f   

fade

faded

There is no compelling reason to consider this anything more than a collection of unrelated forms. Dempwolff (1938) proposed *kusem ‘faded’, noting the phonological irregularity of the Javanese form.

faint

On the basis of the comparison Dempwolff (1938) posited ‘Uraustronesisch’ *leŋeR ‘lame’. The Old and Modern Javanese forms appear to be Malay loanwords, and the Tagalog form (which I am unable to find in my sources) probably is best treated as a product of chance.

fall out

false

Based on this comparison Dempwolff (1938) proposed ‘Uraustronesisch’ *lancuŋ ‘to falsify, forge, counterfeit’. However, the meaning of the Ngaju Dayak form departs radically from this sense (based on a statement in Hardeland (1859) that the rich wear gold instead of brass armbands he speculated that the latter were considered ‘counterfeit’). The Toba Batak form shows an unexplained -m, and the Fijian form also departs widely in meaning.

fan:   vibrate, fan

Dempwolff (1938) compared the forms given here with Malagasy pelika ‘to fan, vibrate, using them to posit Uraustronesisch *pelik ‘to vibrate’. I am unable to find a form that matches his Malagasy citation in any dictionary available to me, and even if it could be found, the semantic latitude allowed in this comparison and the phonological irregularity of the Malagasy form weigh heavily against accepting this as a valid etymology.

fast

Based on the Tagalog and Javanese forms given here Dempwolff (1938) proposed ‘Uraustronesisch’ *tuluy ‘fast’ (schnellsein). It is difficult to see what might have motivated this decision.

fat

father, uncle, elder sibling

feathers:   lose hair or feathers

Tiruray ufaw is assumed to be a GCPh loan; the relationship of Malagasy ofo to the Philippine forms is attributed to chance.

feed

feelings:   harbor bad feelings toward

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.

female:   woman, female

females:   term of address for females

fight

filth:   dirt, filth

Dempwolff (1938) posited *sukar ‘dirt, filth’ on the basis of these two forms, but the sound correspondences are irregular, and there is no known support from other languages.

filth:   dirt filth

Dempwolff (1938) proposed Uraustronesisch *laŋes ‘dirt, filth’ (Schmutz). The Isneg and Ifugaw words form a ‘near comparison’ with Central Philippine forms such as Cebuano laŋsá ‘having a fishy smell or the taste of blood’.

fin

fingers:   take with fingers

finish

fire:   extinguish a fire

firewood

firm:   steadfast, firm

Dempwolff (1938) reconstructed ‘Uraustronesisch’ *tegeŋ ‘steadfast, firm’ (standhaftsein).

firmly:   hold firmly

fish

Chance. One of the most contrived and unconvincing comparisons in Dempwolff (1934-38), this proposed cognate set (which includes Lampung iwa 'fish', Sundanese iwak 'fish') should simply be discarded.

fish sp.

fish sp.

fish sp.

fish sp.

fish sp.

fish sp.

The resemblance of Cebuano baluʔ to the other forms is probably a product of chance. Some of the Oceanic forms (particularly the Polynesian forms) may be related, but a specific common gloss is yet to be found.

fish sp.

fish sp.

fish sp.

fish sp.

fish sp.

fish sp.

The similarity of RALU ive to the Polynesian forms is attributed to chance.

fish sp.

fish sp.

fish sp.

fish:   stench of fish or blood

The Ngaju Dayak, Malay and Javanese forms are cognate, but probably represent a late innovation in western Indonesia. The resemblance of the Tagalog and Malagasy forms to these is assumed to be a product of chance, and although the South Sulawesi forms appear to be cognate, they point to an etymon with penultimate *e, and so do not permit a clear-cut reconstruction.

fish:   to string fish, etc.

Chance. Gilbertese itu may be cognate with Motu Jituri 'a string of fish' (< POc *i-tuRi 'instrument for stringing'), but this interpretation cannot be applied to Nggela itu-itu.

fish sp.

A Dempwolff comparison of very limited distribution and questionable validity

flank:   side, flank

Dempwolff (1938) proposed PAn *tambaŋ ‘side, opposite side, cross side’, glossing the Malay and Javanese forms as ‘überqueren’ (‘cross over’), and the Fijian form as ‘side’, but this meaning does not occur in Capell (1968). The entire comparison is extremely forced, and best treated as a collection of random similarities.

flash:   dazzle, flash

flash:   to flash, of lightning

Ranby (1980) gives NANU iti as a borrowing from English, but Gilbertese iti would appear to be a more likely source. The resemblance of these forms to Proto-Ambon *iti is regarded as a product of chance.

flash:   glitter, flash

flat roof

Chance. Other comparisons such as *buŋbuŋ(-an) 'ridge pole show clearly that the house in WMP society had a sloping roof.

flat

The forms in western Indonesia probably represent a Malay loan distribution, while Rennellese papa is assumed to be a chance resemblance.

flat container:   edge, rim, flat container

This is an example of a type of comparison that is unfortunately rather common in Dempwolff (1938), namely one in which the semantics of forms that would allow a higher-level reconstruction are extremely forced and artificial (the Ngaju Dayak and Javanese forms, which are semantically compatible, cannot safely support a reconstruction, given the longstanding Javanese influence on Banjarese and Ngaju Dayak). Unlike most of the chance resemblances for which Dempwolff proposed PAn reconstructions this one also shows an unacknowledged phonological irregularity, since he cites Fijian i-tabi ‘flat basket’, while Capell (1968) instead gives tabe ‘to hold or carry with the hands under’, i-tabe ‘a small oval basket without handles’.

flattened bamboo

Dempwolff (1934-38) reconstructed *p-al-upuq 'vegetable matter', which he regarded as a doublet of *p-al-aqpaq/. Apart from Malay and Javanese, however, there is little evidence for such an etymon. The similarity of the Tagalog and Iban forms to Malay pelupoh, Javanese plupuh is best attributed to chance.

flattened bamboo

On the basis of the above forms, all of which he glossed as ‘flattened bamboo’, Dempwolff (1938) reconstructed PAn *tepas ‘pressed flat’. However, I have been able to find this gloss only in the Toba Batak form, and the Javanese and Toba Batak forms are phonologically irreconcilable.

float

flow

flow

flow

flower

Despite the striking resemblance of this form among related languages the lack of support in any language outside Bunun and Polynesian strongly suggests that this comparison is a product of chance.

fluid, liquid

fly:   to fly

fold, wrinkle

Dempwolff (1938) posited *kebut ‘fold together, wrap’, but the resemblance of the Tagalog form to the Fijian and Polynesian words seems is best attributed to chance.

follow, go after

food

The Isneg and Sasak forms may be independently affixed forms of *pæn. The similarity of Mentawai iba to these is attributed to chance.

food plant

food plant

forbid

forget

Dempwolff (1938) added Fijian lolo-a ‘unwell; seasickness’ to this and posited Uraustronesisch *leles ‘lose consciousness’ (das Bewustsein verlieren). This is best treated as a chance resemblance.

forgiveness:   pardon, forgiveness

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

form

forth:   wave back and forth

foundation, basis

fraction:   part fraction

fragile:   weak, fragile

Dempwolff cited these forms in support of Uraustronesisch *lawlaw ‘weak, feeble’ (Schwachsein). However, I find no Tagalog form with the meaning given, or anything close to it in any modern dictionary of the language, and the whole comparison is best treated as a product of chance.

fragment, piece broken off

fraud:   deceit, fraud

Also Toba Batak hunsi ‘closed, locked’, a borrowing of Malay kunci ‘lock or bolt’, metathesized in hinsu. Dempwolff (1938) posited *ki(n)cu ‘deceit, fraud’.

free, on the loose

Based on a slightly different version of this comparison Mills (1981) proposed PAn *la(m)baR ‘free; rebellious’. However, almost nothing about this comparison works: the Tagalog word is semantically at odds with the others cited here, Acehnese labaih can only reflect *labas, the Madurese form contains a penultimate vowel that disagrees with the reconstruction, and the other forms are confined to languages in central or southwest Sulawesi, where borrowing cannot safely be ruled out with regard to Buginese or Makassarese influence on Bare'e.

frog

All-in-all this form is best regarded as a loanword. I am indebted to Alexander Smith for bringing this to my attention.

from:   support from below

from:   suspend from poles

fungus