*g   

*gabi taro: Colocasia esculenta

*gábut to pull out or up

*gaCel itch, feel itchy

Note:   Also Kankanaey gáte ‘to have itch, scab mange’, gaté, me-ŋaté ‘itching’, Tagalog katí ‘itch, itchiness; urge or lust for something’, katí, ma-ŋatí ‘to itch, feel itchy, suffer from itchiness’, ma-katí ‘itchy’, Mapun kato ‘an open sore; itchy’, ka-katoh-an ‘be overcome with great itchiness’, Toba Batak gatal ‘itchy’ (< Malay), Sangir katiʔ ‘itch’, ma-katiʔ ‘itchy’, Tae' katiʔ ‘itch’, ma-katiʔ ‘itchy, restless, constantly moving from thing to thing’, Lamaholot gatək ‘itch, feel itchy’.

This comparison illustrates the inherent voicing instability in velar stops, especially the voiced velar stop (Blust 1996a). The weight of the evidence favors *g- in both PAn and PMP, but — as noted by Mills (1975:728) — this was replaced by k- in a number of daughter languages. Although the primary sense of this term, and the only sense that can be attributed to PAn on present evidence, is ‘itch, feel itchy’, it is clear from a number of reflexes in the Philippines and Indonesia that the word also functioned as a double-entendre for sexual lust. Moreover, the reflexes in Aklanon, Iban, Gayō, Karo Batak and perhaps Makassarese suggest that *gatel in the sense of ‘sexually itchy, oversexed’ applied particularly to women, and hence referred to nymphomania in addition to itchiness of the skin.

*gaduŋ kind of tuber: Dioscorea spp.

Note:   Included in Dempwolff (1934-38), but with a more restricted distribution.

*gae to hook, gaff

*gagar brave, plucky

*gagu dumb, mute

*gakgak caw, as a crow; raucous talking of people

Note:   Also Old Javanese gāgak ‘crow’.

*gala the almasiga tree: Agathis celebica (Koord.)

Note:   Possibly a Tagalog loan distribution.

*galamáy finger, toe; tentacles of octopus

*gálaŋ respect, reverence

*galiŋ to grind, crush by grinding

Note:   Also Mansaka galiŋ ‘to grind, as corn or rice’. Toba Batak haliŋ is assumed to reflect *galiŋ with sporadic interchange of *g and *k (Blust 1996a). This variant is far less securely attested than *giliŋ, and the resemblance of the Ngaju Dayak and Toba Batak forms to those in the Philippines may be a product of chance.

*galiŋgaŋ a tree: Cassia alata

*galu stir, mix

Note:   Also Sundanese galo ‘mixed’, Wolio geru ‘stir (e.g. with a spoon)’. Mills (1975:700), Mills (1981:73) attributes Proto-South Sulawesi *garu together with various forms in western Indonesia meaning ‘harrow’, and Fijian kadru ‘scratch, i kadru ‘harrow’ to PAn ?*/ga(n)[dD]u[hØ])/ ‘stir, scrape’.

*galugaʔ a plant: Bixa orellana

*galuŋ curve

Note:   With root *-luŋ₁ ‘bend, curve’.

*gama morning star?

*gamak feel, hold in the hand

*gamat plant from which a dyestuff is obtained

*gamaʔ catch fish or shrimp with the hands

*gambaŋ valued metal

*gamet edible sea animal or plant

Note:   Also Itbayaten gamed ‘an edible, brown, slippery seaweed: Porphyra crispata (Kjellman)’.

*gámit to use

*gamut poison

Note:   Probably identical with PMP *Ramut ‘root’ (some reflexes of which mean ‘medicine’), but the distinctly different sense of these terms and their cross-linguistic agreement calls for a separate entry.

*ganas raw energy, animal appetite

Note:   Also Kapampangan gána ‘appetite’, Aklanon gána ‘appetite, taste for, desire’, Cebuano gána ‘appetite for eating’. K.A. Adelaar has pointed out to me that Kapampangan, Aklanon, and Cebuano forms almost certainly are borrowings of Spanish gana(s) ‘desire, want, appetite’. However, such an explanation is untenable for Old Javanese ganas. It is possible that Pangasinan gánas ‘appetite’ is a blend of an inherited form and an imported meaning that bore a general resemblance to the meaning it replaced.

*ganti to replace, substitute for; to return in kind something given or done

Note:   Also Tausug gantiɁ ‘a replacement, substitute, exchange, successor; surrogate’.

*gánut pull up, pull out, as grass or a tooth

Note:   Also Aklanon gánit ‘to pull young seedlings for transplanting’.

*gaŋa to open the mouth

*gaŋgaŋ heat or dry something near a fire

Note:   With root *-gaŋ ‘dry near a fire’.

*gápas sickle; to harvest rice with a sickle

*gapgap stammer

*ga(m)puŋ float

Note:   Also Iban lepoŋ-lepoŋ ‘floating’, Malay ampoŋ ‘light (in water), cork-like (of wood)’, apoŋ ‘floating on water; driftwood’, ter-kapoŋ-kapoŋ/ ‘bobbing up and down on the surface of water, as driftwood or a floating leaf’.

*gáqud oar; to row a boat

*garaŋ₁ fierce, intense, as of appetite or temper

Note:   Also Kelabit garəŋ ‘roaring, as an animal’, Old Javanese aŋ-garaŋ-gaŋ ‘to attack furiously in great numbers(?)’. Based on the Toba Batak and Malay forms cited here, Ngaju Dayak ga-garaŋ ‘very bright, of the moon’, and Javanese garaŋ ‘to smoke out, fumigate’ Dempwolff (1939) posited *garaŋ ‘hot; ardent, passionate’. However, Pigeaud (1938) glosses Javanese garaŋ as ‘dry, roasted, warmed up (on a grill over the fire); dried (of tobacco over a fire)’, leaving Dempwolff’s cognate set as a collection of forms with far less semantic coherence than those brought together here.

*garaŋ₂ uneven, of edges, jagged, rough

Note:   Also Bikol gásaŋ ‘a jagged stone; broken bits of stone or shell chips’. This is an unusual comparison in that it allows the reconstruction of a pattern of variable base reduplication of the sort seen in English words like hocus-pocus, wishy-washy, or willy-nilly. It is conceivable that the agreement of the Bikol and Sika forms in supporting *garaŋ-gasaŋ is a product of convergence, since it is unknown elsewhere, but this seems unlikely in view of the regular sound correspondences, length, and close agreement in meaning.

*garáw restlessness, wild or unruly movement

*gariŋ mill or grind

*garis to brush past, rub or scrape against; draw a line

Note:   Also Cebuano garás ‘to scratch lightly, put a thin wound on the surface’, Mongondow gares ‘line, stripe’ (< Malay), Wolio garisi ‘line, stripe’ (< Malay), Muna garinsi ‘draw lines in, on something’, (< Malay), Kei garis ‘matchbox, box of matches’ (< Malay). Dempwolff (1938) glossed this form as ‘kratzen’ (‘to scratch’), but the glosses of reflexes suggest that it actually referred to light grazing contact, as of two persons running side-by-side and occasionally brushing against one another, or the contact producing fire when a match is struck, hence the meaning ‘line; draw a line’ in some languages, as opposed to heavy scratching or scraping. The loss of *-s in Yamdena is found in some other items, but is sporadic.

*garit striped, having stripes or streaks of different color

Note:   Also Itbayaten ma-garis ‘striped?’, Ifugaw gallit ‘combination of colors; hence, speckles’. With root *-rit ‘scratch a line’.

*garu stir, mix

*garuC comb

Note:   Also Mansaka karut ‘to scratch (as one’s head)’, Javanese garu ‘harrow; comb’, Totoli kalut ‘to claw, scratch’, Tajio karut ‘to scratch an itch’. Although this comparison is sufficiently well-attested to make the reconstruction likely, Austronesian languages show a statistically significant correlation of *g- or *k- with words meaning ‘scrape, scratch’ and the like, whether they are cognate or not, much like the high-frequency correlation of *ŋ- with words referring to the nose and mouth area, or *-l with words referring to dullness or bluntness. For this reason caution must be observed so as not to confuse genuine cognates with products of chance convergence. It is unclear whether the meaning ‘comb’ in Formosan reflexes of *garuC is original or secondary. Traditional combs were made of bamboo, wood, or bone, and delousing combs had very narrowly-spaced teeth that must have scraped the skin in passing through the hair so as to remove the parasites they were designed to eliminate.

*garus scratch

Note:   Also Itbayaten garas ‘to scratch’, Binukid kahus ‘to scratch (something or someone with one’s nails/claws)’, Malay gərus ‘a shell, Cypraea tigris; giving a gloss to cloth by use of this shell; calendering’.

*gaRami rice straw

Note:   Also Itawis gurámi ‘rice straw’.

*gaRaŋ small edible freshwater crab

Note:   Also Favorlang/Babuza aggan ‘crab’, Kavalan waRaŋ ‘crab (generic)’, Maranao ogagaŋ ‘crab’, Tiruray kagaŋ ‘type of edible crab found in swamps’, Mapun kagaŋ ‘generic term for crab’, Yakan kagaŋ ‘crab in sea’. Zorc (1971) posited PPh *kaRaŋ ‘crawler (of crab); crab’.

The Malayo-Polynesian evidence unambiguously supports *k- in this form, which initially appears to agree with most Formosan reflexes. However, Laurent Sagart has correctly pointed out that all Formosan reflexes with k- are in fact ambiguous for *k- or *g, and that both Puyuma garaŋ (which I had earlier overlooked), and Paiwan gaŋ point to *gaRaŋ, leaving the initial consonant of PMP *kaRaŋ as an unexplained irregularity. For other examples of sporadic *g/k crossover cf. Blust (1996a).

*gaRaw hoarse, raspy (of the voice)

Note:   With root *-Raw ‘hoarse’.

*gaRuk scratch up, scrape up

Note:   Also Iban garuɁ ‘scratch, scrape’.

*gásaŋ coral

*gasgas rub, abrade

*gasiŋ spinning top

Note:   This form shows the voicing cross-over seen in many morphemes that contain a velar stop, especially those with *g. There is no discernible difference of meaning between *gasiŋ and *gasiŋ-an.

*gatak cleave, split

*gateq sap, resin, sticky secretion

Note:   With root *-teq ‘sap, gummy secretion’.

*gatgat to bite something off, to chew something up; to mince

*ga(n)tuk knock against, of body parts

Note:   With root *-tuk₂ ‘knock, pound, beat’

*gatuŋ firewood

*gaut to scratch

*gawaŋ gate, opening

Note:   With root *-waŋ ‘wide open space’.

*gaway₁ tentacles of octopus, squid, jellyfish, etc.

Note:   Given the reduplication in Muna and several Oceanic languages, PMP *gaway-gaway might also be posited. However, since tentacles are inherently plural the chances of historically independent doubling to mark plurality cannot be discounted (as seen in Sa'a ka-kawe).

*gaway₂ religious feast, festivity

Note:   On the basis of Tagalog gáway ‘witchcraft; sorcery; spell’, Ngaju Dayak gaway-an ‘work that is performed to pay a tax’, Malay gaway ‘celebrate a feast’, and Javanese gawe ‘job, task’, Dempwolff (1938) reconstructed Uraustronesisch *gaway ‘(magical) occupation’. However, for Malay Wilkinson (1959) gives gaway ‘tool, instrument, agent’, and in either case the loose semantics of this comparison makes it a poor candidate for a cognate set. The Iban and Mori forms, by contrast, show a close semantic agreement, and point to a much clearer meaning in the proto-form.

*gawed betel pepper

*gayák to plan, prepare to do something

Note:   Possibly a Tagalog loan distribution.

*gayam a tree: Inocarpus spp., possibly the Tahitian chestnut: Inocarpus edulis

Note:   Madulid (1999) also cites ‘Manobo’ kayam ‘a tree: Inocarpus fagifera’, but without specific reference this is of little value. I assume *gayam, but given the tendency to voicing instability in the velar stops the sporadic change in this form could have gone in either direction.

*gayaŋ hunting spear

*gáyat to slice into thin strips, as meat or vegetables

*gayuŋ-gayuŋ kind of seaweed

Note:   Possibly an Ilokano loanword in Itbayaten, although it is difficult to understand why such a borrowing would occur in a language spoken on a small island surrounded by sea.

*gebáq to destroy, demolish

Note:   Also Agutaynen geba ‘for a structure to collapse or fall over’. Malay rəbah ‘fallen to the ground (as in a faint); crashing fall of trees’ may reflect a disjunct *Rebaq, but the Pangasinan and Tboli reflexes point clearly to *g-.

*geCel to pinch

*gedek sound of sob

Note:   With root *-dek₁ ‘hiccough, sob’.

*gek sound of a thud

Note:   This is a root that appears in many longer forms that refer to a dull throaty sound.

*gekgek animal sound

*gelaŋ earthworm; intestinal worm

*gelap₁ lightning that strikes something

Note:   With root *-lap ‘flash, sparkle’.

*gelap₂ dim, obscure

Note:   Also Casiguran Dumagat kələp ‘night, night-time; to be caught out after dark’.

*gelem dark, overcast, cloudy

Note:   With root *-lem₁ ‘dark’. Possibly a convergent innovation.

*geleŋ cut off, sever; ring a tree

Note:   Sundanese keleŋ ‘something cut into small bits; to ring (a tree)’, Bare'e koloŋ ‘hew a tree trunk or beam so as to produce four squared-off sides’ probably reflect the same prototype. Dempwolff's (1934-38) *gelaŋ ‘bracelet’ may ultimately prove to be identical with this reconstruction, the evidence for it being due to widespread borrowing from Malay.

*gelgél to cut with sawing motion

*gem hold in the fist

Note:   This is the simple root found reduplicated in Dempwolff's (1934-38) *gemgem ‘make a fist’. Maranao kem ‘hold inside the palm’, Murik kem ‘hold’, probably derive from the same form.

*gemel take in the hand, clasp, grasp