nervous:   anxious, nervous, restless

Malay bəlisah points to a proto-form with *-q, which Cebuano balísa supports a similar form with final vowel. Malay gəlisah presumably is a secondary development from earlier bəlisah.

1sg oblique

Also Molbog ny-ahiʔ, Lotud j-okiʔ ‘1sg oblique’. The reconstruction of this variant, and *q-final forms of the first person plural inclusive and exclusive oblique pronouns, is based on material provided by Jason Lobel. The reconstruction of all forms of the oblique personal pronouns has benefited from unpublished data provided by Lobel, although they are independently reconstructible without this material.

ant

anxious:   anxious, nervous, restless

Malay bəlisah points to a proto-form with *-q, which Cebuano balísa supports a similar form with final vowel. Malay gəlisah presumably is a secondary development from earlier bəlisah.

arrow

Despite the striking similarity of these forms a reconstruction is impossible. In both languages the last four segments could reflect either *iwal or *iwaR. However, Saisiyat s- can reflect only PAn *C-, which became Kavalan t-, and Kavalan s- can reflect only PAn *s or *S, which became Saisiyat h and ʃ respectively, leaving the proto-form in limbo, despite the near certainty that it points to a phonetically similar form in PAn.

backward:   retreat, move backward

Manobo (Western Bukidnon) endud reflects *endud or *enduD; Malay undur reflects *unduD, *undur, or *unduR.

banana

The forms in northern Luzon reflect *baRat, while Tanjong and Kejaman balat can only come from *balat, and the remaining forms from *balak.

bat sp.

bead (kind of...)

The Isneg form points to *luRut or *lugut, which the Kayan form can reflect only *lukut.

bedbug

According to Wilkinson (1959) Malay basat is of Chinese origin, although this remains to be established.

beetle:   rhinoceros beetle

Osmond (2011:401) uses these and other non-corresponding forms to posit POc *tabuRuRu ‘kind of beetle’. While there clearly is a historical connection between these forms, the sound correspondences fail to agree, and a reconstruction is thus impossible on the basis of known evidence.

biting:   hold by biting carry in the mouth

black

blacksmithing

blind

blind mole (animal)

The Thao and Favorlang/Babuza words are cognate, but point to a lower-order proto-language, Proto-Western Plains *mumu.

blue-black

body

The last vowel of the Admiralty forms (Wuvulu, Aua, Seimat) and that of the Polynesian forms fails to agree, thus preventing a reconstruction.

bone:   collar bone (?)

Mills (1975:813-814) posited Proto-South Sulawesi *salaŋga ‘shoulder(?)’, with the voiced stop preserved in Makassarese, but changed in languages such as Buginese through regular postnasal devoicing. His gloss for both the Buginese and Makassarese forms is ‘breast bone of an animal’, but modern dictionaries give only ‘shoulder’ and verbs based on this noun.

booming sound

borrow

bovid, kind of wild bovid possibly Bos sundaicus

breast

brink:   brink, edge, shore

Also Cebuano taŋpi ‘bank of a body of water; edge of a surface; nearby, close by’.

bug:   rice bug

The final syllable is shared, but nothing else.

busy

Canarium almond

Based on a somewhat larger version of this comparison Ross (2008:317) posited POc *(q)alipa, *lalipa ‘nut sp., possibly canarium almond’. However, the comparison he presents shows no more regularity than the smaller one cited here, and his reconstruction cannot be justified by application of the Comparative Method.

carry:   hold by biting carry in the mouth

chattering:   sound of a monkey chattering

chattering:   sound of a monkey chattering

chewed:   chewed material

Mills (1981:68) proposed ‘Proto-Indonesian’ *kuta() ‘to chew’, but the sound correspondences in the penult are irregular, and the similarity of these forms may be a product of chance.

chicken:   chicken louse

A perfect comparison, except that Philippine reflexes point unambiguously to a trisyllable, while those in northern Sarawak point just as unambiguously to a disyllable, leaving no clear basis for a reconstruction one way or the other.

chiming:   sound of tinkling or chiming

Pangasinan points to *karanciŋ, but Malay to *karənciŋ. This phonological discrepancy may be due to the onomatopoetic character of the word, but in any case a reconstruction is not currently possible.

citrus:   citrus fruit

citrus fruit:   citrus fruit, pomelo

Possibly a loan distribution.

close the eyes

If this comparison is valid the source of Buli -k is unclear, since *k normally is lost in all positions.

close together

cloud

collar:   collar bone (?)

Mills (1975:813-814) posited Proto-South Sulawesi *salaŋga ‘shoulder(?)’, with the voiced stop preserved in Makassarese, but changed in languages such as Buginese through regular postnasal devoicing. His gloss for both the Buginese and Makassarese forms is ‘breast bone of an animal’, but modern dictionaries give only ‘shoulder’ and verbs based on this noun.

comb

The comparison of Tiruray seledoy with Greater Central Philippines forms such as Hanunóo sudláy, or Mansaka sodlay suggests a PPh *suleday, with metathesis of the medial cluster derived from schwa syncope in GCP. However, although Tiruray reduced prepenultimate *a to schwa, it does not appear to have reduced high vowels in the same way, making a direct comparison with the GCP forms problematic. Maranao, together with Tiruray seledoy could be taken as evidence for *saleday, but given their geographical contiguity, and evidence of heavy borrowing from Danaw languages into Tiruray (Blust 1992b), this could easily be a loan distribution.

complete

Borrowing from Tagalog.

contradict:   oppose, contradict

copulate (of fowls)

cross a body of water

cross-sibling

Since at least Milke (1938) there have been proposals for a POc reconstruction *lopu ‘cross-sibling’. However, the evidence for such a form is very limited and contradictory. Hambruch (1908) reported Wuvulu lofu-na ‘cross-sibling’, but my own fieldnotes from 1975 show olofu- ‘elder brother’ and Aua Lovu- ‘elder brother of a man’. The other forms cited here fail to agree phonologically, and there is strong evidence that the POc terms for ‘male’ and ‘female’ (*maRuqane, *papine) also functioned as cross-sibling terms, further weakening the argument for *lopu.

curve

Dempwolff (1938) reconstructed *iluk 'convex curve', but based his etymon solely upon reflexes in Malay and Javanese which appear to contain a common root, but which are not cognate.

dark color

The Polynesian evidence points to a POc form with medial *l, while the Motu evidence indicates *r.

daylight:   morning, daylight

Tolai /l/ can only reflect *l, while Mota /r/ must reflect *r or *R.

dazzling:   shiny, dazzling

deep place in river

Mills (1981) compared these and several other phonologically irregular forms, as Yamdena lifu ‘deep place in a river’ and propsed ‘Proto-Indonesian’ *li(m)buŋ ‘deep or still water’. However, none of the material he presents clearly supports a reconstruction at this stage.

dent

desire:   interest, desire

device:   scarecrow device in paddy field

dew

These words form a near comparison since Isneg lammó points to *lemu, and Tagalog hamóg possibly to *lamuR. In both cases there is a greater than chance resemblance with the established PAN reconstruction *ñamuR ‘dew’.

diarrhoea

digging:   tool for digging

The Ibaloy form points to *saŋgap, while the other forms cited here all support *saŋkap.

dilate:   dilate, open wide

dirt:   dirt, grime

dirty

dog

Agutaynen points to a form with initial *q but final zero, while other languages support the reconstruction of final *q. Since a reflex of PAn *asu ‘dog’ is preserved in some Philippine languages the meaning of this term presumably differed in some still undetermined way.

drizzle

dumb:   dumb, unable to speak

dust

The western Indonesian forms are related, although some may be borrowed from Malay. The similarity of Agta (Central Cagayan) ləppu, Makatea refu and Proto-Polynesian *refu to these is best attribued to convergence, although the similarity of form and meaning is striking.

eagle

Philippine and Bornean forms of this word are irreconcilable primarily because of discrepancies in the initial consonant, but also in some cases because of a lack of correspondence in the penultimate vowel.

east:   east monsoon

This very tempting comparison is imperfect in two ways: 1. languages of the Admiralties point to *u as the first vowel, while those in the Solomons indicate *o, 2. languages of the Solomons suggest a proto-form *koburu, while those in the Admiralties could only have reduced such a trisyllable to a disyllable if the last consonant was *R (hence *kobuRu, with sporadic assimilation of the first vowel to the second, loss of *R and contraction of the resulting sequence of like vowels). Without data from other languages it is very difficult to tell whether this is a remarkable chance resemblance or a valid etymology with irregularities.

ebb:   ebb tide

Malay surut can reflect *suRut or *surut, while Ilokano úgot can reflect *uRut, thus providing no comparative basis for a reconstruction with or without an initial consonant.

eddy:   eddy, whirlpool

Lau /l/ reflects either *l or *R, but Motu loses *l and reflects *R as /r/.

edge:   brink, edge, shore

Also Cebuano taŋpi ‘bank of a body of water; edge of a surface; nearby, close by’.

eel:   eel (small or young)

eels:   trap for fish or eels

The Ilokano form points to *-b, but the Mansaka form to *-p.

embarrassment

Presumably from a PPh *elad or *eled, though reconstruction cannot be achieved with the evidence at hand.

emotional:   emotional pain

The Cebuano form cited here points to *siluq, and the Malay and Sasak forms to *silu. Iban siluʔ is compatible with neither of these, but together with Cebuano síluʔ could be taken as evidence for Zorc’s (Zorc 1982) PAn *ʔ, if it were not for the extremely problematic nature of that proposed proto-phoneme (Blust 2013, sect. 8.2.2.4). Until further cognates are found, then, it seems best to treat this set as a near comparison. Despite this limitation, this word is of some interest, as it is clearly a doublet of *-ñilu and *ŋilu, both of which refer exclusively to the physical discomfort experienced in the teeth after eating something very sour or acid. This meaning also is found in the Iban and Sasak forms cited here, but Cebuano, Iban and Malay all include an additional reference to emotional pain, and this seems to be the distinguishing feature of this variant.

envy

exchange

Maranao points to *sambiq, but Iban to *sambi, with apparently secondary final glottal stop.

exclamation:   exclamation of pain

The variability in this form may be due to its highly expressive character, as it is usually uttered in moments of anguish or shock.

eyelash

eyelash

fade

Tboli budas points to *budas or *bujas, but Kayan bulah can only reflect *bulas. Also Bikol luʔdás ‘faded, discolored’.

fast:   quick, fast

Saisiyat /l/ evidently points to *N (Li 1978:140), while the Sangir and Mongondow forms indicate *l. All of these forms are assumed to contain the root *-kas₃ ‘swiff, agile; energetic’ (Blust 1988:105).

fever:   nightmare, fever

few

Uma kediʔ reflects an undetermined final consonant *-pb, *-td, *kg, or *q, but Paiwan has lost none of these consonants in final position.

field:   scarecrow device in paddy field

field:   scarecrow device in paddy field

finger

fingers:   roll with fingers

finish off (eating... work... etc.)

fish:   trap for fish or eels

The Ilokano form points to *-b, but the Mansaka form to *-p.

flat:   flat surface

flood

Kavalan points to a form with *-p, while Paiwan supports only *-b.

fly

fly off:   fly off, squirt out

follow someone

Yami and Itbayaten point to a form with *-t, while Ibaloy and Pangasinan support *-d.

food:   stuff mouth with food

force:   force, rape

Given the close formal and semantic similarity of the Tagalog and Balinese members of this comparison it is tempting to propose a reconstruction, but the correspondence for the final consonant are irregular. Based on a similar comparison, Dempwolff (1938) reconstructed Uraustronesisch *gagaq ‘to use force’, but he cited the Tagalog form as having a final glottal stop that is not indicated in any Tagalog dictionary available to me.

forget

friend:   friend (male to male)

Despite the distinctive semantic agreement of these terms and their general formal similarity it is impossible to achieve a reconstruction, since Puyuma ariH points to *aliq, Paiwan qali points to *qali, and Tiruray adih points to a proto-form with medial *d, *j or *z.

fruit:   citrus fruit

gleam:   shine, gleam

At first glance these forms appear to share a monosyllabic root *-lak ‘shine’, but Pazeh l reflects PAn *N, while PAn *l > Pazeh r.

gnat

This is a perfect comparison except for the last-syllable vowel, since Paiwan points unambiguously to PAn *e and Kavalan u can only reflect *u.

grime:   dirt, grime

growl

The vowels of Paiwan do not match those of languages in western Indonesia, although these forms undoubtedly are onomatopoetic, and may signal differences of voice pitch.

heart

hen:   strut as rooster around hen

hoarse (of the voice)

Perfect except that Malay reflects *-k or *-g, while Tae' cannot reflect either.

hold by biting:   hold by biting carry in the mouth

infested with lice

insert

Pazeh lezek can only reflect *Nasek, and Bontok lesek can only reflect *lesek or *Resek. Both forms apparently contain the root *-sek₂ ‘insert, stick into a soft surface’.

interest:   interest, desire

interrupted:   interrupted of sleep

invite

The Philippine data support *d as the second consonant, but Puyuma indicates *l.

jackfruit sp.

jawbone

jump down

kind of spear

left (not right)

Also Tagalog kiwáɁ ‘left, left hand side’, Ma'anyan kawiɁ ‘left side’, Malagasy haví-a ‘the left side’, Iban kibaɁ ‘left’, jari kibaɁ ‘left hand’, Sundanese kenca ‘left, left side’. Dempwolff (1938) reconstructed *kiwa ‘left, left side’, despite the far better supported PAn *wiRi in this meaning, and included Malay kiwa ‘be awkward, clumsy’ in his comparison, although no such form appears in Wilkinson (1959). This comparison is highly suggestive, but does not support a reconstruction.

lice:   infested with lice

light tap

With root *-pik ‘pat, light slap’.

like wildfire or a disease:   spread, like wildfire or a disease

loud noise

loudmouth

louse:   chicken louse

A perfect comparison, except that Philippine reflexes point unambiguously to a trisyllable, while those in northern Sarawak point just as unambiguously to a disyllable, leaving no clear basis for a reconstruction one way or the other.

maggot

In both languages *s > t, so this might be derived from *sikay, but the glottal stop in Agutaynen would still be unexplained, as this normally corresponds to Paiwan q.

male:   friend (male to male)

Despite the distinctive semantic agreement of these terms and their general formal similarity it is impossible to achieve a reconstruction, since Puyuma ariH points to *aliq, Paiwan qali points to *qali, and Tiruray adih points to a proto-form with medial *d, *j or *z.

male consanguine

mange:   scabies, mange

mango sp.

material:   chewed material

Mills (1981:68) proposed ‘Proto-Indonesian’ *kuta() ‘to chew’, but the sound correspondences in the penult are irregular, and the similarity of these forms may be a product of chance.

menses

A perfect comparison except that the initial segments of these forms disagree, and Kelabit s in directly inherited forms is normally found only as the reflex of *t before a high front vowel.

mix

Cebuano points to an etymon with *b, while languages of western Indonesia support *p. The forms in at least Old Javanese, Modern Javanese and Sundanese probably are loans from Malay.

mole:   mole, growth on skin

monkey:   sound of a monkey chattering

monkey:   sound of a monkey chattering

monsoon:   east monsoon

This very tempting comparison is imperfect in two ways: 1. languages of the Admiralties point to *u as the first vowel, while those in the Solomons indicate *o, 2. languages of the Solomons suggest a proto-form *koburu, while those in the Admiralties could only have reduced such a trisyllable to a disyllable if the last consonant was *R (hence *kobuRu, with sporadic assimilation of the first vowel to the second, loss of *R and contraction of the resulting sequence of like vowels). Without data from other languages it is very difficult to tell whether this is a remarkable chance resemblance or a valid etymology with irregularities.

morning:   morning, daylight

Tolai /l/ can only reflect *l, while Mota /r/ must reflect *r or *R.

mouth:   hold by biting carry in the mouth

mouth:   stuff mouth with food

move backward:   retreat, move backward

Manobo (Western Bukidnon) endud reflects *endud or *enduD; Malay undur reflects *unduD, *undur, or *unduR.

mud puddle:   mud puddle, pool of water

mute

Philippine evidence points to *amaŋ, but Bornean evidence supports *aməŋ.

naked

Perfect except that a medial b : m correspondence between these languages is otherwise unknown.

nape:   nape of the neck

neck:   nape of the neck

nightmare:   nightmare, fever

no:   no, not

not:   no, not

of sleep:   interrupted of sleep

open:   wide open

open wide:   dilate, open wide

oppose:   oppose, contradict

otter

owl

owl

paddy field:   scarecrow device in paddy field

paddy field:   scarecrow device in paddy field

pain:   emotional pain

The Cebuano form cited here points to *siluq, and the Malay and Sasak forms to *silu. Iban siluʔ is compatible with neither of these, but together with Cebuano síluʔ could be taken as evidence for Zorc’s (Zorc 1982) PAn *ʔ, if it were not for the extremely problematic nature of that proposed proto-phoneme (Blust 2013, sect. 8.2.2.4). Until further cognates are found, then, it seems best to treat this set as a near comparison. Despite this limitation, this word is of some interest, as it is clearly a doublet of *-ñilu and *ŋilu, both of which refer exclusively to the physical discomfort experienced in the teeth after eating something very sour or acid. This meaning also is found in the Iban and Sasak forms cited here, but Cebuano, Iban and Malay all include an additional reference to emotional pain, and this seems to be the distinguishing feature of this variant.

pain:   exclamation of pain

The variability in this form may be due to its highly expressive character, as it is usually uttered in moments of anguish or shock.

peek:   peek, peep at

The Pangasinan form points to *silip, but the Mansaka form to *silib, and so far no arbiter that would allow us to favor one of these choices has been found.

peel

These forms differ in reflecting the presence or absence of *l-, making reconstruction of a Proto-Philippine form impossible.

peep at:   peek, peep at

The Pangasinan form points to *silip, but the Mansaka form to *silib, and so far no arbiter that would allow us to favor one of these choices has been found.

penetrate:   pierce, penetrate

Amis s reflects *S, but Cebuano s reflects *s.

pierce:   pierce, penetrate

Amis s reflects *S, but Cebuano s reflects *s.

plant: black nightshade:   plant: black nightshade, Solanum nigrum

play

plover sp.:   sandpiper sp., plover sp.

Clark (2011:361) proposed Proto-Remote Oceanic *kVlili ‘Wandering tattler: Heteroscelus incanus’, but the evidence he presents does not permit a complete reconstruction.

poisonous insect or snake

pomelo:   citrus fruit, pomelo

Possibly a loan distribution.

pool of water:   mud puddle, pool of water

pool of water:   puddle, pool of water

pregnant

pregnant

puddle:   puddle, pool of water

pufferfish

This comparison was proposed by Osmond (2011:124), who reconstructed POc *toqa ‘kind of fish with toxic flesh, probably Ostracion’.

queasy; urge to vomit:   queasy, urge to vomit

queasy; urge to vomit:   queasy, urge to vomit

quick:   quick, fast

Saisiyat /l/ evidently points to *N (Li 1978:140), while the Sangir and Mongondow forms indicate *l. All of these forms are assumed to contain the root *-kas₃ ‘swiff, agile; energetic’ (Blust 1988:105).

rape:   force, rape

Given the close formal and semantic similarity of the Tagalog and Balinese members of this comparison it is tempting to propose a reconstruction, but the correspondence for the final consonant are irregular. Based on a similar comparison, Dempwolff (1938) reconstructed Uraustronesisch *gagaq ‘to use force’, but he cited the Tagalog form as having a final glottal stop that is not indicated in any Tagalog dictionary available to me.

raptorial bird

relationship between spouses of siblings

resound

With root *-tuŋ ‘deep resounding sound’.

restless:   anxious, nervous, restless

Malay bəlisah points to a proto-form with *-q, which Cebuano balísa supports a similar form with final vowel. Malay gəlisah presumably is a secondary development from earlier bəlisah.

retreat:   retreat, move backward

Manobo (Western Bukidnon) endud reflects *endud or *enduD; Malay undur reflects *unduD, *undur, or *unduR.

rhinoceros beetle

Osmond (2011:401) uses these and other non-corresponding forms to posit POc *tabuRuRu ‘kind of beetle’. While there clearly is a historical connection between these forms, the sound correspondences fail to agree, and a reconstruction is thus impossible on the basis of known evidence.

rice bird

Tagalog pírit must reflect *pidit, while Kelabit pirit cannot.

rice bug

The final syllable is shared, but nothing else.

river

The Hanunóo word may be identical to likúɁ ‘crook, bend (as in a stick, river, etc.)’.

river otter

roll with fingers

rooster:   strut as rooster around hen

rose-myrtle

Also Malay kərəmuntiŋ ‘evergreen shrub with an edible fruit, the rose-myrtle: Rhodomyrtus tomentosa’.

sand

sandpiper sp.:   sandpiper sp., plover sp.

Clark (2011:361) proposed Proto-Remote Oceanic *kVlili ‘Wandering tattler: Heteroscelus incanus’, but the evidence he presents does not permit a complete reconstruction.

sarong:   sarong, wrap-around waist cloth

scabies:   scabies, mange

scar

scarecrow:   scarecrow device in paddy field

scrape:   scrape, scratch

Ibaloy KabKab points to *gabgab. Both of these forms show the common association of initial velar stops in Austronesian languages with the notions ‘scratch, scrape, rub’ and the like.

scratch:   scrape, scratch

Ibaloy KabKab points to *gabgab. Both of these forms show the common association of initial velar stops in Austronesian languages with the notions ‘scratch, scrape, rub’ and the like.

shallow (of dishes)

shark

Despite the generic similarity of these words the only forms that show sufficient agreement in sound correspondences to permit a reconstruction (*maŋiwaŋ) are Sasak maŋiwaŋ, which appears to be a Buginese loan, and the similar forms in Buginese and Makassarese, which are very closely related. These, together with POc *maŋewa strongly suggest a PMP form similar to *maŋiwaŋ, but its precise form cannot be reconstructed based on currently available evidence.

sharpen

shine:   shine, gleam

At first glance these forms appear to share a monosyllabic root *-lak ‘shine’, but Pazeh l reflects PAn *N, while PAn *l > Pazeh r.

shiny:   shiny, dazzling

shore:   brink, edge, shore

Also Cebuano taŋpi ‘bank of a body of water; edge of a surface; nearby, close by’.

short-necked