*e   

*e numeral prefix

Note:   Also Tongan ʔe ‘untranslatable particle used before numerals’, Hawaiian ʔe ‘general numeral classifier which precedes numerals from from 1-9’ (Elbert and Pukui 1979). In addition to the indisputably cognate forms cited above it is tempting to compare Palauan e ‘prefix used to form number words that refer to units of time’. However, cognates are unknown in any other non-Oceanic language, and PMP mid vowels *e and *o are rare and problematic. For these reasons I treat Palauan e as a chance resemblance.

In the standard grammars of Fijian and several of the Polynesian languages the numeral marker e is equated with the predicative or existential marker of similar shape which is widespread in Oceanic languages. However, it is noteworthy that reflexes of *e ‘numeral marker’ are known to be distributed only from the western Solomons to Polynesia, whereas reflexes of the variant *i ‘numeral marker’ are found both in Central Malayo-Polynesian and in Oceanic languages.

One hypothesis worthy of further investigation is that a numeral marker *i was innovated in PCEMP and a predicative or existential marker *e was innovated in POc, with a subsequent tendency for the former (which was syntactically more restricted) to be equated with the latter (which was syntactically freer). Just how restricted the distribution of the numeral marker *i, *e was remains unclear, but in most languages it does not occur with ‘one’, and in many others it does not occur with numerals above ‘five’.

*embaq bring, take along

Note:   Also Acehnese ba ‘bring, take along’.

*embaw high, on top

Note:   With root *-baw₁ ‘high; top’.

*embej bind the waist; band around the waist

Note:   With root *-bej ‘wind around repeatedly’.

*ebek thudding sound

Note:   With root *-bek₁ ‘dull muffled sound’.

*embeŋ a dam; to dam a stream

Note:   Also Iban baŋ ‘thwart in a longboat, between the gunwales’, Malay embaŋ ‘thwart, crossbar’, Sasak embaŋ ‘dam up’. With root *-beŋ₂ ‘block, stop, dam’.

*embet bind the middle

*ebuk dust, powder

Note:   With root *-buk₁ ‘decay, crumble; powder’.

*ebun heap, pile, swarm, flock; to swarm, assemble, gather

Note:   With root *-bun ‘heap, pile; cover with earth; assemble, gather’.

*ebuq gut a fish, eviscerate

Note:   Kelabit ibhuʔ is assumed to derive from underlying **ebhuʔ through affixation with *-in-, a derivation which enables us to explain the voiced aspirate as the regular reflex of *b (following shwa). A parallel example is seen in ebhaʔ ‘water’, ibhaʔ ‘kind of bark bowl used for emptying water’.

*ebuR sow, strew, scatter

Note:   With root *-buR₂ ‘strew, sow; sprinkle’.

*ebus finished, gone

Note:   With root *-bus ‘end, terminate, finish’.

*embut to pound, of the pulse

Note:   Also Old Javanese embun, Madurese embun ‘fontanel’. PWMP *embut is inferred by extraction from the reduplicated form.

*edaŋ light, radiance; shine

*endaŋ stop, come to a stop

Note:   Also Bikol udóŋ ‘stop, come to a stop’.

*edaʔ vetative; don't

Note:   Also Malay endah ‘no, not (Brunei, Sarawak)’, Javanese aja ‘don't!, may (it) not’, Buginese ennaʔ ‘no, not’. Bare'e ja ‘only, just, not then’, Makasarese -ja ‘enclitic with a limiting, and therefore sometimes exclusive, accentuated or contrastive character; only, but’ probably are distinct.

*edeg back (of humans or animals)

Note:   Also Casiguran Dumagat adég ‘the upper back portion of a person's body; turn your back on someone or something’, Singhi dug ‘back (anat.)’. If Singhi dug (expected **dog) is in fact cognate with the Philippine forms cited here *edeg must be assigned to PWMP. How its referent might have differed from that of *likuD remains unclear.

*edem₁ overcast; dull lustre

Note:   With root *-dem₁ ‘dark; overcast’. Old Javanese meḍem is cited under eḍem, but the available textual evidence does not rule out the possibility that it is the morphologically complex surface realization of underlying peḍem.

*edem₂ soak, dye

*eden to brood, sit on eggs

Note:   Also Bontok eleŋ ‘to rest, stop working, put down one's load for awhile’, Cebuano ugaŋ ‘mother hen’, Tiruray eran ‘(of a hen) to sit on eggs’, Balinese edem ‘rest, be at rest, sleep’.

*enduR thunder

*endut resilient, springing back when pressed down

Note:   Old Javanese eṇḍut ‘mud’, Balinese eṇḍut ‘mud, clay’, eṇḍut pasih ‘quicksand on a beach’, Sasak endut ‘muddy, slippery’ appear to be distinct.

*egaŋ dry, dried up

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

*egap gasp, open the mouth to breathe

Note:   Also Tagalog higáb ‘yawn’, Sundanese ɨŋap ‘feeling of suffocation, choking’, Sundanese eŋap ‘gasp for breath’, Old Javanese egep ‘gasp for breath’, Palauan chisb ‘hiccough’.

*eŋgaʔ no, not

Note:   Also Aborlan Tagbanwa ɨga ‘not’, Malay (Jakarta) eŋgaʔ, eŋgah ‘no, not’.

*eŋgem hold something inside the mouth

Note:   Also Toba Batak oŋom ‘hold something in the mouth without swallowing’. With root *-gem ‘grasp in the fist’.

*egik high-pitched throaty sound

Note:   With root *-gik ‘shrill throaty sound’. PWMP *e > Bikol u in the penult is exceptional, but not unparalleled (cf. e.g. Dyen 1953:fn. 67, and the Bikol reflex of *edeg, *Reken, etc. in this dictionary).

*e(ŋ)guk make a gurgling sound

Note:   With root *-guk ‘deep throaty sound’. Pending further information, I treat Manggarai NC- and eNC- as diachronically equivalent.

*ejaʔ vetative; don't

*e(ŋ)juŋ nose

Note:   Also Ilokano agóŋ ‘nose; the funnel-shaped entrance to a bow net’, Kakiduge:n Ilongot ɨgɨŋ ‘nose’, Iban (Lemanak) ɨdɨŋ ‘nose’, Ifugaw olóŋ ‘nose’. It is possible to reconstruct a Proto-Philippines disjunct *ajuŋ, but this possibility -- because it is only weakly attested -- is ignored here.

*ekab open, uncover

Note:   Also Tagalog hikáb ‘yawn’. With root *-kab ‘open, uncover’.

*e(ŋ)kak choke, gasp, struggle for breath

Note:   With root *-kak₁ ‘cackle, laugh loudly’. Possibly identical to *Sekak.

*ekaŋ hoot, croak

Note:   With root *-kaŋ₂ ‘bark, croak’.

*eŋkaŋ spread the legs

Note:   Also Balinese eŋgaŋ ‘stand apart like a pair of compasses, legs’. With root *-kaŋ₁ ‘spread apart, as the legs’.

*ekap gasp for air

Note:   Also Ilokano rikáb ‘gasp, respire convulsively’.

*eŋkaq breathe quickly or deeply

Note:   Also Old Javanese eŋkeh, eŋkuh ‘sigh’.

*ekeb cover

Note:   With root *-keb₁ ‘cover’.

*e(ŋ)kek subdued laughter; sobbing

Note:   With root *-kek ‘shriek, creak, cluck, chuckle’.

*eŋkel bent, stooped

Note:   Also Bare'e ogo ‘bowed by age or sickness’, oŋgo ‘bent, bowed, as someone in walking’. With root *-kel ‘bend, curl’.

*ekeŋ stiff

Note:   With root *-keŋ₁ ‘cramps, stiffening of limbs’.

*eŋkeŋ moan, hum, howl

Note:   With root *-keŋ₃ ‘hollow, resounding sound’.

*ekep brood, sit on eggs

Note:   Also Nias uxu ‘brood, hatch eggs’, Bolaang Mongondow oob ‘brood, sit on eggs’, Uma ooʔ ‘brood, sit on eggs’, Bare'e oʔo ‘brood, sit on eggs’. With root *-kep₁ ‘cover; fold over’.

*eŋker low guttural or rumbling sound; snore, roar

Note:   With probable root *-ker (cp. *-kur ‘coo; turtledove’).

*eket stick, adhere

Note:   With root *-keC ‘adhesive, sticky’.

*e(ŋ)kik shrill cry

Note:   With root *-kik ‘shrill throaty sound’.

*ekit squeak, shriek

Note:   Javanese eŋkèt probably is a loan from Malay.

*ekuk make a croaking sound

Note:   With root *-kuk₂ ‘sound of a sob, cackle, etc.’.

*ekup enclose, cover

Note:   With root *-kup ‘enclose, cover’.

*eŋkus snort, sniff; blow air through the nostrils

Note:   Also Sasak eŋkos ‘to spit, hiss (snake)’.

*eŋkut guttural sound

*elak separate two things

Note:   Also Fordata elat ‘frontier, border, line of division’.

*eleb knee

Note:   Also Kanakanabu ə'və ‘knee’. Probably *qeleb.

*elem₁ shade, darkness

Note:   With root *-lem₁ ‘dark; obscure’. To judge from the Bolaang Mongondow contrast of oḷom and olinow, and from the reconstruction of PMP *qalinu ‘shade, shadow’ next to *elem, this item probably referred to formless shade in contrast to the well-defined shadow of something. The similarity of the Manggarai word to the other two cited here may be a product of convergent evolution operating on a common monosyllabic root.

*elem₂ cautious, suspicious

*elet₁ borrow, be in debt

*elet₂ exert strength or effort

*eliŋ high-pitched sound

Note:   With root *-liŋ₁ ‘clear ringing sound’.

*eluŋ bend, curve

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

*ema₁ father's sister, possibly address term

Note:   Also Iban mak ‘mother’, Javanese emak ‘maidservant who takes care of the children; mother, mom’. The meaning of this form is problematic. If we simply conjoin the attested glosses it would seem to follow that *ema meant both ‘mother’ and ‘father’. Such a result, however, would be anomalous, since (1) the meanings ‘mother’ and ‘father’ are represented by PAn *ina and *ama respectively, and (2) reflexes of *ina apparently never mean ‘father’ and reflexes of *ama apparently never mean ‘mother’. If we posit *ina and *ama as referential terms, and *ema as vocative, we can eliminate the problem of synonymy raised in (1), since then *ina = ‘mother (ref.)’, *ema = ‘mother (voc.)’, *ama = ‘father (ref.)’, and *ema = ‘father (voc.)’. Nonetheless, we would still have no explanation for the unique occurrence of a single vocative form representing two address forms. Similarly, we cannot support an inference that the material cited here reflects *ama with oxytone stress, since the divergent semantic evolution of cognates assigned to *amá would remain problematic in any case.

Alternatively, we might assume that *ema was used in teknonyms, as reported for Tboli and Manggarai. Even so, the problem of a single teknonym meaning both ‘mother’ and ‘father’ would remain unsolved. The only parallel to the semantic range of *ema known to me is that of *aya ‘FZ’, which is reflected in some languages with the meaning ‘mother’ and in others with the meaning ‘father’, but which also appears in a few languages with the meaning ‘aunt’, or specifically, ‘FZ’. Perhaps *ema was a vocative/address form of *aya, just as *mama was a vocative/address form of PMP *(ma)tuqah ‘MB’. If so, however, the absence of any reflex which refers exclusively to female collaterals of the parental generation is puzzling, as is the use of the apparently redundant vocative marker -ʔ (for further discussion of vocative marking in PMP cf. Blust 1979).

*ema₂ kiss

*emeg damp, moist

Note:   Also Tagalog hímig ‘humidity’, Bikol damóg, dumóg ‘saturated, Cebuano humúd ‘wet’, Bolaang Mongondow omok ‘make wet, dampen’.

*emis sweet taste

*emit small in quantity, few; trifling amount

Note:   Also Ngadha emé ‘small, few, insignificant’.

*emun if

Note:   Also Ma'anyan, Kapuas, Siang amun ‘if’, "Proto-Malayo-Javanic" *lamun ‘provided it be’ (Nothofer 1975:97), Brunei Malay (l)amun ‘if, provided that’.

*emuŋ gather, collect together

*emuR hold in the mouth

Note:   Also Sundanese ɨmɨy ‘hold something in the mouth for awhile, as the saliva from the betel chew’ (< *emeR), Tae' ammo ‘put something in the mouth and conceal it by closing the lips’ (< *emuq). With root *-muR ‘gargle, rinse the mouth’. This comparison may be a product of chance.

*-en marker of direct passive

Note:   *-en is one of the most important affixes that can be reconstructed for PAn, PMP and PWMP. Like other affixes in the ‘focus system’ it evidently marked voice in verbal constructions, and at the same time served to create deverbal nominals. Both functions are widely attested, the latter often fossilized in reflexes of *kaen-en ‘be eaten/what is eaten = cooked rice’. The use of *-en to mark passive imperatives has not been noted in Formosan languages to date, but is widespread in the Philippines and western Indonesia. In some languages, as Bikol, there is a synchronic contrast between imperative marking with a reflex of *-en, and an alternative marking with a reflex of *-a. The former alternative occurs when a short (agent/possessor) pronoun immediately follows the suffixed verb, while the latter occurs without the pronoun. In other languages (as Kelabit) imperatives which use a reflex of *-en are felt to be ‘softer’ or more polite than imperatives which do not make use of passive verb morphology (in this connection cp. also Malay passive imperatives/dehortatives in di-, as with Jaŋan di-gaŋgu guru-mu! ‘Don't disturb your teacher!’).

Since at least two forms of imperative marking can be reconstructed for PWMP, it would be worth trying to determine whether a similar contrast existed in PAn. In addition to the above uses several languages in the Philippines and western Indonesia use a reflex of *-en to mark various types of usually visible bodily afflictions (skin diseases and the like), as well as comparable conditions of decay or deterioration in non-human subjects (e.g. *bubuk-en ‘crumbling, of something infested with weevils’). Finally, reflexes of *-en are used in some kin terms to mark collaterality, as with *anak ‘child, offspring’, but *anak-en ‘niece, nephew’, *ama ‘father’ but *ama-en ‘uncle’, *ina ‘mother’, but *ina-en ‘aunt’, etc.

*ena catch or get caught in a trap; hit mark with spear, etc.

Note:   Ifugaw (Batad) ona may reflect *kena.

*enap scale of a fish

*enem six

Note:   Also Paiwan unem ‘six’, Balinese eŋem ‘six’, Banggai noom ‘six’, KOM nemu ‘six’, Lamaholot neme ‘six’, Tetun neen ‘six’, Leti nema ‘six’, Selaru nem ‘six’, Yamdena nem ‘six’. I assume that the initial /h/ which appears in Itbayaten ha-ʔnem, Manobo (Western Bukidnon) he-ʔenem and Dampelas h-onoŋ is secondary, but it is possible that a contrast between *a-enem and *ha-enem was found in PMP (cf. Yamada 1991:124, where Itbayaten a-ʔnem signals ‘six’ with reference to things, while ha-ʔnem signals ‘six’ with reference to animals or humans). The morphological agreement of Paiwan ka-unem-an ‘a June typhoon (first of season)’ with Manobo (Western Bukidnon) k-enem-an ‘sixty’ is attributed to chance.

*enép keep to oneself

Note:   Also Balinese enek ‘be silent’.

*eñat stretch

Note:   Also Bintulu p-enat ‘stretch oneself’. With root *-ñat ‘stretch’.

*eñep sunken, submerged

Note:   Also Old Javanese (he)neb ‘that which has settled on the bottom, is precipitated’. With root *-ñep ‘dive, sink, disappear under water’.

*eŋap gasp for breath

Note:   With root *-ŋap ‘open, of the mouth; gaping’.

*eŋaq wide open (mouth)

Note:   Also Tagalog hiŋá ‘respiration’, Sangir éŋa ‘beginning stages of asthma’. With root *-ŋaq ‘gape, open the mouth wide’. Possibly a convergent development.

*eŋeŋ buzz, hum

Note:   With root *-ŋeŋ ‘buzz, hum’.

*eŋes sniffle, snore

*eŋik high-pitched sound of complaint

Note:   With root *-ŋik ‘shrill throaty sound’. Both Minangkabau and Tontemboan show unexplained lowering of *i, but the root *-ŋik (with *i) is otherwise well attested.

*eŋus sniffle, pant

Note:   Probably with root *-ŋus ‘snout’.

*eŋut complain loudly

Note:   Also Manobo (Western Bukidnon) eŋit ‘laugh aloud’. With root *-ŋut ‘mumble, murmur, mutter’.

*epak break, crack, split

Note:   With root *-pak₂ ‘break, crack, split’.

*epan start a fire

*epaŋ notch; notched

*epap clap, flap, slap

Note:   Also Bidayuh bab ‘flapping sound (as on a flat surface)’, bap ‘flop, dull sound’.

*empas fall away, crumble (as a wall)

*empeŋ block, obstruct, dam a stream

Note:   With root *-peŋ ‘plug up, dam; cover’. Dempwolff (1934-38) included Fijian bā ‘fish fence’, Samoan pā ‘fence’, but it now seems clear that these forms are not to be compared with reflexes of *empeŋ.

*epeŋ make a dull sound in the ears

*epid braid, intertwine

Note:   With root *-pid ‘braid, wind together’.

*epiq burnt food which adheres to the pot

*epit₁ burnt food which adheres to the pot

Note:   Also Cebuano húpit ‘have something sticking to it such that it is hard to get off (as feces on rump)’.

*epit₂ fibrous material at base of coconut frond; bag

*empu grandparent/grandchild (recipr.); ancestor; lord, master, owner