikol, Aklanon, and Hiligaynon affixes appear to be cognate. The similarity of these forms to Kayan pala-, RALU wara- is attributed to chance.

pregnant

prepared:   ready, prepared

Based on forms in Toba Batak and Javanese Dempwolff (1938) posited *sikep ‘ready, prepared’. However, the Karo Batak form shows that Proto-Batak had *siŋkeb, which is phonologically incompatible with the forms in Old Javanese, Javanese, Balinese and Sasak, leaving either a comparison with relatively little time-depth, or a loan distribution.

preserved, protected

None of these forms appear to be related, yet Dempwolff (1938) combined them under Uraustronesisch *le(m)bu ‘preserved, protected, hedged in’ (Eingehegt), in a proposed etymology that is best dismissed as a product of chance.

press

press down

pretend:   feign, pretend

Dempwolff (1938) proposed *sisip ‘to feign, pretend’ based on data from Tagalog, Toba Batak, Javanese and Samoan. I am unable to find an equivalent of his Tagalog form in either Panganiban (1966) or English (1986), or an equivalent of his Samoan expression in either Pratt (1893) or Milner (1966). However, even if these words were taken at face value this comparison would remain unconvincing without further support.

pretty

project:   plan, design, project

preserved, protected

None of these forms appear to be related, yet Dempwolff (1938) combined them under Uraustronesisch *le(m)bu ‘preserved, protected, hedged in’ (Eingehegt), in a proposed etymology that is best dismissed as a product of chance.

provided that, in spite of

proximal deictic:   this, proximal deictic

pry up

The Hanunóo and Gorontalo forms presumably are related. Although both exhibit regular phonological correspondences with the similar words in Ngadha and Rotinese , this is not the case with Manggarai Nuʔa (where a final l is expected). Since the Manggarai, Ngadha and Rotinese forms almost certainly are cognate with one another (but not with Asilulu huʔak) this leaves two low-level cognate sets (one GCPh, the other Lesser Sundas) and a potpourri of chance resemblances.

pry up

pull up weeds

pulpy leaf

Chance. Although this item might be considered a doublet of *palacipaq, Tae' balaʔbaʔ appears to derive from an infixed form of *bejbej (cf. Tae' baʔbaʔ).

pumpkin:   gourd, pumpkin

Malay non-cognate, Ngaju Dayak possibly an early borrowing from Javanese.

pure

purpose:   intention, purpose

pursue

This comparison was the subject of a tortuous discussion in the earlier literature based on: 1. an approach to linguistic reconstruction which circumvented the normal scientific procedure of triangulation in the justification of inferences, and 2. a failure to distinguish directly inherited from borrowed forms (Dyen 1953:363ff, Hendon 1964:377ff). The Ngaju Dayak, Iban and Malay words clearly reflect PAn *buRaw 'drive off, chase away, expel' with the atypical, but recurrent change *-aw > u. All other forms in island Southeast Asia appear to be Malay loans, as attested by lexical doubletting in Karo Batak, Toba Batak and Dairi-Pakpak Batak, where a directly inherited reflex of *buRaw is found next to the Malay loanword, with semantic differences that are consistent with a borrowing hypothesis. Given the relatively complete integration of this loanword into the affixational systems of some of these languages (e.g. Toba Batak, and the appearance of buru in Old Javanese texts it appears likely that this word was borrowed fairly early in most languages (perhaps during the classic Sriwijaya period, circa 7th century AD). The similarity of the Southeast Solomonic forms to those in western Indonesia is attributed to chance.

pursue, chase

push aside

push shove

Probably a chance resemblance. Dempwolff (1938) posited ‘Uraustronesisch’ *tunzaŋ ‘push, shove’ (stossen).

put in place

putrid