Volume 59 1950 > Volume 59, No. 2 > Some notes on Nggela grammar, by C. E. Fox, p 135-169
                                                                                             Previous | Next   

- 135
SOME NOTES ON NGGELA GRAMMAR

1. Nggela 1 is the central island of the Solomons and was called Florida by Mendaña, the Spanish discoverer. It consists of three main islands divided by narrow river-like straits. The most easterly is known as Little Nggela, Nggela pile; the large central one as Big Nggela, Nggela sule, and the westerly one as Mboko ni Mbeti. There is one common language but the letter dh of Nggela pile is pronounced h in Nggela sule, and z in Mboko ni Mbeti, though in printing h has been used throughout. There are some differences in vocabulary between the three dialects, but not more than twenty words (out of 20,000) are different in Nggela pile and Nggela sule. Mboko ni Mbeti has a larger number of words peculiar to it, about fifty in all. For example it uses voli, to buy or sell, for pelu; kausa, a garden, for legai, and toa, go, for tona. But the great majority of the words are the same in the three dialects, and they form one language, Nggela. It is spoken by about 5,000 people. The Ruavatu language of the coast of the large island Guadalcanar opposite Nggela is not very different. Nggela is also known and spoken by the people of Savo island (in addition to their own largely non-Melanesian speech) and by the people of Russell Island, whose native tongue is quite non-Melanesian; and it is understood by many on Santa Isabel and on the coast of Mala opposite Nggela. Probably it is known by some 10,000 people.

2. It is not intended in this paper to write a full grammar of Nggela. Bishop Patteson was the first to study the language. He called it Anudha, which is the name of an island close to Nggela sule, to which Patteson's informant probably supposed he was pointing. A short study of - 136 “Anudha” language was printed. This was used by H. T. von der Gabelentz in 1873 for his “Anudha” in Melanesischen Sprachen, p. 130-135, a vocabulary and grammar of Belaga-Gaeta (Nggela pile) and a study of “Anudha” pronouns and numerals was published by Mueller. But the first thorough study was made by Codrington in his Melanesian Languages in which he included a grammar of Florida. This was done with the aid of schoolboys at Norfolk Island, using Mota as a medium, for Codrington had little practical knowledge of Nggela. Like all Codrington's work it is of high excellence, but naturally limited and incomplete though correct as far as it goes. Gradually the New Testament was translated from Mota by Nggela teachers, and the Rev. John Penggone, a native of Nggela pile, translated the Pilgrims' Progress. The Prayer Book was also translated, and Reuben Bula, of Nggela pile, wrote some folk-tales for Codrington which were translated and published by S. H. Ray in 1897. In 1937 Dr. Ivens, using the material of the translations, published “A Grammar of the Florida Language,” 2 which is a re-writing of Codrington's with some fresh material. There are, however, many misprints in the translations which sometimes mislead Ivens, and the value of the New Testament translations is diminished by the fact that they were done from Mota, by Nggela teachers who spoke Mota and followed Mota idioms too literally; in Mota for example there is no way of writing passives except to turn them into active forms. In 1941 I published “Nggela for Beginners” 3 following Codrington but with some corrections and additions. In this paper the grammar is approached in a different manner, and only by way of some notes on word forms.

3. Nggela is an important language because of its central position in the Solomons; it has links both in grammar and vocabulary with the languages of the eastern Solomons such as Sa'a and Arosi and also with those of the western Solomons such as Roviana. If any Solomon island language should become a lingua franca for the whole group Nggela would have strong claims. Besides being known by more people than any other language of the group it is probably the easiest language for white people to learn and is a rich one. Government officials in the Solomons have - 137 never learnt a native language because there are so many, and so have never been in close and intimate touch with the people. Pidgin English is what they use even in law-courts, or else in the courts there are very indifferent interpreters. A much wiser course was followed in Fiji where one language, Bau, was used and spread, and is now called Fijian; leading to greater unity, and to a greater friendliness and understanding between Government and people. However the Solomons are under the High Commissioner for the Western Pacific who is in favour of basic English as a common language for the Solomons and not a native lingua franca; and when a grant was sought by the Resident Commissioner of the Solomons for the publication of the writer's Nggela Dictionary the reply was that the authorities were “not interested.”

4. There are traditions in Nggela of an earlier people who spoke a language different from the present one. The Nggela word for these people is Mumutambu or Mumulou; and the same name is used in other parts of the Solomons for wild men of the woods, or for what we should call fairies, but in Nggela they are thought of as a real people, but uncivilised compared with the present people. They had straight hair, hairy bodies, no arts or crafts. They tried to talk the language of the present people when the latter arrived, but they mispronounced the sounds. They generally used k for t, r for n, and t for s; they could not pronounce d, h, g or p and confused l and r. The following are given as some of their attempts at Nggela words.

English Nggela Mumutambu
man tinoni kirori
shore sapa tapa
  longa ronga
down horu roru
mine ganggua gakua
child nggari ngari
sleep maturu makuru
arrive poso koso
fall rupandu lukatu

5. The people of Nggela almost all live on the coast, and there are only two or three hill villages; but formerly there were a number of people living on the hills, and in these hill villages there were differences of speech. The hill people of the Nggela pile were called Komu levu and had the curious habit of infixing tn in Nggela words:

- 138
English Nggela Komulevu
man tinoni nitnoni
moon vula vutnula
one sakai saktnai
down horu hotnoru
mine ganggua gatnagutna
wake rarai ratnai
arrive poso posnoso
teach tarai tatnaratnai

6. In Nggela sule in the Boli district metathesis of Nggela words was once so firmly established as to make Boli talk unintelligible to the rest of Nggela. With the translation of the New Testament and Prayer Book into normal Nggela, uniformity has come and the metathesis has largely died out from constantly hearing the normal Nggela read; but it was once as prevalent as it now is in the spoken language of Kwara'ae on Mala (as contrasted with the printed language there). But some transposed forms are still retained in Boli in everyday speech. Such are:

English Nggela Boli
man tinoni nitoni
moon vula luva
heart lio iro
shore longa ngola
one sakai kasai
down horu rohu
foolish mbule lumbe
child nggari ranggi
sleep maturu ratumu
fall rupandu nduparu

7. In Nggela as a whole metathesis is only sporadic, but seems to be deep-seated.

For example in Tagalog (Philippines) the large reef crab is called alimango, and Lau (Malaita) has the same word for the large reef crab, alimango, but in Nggela the same crab is lingamo (na limgamo).

A Nggela word for ten is salage, though the ordinary word is hangavulu. Salage is ten also in Santa Isabel. Now the Lau (Malaita) word for a hundred is tangalau (Nggela, hangalatu), construct tangale and by metathesis talenga. Salage seems to be a form of this, with the meaning confused, and really another form of Nggela hangalatu, hundred, shortened and transposed.

Metathesis in Nggela is seen to be deep-seated by the presence of infixes, which as Brandstetter pointed out, are due to the transposition of syllables. There is in Nggela an infix ni, not recognised as such by the Nggela people. The modern article in Nggela is na, na tinoni, a man, but the article in some Solomon Island languages is ni as in Arosi, and this may have been the older form in Nggela for ni has - 139 become infixed and then the article na used with the word so formed. Ni is usually infixed to verbs to form nouns, just as the Achinese language of Sumatra ne is infixed to verbs to form nouns:

  • Achinese pegah, to say; penegah, a saying; seut, to reply; seneut, a reply.

So in Nggela:

  • hage, to embark; hanigela, a crew.
  • haro, gather food; hinaharo, garden food.
  • hola, to support; hinola, 4 a beam.
  • saga, to wither; sinaga, a burnt garden.
  • toga, to dwell; tinoga, a visit.
  • tete, to cross a stream; tenete, a bridge.
  • kova, to cross a hill; kinokova, hill garden (ni kokova).
  • kabu, to reside, dwell; kinakabu, kingdom.
  • vahuhu, bear a child; vinahuhu, relatives.
  • vano, to go; vinano, a journey.

Ni kakabu is still found in Boli for kingdom (kinakabu, elsewhere). A similar infix is found in Santa Isabel as for example in the word for teacher, tinarai, the Nggela word for teacher being tarai. 5

  • bangara, king; binangarana, kingdom.
  • mate, die; minate, death.
  • hena, take; hinena, booty.
  • toa, live; tinoa, life.
  • sara, scoop; sinara, handful.
  • tago, to have; tinago, possessions.

8. It is remarkable how deep is the instinct for metathesis in Melanesian languages. Dr. Brandstetter has been quoted as saying that in Indonesian languages also “metathesis turns up in those languages in all sorts of forms, either sporadically or in regular series.” He might have added “both in vocabulary and in grammar.” The same is true of Melanesian languages. Deck in his grammar of the Kwara'ae language of Mala gives examples of the metathesis which abounds on that language, but gives the reader no idea of its extent in spoken, as opposed to printed, Kwara'ae. If we call the printed form classical Kwara'ae, then the spoken Kwara'ae differs in fifty per cent of its words, which - 140 are transposed forms of the printed ones. When the printed form of Kwara'ae of the New Testament translation is read to a congregation of Kwara'ae people it is like reading Chaucerian English to a modern congregation of English people. In time from familiarity the printed form may prevail, as it has in Boli on Nggela. In spoken Kwara'ae not only is there much metathesis, but often final vowels and glottal stops disappear altogether. In speech the word printed oli is pronounced oil; buli, buil; luma, luam; leka, leak; siramu, siraom; taifiliaku, taifilauk; tala ana, talan; sasi akau, saisako; and so on.

The following is an example of printed Kwara'ae followed by the same as it is spoken.

  • 1. Ta'ena ofodangi nau ku fanga ka sui nau ku leka karangia kafo suungia na talafo alo; nau ku leka fainia saena luma.
  • 2. Taen ofdaing nauk fang ka sui nauk leak kareng kaf sungia talafalo; nauk leak fainia sean luam.

The same sort of thing is found in the language of Rowa in the Banks Islands, where though the language does not sound to the ear like Mota it is really Mota transposed, metathesis appearing in a majority of the words, e.g.:

  Mota Rowa
tooth liwo liew
firewood lito liet
shine singa sieng
down siwo siew

In my book, The Threshold of the Pacific, I gave examples of the language of Marau wawa, a small island off the coast of the Bauro district of San Cristoval. Some words of this language appear to be non-Melanesian, but a large number seem to be Bauro words transposed, but with the article of the Bauro word not recognised as an article and taken to be part of the word.

  Bauro Marau wawa
come here siri mai miri sai
hungry hioro rioho
yam na ufi fauni
water na wai wanai

9. In Polynesia the case seems to be different from either Indonesia or Melanesia as regards regular or grammatic metathesis, though Pratt in his Samoan Grammar - 141 remarks that “the Samoans transpose consonants almost unnoticed by their hearers, saying namu for manu and so on.” I do not know whether since the language has been printed this metathesis has been lost in Samoa. Samoan is the Polynesian language closest to Melanesian languages in several respects. In the Solomons the children often transpose words to make what they fondly believe to be a secret language of their own. European children sometimes do the same, though not to anything like the same extent. There must be some explanation of this general inclination to transpose syllables which is so characteristic of Indonesian and Melanesian languages, and is found at least sporadically in Polynesian languages. 6

A common explanation is that it is simply a natural mental carelessness and indolence. This seems to be Sweet's view, but hardly seems to account for the facts, e.g., grammatical metathesis.

Another explanation is based on the original character of Austronesian speech. There seems to be evidence that in the original stages of Austronesian speech formative mono-syllablic particles modifying the word-bases could be either prefixed or suffixed, and only at a later stage in each language became, by custom, prefixes or suffixes. From a root lumu, soft, you could with the formative ta make either talumu or lumuta, something soft, as moss. But this explanation also hardly appears sufficient.

A third explanation is that when Austronesian languages were introduced among alien people they, in their endeavours to pronounce the words, transposed some of the syllables; as the Nggela people say the former people of the Solomons did. In that case where there is much metathesis as in Indonesia and Melanesia it arose from the introduction of an Austronesian language among peoples to whom it was a foreign speech. Where there is little metathesis as in Polynesia it is because the Polynesians are the original speakers of Austronesian, and it was their speech that gave us the modern languages of Indonesia and Melanesia, modified by contact with the older languages spoken there.

- 142

But the reader of these notes may be able to suggest a simpler explanation than any of these. Metathesis, “either sporadically or in regular series,” as Brandstetter puts it, is a fact in these languages, which must have an explanation.

10. Nggela morphology. Nggela words consist of four sounds of which the second and fourth are vowels, and the first and third consonants. No Nggela word ends in a consonant: bosa, to speak.

Such forms are the word-bases of the language but they have been modified in various ways, and they have also decayed in various ways. They are modified by the use of formatives, monosyllabic particles, consisting of a consonant followed by a vowel, which usually have no meaning in themselves. These can be either prefixed or suffixed to the word-bases, and occasionally infixed to alter or limit their meaning, and are sometimes grouped together, two, three or four of them, to make compound prefixes and suffixes.

The following are examples of Nggela formatives:

  • (1) ka, a prefix to a verb making a past participle passive; luba, to loose; kaluba, loosened.
  • (2) vi, a suffix showing that the action of the verb is directed on a particular object (making a verb transitive) and often taking the place of an English preposition; hora, to spread; horavi to spread over (something).
  • (3) lagini, a group of three formatives, making a compound suffix and giving a causative sense to a verb; kata, to stick; katalagini, to cause to stick.

In Nggela the accent is always on the penultimate syllable so that the addition of a suffix throws the accent forward; bósa, bosáva (speech); mána, manáha.

The word-bases also decay in various ways. The two most common are:

  • (1) The loss of the first consonant of the word-base; ambe, to support; tambe, to hold; asi, wild (Arosi, wasi, wild).
  • (2) The loss of the second consonant tona, to go, move quickly (Lau, tona, to start); Mboko ni Mbeti, toa, go (Mota, toa, run).
- 143
  • (3) Either of the two syllables of the word-base may be lost, but this is rare in Nggela; mā, to be ashamed (Arosi, masa); either vake or he, to give (the va of vake may be a prefix, cf., however, Sa'a, Arosi, wate, to give).
  • (4) Both consonants may be lost, though this is rare in Nggela; ao and vao both mean virgin forest.
  • (5) A good example of the decay of a word in Nggela is the word sakai, one (Arosi, ta'ai) from which are derived four other forms with the same meaning: sekei, sikei, siki, si and ki; all five forms are in use.

11. Reduplication. Besides being modified by the addition of prefixes or suffixes or infixes, or by the decay of the word through the loss or change of one or more of the four sounds of the word-base, there is another form of change of the word-base which plays a great part in Nggela, reduplication, which is of several kinds.

  • (1) The most usual form of reduplication in a word-base of four sounds is to reduplicate the first two and the last. This method is almost always used when the vowels of the word-base form a dipthong or semi-dipthong: sari, saisari; tore, toetore.
  • (2) But if the two vowels of the word-base are the same only the first two sounds are reduplicated. Ivens supposed the vowel should be lengthened, and in revising in England the Nggela New Testament had the Nggela words so printed; tootobo, tiitili and so on, but actually this is not the case: tombo, totombo; tili, titili.
  • (3) Sometimes even when the vowels are different only the first two sounds are reduplicated. This is not incorrect and in the translations often occurs, but is only rarely heard in spoken Nggela: toka, totoka.
  • (4) Both methods may be followed with the same word-base with a slight difference in meaning. Thus from kolu, to be with, there are two reduplicated forms: koukolu, the usual one, and kokolu, which is said to mean cohabit, a special limited meaning.
  • (5) The whole word may be reduplicated: keha, different, foreign; kehakeha.
- 144
  • (6) If there is a prefix to the word-base the reduplication affects only the first four sounds: ta-vike, taitavike; ma-tagu, mamatagu; tondore, tondo-tondore.
  • (7) With decayed word-bases of three sounds, either the whole word is reduplicated, or the first two sounds: tao, taotao; liu, liliu; koe, kokoe; tioko, tiotioko.

But much the most common form of reduplication in spoken Nggela is the one mentioned first: logu, loulogu, except when the two vowels are the same.

12. Single sounds in Nggela. The ones to be considered are a, e, i, o, u, k, ng, m, p, t.

a. 1. A pronoun, when it is the object of a transitive verb, him, her, it. This is modified to e when the pronoun is the subject of the verb: tu ndolovi-a, I love him, her it; e ndolo, he loves.

At the end of a Nggela sentence the forms u a, o a, e a and so on are used, meaning “so I say” (u a), “so you say,” “so he says,” and so on. In these expressions the a may stand for the pronoun it, or may represent some other decayed word: te mate tua, e a, he died, so he says; e uto le, ra a, excellent, so they say.

2. A personal article, which is used with nouns which are the names of persons, or with pronouns, but not with nouns which are the names of places, which take na or i: a Piluka; a John; a goe, thou; a hei, who; a hanu, 7 so and so. This personal article a can be used with all the personal pronouns except those of the first person when an older personal article i only is used, i nau, I; i gita, we. I is the personal article in the eastern Solomons and in the Banks Islands, and a appears to be later, because both can be used together and then a precedes i: a goe or ai goe, thou. I can be used in the second person, i goe, i gamu, but a only in the third person: a nggaia, a nggaira, he they; however, in these last forms the i is probably the older personal article.

The personal article a can be used with relationship terms: a tamana, his father.

It can also be used to personify, when the word it precedes really becomes a name, and would have a capital letter - 145 in English: a Kiko, the Deceiver (kiko, to deceive); na kiko, a deceiver.

The personal articles can be separated from their noun by the negative mua, not: a mua Piluka, not Piluka; i mua nau, not I.

3. a is a prefix to many words with no apparent function; this may possibly be a case of euphony: one, sakai, asakai; his, nina, anina; then, ge, age.

4. a is a suffix forming adjectives and sometimes nouns from verbs: vonu, to fill, vonua, full (vonugia, filled); maemane, to be straight, maemanea, straight; voro, fold a leaf, na voroa, drinking cup.

5. a is a preposition, or according to Codrington and Ivens a gerundive suffix to verbs; both Codrington and Ivens class it with the suffix a which makes verbal nouns; but the question arises whether this a must not be classed as a preposition, as in the languages of Mala and San Cristoval, always followed by the personal pronouns nggu, mu, na. Codrington and Ivens always call these “possessive” pronouns, and it is true that they are often the equivalent of the English possessives my, thy, his, etc., nevertheless they are in reality personal pronouns and can be used sometimes in Solomon Island languages a subject or object of a verb or can follow prepositions, e.g., muri-mu, after you.

Of the preposition a in the Mala language of Kwara'ae (and the same holds good in other Mala languages and in the languages of Arosi, Bauro and Santa Ana) Deck writes: “It means ‘of, from, out of,’ pidgin English along, and expresses genitive relation and association.” “This preposition,” he adds, “is also used with the displaced object of a transitive verb when an adverb comes between it and the verb stem,” i.e., after a compound verb. This use seems to bring the Nggela a in line with the Kwara'ae a, and to form an important link between thelanguages of the central and eastern Solomons.

It will be useful to give examples of the use of this a in Nggela: lavi-mbule, to tease (vb. and adv.); te lavi-mbule-a-mu, he teases you, or na lavi-mbule-a-mu.

In the first, te is the verbal article, and in the second, na is the noun article, and both Codrington and Ivens consider a to be a suffix to the verb, making it a gerundive, and - 146 they write na lavi-mbulea-mu. If, however, this a is a preposition, as in Mala and San Cristoval, the word should be written na lavi-mbule-amu, “the teasing along you,” and this is how Nggela people actually do write it in their letters.

It will be useful to give some further examples, all with the use of the noun article na, which is the usual form of such words in Nggela. It is only occasionally that the verbal article te is used, though it is not incorrect. Codrington and Ivens seem to be unaware of its use: lambu-langga, to hit hard (vb. and adv.); na lambu-langga ana, he was hit hard (“the hitting hard along him”); mbosa-vani, to speak to; na mbosa-vani anggu, I was spoken to; mbosa-ndika, to speak evil; na mbosa-ndika ana, he was evil spoken of; tuguru, to stand; tuguru-vagi, to cause to stand, raise; na tuguru-vagi ana, it was raised; na tuguru-vagi oli ana, he was raised again; te tuguru-vagi oli ana, is also a possible form of the last.

Codrington and Ivens consider all the above forms to be gerundives, and take the a of ana with the verb, calling it a gerundive suffix, and the na a possessive pronoun.

If in such phrases the verb is intransitive the sense is never passive: toga, to dwell; na toga ana, he dwelt.

If, however, the verb is transitive, but has no special transitive suffix, the meaning may be either active or passive, according to the context: lambu, to kill; na lambu ana, he killed or he was killed.

If the Nggela verb has its proper transitive suffix, always used when there is an object following, the meaning will vary according to the use or not of this suffix: hanga, to open; hañga-vi, to open something (vi trans. suffix); na hanga ana, it opened; na hanga-vi ana, it was opened. These uses are very common in spoken Nggela, but much rarer in the translations, because in these, Mota idiom has been followed, and Mota has no way of giving a passive turn. They occur, however, chiefly in subordinate sentences, and must often be translated by an infinitive in English: tu tamboa na goni ana, I tried to do it; tu rigia na hage ta na vaka ana, I saw him embark on the ship. Codrington and Ivens take na hage ta na vaka to be a compound verb, and the gerundive to be na hage-ta-na-vaka-a: anggaia te mua tangomana na lambu amu, he was not able to hit you. Such forms also occur after prepositions, nouns and adjectives: - 147 I murina na sumba ana na legai, after the garden was burnt (sumba); na vaugilala na mai ana, the sign of his coming; na pukuna na lambu-tambo ana, the reason he was murdered; te sule na kutu ana, great was its fall. When such forms occur in Mala languages the ending of the verb is la, which Ivens takes to be a gerundive suffix. Nevertheless, though it is la, not a, and distinct from the a referred to by Deck, it may be the same in origin and a preposition, not a gerundive suffix. The Mala forms are laku, lamu, lana, where Nggela has anggu, amu, ana: Lau (Mala) sau, to kill; na sau lana, he killed; na saungi lana, he was killed (ngi is the transitive suffix). Ivens would write these na saula-na, na saungila-na, and call them gerundives, but they cannot be used without the pronouns. Similar forms in Nggela are used with abstract nouns: nggari, a boy; na nggari ana, his boyhood; na nggari-mani ana, his youth; sule, big; na sule ana, its size. So in Kwara'ae: doe, big; na doe lana, its size. And in Mota (Banks Island): tole, long; na tole ana, its length. All such forms Codrington and Ivens class as gerundives with possessive pronouns. However, they must always be followed by pronouns and cannot stand alone, whereas the true abstract nouns can be used without pronouns and stand alone; and in Nggela are formed differently, by suffixing the formative ga: nggari, boy; nggariga, boyhood; sule, big; sulega, bigness, size.

The question is whether the la forms of Mala, which are identical with ra forms in Santa Isabel 8 should (along with the Nggela a forms) be identical with the preposition a of Mala and San Cristoval languages, or whether la (Mala) ra (Santa Isabel) and a (Nggela) are verbal suffixes forming gerundives.

6. a is a suffix added to verbs with a transitive ending to form a past participle passive: ndolo, to love; ndolo-vi, to love (someone or something); ndolo-via, loved; aho, to tie, fix (a hook); aho-ri, (trans. ending); aho-ria, tied, fixed; na halili te ahoria tua, the hook is fixed.

Where the verb has no transitive suffix the a is suffixed to the word-base: ambe, to support; ambea, supported; ngiti, to break; ngitia, broken. Sometimes the suffix is not - 148 used and the past participle is merely the verb with its transitive ending.

In Nggela there is a considerable number of past participles passive formed in this way, and a passive is sometimes, though rarely, so expressed. The transitive suffix does not always end in i: dura, to destroy, break up; dura-ke (trans. ending); dura-kea, destroyed. This method of forming a past participle passive, which tends to lead to a regular method of forming passives, is more common in Arosi and Lau, and occurs in Santa Isabel also. It consists of the word-base and trans. suffix and a. It seems probable that the a represents an original na and is probably a demonstrative pronoun. Na is used in this way to show completed action in Santa Isabel, and quite regularly in Lau and Arosi: Lau sui, to finish, sui na, finished; Arosi rei, to see, reia na, seen (it has been seen). Such past participles in Nggela have the endings gia, hia, kia, mia, ria, sia, tia, via. By comparing Malagasy passives in which the endings are fina, rina, tina, fana, etc., and the Maori, with such endings as ria, tia, ina, na, it would appear that both the Malagasy and Maori possessives were formed in the same way as the Nggela, i.e., by a word-base and a trans. suffix and a demonstrative na, or a; but the intermediate stage of adding a transitive suffix to a verb has disappeared in those languages.

e. 1. e is the numeral article with numbers from two to ten, and also with ngiha, how many: e tolu, three; e ngiha, how many.

2. e is the short form of the pronoun 3rd pers. he, she, it, when this is the subject of a verb. When the pronoun is the object of a verb the form a is used, of which e appears to be a decayed form. The full form of this pronoun is anggaia which is a compound of a personal article ngga a demonstrative, i probably a fossil personal article (since it is used as such with other pronouns i nau, igoe, I thou, etc.) and a the pronoun. The short forms of the pronouns when they are the subject of a verb are therefore u, o, e and u, o, a when they are the object: u lutu, I work; o lutu, you work; e lutu, he works. Thus the verb in Nggela can be used without a verbal article, as it is in Mala languages, another grammatical link with Mala and San Cristoval languages. (cf., a'ua, how, Arosi, with e gua, how, Nggela; the a in Arosi is - 149 the pronoun). Codrington and Ivens, however, consider the e to be a verbal article (or particle as they call it) and u and o only to be pronouns. Cf., later under t.

i. 1. i is no longer a personal article in Nggela with the names of persons, as it is for example in Mota, but it appears as a personal article with all the personal pronouns except in the third person, and is always used exclusively in the first person: i nau, I; i gita, we (inclus.); i gami, we (exclus.). It can be dropped and a substituted: a goe for i goe; or combined with i: ai goe; and it can be separated from the pronoun by the negative mua, not; i mua nau, not I.

2. i is an article with words of place or time, which fall into the same category in Nggela, e.g., hau means far off, either in place or time. With words of place or time i replaces the usual noun article na: i vei, where; i ngiha, when; i ani, here; i ngga, there; i ndania, long ago; i lau, at the shore; i pari, on the ground, outside, on the low land; i Belaga, at Belaga (na Belaga, Belaga); i voroa, the place of shellfish. It is also used with prepositions, including ta, the preposition of general relationship, though with the latter the i may be omitted: i sara, under; i loka, within; i muri, behind; i ta na vale, in the house; i ta mua, with you. The above words are usually written as ivei, iani, iloka, etc., in printed Nggela.

3. i is a preposition, meaning belonging to, of, for, the same in meaning as the preposition ni, from which it is probably derived. When it follows a noun ending in a the a usually becomes e by umlaut. It is written as part of the word it follows: malei gime, a bed (mala, place); vatei sopou, a chair (vata, sort, kind); matei gambu, price of blood.

4. i appears in some words in Nggela as a formative prefix though it is not as common as in some Melanesian languages, above all in Fijian. It is probably the same as i 3. Nggela people do not recognise the i as a prefix and write it with the word that follows: ikaru, a baler, karu, to bale; ihola, a support, hola, to support; imboro, a lever = gai ni boro; indale, a stone for cracking nuts; igaho, a digging stick; igula, sandpaper (gula, a tree with rough leaves); ikona, a crook; isile, an incision; cf., isele, a knife, probably an introduced Fijian word.

- 150

5. i is a transitive suffix to some verbs: kapu, to grasp; kapui, to grasp (something).

6. i is a personal pronoun, 3rd person plural neuter, taking the place of ra used for persons: mbosa, to speak, speak of; mbosaa, speak about it; mbosara, speak about them (persons); mbosai, speak about them (things).

i is used after verbs ending in a, e, o, u, but after verbs ending in i its place is taken by gi: rigi, see; rigigi, see them. In Mala langauages gi is added to nouns to make a plural; and in Arosi 'i is a pronoun neuter plural, and is suffixed to pronouns. Lau: beu gi, houses; Arosi: au omesi'i, I saw them (things); ana, his thing, ana'i, his things.

7. The same i is always added to some adjectives and adverbs in the plural; probably correctly to all adjectives when they make a predicate: uto, good, pl. utoi; taga, lost, pl. tagai; le (or lea), merely, pl. leai; kiki (or kikia) small, pl. kikigi; e rua na vatu te paparai, two stones are hot (papara, hot); te mukui, they spring up (muku, spring up).

Nggela uses the sing. pronoun as subject with neuter plurals.

o. This is the short form of the pronoun, 2nd pers. singular when it is the subject, but the form is go as the object: o mbosaa, you said it; te rigigo, he sees you.

u. This is the short form of the pronoun, 1st pers. singular, whether as subject or object. When these short forms are used as the object they must be considered as suffixes since the accent is altered: rígi, see; rigíu, rigígo, rigía, but as subject these forms are written separate: u rigia.

t and k. It is convenient to take these together because they are the two verbal particles in Nggela, that is to say, the two living ones. They always precede a word which is a verb (except when the short pronouns are used) and mark it out as such, just as the noun particle precedes a word which is a noun and marks it as such. Codrington and Ivens call such forms “verbal particles.” Codrington says that any word used with them is a verb, and gives their form as te, ke; he adds that they change with person and numbers by the taking into the particle of the vowel of the pronoun. Ivens criticises this, saying they are “not merely verbal particles.” But both Ivens and Codrington consider e a verbal particle, and not a pronoun, and Ivens adds that e is “not used with - 151 a subject but impersonally.” On my view this e is a pronoun, not a verbal particle, and is the subject of such sentences.

The view taken here is that the original form of the verbal article was ta, ka, the form still retained with the pronoun 3rd pers. plural ra: ta ra, ka ra; but that when used with the short forms of the pronouns the a is elided, and we get the forms, in the singular, tu, to, te, and ku, ko, ke, and in the plural the forms ta, tai, tau, and ka, kai, kau, i.e., t or k prefixed to the short forms.

Note that in Nggela the verbal article does not immediately precede its verb, but the short form of the pronoun comes between article and verb, tu lutu, I work. In Mota the order is reversed, the pronoun being always separated from the verb by the verabl article: nau we ilo, I see. However, the longer form of the pronoun in Nggela does precede the article, only then the shorter form of the pronoun must always follow it: inau tu lutu, I work; anggaira tara lutu, they work.

The verbal article t is used with verbs in the present or past tense, and k with those in the future tense.

ngg. Like t and k this form varies, either by writing with the short forms of pronouns, or by changing its vowel to agree with the vowel in the word which follows it. Its original form was probably ngga, and in origin it appears to be a demonstrative, and allied to the demonstrative ka which appears at the end of sentences in Nggela. It resembles the pronoun 3rd pers. sing. (a or e) in having two forms, ngga and ngge, the second a decayed form of the first. It is usually translated: then, if, in order that, at the beginning of a sentence, and the form is then ngge; at the end of a sentence the form is ngga.

There is some inconsistency in its use. The usual forms with the short pronouns are ngge u, nggu tu; ngge o, nggo to; or nggu, nggo alone; in the third person seldom ngge te, almost always ngge alone; in the plural ngga, nggai, nggau, nggara, like the verbal article: tu liona ngge mai, I wanted him to come; te nea ngge ko rigia, he did it that you might see it.

This formative or demonstrative also occurs in ngge a, parallel to forms u a, o a, e a, previously mentioned. Ngge a means “then” at the beginning of a sentence; “perhaps” at the beginning or end of a sentence, when it is usually - 152 combined with the word sō, still: sō ngge a; and at the end of a sentence the meaning is: “in that matter,” “on account of that,” “therefore,” or “thereby”: ko mbei rutu ngge a, don't be angry on that account; ko mbei goi taho nggea, don't therefore fail again; ngge a puputu, and then it appeared. Ngge often precedes the conjunction ma: ngge mo 9 to ko rigia, and if you see it. E nggea and sō nggea are both used for “perhaps,” “possibly.” Ngge also occurs in ngge gua, by and by; ngge ni, today (past time) and ngge vā, today (future time).

I ngga translates the English “there is, there are”: na vavuru i ngga? is there any tobacco; i ngga, there is. Nggea, nggegua, nggeni, nggevā, enggea, sōnggea, and ingga are all usually written as single words.

m. This, like t and k, is followed by various vowels, but its original form is evidently ma; it is a connective, but far more widely used than any English conjunction, and often used where we have no conjunction in English. It often begins independent sentences. It drops the a of ma before the short pronouns, and changes its vowel to conform to that of the following syllable, but at the beginning of a sentence it is usually ma: mu tu rigia, mo te rigia, me te rigia, etc. So also mi tatana, with him; mi ani, and there; mi nau, and I; mi goe, and you, etc.

p. This also is a connective, usually translated “but.” Its proper form is pa, but it changes the vowel just as m does: anggaia pi goe, he or you.

13. We come now to particles which consist of two sounds. The following occur in Nggela: mba, nda, ga, ngga, ha, ka, la, ma, na, nga, pa, ra, sa, ta, va; ngge, ke, me, te; ndi, gi, nggi, hi, ki, li, mi, ngi, pi, ri, si, ti, vi; nggo, ko, mo, ro, to; mu, tu. Some of these have been mentioned and most of them need not be discussed; some are prefixes or suffixes or both and are of great importance in the building up of the words of the language. They often come together and make compound forms, e.g., lavagini is a compound of four of these formatives, and is a verb suffix. Many of these formatives have more than one function in word-building.

Those referred to here are ga, ka, ma, na, ta, va, ni, si, ro; all of which are important.

- 153

ga. 1. Ga makes abstract nouns from verbs and adjectives. Since it is suffixed the accent is thrown forward: ndólo, to love, ndológa, love; tambu, holy, tambuga, holiness; vola, to live, volaga, existence; ganagana, to think, ganaga, thought; kula, a friend, kulaga, friendship.

2. Ga prefixed to pronouns makes possessive pronouns. It is prefixed to the forms nggu, mu, na, etc., but a is suffixed to the first two, so that the forms are: my, ganggua; thy, gamua; his, gana; our, ganda (inclus.), gamami (exclus.); your, gamiu; their, gandia or gandira. The nouns with which these possessives are used are the names of things to eat or drink, and some other nouns, less than fifty in all. Some of these are: kema, clan; komukolu, neighbour; tako, shield; kana, enemy; mate, death; ndolo, love; undu, friendship; hohoga, property; mbore, armlet; una, earring; susumalagaura, frigate bird tattoo (a Mala word, the frigate bird is ndaula not gaura in Nggela); mbulao, shell ornament; tiatia, wristlet; huguru, clump of coconuts; tindalo, ghost; and the names of trees whose fruit is prized such as mbeta, breadfruit; ngali, almond; kola, mango; gaviga, Malay apple and mbua, betel palm.

3. Ga is suffixed to any common noun to make an adjective giving the sense of having, or possessing, or being full of the thing signified by the noun. Some of these are in common use, but fresh ones can be made by anyone; it is a living prefix. It can also be suffixed to some other parts of speech: vale, a house, valega, owning houses; rongo, money, rongoga, possessing money; ndolo, love, ndologa, loving; gambu, blood, gambuga, bloody; gavu, mist, gavuga, misty; horu, descend, horuga, deep; vanga, fruit; vangaga, fruitful.

4. Ga when suffixed to a reduplicated verb makes a past participle passive. This is also an Arosi use: geli, to dig, geligeliga, dug; lutu, to work, lutulutuga, worked, as a garden.

5. Ga is sometimes, though rarely, a transitive suffix to a verb: tara, to twist; taraga, to twist on (to something).

6. Ga is added to numerals to give the meaning so many fold: rua, two; ruaga, twofold

7. Ga is a prefix to some words, but with no apparent function, a fossil not a living prefix: ga-laga, up, east; ga-lau, to cross over; ga-lago, warped; ga-lolo, to roll fibre on the thigh; ga-gua, of what sort. Nggela has two forms, - 154 galau and galao, to cross over; and the second seems to connect it with lao, lago, to go. Both take the trans. suffix mi; galaumi or galaomi, to cross over something. In Dr. Ivens' Bugotu Dictionary gatha-umi should be gathau-mi, to cross over (trans.). Ivens compares it with Nggela gala-omi! The Mota form is lagau, a case of metathesis which disguises its origin.

ka. 1. Ka is a prefix which makes past participles passive and so forms passives. Dr. Ivens could find only one example in the texts (kalumba, from lumba 10) when writing his Grammar of Florida. He found only one example also in Bugotu and Vaturanga (Santa Isabel and Guadalcanar), while Dr. Codrington says that this prefix is only found in Fijian. 11 Since the prefix is fairly common in Nggela this is surprising and shows the difference between translations (of Mota) and the spoken language. It is given in the Fijian Dictionary as a prefix which makes active verbs passive, and this describes its function in Nggela: mbihu, to twist, kambihu, twisted; mbilu, to pluck, kambilu, plucked; mboha, to break, kamboha, broken; hui, to ransom, kahui, ransomed; kovo, to bend, kakovo, bent; mindi, to split, kamindi, split; rango, to wither, karango, withered; mboki, to overturn, kambokili, overturned (li, trans, suffix); te kambokili tua na vatu, the stone was rolled aside. Ka resembles the prefixes ma and ta in forming passives, but these have other functions.

2. Ka is a demonstrative following nouns or verbs. It is rightly written separately as its presence does not affect the accent of the word which precedes it. With verbs its function is to make a past tense. It combines with other demonstratives ke, ri, to form compounds kerika kakeri: i mbongi ka, on that night; August 8 ka; tu rigia ka, I saw him.

ma. 1. Ma has already been described as a connective.

2. Ma is sometimes, though rarely, a transitive suffix to verbs: sago, to scrape, sagoma (trans.); te sagomā na luhuna, he scraped his chest.

3. Ma is a fossil verbal article. It is no longer a living one like t and k (ta, ka) because it can no longer be used to mark a verb with introduced words, as t and k can be. - 155 t and k are used with introduced English words like kuki, to boil, te kukia, he boiled it, or te turati, it is stale (“too late”); but ma cannot be so used. Ma is always written as part of the word, and sometimes the word-base to which it is prefixed is no longer found in Nggela, though it may occur in neighbouring languages. The Nggela word for “living” is mauri but uri is not found in Nggela, though it occurs in the Areare language of Mala. Ma, however, often retains its ancient function which is simply to mark a word as a verb, i.e., it is a verbal article. In Arosi more than 800 words have the prefix ma and in Nggela there is also a large number. It is a characteristic mark of an Austronesian language.

The following are examples of ma prefixed to word-bases which, so far as I know, do not otherwise occur in Nggela: mauri, to live; maturu, to sleep; marama, to shine; magogo, to tremble; maolo, to lean back; matagu, to fear; madeve, 12 thin.

Usually the word-base does occur, as well as the ma form, and there is no difference in meaning between the two, ma merely marking the word as an active verb: mbūmbū, to stare, mambūmbū; ngoli, to come to nought, mangoli; sua, to come out, masua.

Sometimes the ma form is both an active verb and can also be used as a past participle: mautu, to cease, or ceased; marutu, to crush, or crushed.

Often, however, the addition of ma makes a past participle passive, and passives can be so formed. In such cases ma has the same function as the prefixes ka and ta: geli, to dig, mageli, dug; mbita, to crush, mambita, crushed; posa, to burst, maposa, burst; rosi, to tear, marosi, torn; kolu, to gather, makolu, gathered together.

Ma very frequently marks a condition, state or process, and for this reason Codrington called it “the conditional ma,” though this is only one of its functions. In such cases the word to which it is prefixed is often reduplicated: masake, to be joyful; masua, to be slipping; maluluka, to be loose (rather than “loosed”); malovolovo, to be tottering; matanga, 13 to be in the middle, halfway. The last word is - 156 used of the low hills between the sea and the mountains, used for gardens.

Ma forms many adjectives: malumu, soft; malaho, cold, malaria; manilu, sweet; mandolu, round; magora, pure.

Ma also forms nouns and adverbs: magutu, master; malagai, strong man; marara, a flower; maramana, world of light; mandolo, right hand (also “habitual”); malili, sideways; mangoli, vainly; malobu, light rain (kolobu, tears).

The vowel of ma may alter to, o or u: madiri, stale; modiri, stale; mudiri, stale.

The initial consonant of ma sometimes replaces the initial consonant of the word-base: maro, surprised (haro, 14 to be surprised); manga, mouth, opening (hanga, to open); mava, heavy (Lau, bafa, to be heavy).

Ma may be joined to other formatives to form a compound, such as mala or manga. Ivens in his Grammar of Florida gives only one example of ma in manggoti, broken, but it is a question whether this is not really mang-koti; like manggili, squinting, from kili, to move from side to side. These may be examples of a compound formative mang, rather than of ma.

na. 1. Na is the article with all common nouns, 15 which in Nggela, though not in English, includes the names of islands and villages: na vale, a house; na Nggela, na Vatupura, Nggela, Vatupura (village): It can be translated as the or a, and is used before nouns both in the singular and plural. It marks a word as a noun,15 just as t and k mark it as a verb, and therefore it seems right to call these articles, nominal and verbal, because they differ from other formatives in that they define a word as a particular part of speech. In na lutu, work, lutu is a noun; in te lutu, he works, lutu is a verb. All other formatives have more limited functions, such as marking a verb as an active or transitive verb; showing relation between two nouns, and so on. Na may be a demonstrative in orgin.

- 157

2. Na means of or belonging to. It may be short form of ana, already discussed, or related to Mota, ta, belonging to. Ivens says that S. H. Ray connects this na with the ligative article na in Indonesian languages: na Nggela, a man of Nggela.

3. Na is a transitive suffix to some verbs: vetena, to command (Mota, vet, command); te vetenaa na nggari, he ordered the boy.

4. Na is a suffix to verbs, its function being to make the verb a noun, the agent of an action: kiko, to lie; kikóna, a liar.

5. Na is an adjectival suffix, in rare use: meto, dirt; metóna, dirty.

6. Ivens gives na as a demonstrative like ka, used to denote a preterite. This would be in keeping with its use in other languages, but is unknown to the writer, and the two examples Ivens gives can be translated otherwise: te hage iloka na, te taho nina langga na. In the first the na is the suffixed pronoun, he goes inside, literally “inside of it,” the usual use. In the second na is also the pronoun nina, meaning “for his part; “there was nothing as for him his strength,” i.e., he had no strength. The examples given do not seem to establish this use of na, since Ivens does not mark where the accent falls.

7. Na is rarely found as a suffix to verbs making a past participle, but with this function it is purely a fossil, not a living formative as in Arosi or Lau where it is living and can be used with introduced English words, such as kuki, to boil, kukina, boiled: nggere, to make lines, nggere-nggerena, striped; tao, to follow, taona, will o' the wisp, ignis fatuus.

8. Na is added to the numerals to make ordinals: rua, ruána.

ta. 1. Ta is a verbal article (t) has been already discussed.

2. Ta is a subjunctive particle used with verbs. Like the verbal particle and ma and ngga it changes its vowel to conform with the vowel sound which follows. It does not change its vowel before kai, we, and kau, you. It is usually preceded by ngge. It is probably the same as the Mota ta, used in the same way to imply doubt. Codrington and Ivens take it to be the verbal article t, and the subjunctive sense - 158 to reside in the ngge which usually precedes it, but when ta is used without ngge and with the future verbal article k some doubt seems to be implied, and it can be translated by the English “should”: ngge to ko liona ke, if you should wish it; ngge ta kau rigia, if you should see it; te ke rigia ke, should he see it.

3. Ta is a preposition of general relationship as regards place or time, and can be translated in, on, at, to, near, with or from. In and from are its usual meanings. It is often preceded by the article i. By an extension of its preposition it may mean than (from) and while (in): ta na komu, in the village; rugu horu ta na vale, go out from the house; ta na bela, on the table; kabu ta miu, stop with you; ra na vahagi ta dira, sick among them; kaekangge ta mua, enough for you. In the translations ta is often used for the Mota ape, about, concerning, but Nggela uses no preposition in such cases: ganaganaa, think about it.

Ta combines with the pronouns nggu, mu, na, etc., but they become slightly altered in the combination: tanggua, tamua, tatana, tatada, tamami, tamiu, tadia or tadira.

4. Ta is in rare cases a transitive suffix to verbs, and also a noun-forming suffix.

5. Ta is a prefix to verbs, almost as common as ma, and more common than ka. It cannot be prefixed to introduced words, i.e., it is a fossil formative. It resembles ma very closely in its uses. The chief difference is that it does not form many adjectives, whereas ma does, nor does it form so many words expressing condition, state or process. On the other hand it forms more past participles, and so passives, than ma does, or even than ka does. But like ma it is often merely the mark of an active verb, the sense of the word-base remaining unaltered; but it is always written as part of the word to which it is prefixed, unlike the living verbal article t. As in the case of ma some word-bases which would otherwise be lost in Nggela are preserved in the ta forms. And like ma it combines with other formatives to make compounds such as tanga, tara and tava.

It is prefixed to many verbs to make a past participle and so can be used to form passives in Nggela. With a good many of these ma or ka can also be used without any change in the meaning. Codrington and Ivens supposed that in all these ta forms there was an idea of spontaneous action, the - 159 thing happening of itself, but this does not seem to be so in many cases: mbita, to squash, tambita, squashed; koso, to sever, takoso, severed; mindi, to split, tamindi, split; ruha, to crush, taruha, crushed; vuti, to root up, tavuti, uprooted; mboha, to break, tamboha, broken; te takoso tua na alo, the string has been cut.

It does, however, sometimes imply spontaneous action, and then implies that the thing comes into the condition signified by the verb, whereas ma implies that it is or remains in that condition: tandeke, fallen, of leaves; tapese, fallen apart; tasipa, come out (of itself); tambihu, fallen out.

More rarely the prefix ta tends to make the word-base an adjective: tambohu, in a frightened condition, fearful; tapele, staggering; tarusu, miserable; talau, everlasting; talope, immature; tatohu, dead.

Quite often ta, like ma, merely marks a word as an active verb. Sometimes the form with ta is used both as an active verb and as a past participle: tambokili, to remove or removed; tavongo, to grope; tavugi, to bury (Mota, tavig); tagao, to steer; tambiru, to turn round; tagoro, to look after.

As with ma, some of the ta forms are now nouns, and take the article na: tambili, a mortar; tambuto, a boundary; tandau, a cross beam; tameme, powder; tavuvu, a gust of wind.

The vowel of the prefix may change: tororo, to slide; tosousondu, smashed; tovongo, with the meaning of tavongo modified.

Such forms as tanggoti, crushed, and tanggoru, crunched, seem to show traces of a compound prefix, tang. Other compound prefixes are tama and tara, but the most important is tava, which always shows spontaneous action, although it only occurs in a few words in Nggela, and is fossil. Codrington and Ivens do not mention it, but both give tapa as a prefix of spontaneous action. Codrington citing tapatuguru, to stand up, and Ivens tapataligu, to go round. Two other examples might be cited: tapatao, to chase, and tapatau, to seek food; but there does not seem to be the notion of spontaneous action in any of these words, and tapa is itself a verb meaning to rise or move swiftly, so that the four words may be combinations with tapa: tavaguliti, to - 160 come off, of itself, of skin (guli); tavahihi, to drop off, of itself; tavanonori, to peel off, of itself (after sunburn); tavaole, to wander; tavasuala, to slip or slide of itself; tavatogi, loosed of itself (tatogi, loosed by an agent).

va. In most Melanesian languages causative verbs are formed by prefixing va, or its compound vaka, or some form of these. Mota has both va and vaga. San Cristoval languages and Mala languages have the second haga, ha'a, faga, fa'a and faa. In Indonesian and Polynesian languages both these causatives appear, the simple and the compound. They are one of the marks of an Austronesian language.

Nggela, however, is peculiar, only va occurs as a causative prefix, and that only in a fossil form and in only a few words, and not as in other languages as a living causative prefix. Yet va and vaga occur in Nggela, but with a changed function.

The causative is expressed in Nggela in a different way, by suffixes instead of by prefixes, though these suffixes are thoroughly Austronesian in character. Codrington and Ivens both omitted them altogether in their grammars of Nggela, so that it will be useful before describing the use of va and vaga in Nggela, which is not to form causatives, to describe the way in which Nggela makes its causatives, something which distinguishes it from the languages of the eastern Solomons.

Nggela makes transitive verbs, as the other languages do, by means of suffixes. Sometimes these are a forms, ga, ma, na, etc., occasionally e forms, but almost always i forms; the transitive suffixes in Nggela are i, gi, hi, li, mi, ni, pi, ri, si, ti, vi; they have the same function as in other Melanesian languages.

Besides these Nggela, again like other Melanesian languages, has compound forms of these, the first half of the compound suffix the same as the a transitive suffixes, the second half the suffix gi: gagi, hagi, lagi, magi, nagi, pagi, ragi, sagi, tagi, vagi. Most Melanesian languages use these compound forms also to make transitive verbs, but in Nggela they are generally used to make causatives, and one of these suffixes lagi has the regular function of making a causative, and can be used with almost any verb to do so. Ordinarily, when the transitive suffix is mi, the causative suffix is magi, when the transitive is si, the causative is sagi, and so on; but - 161 lagi can be used instead of the others to make a causative; sometimes there are two causative suffixes to a verb, lagi and another, sometimes only lagi, which seems to have been steadily ousting the others and may now be called the typical causative suffix in Nggela; it can be used in this way with introduced English words: kuki, to boil; kukilagi, cause to boil. Nggela has lost the use of the prefixes va and vaga as causatives and substituted the use of a causative suffix: sara, to arrive; saravi, to arrive at (trans.); saravagi, to cause to arrive (caus.).

It is easy to see that a causative may develop from a transitive form. Melanesian transitive suffixes often give a causative turn to the word: kukilagi could be translated “boil it” as well as “cause it to boil,” and so with other verbs that are intransitive in English but can be made transitive in Melanesian languages with suffixes: na guri te saravagi nia i longa, the wind arrived it (drifted it) ashore. The short forms often take the place of English prepositions, the compound give the verb a transitive but different meaning: sara-via i longa, arrive at the shore; sara-vagi nia i longa, arrive it at the shore. But in the second the sense tends to be causative: na guri te saralagi nia i longa. Here, with the use of lagi instead of vagi, the sense is definitely causative. All these Nggela causatives are followed by the preposition ni.

An indication that va and vaga are no longer both causatives in Nggela is that multiplicatives can no longer be formed from them, as was pointed out by S. H. Ray. Multiplicatives are formed with va and vaga in languages which use those prefixes to make causative verbs. Yet Codrington and Ivens both state shortly that in Nggela “the causative prefix is va.

Va and vaga are no longer causative prefixes in Nggela, nevertheless, va is found in words in which va must once have been causative in function, but where va is now a fossil causative: vagala, to light a pipe, gala, burn; variu, to turn (trans. or intr.), of a door, riu, turn (intr.); vasuka, to throw a dart (make it go in (suka)); vatoga, to watch, toga, to dwell; vatori, to build (Mota taur, build).

Va appears to have a quasi-causative force in: vasiu, to bathe; vahagi, to be sick (Mota matag; Arosi mata'i; IN. saki).

- 162

Va is embedded in a few words, which are not quite causatives: vauvasuka, to throw darts; vaovarongo, to listen.

Va sometimes occurs in words that are now adverbs or nouns: vakale, clearly, kale, to hit; vakodo, clearly; vautu, a sign or omen.

There are some words in the texts of the New Testament, from which Ivens constructed his grammar, in which va is a causative. I mistrust these texts which are a translation of Mota, by Nggela teachers who spoke Mota, because I think they were influenced by the Mota forms, and the teachers' familiarity with va as a causative in Mota. These words seem to me to be imitations of Mota words, not real Nggela. The example Ivens gives of va as a Nggela causative is va-marara, to enlighten, which is a translation of the Mota va-marmararan. So the translators use vavola, to save (translating Mota va-esu), but Nggela speakers use vola as either intransitive or transitive, either live or make live, te volaa, he saves him. So also in the translations vamate is to kill, but a Nggela speaker says matea, kill him, matea na mbulu, put out the light. Va-uto, to bless, is from Mota va-wia, to bless, “cause to be good.” It was easy for Nggela teachers, familiar with the Mota va, to use va in such words, because there is a use of va in Nggela, probably derived from a former causative va, which will be described presently. I feel these words are an imitation of Mota, and an example of how one language may be affected by translations from another, and its imitation; and then these forms, continually heard and read, come to be accepted as genuine.

Va is a living prefix in Nggela with a different sense from a causative. It is what Williams has called in Maori an “inceptive,” used with a verb to express what one is beginning or intending or purporting to do. The same sort of particle is found in Lau, though there it is ka as in Maori, and used as ka is in Maori. It is possible that it is derived from vā, to go, as Ivens thinks; but there are objections to that view. A person intending to go to work says: Nau ku va lutu; if va is from vā, go, one would expect: Nau ku vā ni lutu. It may, in fact, be derived from a former causative va. Williams writes in his Maori Dictionary: “with an intransitive verb, adjective or particle whaka (the causative) signifies a beginning or approach to the action or - 163 condition indicated.” This describes quite well the Nggela use of va with verbs; as though the first syllable of vaka is used in Nggela. This inceptive va is in continual use in Nggela: ku va lutu, I am off to work; te va tona, he is going (tona, go).

The only occurrence of vaka in Nggela seems to be in the word vakalio, see clearly, where the meaning of vaka is not obvious. The real form of the Austronesian causative is vaga in Nggela, though it has lost its causative sense, and is translated: like, as, though: vaga meomeo, behave like a child, childish. Cf., Maori whaka-tangata, behave like a man, manly.

With this meaning “as, like,” vaga is very common in Nggela speech: vaga eni, like this, thus; vaga keri, like that; te vagaa, it is like that, so. Vaga can be used as a verb and take either set of pronouns after it: vagau or vaganggu, like me; vagago or vagamu, like you; vagaa or vagana, like him, her, it. As a verb vaga can have a trans. suffix, ha: vagaha, to liken or compare with. Vagi is another form of it, though much less common than vaga: vagi meomeo, childish; vagi liu, disobedient (liu, turn aside); vagi anggaia, like him. This also can become a transitive verb, with the suffix ni: te vaginia na lake, like fire, as though fire.

MN. vaka, IN. paka, and PN. whaka, all tend to develop other than a causative meaning. Often in MN. vaka makes adverbs and nouns. In Maori wha and whaka (Maori has both forms) are not merely causative prefixes, e.g., Mao. whawha-tanga, to feel with the hand (cf., IN. tanga, hand, MN. tanga, a handle). The same is true of pa and paka in IN.

In Fijian the use of va and vaka is in some respects like that of va and vaga in Nggela. The difference is that the causative force has been kept in Fijian as the normal use, and has been lost in Nggela, though it appears (va) in a fossil form in a few words. In Fijian also vaka can be a verb: “to resemble, be like,” and also (as in Nggela) a conjunction “like, according to.” Fiji: sa vaka onggo, like this; Nggela: e vaga eni, like this. These quasi-causative uses of vaka in Fijian, and wha and whaka in Maori, show that the Nggela vaga at any rate, and probably va also, were once causatives.

- 164

In Arosi (San Cristoval) ha'a is more strictly a causative prefix, and there are more than 1,500 words in the Arosi Dictionary in which ha'a appears as a causative prefix: dadao, lie down; ha'adadao, cause to lie down.

And yet in Arosi also ha'a is not solely a causative prefix. It can be used with nouns and adjectives, and makes nouns and adverbs as well as causative verbs: rago, many, ha'a rago, often; bae, a pole, ha'a bae, set up poles; uru, 16 revenge, ha'a uru, take revenge; wasi, wild, ha'a wasi, hunt wild pigs or train dogs to do so; didiri, slide, ha'a didiri, surf-board.

The Arosi ha'a may also have the meaning: “as it were, as”: ruani, a partner, ha'a ruani, to be partners; samo, speak like a foreigner, ha'a samo, imitate a foreigner; tawa, a stranger, ha'a tawa, treat as a stranger; 'ua, how, ha'a'ua, to behave.

Probably va and vaga lost their causative function in Nggela owing to the development there of causatives from suffixes. The general picture is that of languages far apart, in IN., MN. and PN., developing a common stock of formatives, each in its own way, according to its individaul genius. The stock is common; the development is different.

ni. Ni is the most important formative in Nggela, and this formative is common to IN., MN. and PN. In Nggela it occurs with many functions, the chief of which are:—

  • 1. A verb.
  • 2. A pronoun or article.
  • 3. An article.
  • 4. A preposition, its main function in Nggela.
  • 5. A transitive suffix.
  • 6. A suffix to relationship terms.
  • 7. A suffix to numerals.

1. Ni is a transitive verb meaning to do, act in any way, think, say. The form ne is also found with the same meaning.

2. Ni is a pronoun, or perhaps an old personal article used with a pronoun. In Lau the 3rd pers. pronoun is nia, Arosi ia; in asking a question in Lau this is shortened to ni? - 165 pidgin English “him”? In Nggela in such a question the form is ina? is that so? “him”? which appears to be a metathesis of the Lau form, though it might be an independent form, na, of the pronoun. Codrington has shown that the Nggela anggaia, he, she, it, and the Fijian koya (ko ia) both reduce to ia. He thinks i is a personal article and a the true pronoun. If Nggela ina is the equivalent of nia elsewhere, and Codrington is right, ni would appear to be an old personal article. 17 Nia also appears in Nggela as a connective, an introductory word at the beginning of a sentence, and is equivalent to such English expressions as “that being so,” “this also,” “and with that.” This again seems to be the pronoun it. Mota uses ni as the short form of the pronoun 3rd pers. sing.

Ni is also found in the second set of pronouns nggu, mu, na, as the plural of na when that is neuter, it is thus the equivalent of the i, gi, plural of the first set. The verb lio, to like, takes the second set of pronouns as object, thus we get: tu liona, I like it; tu lioni, I like them (neuter).

3. Ni has already been described as an infix in words like hanigela, crew, nihagela; hage, to embark; hagela, a collective noun, and the article ni which is infixed; and since the present Nggela article na is prefixed to such words, na hanigela, we have two articles, the second unrecognised as such. This ni also occurs in one Nggela word, niala, a row, the same meaning as alaala, a row, but the ni is not recognised as an article and the word takes na, na niala or na nianiala, a living article followed by a fossil one.

4. The main use of ni in Nggela is as a preposition. Its use as a preposition is peculiar and one of the chief characteristics of the Nggela language.

Ni means belonging to, of: mane ni Nggela, a man of Nggela, and though it often has to be translated in English by other prepositions, and sometimes cannot be translated at all, this is its main, fundamental meaning: alo ni pitianggu, rope for tying me; vitolo ni vanga, hungry for food; tuguru ni gonindila, stand up to get ready; mbutu ni tarai, come to teach; vuni ni lutu, begin to work; ndutu ni mate, near death; kambu ni vunagi, behave like a king.

- 166

With the same root meaning ni is prefixed to the second set of pronouns (the first two then suffixing a as with the preposition ta). Thus, what are called possessives are formed: ninggua, my; nimua, thy; nina, his; dida (for nida) ours (inclus.); nimami, ours (exclus.); nimiu, yours; didia, didira (for nidia, nidira), theirs. The formative a is sometimes prefixed to these forms, but with no alteration in meaning. When the article na comes before these forms the meaning is mine, thine, etc. Codrington and Ivens consider this ni to be a noun, and so distinct from ni the preposition. There seems to be no real reason for thinking so: nina Piluka, Piluka's. We can translate this as “belonging to (ni) him (na) Piluka,” the na being an anticipatory object, which is not only a regular use after a verb (te rigia Piluka), he saw him Piluka, but also occurs after prepositions: sarana na vale, the na of sarana meaning “it,” an anticipatory object, and the second na is the article, “under it the house.”

Deck has shown that in Kwara'ae the a of aku, amu, ana, my, thy, his, is a preposition of just the same kind, taking an anticipatory object, ana John, belonging to him John, John's.

If the ni of Nggela in such possessives, and the a of Mala and San Cristoval in similar forms, are both prepositions, and not nouns, as Codrington and those who have followed him supposed, then analogy would lead us to expect such forms as no and mo in Mota, or the Fijian forms, or to and ta in Maori, to be prepositions also.

These possessive forms in Nggela aften have the meaning “for my part,” “as for me,” and are common at the end of a Nggela sentence, or they may come at the beginning and have the meaning “I ought, I must” (belonging to me): Inau tu ganagana ninggua, I think myself, for my part; ninggua nggu tona, I ought to, must, go.

5. Ni as a preposition is normally followed by the first set of pronouns u, go, a, and can be translated in various ways. It usually comes, with its pronoun, before the verb, a peculiarity of Nggela grammar; and it is equivalent to a transitive suffix added to a verb 18: te nia hare, he is astonished at it. This is equivalent to hare with a transitive suffix ni: te harenia. The position before the verb is worthy of remark, because there is a tendency in Nggela to throw - 167 words back. Sometimes the object comes before its verb, very often before the subject, which frequently comes last in a Nggela sentence: te rigia na vale na nggari, the boy (nggari) sees the house. In the same way ro (a contracted form of rua), which is the sign of the dual pronoun, comes before its pronoun, not after as in most Melanesian languages: ro dida, ours; ro nimiu, yours; ro gamu, you two; cf. Mota kamu-rua. So also the article comes before the negative, separated from its noun, in such a sentence as: “I am not a white man,” na mua manepura inau. It is therefore not uncharacteristic of Nggela to put ni with its pronoun before the verb. There are reasons for thinking that the preposition ni is in fact a transitive suffix which has become separated from its verb, like ki in Maori.

The following are examples of the use of ni in Nggela, the 3rd pers. pronoun being added to it in the examples: nia mbule, drunk with it; nia ndikalio, sad at it; nia gogo, flee from it; nia hove, forbid it (no Eng. prep.); nia huru, accuse of it; nia liliu, change into it; nia pitu, wait on account of it; nia pelu, buy with it; nia ponolio, forget about it; nia sika, hate it (no Eng. prep.); nia vola, live by it; nia vugu, cough because of it; nia mai, come with it; niu sara, come on my account. There is another form of the preposition, the equivalent of ni, which can also be used with pronouns in the same way, coming before the verb. This is a compound nigi: te nigia hare, he is astonished at it.

There is an instrumental preposition in Santo, New Hebrides: gini. 19 In Bauro, San Cristoval there is also a preposition gini which is the same as the transitive suffix gini used after most verbs with the, the great variety of meanings which transitive suffixes have in Melanesian languages, so that they have to be translated by various English prepositions. The Arosi equivalent of the Bauro preposition and suffix is 'ini. It follows hundreds of verbs and has to be translated by as many English prepositions as the ni before a verb in Nggela, and like Bauro gini it can be separated from the verb and used as a preposition. In Arosi it certainly seems to arise from the transitive suffix, which has got free from its verb and begun an independent life as a preposition: ia songa'inia, he asked about it; 'inia ta? about what; 'inia na haka, about the ship.

- 168

In Nggela, in the causative suffixes, we have the equivalent of the Arosi transitive suffixes with 'ini, e.g., harelagini, cause to be astonished, which can be analysed as hare (astonish) and la (transitive suffix) and gini (compound trans. suffix).

In Mota, nia is an instrumental preposition, and comes last in the sentence: o gasal ni me vusia nia, the knife he killed him with. In Nggela this would be: na isele te nia lambua. They are evidently the same, but in Mota the origin is lost and Codrington supposed a to be part of the preposition. Mota also has ris nia, change into, where Nggela has nia lia, with the same meaning.

The preposition follows the verb in a few cases in Nggela: holoa nia John, call him John (but nia aha, name him) and usually does so when it is instrumental, te lambua nia na ivi, he killed him with a knife (ivi = isele), but need not: te nia kalea na vatu, he hit him with a stone.

This is the normal order: preposition, pronoun, verb, pronoun, article, noun, and it is characteristic of Nggela.

It seems reasonable to suppose that Arosi 'ini, Bauro gini, Nggela nigi and ni, are all one and are prepositions which were once transitive suffixes, and have got free from the verb. Though ni precedes the verb in Nggela it has all the force of a transitive suffix. And it is reasonable to hold that this ni is one and the same as the preposition ni, of, belonging to, which occurs in IN., MN. and PN. so very widely. Why should not nouns have transitive suffixes as well as verbs to express what we express by prepositions?

6. Ni is a transitive suffix to a few verbs.

7. Ni is a suffix to collective terms of relationship: hogo, clansman; na lei tama-hogo-ni, clansmen.

8. Ni is added to numerals to make ordinals: rua, ruani.

si. 1. Si is a genitive preposition joining nouns. The other genitive prepositions in Nggela are i, li, gi, ni, ri, ti.

2. Si is a transitive suffix to verbs. The other similar transitive suffixes in Nggela are i, gi, hi, li, mi, ni, ngi, ri, ti, vi.

3. Si occurs as a prefix in si-vuraga, to be born.

4. Si is a short form of sakai, one. Cf., Lau indefinite article si.

- 169

5. Si is a suffix forming superlatives of adjectives, nouns, and verbs: mbona, poor, mbonasi, very poor; mbili, black, mbilisi, very black; nggari, child, nggarisi, very much a child; garu, to desire, garusi, lascivious; nongi, to beg, nongisi, habitual beggar. Why should only adjectives have superlative forms?

NOTE

[562] Stone Weapon from Borneo.

The attached photograph illustrates a heavy stone tool, apparently of serpentine, which is item No. 1479 in the Sarawak Museum, Kuching, Sarawak, Borneo.

The implement is 356 mm. long and 96 mm. across at its broadest. It is heavy, and finely finished. The edge is chipped at several points. The head is drilled to form a hole tapering conically inwards. The top of the “handle” is worked into concentric grooves or ridges.

This is the only implement of its kind so far known from Borneo or the area generally, and it seems to show strong similarities to items familiar much further south-east in the Pacific, notably in New Zealand.

It was collected in 1881 at the house of Pengeran Samah Digadong, Chief of the Buludupi, a Mohammedan group living at Melapi in the lower Kinabatang River, North Borneo.

The occurrence of this item in North Borneo is not merely of intrinsic interest. If it is “Maori,” then it is also an indication of the degree of mobility of objects. I have already suggested, in this Journal (Vol. 58, pp. 91-111) that rather too much theorising has been done with regard to waves of culture supposedly represented by different types of stone implements. Here is one more example of the way a material item moves.

—Tom Harrisson, Curator, Sarawak Museum.
Illustration
1   Ngg is written for the sound of ng in finger; g is the Melanesian g, almost v before u; b and d occur in the translations but are here written mb and nd.
2   Bulletin of the School of Oriental Studies, Vol. VIII, Pt. 4.
3   Melanesian Mission Press.
4   Also ihola.
5   Roviana, a language of New Georgia with a non-Austronesian vocabulary, has the same infix ni, making nouns from verbs; from metathesis of the two consonants in e.g., ni mate, mi nate.
6   Polynesian languages have been less studied than either Indonesian or Melanesian languages, so there may be more metathesis than is at present recognised; for example there may be unrecognised infixes in Maori.
7   i hanu, at such and such a place.
8   Ivens, Dictionary of Bugotu.
9   The a of ma alters by attraction to the vowel following.
10   Ivens, Grammar of Florida, p. 1094.
11   Codrington, Melanesian Languages, p. 187.
12   Arosi, manihi. Mota has the metathesis mavinvin.
13   Lau has the metathesis mangata, in the middle, and also matanga or matonga.
14   Or hare.
15   It is also used with the pronouns hanu, such and such; hava, what; nina, his; na hanu, na hava, na nina; but not with vei, where (Arosi, na hei).
16   Nggela utu.
17   The di of the pronouns dia, dira, they, in the second set of pronouns, may be a form of this ni.
18   Ivens called this a prepositional verb.
19   Codrington, Melanesian Languages, p. 295.