*z   

*zakan to cook in water, cook rice

*za(ŋ)kit contagious, spreading by contact

*za(ŋ)kuC grab, grasp, seize

*zalan path, made by a human as opposed to an animal; way or means to do something

Note:   Also Sa'ban alin, Kenyah (Long Anap) janan, Kayan (Uma Juman) alan, Berawan (Long Terawan) ilan, Lemeting jalin, Karo Batak dalin ‘path, road; means to achieve something; cause of something; way of doing something’, Buli laliŋ ‘footpath’, Moor ràrina, Yeresiam jáàra, Serui-Laut raŋ, Munggui layan ‘path, road’, Puluwat yel- road, path, Toqabaqita tale ‘path, track, road; a row of something’.

A number of words in South Halmahera-West New Guinea languages resemble the expected reflex of PAn *zalan, but with various irregularities. It is unclear whether these should be treated as inherited forms with sporadic changes, as loanwords, or as chance resemblances. The variant forms from Lemeting, Karo Batak , Buli and Moor might be taken to support a double *zalin, but since Sa'ban alin, Lemeting jalin and similar forms in other North Sarawak languages appear to be products of low vowel fronting it seems best for now to treat the minority pattern with last-syllable *i a product of convergence.

*zalateŋ stinging nettle: Laportea spp.

Note:   Also Sangir əŋgalatiŋ, Bare'e gelata ‘stinging nettle tree: Laportea spp.’.

*zaluaŋ bark cloth

*zaŋan₁ handspan

Note:   Also Maranao raŋao ‘measure by using stretched thumb and finger as measuring tool’, Mongondow yaŋow ‘span with the fingers’, toŋo yaŋow ‘one handspan’, Sa'a taŋaa ‘span; to span with the hand’.

*zaŋan₂ negative

*zaŋaw insect destructive to rice

*zaŋkal handspan

Note:   Also Tae' daŋkan ‘a span’, Manggarai daka ‘small unit of measure’, se-daka ‘handspan’. Mills (1975: 719) reconstructs Proto-South Sulawesi *jaŋka[l?] ‘handspan (unit of measure)’, a form which he proposes to derive irregularly from Dempwolff's (1934-38) *ze(ŋ)kal.

*zaŋkaŋ wide apart, as the legs

Note:   With root *-kaŋ₁ ‘spread apart, as the legs’.

*zapit pinch, squeeze, press

Note:   With root *-pit ‘press, squeeze together; narrow’.

*zaqat bad, evil; ugly

Note:   Also Mapun maat ‘evil; badness’, mag-maat ‘to do bad; to have bad intentions’, ka-maat-an ‘dislike something’ (lit. ‘be affected by the badness of something’), Tausug kaat ‘(for a person) to go wrong, do things that are considered bad; (for someone’s behavior) to become bad’, Murung jeʔet ‘bad, Iban jaiʔ ‘bad, spoilt, broken, out of order, no more use’, Jarai sat ‘bad’, Toba Batak jahat ‘bad; vulgar’, Lamaholot daʔ ‘bad’, Toqabaqita taʔaa ‘be bad, no good; general term for expressing negative evaluation’.

*zaqet bad, evil; ugly

Note:   Also Kayan (Uma Juman) jaʔək, Bintulu jaʔəs ‘bad, evil’.

*zaqit to sew

Note:   Also Itawis dahét ‘stitch’, mad-dahét ‘to sew’, Jarai jhik ‘to sew’, Muna dhai ‘sew an ornamental seam (to trouser legs; a practice in former times)’, Molima saila ‘to sew, make nets’.

*zariaŋaw a plant: Calamus spp.

Note:   Dempwolff (1934-38) reconstructed doublets *zaraŋaw, *zariŋaw.

*zariŋ noose or net trap

Note:   Also Malay jaliŋ ‘barrier net (said to be rare)’. In addition, Dempwolff (1938) included Tagalog dáliŋ ~ dálin ‘fastening; binding for bamboo floors’, but the divergent semantics raises questions about whether this form is cognate. This comparison, which was first proposed by Dempwolff (1938) in a somewhat more restricted form, remains problematic. I assume that Maori tari shows the ‘third palatal reflex’ found in a number of other forms, but the connection of Oceanic forms to those in western Indonesia may be due to chance.

*zaRami rice straw

Note:   Also Kavalan Rami ‘rice straw’, Bunun dumali (< met.) ‘millet straw’, Tagalog dayámi ‘straw; hay; dried rice or grain stalks used as fodder for cattle’ (Kapampangan loan), Binukid lagami ‘stalk that is left after the grain or sugarcane has been harvested; straw’, Bare'e deami ‘rice stubble; fallow ricefield’.

The basic referent of this term appears to have been that part of the rice stalk that remains in the field after the harvest (to fertilize the soil for the next planting). In at least two widely-separated languages (Old Javanese and Tontemboan) this word also refers to a flute made from a rice stalk, and in parts of northern Sulawesi the meaning has shifted from the stalks of the planted rice left after the harvest to the invasive weeds and brush that sprouts in a newly fallow field.

*zaRaŋ having wide intervals; seldom

Note:   Also Tagalog dálaŋ 'sparseness; scarcity; infrequency'. Dempwolff (1934-38) has *zaraŋ ‘seldom; scarce’. The Bisayan form shows secondary prenasalization of g from *R.

*zaRem needle

*zaRum needle

Note:   Also Puyuma daum ‘sewing needle’ (< Paiwan), Agta (Dupaningan) dagum ‘needle’ (< Ilokano), Bontok dágom ‘needle’, Ibaloy shaKom ‘needle (as for sewing, injection)’, Pangasinan dágom ‘needle’, Tagalog ka-rayom ‘needle’ (< Kapampangan), Hanunóo dágum (< a Bisayan source?), Tausug jawm ‘a needle; any needle-like object, as a compass needle or the hands of a watch’ (< Malay), Bare'e jaru ‘European fabric needle (< Malay or Buginese). Manggarai jaruŋ may be a Makasarese loan.

*zauq far

Note:   Also Toba Batak dao ‘far, distant’. This morpheme appears to have replaced PAn *dawiN ‘far, distant’ in PMP. Favorlang/Babuza ro, ma-ro is a superficially similar form that, if accepted, would establish both *zauq and *dawiN as PAn, and raise fundamental questions about homonymy. However, the Favorlang/Babuza reflex of PAn *z is clearly t in etymologies such as *zalan > tarran ‘road’, and *quzaN > utas ‘rain’, and it therefore seems best to treat the similarity of Favorlang/Babuza ro, ma-ro to such non-Formosan forms as Lawangan oroʔ, Mambai ro or Waiyewa ma-ro as a product of chance convergence.

*zawa₁ daybreak, dawn

Note:   Also Malagasy (Provincial) javanjavana ‘illumination’.

*zawa₂ millet sp.: Setaria italica (?)

Note:   Old Javanese jawa, after which the island of Java may have been named, is sometimes thought to be a loan from Sanskrit yawa ‘barley’, a word that reportedly derives from the more general term ‘grain’ attested in both Avestan and Sanskrit (Buck 1949: 513). Adopting this interpretation requires the assumption that Sanskrit yawa ‘barley’ was borrowed as pre-Javanese jawa ‘millet’ within the past two millennia, as the early Indian name for Java was Yawadwipa, presumably because millet was widely cultivated on Java at that time (barley is not a native Southeast Asian crop --- see the doubts expressed in the gloss of Old Javanese jawa).

This putative loanword then spread northward with substitution of d for j in languages that lacked palatals, presumably replacing whatever term applied to the type of millet that this word represented. One possible scenario, then, is that jawa > dawa replaced reflexes of *beCeŋ ‘foxtail millet: Setaria italica’. However, the specific application of reflexes of *zawa to foxtail millet in southern Taiwan and much of the central Philippines (but not in intervening areas) argues against this interpretation and in favor of *zawa as a PAn term that has been lost over wide areas but retained in scattered pockets in its original meaning. Various references (Aklanon, Cebuano, Agutaynen, Yakan, Iban) suggest that millet was secondary to rice as a daily staple in virtually all language groups, a factor that may have played a role in the far more robust attestation for terms like *pajay or *beRas as compared with *beCeŋ, or *zawa.

*zeket to burn (fields, etc.)

Note:   Miri jukat is assumed to be a u ablaut form of an unrecorded base **jakat.

*zelaq tongue

*zelay a cereal grass with edible seeds that are also used as beads, Job’s tears: Coix lachryma-jobi L.

Note:   Also Malay henjelay ‘millet’, Old Javanese jahəli ~ jahli ~ jali ‘a kind of millet: Eleusine coracana’, Javanese jali ‘a plant that produces corn-like (barley-like) kernels’, Balinese jali-jali ‘a species of plant’.

*zeliŋ cross-eyed; squint

Note:   Also Manobo (Western Bukidnon) siliŋ ‘cross-eyed’, Old Javanese jilak ‘partial eye defect (cross-eyed?)’, Wolio jilo ‘cross-eyed, squint’, Rotuman jila ‘squint, be crossed (eyes)’.

*zelzel stuff or fill with material

*zemak feel, grope

Note:   Also Toba Batak gomak ‘grasp something, take a handful’, Mongondow domok ‘catch, grasp, seize; fight, come to blows or grips’.

*zeŋkuŋ curved

*zepit pinch, take up with the fingers

Note:   Uma jopiʔ may be a Kulawi loan. With root *-pit (or *peqit?) ‘press, squeeze together; narrow’.

*zeRa hurt, frightened, or discouraged by previous bad experience

Note:   Also Tagalog daláʔ ‘wary, cautious or scared because of previous painful, unpleasant, unfortunate or unsatisfactory experience’, ma-daláʔ, dalaʔ-ín ‘to have an unpleasant experience caused by the attitude of others or by other circumstances; to be taught by painful experience (in consequence of which one is wary about returning to a certain place)’, Cebuano dulá ‘for animals to be shy and not allow anyone to approach them’, Mapun dallaʔ ‘to give up doing something (because of having a bad or painful experience; to begin to give up’, dallaʔ-un ‘cause someone to give up doing something’.

It is unclear whether the Tagalog and Mapun forms justify a doublet *zeraq, but this seems doubtful, since the expected Tagalog reflex of such a form would be **diláʔ. Rather, it seems more likely that forms such as those listed in this note are loanwords from Malay. The same cannot be said, however, for Manobo (Western Bukidnon) deǥa, and so the reconstruction stands despite the questionable nature of some of its members.

*zeRet noose trap

Note:   Also Toba Batak jorat ‘noose trap for catching animals’ (< Malay).

*zilak cross-eyed

*zilat to stick out the tongue

*zilateŋ stinging nettle: Laportea spp.

Note:   The Oceanic members of this comparison are taken from Ross (2008:232).

*ziŋkit touching or joined along the length

Note:   With root *-kit₂ ‘join along the length’.

*ziRet noose trap

Note:   Iban jərat ‘running noose, snare’ is assumed to be a Malay loan.

*ziteq vine from which poison is obtained

*zitzit to hiss

*ziuq bathe