Volume 74 1965 > Volume 74, No. 2 > An outline of the structure of the language of Nukuoro, by Vern Carroll, p 192 - 226
AN OUTLINE OF THE STRUCTURE OF THE LANGUAGE OF NUKUORO
University of Chicago
0.1. Nukuoro is a ‘Polynesian Outlier’ in the United States administered Trust Territory of the Pacific Islands. It is located at approximately 4° N 155° E; 164 miles to the south is Kapingamarangi, another Polynesian Outlier; 113 miles to the north-west is Satawan, the southernmost of the Mortlock Islands, in all of which a dialect of Trukese is spoken; and 176 miles north-east is Ngatik, where Ponapean is spoken. Nukuoro and all of these nearby islands are small coral atolls.
The speech community at present (1965) numbers a little over 400 persons: about 260 resident on Nukuoro; close to 125 located on Ponape, the District centre; and a few others on each of the islands in the District. Many older people speak considerable Ponapean and there are a few who know Kapingamarangi, one or another of the Micronesian languages, some German, Japanese, or English. The bible and hymnal used in church services on Nukuoro are in Ponapean; otherwise the Nukuoro language is used exclusively on the atoll. There appears to have been no wholesale adoption of lexical and grammatical forms from outside. No severe demographic or cultural upheavals have been known on Nukuoro. The number of outsiders resident on the atoll at any one time has always been very small and relationships with foreigners have not been traumatic. There was no regular elementary school during the German and Japanese administrations. Since the War, the American administration has employed local teachers to conduct six grades of elementary school four hours each day in the Nukuoro language.- 193
In short, there is little bilingualism on Nukuoro. School children and adults usually have some facility in another language, but its use is limited to the classroom, bible reading, conversation with visitors, and the like.
Nukuoro is not mutually intelligible with Kapingamarangi, although speakers of one of these languages can make themselves understood to a speaker of the other with some difficulty. Those who have heard other Polynesian languages, such as Ellice and Maaori report that they are very much aware of lexical similarities, especially names for plants, fish, construction terms, and the like. A monolingual speaker of Nukuoro understands nothing of Ponapean, Trukese, Kusaean, or any other Micronesian language.
The historical derivation of Nukuoro language is not clear. Local traditions place the point of origin of the first settlers in Samoa. A small group from there is thought to have visited many islands during the course of the wanderings which brought them eventually to Nukuoro. Some of the most frequently mentioned way points are Nukufetau (Ellice), Kapingamarangi (TTPIS), and Oneap (Mortlock Islands, TTPIS). However, the place names in Nukuoro legends have been greatly corrupted by the substitution of modern place names for archaic ones, using only rough phonetic similarity as a guide. There are no other data at hand which give a clear outline of the historical picture. After a period of initial settlement, no regular contacts were maintained with other islands, although castaways occasionally arrived from distant points. Although the cumulative effect of periodic additions to a small society must have been great, the overall cultural-historical pattern appears to have been one of isolation until the coming of European ships in the mid-18th century. The depth of occupation sites indicates that Nukuoro have lived on the atoll for many centuries.
The following table indicates the percentage of shared cognates on lexico-statistic lists for Nukuoro and other Polynesian languages.
It will be noted that Nukuoro appears significantly more closely related to Kapingamarangi than to the other Polynesian languages. It occurred to me that my list for Kapingamarangi might have produced a false impression since it was elicited from a Kapinga living on Nukuoro, using the Nukuoro language as the medium of communication. But two other lists for Kapingamarangi produce roughly comparable figures (50% and 58%).
One has the feeling in working with these lists that a lexico-statistic list completed by a linguist who is compiling it with the assistance of native speakers is a different thing from a list made up from a published vocabulary. Even though the user of a vocabulary list must credit a test item with cognation if any one of the listed items is cognate, one gets in the habit of expecting fewer total cognates from such a list. It is therefore suspected that the figures for Rennell, Pileni, Maaori and Kapingamarangi are somewhat higher than they would have been if vocabulary lists had been used for these languages. When, on the other hand, a comprehensive dictionary is used to compile a list, it seems reasonable to suppose that the number of cognates will exceed the number on a list produced by direct elicitation. Thus the figure for Tongan might well be higher than it should be. The larger societies (Tonga, Samoa and Tahiti) apparently diverge rather more from Nukuoro than the smaller ones. Hawaii and Maaori do not fit this generalization very well, possibly because of the comparative recency of their attaining a large population. Tahiti and the islands in its cultural ambit (Tuamotus) are especially divergent, as Elbert has noted. It is hypothesized that larger Polynesian speech communities have a higher rate of linguistic change than smaller ones, owing to the general Oceanic cultural tendency for small status and residential groups to discriminate themselves from each other on the basis of language. It is further hypothesized that very remote speech communities, such as Nukuoro and Kapingamarangi will have a lower rate of linguistic change than those more in contact with the centres of innovation.
If all of these hypotheses are entertained seriously then we would have to say that, in terms of time-depth separation of Nukuoro from the languages above, Nukuoro is about equally distant from all of them.
0.2. The previous work on Nukuoro language is easily summarized. Christian published a Nukuoro vocabulary of about 300 items in 1898. 1 Kubary and the anthropologists with the German Südsee Expedition of 1908-1910 added several hundred more items to the list. 2 Elbert compiled a vocabulary of about 1,500 items with a few grammatical notes, which appeared in mimeographed form in 1946. 3 - 195 0.3. The present work does not aim at theoretical innovation or descriptive ingenuity. The descriptive format is adopted from Biggs's description of New Zealand Maaori, which is to my mind the most satisfactory description of a Polynesian language in print. 4 The description which follows was produced in the field during the academic year 1963-64, while the author was engaged in sociological research on Nukuoro Atoll. Assistance was mainly provided by a local school teacher, Tobias, but many other members of the community were helpful on linguistic matters.
The principal text from which the present analysis derives is included as an appendix to this study. Use was also made of the New Testament books of Mark and Matthew, which were translated into Nukuoro from Gilbertese by the late Chief, Leka. Several printed and mimeographed pamphlets prepared by local school teachers and produced by the District Department of Education, Ponape, were also consulted. These few items constitute the entire literature in Nukuoro. Many examples were culled from conversations concerning other matters or were made up by the author, who acquired some facility in spoken Nukuoro during his stay.
0.4. The orthographic system employed here was developed locally by the late Chief, Leka, perhaps with the assistance of resident Europeans or missionaries in Ponape. It is known and used by nearly all Nukuoro and has been the educational standard for forty years.
It will be noted that the Nukuoro do not “occasionally lapse” to p, t, k, for b, d, g, as Elbert asserts, 5 but use these letters for phonemically distinct sounds.
The linguist may find Nukuoro orthography inelegant but I insist on using it so that this work will be useful to the Nukuoro people. Aside from them, I am addressing the specialist in Polynesian languages, who should have no difficulty with this orthography. Most linguistic work in the Trust Territory has not been useful to the local population just because the linguist insisted on using an orthography that made no concessions to the long-established indigenous writing system.
1.0. Nukuoro utterances are divided into shorter stretches on the basis of juncture. The phonetic character of juncture in Nukuoro is described in section 1.4. Subsequent divisions are into syllables and phones. The present work is limited to a summary discussion of the phonology of Nukuoro and the grammatical analysis of the ‘contour word’, the morphological material in a contour span of minimum length. These notions are defined in section 2.
1.1. There are fifteen Nukuoro segmental phonemes which are represented orthographically as follows:- 196
all consonants and vowels may occur lengthened (phonemically double). A double vowel is written as such. Double consonants corresponding to the single consonants b, d, g, m, n, ng, v, s, h, l are written p, t, k, mm, nn, nng, vv, ss, hh, ll. It will be noted that lengthened stops are written /p, t, k/ and lengthened /ng/ is written /nng/; otherwise all lengthened phonemes are represented by writing their symbol twice. This orthography is used in all Nukuoro forms cited.
1.2. The phonemic status of the vowels is established by the following minimal contrasts:
The phonemic status of the consonants is established by the following minimum contrasts:
The phonemic contrast of long vowels with short vowels is illustrated below:
There are no semivowels phonemically. The unstressed members of vowel clusters maintain a five-way vowel contrast, as is illustrated by the following examples:
The following examples illustrate the phonemic status of the long consonants:
1.3. The following presents a brief phonetic guide to Nukuoro phonology with some indication of the allophonic range of each phoneme. As one would expect in a system involving relatively few contrasts, the number of distinguishable allophones of each phoneme is very large. We are not prepared to go greatly into detail on this subject, but some good tapes are available to any one interested in pursuing the articulatory or acoustic analysis of Nukuoro.
Single vowels in the syllable following a stressed syllable are devoiced when contour final.
Double phonemes are realised as follows:
1.4. Three types of juncture are distinguished for Nukuoro.
Open transition between vowels (symbolized +) occurs at morpheme boundaries within bases, between bases in sequence in the same nuclear position, and between pronouns and other morphemes. Within the same morpheme, or between bases and minor morphemes, or between minor morphemes, adjacent vowels are in close transition. Vowels of the same quality in close transition are never rearticulated; in open transition there is a slight rearticulation, just noticeable in the slowest speech. It is a feature of the morphophonemics that between bases and minor morphemes, two adjacent vowels of the same quality become a single vowel of normal length in fast speech.
Non-final juncture characterizes a point of possible momentary pause, sometimes phonetically marking the boundaries of the grammatically determined contour word. While some of these possibilities may in fact be actualized in slow speech, all possibilities would never be actualized even in the slowest possible speech. However, the “psychological reality” of these juncture points was demonstrated by dividing a long text into contour words and asking an informant to pause momentarily after each contour word. He found no difficulty in doing so and other Nukuoro listening to the recording that was made of this drill found the intonation patterns and pauses perfectly natural, although much slower than they would ever have occasion to speak. Our tapes illustrate these points quite nicely. When a pause occurs at a point of non-final juncture, pitch is sustained through the pause; loudness is sustained through the syllable next before the pause, at which point there is some diminuendo; vowels in the syllable next before the pause remain voiced.- 199
Final juncture is always characterized by pause. The prejunctural syllable is characterized by decreasing loudness, decreasing pitch and the devoicing of vowels, except following /l/. In any connected discourse only a limited number of non-final juncture points can occur before a final juncture is required; thus at a point which might otherwise be marked by non-final juncture, final juncture may occur, depending on what has occurred previously in the discourse. A final juncture to final juncture stretch containing a large number of non-final junctures is normally followed by a stretch with fewer non-final junctures between sequential final ones.
The stretch between any two junctures or between silence and juncture is called a contour span. As defined the boundaries of contour spans and of contour words (2.1) coincide in most cases.
Primary stress (′) occurs at least once in each contour word, predictably on the penultimate syllable of each base, pronoun or other morpheme occupying the nuclear position. Reduplicated bases behave as two separate bases in this respect. Primary stress is phonetically defined by rising pitch.
In the examples quoted in the following sections, final and non-final junctures are omitted to save cluttering the text. These phenomena are amply illustrated in the appended text. Open transition and stress are marked neither in the examples nor in the text material, as it appears that their occurrence is predictable when morphophonemic notation is employed.
1.5. Syllables take the shapes V, VV, VVV, CV, CVV and CVVV. All possible V and VV combinations occur. All possible CV combinations occur except /vu/.
The first member of a diphthong is always the syllabic peak when the syllable is stressed; elsewhere there is little difference between members, the peak of sonority tending to occur on the most naturally sonorous vowel.
1.6. The appended text contains 11,420 single segmental phonemes (place names and personal names being excluded from consideration), and 457 double segmental phonemes. Thus single phonemes are 25 times as frequent as double ones.
Single phonemes in rank frequency order with number of occurrences expressed absolutely and as a percentage of the total number of phonemes: a 2,721, 23.8%; i 1,374, 12.0%; g 1,206, 10.5%; e 1,085, 9.5%; d 954, 8.3%; o 841, 7.3%; u 722, 6.3%; 1 700 6.1%; h 531, 4.6%; n 428, 3.7%; ng 294, 2.5%; m 276, 2.4%; b 114, 0.9%; v. 92, 0.8%; s 82, 0.7%.
A total of 457 double phonemes occurred in the following rank order: t 133, 29.1%; aa 129, 28.2%; oo 54, 11.8%; k 32, 7.0%; ee 31, 6.7%; nn 18, 3.9%; uu 18, 3.9%; ss 16, 3.5%; ii 12, 2.6%; mm 5 1.0%; ll 3, 0.6%; p 3, 0.6%; hh 2, 0.4%; vv 1, 0.2%; nng 0, 0.0%.
The data on single segmental phonemes may be summarized by noting the frequency of occurrence of each type of phoneme: vowels 59.0% (a: 23.8%), consonants 40.9%. Stops account for 19.9%; nasals - 200 8.7%; fricatives 6.1%; flap 6.1%. Alternatively, dentals 18.9%; velars 17.7%; labials 4.2%. These figures show that vowels account for almost 60% of the segmental phonemes in Nukuoro with /a/ accounting for 40% of the vowels. Stops account for half the consonants with nasals, flap and fricatives together accounting for the other half. Alternatively, one might note that dentals and velars each account for almost half of the frequency of consonants, with labials being relatively infrequent.
1.7. In addition to the contractions in spoken Nukuoro mentioned under the heading of “juncture” it should be noted that an unstressed vowel or two is lost in the environment /d . . . d/, the result being heard as a phonetically long d (/t/). As the Nukuoro are not aware that the /t/ occurring in nouns and pronouns derives from contraction, they (and we) do not analyse such forms when they occur. It will be noted that nouns do not normally have any ‘number marker’ (see section 2.6.7) except by the addition of a member of decade 70. Nouns beginning with /d/ however are thereby marked for plural:
The singular we take to be a contraction of the plural form plus the singular article /de/, the plural marker being ø (see section 2.6.7). In the appended text, all such instances of de + d > t are underlined.
2.1. The grammatical analysis proceeds by listing the relational words of high frequency (minor morphemes) in decades according to the positions in which they occur with respect to words carrying the bulk of the semantic load, words which occur individually much less frequently (bases). Minor morphemes are classified as ‘preposed’ if they normally precede a base and ‘postposed’ if they normally follow a base.
The unit within which both bases and minor morphemes distribute, the internal structure of which is described in the following section, is called the contour word. A contour word consists of a nucleus and a periphery. The nucleus may be filled by one or more bases with or without minor morphemes of decades 10 and 210; or by certain combinations of minor morphemes acting as base surrogates. The periphery may be filled by minor morphemes (other than decades 10 and 210) occurring before the nucleus (preposed) or after the nucleus (postponed), or in both positions.
A contour word may consist of a single base only, e.g. hanno ‘go!’, uaa ‘yes’! but typically morphemic material also occurs in the periphery, e.g. o de laa ‘of the sun’.
From the Nukuoro point of view, all morphological material within the contour word is ‘bound’, except to the degree that the speaker undertakes some grammatical analysis to separate words when writing. Therefore, it will not do in Nukuoro to consider only “the minimal forms which an informant will offer in citation” 6 as the isolable constituents of contour spans. In the appended text, we separate all minor morphemes from each - 201 other and from bases by a single space; the constituent morphemes of ‘pronouns’ (certain combinations of minor morphemes from decades 40-80), and ‘affixes’ (decades 10 and 210), as well as the constituent morphemes of bases, base surrogates, and other clusters of minor morphemes which behave like bases, are separated by hyphens. In the examples following there are no hyphens and no spaces except between minor morphemes and bases unless the example is more informative with than without. Normal punctuation is also retained in the Nukuoro examples for ease in reading.
2.2. In any utterance a contour word is initiated by:
This definition is formally equivalent to that Biggs has proposed for Maaori, 7 once allowance has been made for the fact that, in Nukuoro, pronouns serving as the subject of a verbal expression usually precede the verb.
While this definition covers the vast majority of the cases at hand, it should be noted that some disparities were observed between the gramatically defined contour word and the phonetically defined contour span. A study of these variant cases was most instructive in that it indicated the points at which final juncture would never be possible. We list here the additional rules that would have to be invoked to establish contour word boundaries only in those places where non-final juncture in fact occurs.
Non-final juncture also occurs before
The sum of these additional rules gives us a much more powerful set of specifications than is minimally necessary to define the boundaries of the non-final contour. Many cases which fall under one or another of the above rules will be found in the appended text to be divided according only to rules (a) and (b). This is because my assistant found no difficulty in pronouncing contour words established on the basis of rules (a) and (b) only, except for a case of (c) and a few cases of elision across contour word boundaries. Only those places where he refused to separate a contour are noted in the text material. Nevertheless, the recordings which we made before he learned about contour words show no such willingness to divide contours, even when he was instructed to speak as slowly as he possibly could. It is from these first recordings and from the speech of island people that the above observations were culled.
2.3. Minor morphemes are listed below. Base classes are subsequently defined with reference to their occurrence with various minor morphemes. A note on the classification of complex bases and detailed discussion of each minor morpheme complete the present study.
Decades are numbered outward from the base. Members of decade 10 occur only immediately before bases; members of 20 may occur before members of 10 and so forth to decade 90, which contains those minor morphemes which may precede all other preposed minor morphemes. Minor morphemes which occur only immediately following bases are assigned to decade 210, with postposed minor morphemes which can occur subsequently being assigned to higher numbered decades. Thus the decade to which a minor morpheme is assigned represents an assertion that, in any contour word in which morphemes from different decades occur together, the lower numbered form will always occur closer to the base than the higher numbered form. Any exceptions to this are noted in section 2.6. The numbering system also asserts that no two members of the same decade can occur together in the same contour word, except as specified in section 2.6. Also, except as specified, no postposed minor morpheme can occur preposed to a base (nor can a preposed minor morpheme occur following a base). All minor morphemes can occur immedi- - 203 ately in sequence with bases, except as noted. Within the decade, numbering is entirely arbitrary. For convenience of reference, decades are listed in the order in which their members normally occur in the contour word.
90. Introductory words; phrase and word connectives.
60. Possession markers
50. Pronoun person markers—plural
40. Pronoun person markers—singular
Pronoun number markers
30. Verbal aspects
240. Directional I
250. Directional II
2.4. Bases are classified as either nouns (N), adjectives (A), or verbs (V).
All possible shapes for contour words containing an N-class nucleus are given by the formula:
± 90 ± 80 ± (70 ± 60 ± p) or (23) ± 10 + N ± 210 ± 220
± 240 ± 250 ± 260 ± 270 ± 280 ± 290 ± 300- 205
Three subclasses of nouns may be distinguished. Personal names (Nn) are those nouns which may be preceded by 23, and any longer forms from which the latter derive (see section 2.6.2). Place names (Np) are those N which cannot be preceded by 71-3 and which are not Nn. Location words (N1) are those N which can be immediately followed by the expression 71 + N (when preceded by 81 or 82).
All possible shapes for contour words containing verbs are defined by the following formula:
± 90 ± 80 ± 70 ± 60 ± p ± 30 ± 21 ± 10 + V ± 210 ± 220 ± 230 ± 240 ± 250 ± 260 ± 270 ± 280 ± 290 ± 300
Contour words containing adjective class nuclei are defined by the following formula:
± 90 ± 80 ± 70 ± 60 ± p ± 30 ± 21 ± 10 + A ± 210 ± 220 ± 240 ± 250 ± 260 ± 270 ± 280 ± 290 ± 300
A sub-class of adjectives comprising the numerals (An) is defined by their occurring directly following 16-18.
It will be immediately noted that the difference between the various classes of words is very slight indeed. The distributions of V and A are identical except that the former may take 230 (ina) while the latter may not. Whereas Biggs separates V from A in Maaori on the basis of the former's occurrence with 211, this is not practical in Nukuoro.
NI in Nukuoro include the following: uda ‘towards home’, dai ‘lagoonward’, dua ‘seaward’, ngaiho ‘downtown’, ngaage ‘uptown’, lodo ‘inside of’, lalo ‘under’, lunga ‘over’, honga ‘on top of’, daho ‘at the side of’, gaogao ‘beside’, tege ‘at the outside corner of’, lotedege ‘at the inside corner of’, lubaasi ‘all around’, teungalodo ‘middle of’, bido ‘end of’, lubido ‘both ends’, magava ‘between’, mua ‘front’, muli ‘back’, ngudu ‘rim of’, gabugabu ‘shore’, ulu ‘uppermost part’, mada ‘front’, aloalo ‘part out in front’, hiihii ‘edge’, tahido ‘base’ or ‘stem’, logunga ‘inside corner’ (of a house).
An are dahi ‘one’, lua ‘two’, dolu ‘three’, haa ‘four’, lima ‘five’, ono ‘six’, hidu ‘seven’, valu ‘eight’, siva ‘nine’.
2.5. ‘Complex bases’ are combinations of independent bases which behave as a single base by taking a single contour accent, a single affix, or the like. They are to be distinguished from bases in sequence.
The following rules apply: N + A, N + V, N + N behave as N; A + N, V + N, A + A, V + A, A + V, V + V behave as V.
Examples of the above permutations are listed below.
11. haga- (‘make’) occurs in sequence with bases of all classes. For A and N it is transformative, both becoming V.
12. ada- (conjectural) occurs in sequence with V and A. It is not transformative. The most usual sense is ‘if’ or ‘can’.
ada also occurs in the nuclear position:
13. ma- (resultative) occurs in sequence with V, transforming them into A. It indicates the product of the action.
14. hii- (‘want to’) occurs in sequence with V and A. It is not transformative.- 207
51. vaa- (able to') occurs in sequence with V and A. With V, it is transformative to A.
16. 16-18 are numeral classifiers, occurring only before An (and divisive for this sub-class).
17. hua- counts breadfruit, coconuts (ripe and drinking) and huahu (a cooked food made from taro and grated coconuts) by tens: hualua ‘twenty cocounts, etc.’
18. ua- counts pieces of coconut frond thatching by ten: ualua ‘twenty pieces of roofing’.
Four kinds of reduplication are found in Nukuoro. R1 occurs in bases of all classes. R2, R3 and R4 occur only in A and V.
(R1) Reduplication of the entire base.
(R2) Reduplication of the first syllable of a base. Very few bases in Nukuoro are reduplicated in this fashion.
(R3) Reduplication of the initial consonant of a base. It might be noted that R3 is the citation form for all bases which can be reduplicated in this way. Many, but not all, ‘verbals’ (A + V) agree in number with the “subject” of the verbal expression.
(a) The “subject” in ‘intransitive’ verbal expressions is normally the actor.
(b) The ‘subject’ of a ‘transitive’ verbal expression is normally the goal. All forms are V.
(c) In some transitive verbal expressions, however, the distinction between the reduplicated and non-reduplicated forms of this type is the same as that in R2. All forms are V.
All non-reduplicated forms above can not be used with singular goals and thus fall also under the heading of R3b. Likewise R3a and R3c may both be equally applicable to the distinction between reduplicated and non-reduplicated forms of type R3.
R3 should not be confused with contracted forms consisting of an unreduplicated base + 70, e.g. ssugi ‘light touch’ (< se + sugi ‘touch lightly’).
(R4) Reduplication of the initial vowel of a base.
In certain cases, R3 and R4 both occur, almost surely because the base is composed of two different root morphemes.
R4 should not be confused with the doubling of the initial vowel that many bases undergo when 210 is added.
Imperative forms lengthen phones in all positions, but this is perhaps best considered as a result of ‘imperative stress’ (i.e. as a matter of suprasegmentals, rather than as R, although we note that the imperative form of many bases doesn't undergo this process.
Some imperative forms double the first vowel of the base, e.g. gaave! ‘take’ not to be confused with the cases described under R4). Some imperatives double a medial consonant, although these are always in morph-initial position, e.g. hanno ‘go’.
There is no incompatibility between R2 and R3, although most B which take R2 do not also take R3.
2.6.2. Decade 20 contains preposed minor morphemes as follows:
21. de (negative) is obligatorily paired with some member of decade 30, occurring with all members of this decade. It thus enjoys the same freedom of occurrence as decade 30. It may immediately precede 31, usually emphasizing the negation slightly thereby. It also occurs in the form te + 31 when the action was completed in the past.
23. a (personal) occurs directly before the familiar form of personal names. A person's name which has not been contracted and thus rendered euphonous can not take this article. So, for example, if a “Mister Smith” is a government official and is always so called and so referred to, then a Mister Smith is considered euphonous; but one could not say in referring to a local person named Anton, *a Anton, because the familiar form of his name is Dono; a Dono would be proper. A is also used before kinship terms and pronouns and other words of similar import when these are used as euphemisms for a proper name (it being impolite to mention a person's name within his hearing).
a dangada la ‘those people’ (referring to specific ones).
a is never required, although it is considered better style to use it whenever appropriate.
25. ha. This article is known to occur before only one base in Nukuoro, hine ‘female’. One may say ha hine or de hine or se hine, although one occasionally hears de hahine, se hahine.
2.6.3. Decade 30 (verbal aspects) contains preposed minor morphemes occurring in sequence with V and A. 31-33 especially can quite definitely not be regarded as merely “tenses”.
31. e (general aspect) is the most frequently occurring member of this decade, marking only the verbal function of a V or A. It is neutral with respect to time and intention, often merely definitional. Au e unu ‘I am drinking’ or ‘I am going to drink’ or ‘I was drinking’ or ‘I will drink’ or ‘I drink’, etc.
Verbal expressions in sequence repeat their aspect markers: la e hano e unu e hebagi ai ‘he went and drank and fought’ or ‘he will go and drink and fight’.
Before An, 31 indicates the cardinal, as opposed to the ordinal function of the numeral: Tolulaangi ‘Wednesday’ (i.e. the third day of the week) e dolu laangi ‘three days’.
E is also joined to V and A to produce a citational form. Often, but not always, an informant asked for the Nukuoro word for, e.g. ‘climb’ will reply e gage.
32. ga (anticipatory) is used where the verbal idea is indefinite, where permission or assent is being sought, or to indicate the passage of time. Au ga unu ‘may I drink?’ or ‘I have been thinking of drinking’ or ‘I will drink (later)’, etc.
De vaga ne malanga mai i Bonibei galava ga humai ai. ‘the ship has left (coming hither) Ponape and then will come here’ or ‘. . . and then came here’.
Ga sometimes serves as a member of 90, being equivalent to 97: Ga mailele na hu ia, gai au ga dangi ‘if he dies, I will cry’; Ga de na hu dau hai ma goe e de dau ‘if this is your way of doing, then you are not included (in the group)’.
Interposed between An, ga serves to enumerate: ga dahi, ga lua, ga dolu . . . ‘one and two and three . . .’. There is occasionally a non-final - 211 pause after ga in the narrative style. This is either hesitation or pause for dramatic effect.
33. gu (decisive) indicates that one's intentions are firm, that the action has continued for some time in the past or has been accomplished, or that some state has existed for some time, or to indicate an increment to the verbal expression. Ga daa na hu de ono, gai au gu hano ai lo ‘at six o'clock, I will go’; Godou gu hiikai ‘you have been hungry for some time (you are famished)’; Au gu mmae ‘it really hurts me’. cf. Au e mmae ‘it hurts me’.
34. ne (perfect) indicates completed action, rather underlining the aspect of completion. De hale nei ne baguu anaahi ‘this house collapsed yesterday’.
35. gi (prescriptive) indicates that someone else wants the actor to act thus. It is usually translated ‘should’. Gi dabu Doo ingoo, ‘hallowed be Thy name’; Bolo i de ia, goe gi seesee ange ‘he says that you should walk up to his place’.
There is little essential difference between 35 and 81. They are separated here only for convenience, in anticipation of our description of the whole utterance. In dependent clauses gi often means ‘in order to’. Ne hai de Laangidabu gi hagammabu ai dangada ‘Sunday was made for the people to rest’.
In the following example it is quite impossible to decide whether the final gi is 35 or 81. Tadeu dagi e hagadaba bolo godou gi hebagi gi kii ai godou ‘our leader says that you should fight and should (also) win’ or ‘. . . fight in order to win’.
Gi behaves as a member of 90 in introducing dependent clauses: Gi go au na go goe, au ne hano donu ‘if I were you, I would surely have gone’.
The freedom of movement in the periphery which characterises 33 and 35 is especially characteristic of 36-39x. These aspects, of more circumscribed semantic, appear to have been combinations of minor morphemes which began in the extreme periphery and worked themselves inward until they now function as the other members of 30. There is normally a slight difference in meaning associated with any difference in the periphery e.g. kana giladeu e hiigai ‘It is feared that they are (always) hungry’ but Giladeu kana hiigaii ‘It is feared that they are hungry (at this time)’.
36. nogo (past progressive). Anaboo nogo bala-langi ‘last night it was raining’.
37. tigi (‘not yet’). Denga mamu tigi loomai ‘the fish have not yet come’.
38. kana (warning). Preceded by a subject this aspect is usually best translated ‘don't ever’. Without a subject it is usually best translated by ‘lest’. Goe kana hagasenga donu au ‘don't you ever deceive me’; Au de e dugua dau dama gi gaugau sogosogo, kana malemo hu ia ‘don't ever let your child go swimming alone, lest he drown’.- 212
39. mele (hypothetical) is often translated ‘one assumes’ or ‘it would appear that’; Mele hanu mamu i kinei ‘one has the feeling that there are fish here’; De henua nei mele hanu magi iai ‘this country appears to have some sickness’.
39x. goi (‘still’) Goi lagolago hu agu baibu ‘I still have lots of cigarettes’.
As mentioned above, 36-39x are incompletely assimilated to their status in this decade. One clear indication of this is the fact that they can occur in sequence with some members of 31-35. The following exhaust the possibilities:
2.6.4. Decades 40 and 50 contain morphemes that combine with each other and with 60 and with 70 to yield words we choose to call ‘pronouns’. Our definition of the contour word leaves all pronouns except those immediately preceding bases (in practice N only) in the nuclear position where they need not be considered as peripheral minor morphemes at all. It is, however, convenient to consider them as minor morphemes and to analyze the combinatory possibilities of all pronoun forms at the same time. The formulae below express the possibilities for co-occurrence of minor morphemes of decades 40-70. Subsequent paragraphs indicate the possibilities for the co-occurrence of pronouns with other minor morphemes and with bases. Finally, all the combinations of 40-70 are spelled out for convenient reference
Pronouns are thought of as
a. (40.1) or (78 + 50 + 45-6) or (78 + 48.2) or b. (71.3 or 72 or 73.3) + 60) + ((40.3 or (50 + 45-6) or d. 60 + 48
For future convenience, we shall isolate a subtype of (b), consisting of (b) minus the term italicised; this sub-type we label ‘c’.
The following formulae indicate the possibilities for co-occurrence of each of the above types with other preposed minor morphemes within a single contour word (marked /.../) and within certain longer spans.- 213
/ + (81* or 82* or 83* or 86) ± 77** + pi / (*40.1 obligatorily preceded by 71 if 81 or 82 present)
± 77** + Pi / + 30 ± 21 + (V or A) / (except before gida) (V or A) ± Pi
... Pp + N /
/... 74 + Pd ± N /
/ 84 or 85 + Pd ± B /
/ Pd + 30 ± 21 + (V or A) /
... Pt + B
81 + Pt
It is understood that all members of 90 and 83 can occur in sequence with all pronoun forms.
When we spell out all of the possibilities expressed by the formulae above (considering the notes on allomorphs in the next section) we are given the following forms:
Pd are isomorphic with the plural forms of Pp.
84 + Pd—possessive I
84 + Pd—possessive II
85 + Pd—possessive I
85 + Pd—possessive II
In contour final position
74 + Pd ...—possessive I
74 + Pd ...—possessive II
74 + Pd // When immediately followed by a base, the pronoun forms combining with se are as above. When the pronoun + se are in contour final position, the forms are se + Pd.
84 + Pt ...
84 + Pt //
85 + Pt ...
85 + Pt //
74 + Pt (in fact, dada and dodo are used in many places that would appear to require 74 + Pt). This form lacks plurals, the plural of Pt substituting for it in appropriate cases.
74 + Pt ...
74 + Pt //
2.6.5. This section lists the supplementary notes on allomorphs necessary to generate the above forms from the listed minor morphemes. Notes are arranged according to the index number of the minor morpheme concerned.
85. (85 + Pp + N) is about equal in meaning to (84 + Pd + N)
ni agu biigi
‘for my pigs’
= ma agu biigi
only (N + 85 + Pd//) carries the sense ‘belonging to’.
71. Alternant 71.3 is invariably selected by singular Pp. First and third person plural (singular object) select 71.1. As remarked elsewhere de + d . . . is always contracted to t . . . Alternate 71.4 is selected by second person plural.
73. 73.3 is selected by all P.
74. Like 71, alternant 74.3 is selected by singular P. The first and third person plural select 74.1. The second person plural selects 74.4.
75. In contour final position (ni + Pd) occurs. With a base following, ni is omitted.
78. 78.1 occurs in the first and third person. 78.4 in the second person.
60. For a singular object, 60 combines with Pp only in the singular. For dual or plural object, 60 combines with all forms. The second person invariably selects 61 in the plural. In the case of 74 + Pd, 60 is omitted in the plural. Following 84 and 85, 61 and 62 are doubled in all forms; where this results in *ooo, rewrite as oou.
42. 42.3 occurs after 61. 42.4 occurs after 62.
43. 43.3 occurs after 61. 43.4 occurs after 62 and after 61 when in contour final position.
45. 45.1 occurs in the first and third persons. 45.2 in the second person.
46. 46.1 occurs in the first and third persons. 46.2 in the second person.
48. 48 is obligatorily paired with 60 and either 70 or 84-5. 48.1 is selected by 61. 48.2 is selected by 62 and in all cases where there is a following base.
2.6.6. Decade 60.
Alone, 60 occurs only preceding N. It is obligatorily paired with a number marker (a member of decade 70) when preceding all N except Nn and Np. In the construction N + 60 + N, it agrees with the noun preceding (which by our definition is in the preceding contour word) as in De naivi a Soan ‘John's knife’; Tama a Deagu ‘Deagu's child’; Se malo ni o tangada la ‘it's a piece of clothing of that person’. In (± 70 + 60 + (50 ± 40)), 60 occurs in the sequences 60 + N, 60 + A, 60 + 30 ± 21 + V. In these three cases 60 agrees with the nucleus of its own contour word. When following 85 (ni), 62 may be omitted (see section 2.6.8), except when 50 follows.
Many words take both 61 and 62 with some difference in meaning. De kai o Vave ‘the story about Vave’; De kai a Vave ‘Vave's story (i.e. the one he made up or tells)’.
The following examples of constructions in which both 61 and 62 might occur shed further light on their semantic. dogu hai ‘how it should be done with regard to me’, dagu hai ‘my way of doing’; oo muna ‘words said about you’, au muna ‘your words’ (i.e. what you said); ogu lodo ‘my will’, agu lodo ‘will of others concerning me’; dogu dao ‘my spear’, dagu dao ‘the spear made by me’; dogu bodu ‘my wife’, dagu bodu ‘the wife’ e.g. that I am looking for for someone else; dogu ngavesi ‘my coffin’, - 217 dagu ngavesi ‘my storage box’; de naivi ogu e tuu ai ‘the knife used for cutting me’, de naivi agu e tuu ai ‘my knife that I cut with’; de nui ogu e gage ai ‘the coconut tree for me to climb, de nui agu e gage ai ‘the coconut tree which I climb’; dono bodobodo ‘his shortness’ (a permanent characteristic), dana bodobodo ‘his shortness’ (a temporary condition, e.g. when hunched over).
2.6.7. Decade 70. The preposed minor morphemes of this decade are articles of one kind or another (except 77 which is considered with the pronouns in sections 2.6.4-5). All of them occur in sequence with every base class, excepting 72 and 73 which do not occur in sequence with V N1 can not be preceded by members of 70 other than 74-75.
71. de (‘the’—sing).
71 + N: de nui ‘the coconut tree’
71 + A: De laanui o de hale la se lau dangada e oo ange gi ono lodo ‘the bigness of that house is (enough for) 100 people to be accommodated inside’.
71 + V: De haula de laama i de laangi e de heohi ‘lighting a lamp in the daytime is not right’.
72. lu (‘the’—dual).
72 + N: lu daane la ‘those two men’.
72 + A: lu mmea la se dagodo gege ‘those two reds are different’ (from each other).
73. denga (‘the’—plural). Although it is better form to use 72 wherever it is appropriate, 73 can be used in referring to two or more. Denga also means ‘each’.
73 + N: Denga huaabodu nei donu hu ma de hai hegau ange gi aladeu gai ‘each family will be responsible for for its own food’.
73 + A: Denga gelo agu e sulu gai goe e de maua ‘the deeps my diving but you can't (i.e. I can dive to greater depths than you can)’.
a (73.5) occurs before minor morphemes as follows: agai (91) ‘and then, finally . . .’, ama (93) ‘and, among other things . . .’, abe (95) ‘or, to cite only one more example’. Note also anei (271) preceding N is the plural of de nei ‘this’, ana (272) preceding N is the plural of de na ‘that (near you)’, ala (273) preceding N is the plural of de la ‘that (away from us both)’. a also occurs following 93. This construction invariably follows a N, excluding other postposed minor morphemes and ending the contour.
ø (73.3) occurs following members of 80, in pronouns, and preceding the object of a verbal construction. i de holau ‘in the canoe house’ but i holau ‘in the canoe houses’; e tilo tahola ‘look at the whale’ but tilo dahola ‘look at the whales’.- 218
74. se is a general article, often best translated ‘it is . . .’ Se aha de me la? ‘what is that thing?’ Se hala ‘it is pandanus’. Deai donu se mamu anailanei ‘there are no fish at all today’; Ga se hano hu goe e hagadahao ma e deai ‘if you are going to fool around, then (the answer is) no’.
Se is often used as the citational form of a noun. If one asks a Nukuoro to cite a noun, he will usually reply ‘se . . .’.
Se also acts like a member of 30: Au se hano donu hu e hagadahao ‘I went planning just to amuse myself’. Au e hano would imply the lack of a definite plan.
75. ni is the plural of se. Ni aha agu e hai? ‘what (things) are there for me to do?’; Ni maanadu baubau i ono lodo ‘there are evil thoughts inside him’.
Ni is used frequently in citing something as a member of a class. Ni me hhadu ‘made-up things’ (referring only to a single thing) i.e. ‘it's a lie’.
76. dahi (‘one’, ‘a’) is considered an article when it is not preceded by 31 (in which latter case it is An). Dahi laangi hu dahi dangada ge gu doo age agina ‘one day a person from elsewhere beached there’.
77. hanu (‘some’) is considered an article when it is not preceded by 31 (in which case it is An).
78. is considered in section 2.6.4 with other pronoun particles.
It is at this point convenient to define a notion which we label “number marker”.
Every N base (except N sub-classes) not occurring in sequence with another base is either preceded by a member of decade 70 (including 73.3 (ø)), or 83, or followed by an A base whose form indicates whether it is singular or plural. Collectively, these indications of number are referred to as “number markers”. A rigorous definition of the contour word (section 2.2) would note, that for the purposes of determining contour word boundaries, all number markers are considered members of decade 70. Under certain circumstances, specified passim, one sort of person marker may be omitted when another sort occurs in sequence with a N.
2.6.8. Decade 80. This decade comprises preposed minor morphemes serving a wide variety of relational functions. All of them except 84-5 occur in sequence with bases of all classes.
81. gi ‘to’ (see also section 2.6.3).
82. i ‘at’.
Before N, 81 may express motion toward and 82 may express location. Except in the case of Nn, Np and N1 70 (71-3 or 76) must intervene. 73.3 is the allomorph of 73 most often selected for plural N. gi de vaga ‘to the ship’; i de momme danua ‘in the good place’; gi Soan ‘to John’; i Luugu ‘at Truk’.
Before V (and following a V or A), gi means in order to’. Au de e savini gi hagailoo ina ‘don't run to make this known’.- 219
After certain verbal expressions (82 + N) or (82 + P) indicates the referent of the action, or the means of the action. e aloha i de goe ‘loves you’; lili i manu-llele ‘mad at the birds’; pili i de tala ange ‘bored of telling’.
Both 81 and 82 are used after certain A with 81 indicating the goal of the action. De hodooligi ne basa ange i de hainga gi de gau la ‘the chief spoke about the law to the people there’.
82 + N1 + 70 + N is a common locative construction. I honga de hale daumaha ‘on top of the house for church services’ (i.e. the church).
81 acts as a member of 90 with the meaning ‘if’, as indicated in section 2.6.3. As such it may precede 80.
83. go precedes bases of all classes and P serving to emphasize or set off the word it precedes. It is usually best translated by ‘it is . . .’. Non-final juncture may occur just after 83 when N or P follow. Juncture is especially frequent preceding Np. Juncture in these cases serves to emphasize the identity of that referred to by the N or P following. Go Soan gu hadu muna ‘it was John who lied’. We might easily have construed 19x goi as 83 + 82. Goi muli hu oou iai ‘you are still behind’ (lit. it is at the late time, your being). Although this was surely the origin of 19x, we find it more convenient to consider goi as a member of decade 10 than as 83 + 82 because it occurs only immediately before A or V.
84. m- is obligatorily paired with 60 ± (± 70 + 50 + 40), the various permutations of which are discussed in section 2.6.4. The general sense of this morpheme is one of future possession and it is usually best translated ‘for’. It occurs only immediately before or after N. Dahi ubu maagu ‘a cup for me’; Giladeu e vange me maimai maaladeu dama ‘they gave sweets to their children (for their use)’; Ia e vange dono hale mo de nohoangadabu ‘he gave his house to the church for its use’; Tama la gu tili ange dana masese balanga ma Soan ‘that kid gave away his lighter to John’.
Ma between N means ‘and’ and is indistinguishable from 93 which could be construed as (84 + 62).
85. ni is obligatorily paired with 60 or 70. Its general sense is one of present possession and it is usually best translated ‘belonging to’.
The difference between (85 + 60) and 60 is indicated by the following examples. De me pasa a Soan ‘John's radio’; De me pasa ni a Soan ‘the radio belonging to John’.
Note that there is no difference in meaning between ma and ni when immediately followed by N. ni tangada la=ma tangada la ‘for that man’.
In the construction (85 + 60 + N), wherever 61 is appropriate, it must be expressed. However, when 62 is appropriate it may in some cases be omitted and in other cases must be omitted. ni o tangada maolunga de bodu la ‘that wife of an important man’; ni (a) tangada maolunga tama nei ‘this child of an important man’; de hale ni tangada la ‘his house (that he built)’; de hale ni o tangada la ‘his house (belonging to him)’.- 220
86. e (agent) indicates the actor in a passive construction. Ia e gage de nui or E gage ia de nui ‘he climbs the tree’; De nui e gagea e ia ‘the tree was climbed by him’.
When in sequence with a phrase introduced by 86, 211 is normally affixed to V.
87. ee (vocative) occurs before bases of all classes, also in the expression: Ee la! ‘So!’
2.6.9. Decade 90. This decade contains introductory expressions, and word and phrase connectives. Non-final juncture may occur following 91, 99 and sometimes 95; final juncture normally precedes them. 91 and 99 must either be preceded or followed by final or non-final juncture. 93 may be followed by non-final juncture. All members of this decade can occur in sequence with bases of all classes.
91. gai (‘then’, ‘and so’, ‘and’, ‘but’) is a loose connective which occurs only in phrase initial position. Its use is amply illustrated in the appended text.
93. ma (‘and’) is a close conjunctive. N and A (except Np and Nn) so connected are obligatorily preceded by 70 (or a number marker); this feature is optional when V are so connected. Occasionally in narrative non-final juncture occurs after 93, either as hesitation or for dramatic effect.
N + 93 + N: de gahudi ma tolo ‘the banana and the sugarcane’.
N + 93 + A: de maalama ma de gohu ‘light and darkness’.
V + 93 + V: Ssulu ma ssiisii ni hegau laosedi ‘diving and fishing are aquatic occupations’.
Although 93 normally joins words it may be used as a phrase connective with the meaning ‘and so’, ‘and thus’, ‘and then’. Ia gu magi ma ia de e humai ai lo ‘he was sick and so he didn't come’.
Following P, ma means ‘including’. Golu ma ai e hulo? ‘you and who else went?’
Paired with 35, the expression is best translated ‘in order to’. Gaavee dagu denggi ma gi hano ‘take my flashlight to go’.
94. matali (‘along with’, ‘accompanying’) occurs in sequence with Pi, Nn, and N of similar import following verbal expressions of motion. Following verbal expressions not implying motion, it means ‘in addition to’ or ‘as well’, or ‘at the same time as’. Seesee matali gidadeu ‘walk with us’; Au e iloo matali ia ‘I know in addition to him’ (i.e. I also know!); De hale nei ne hai matali taonga ‘this house was made at the same time as the party’.
95. be (‘or’) is a word disjunctive. N (except Np and Nn) so connected are obligatorily preceded by 70. In addition the second member of the conjunction must be preceded by 83.
N + 95 + N: de gaago be go de biigi ‘the chicken or the pig’.
A + 95 + A: mouli be magau ‘living or dead’.
V + 95 + V: daadaa be solo ‘scrape or grate’.- 221
95. . . . 95 forms an ‘either . . . or’ construction. Be ia e lodo be deai ‘either he wants to or he doesn't’.
97. no ('if). No hale nei ni ou ‘if these houses are yours’.
Note that N following 97 select 73.3 (ø) as the plural article.
99. bolo (‘that’, ‘according to’, ‘“. . .”’). Following such verbal expressions as maanadu ‘think’, haiange ‘say’, ssili ‘ask’, etc., bolo introduces the predicate. Au e maanadu bolo de vaga e humai ‘I think that the ship is coming’.
Preceding P, Nn or equivalent expressions, it means ‘according to’ or ‘says’. Bolo i de ia, goe de e hano ‘according to him, you are not going’.
Preceding other expressions, bolo indicates that the material following is a quotation. Bolo uu ‘he (or she or I, etc.) said, “yes”’.
Followed by i, the expression means, ‘it is believed’. Bolo i dangada e dahi eidu i lo te lodo ‘people believe there is a ghost in the lagoon’.
Bolo can also mean ‘let's pretend’. Bolo goe se dangada abasasa ‘let's pretend you are a European’.
Members of 90 occur in sequence within the same contour word as follows. Where different phrases are glossed the same, informants insisted there was “no difference”.
gai be; be no: gai be no ia gu magi ‘or perhaps he is sick’; be gai; gai no: be gai no ia gu magi ‘or perhaps he is sick’; gai bolo: gai bolo au gi hanange ‘and it was said that I should come’; bolo gai: Bolo gai goi hanu ange donu ‘and according to him, there are still some more’; bolo be: Bolo be hanu la au baibu ‘he wants to know if you have cigarettes’; ma no: Ma no se eidu ‘and perhaps he is a ghost’; no be: no be ia tigi iloo ‘if perhaps he doesn't yet know’; bolo no: Bolo no e hanu i oo daha ‘he wants to know if there is any at your place’; no bolo: No bolo e hanu i oo daha ‘he wants to know if there is any at your place’.
Other combinations of members of decade 90 are not possible.
2.6.10. Decade 210. Suffixes. Bases of all classes may take 210. Any base + 211 usually becomes V. Any base + 213 always becomes N. Allomorphs of 211 and 213 are listed in section 2.3.
211 -a and its allomorphs. Suffixes of this sort are more profitably considered ‘transitivizing’ than ‘passive’. A passive is provided by 13 + V. Examples: gaadia (kadi ‘bite’); masaia (mmasa ‘dry’); hagasegedia (hagasege ‘let slip’); boogia (poo ‘grab’); sigohia (sigo ‘catch’); daulia (dau ‘count’); unumia (unu ‘drink’); daadaangia (dada ‘pull’); milosia (mmilo ‘flow’); gaina (gai ‘eat’). The addition of a transitive ending to a stem may entail the reduplication or un-reduplication of an initial consonant or a following vowel, as some of the above forms illustrate. Occasionally more radical changes are entailed: laangia (<lala ‘roast’); suungia (<ssui ‘wet’). More than one allomorph of 211 may be selected by the same stem. In some of these cases individual Nukuoro disagree as to which is the correct form while agreeing that there is no difference in meaning between the two forms, e.g. hulia (or hulisia) ‘to turn something over’; oga (or ogohia) ‘to husk’. In some cases people agree that two - 222 forms are equally correct and that there is no difference in meaning between them, e.g. galohia (or gaaloa) ‘to stir’; gabidia (or gabia) ‘a way of carrying something’. In still other cases different transitive endings are selected by the same stem, each resulting form having a distinct maning: ulu ‘to go inside’, ulua ‘the condition of entering’, uluia ‘turning the inside out’; kumi ‘squeeze’, gumidia ‘grab it’, guumia ‘to figure out’. The transitive form of many intransitive verbals conveys a meaning quite different from that of the stem: magi ‘sick’, magia ‘jealous’; gai ‘eat’, gaia ‘steal’; lele ‘fly’, leengia ‘to be hit by flying fish’; sao ‘escape’, saohia ‘to have free time’; mmala ‘bitter’, malaia ‘cursed’.
Most usually the addition of 211 to a stem indicates that the action is or will be definitely accomplished, in that it is the responsibility or intention of someone to do so. Ia e gage de nui ‘he climbs (or will climb, or did climb) the coconut tree’; Ia e gagea de nui ‘it is for him to climb the tree’.
With 272, which indicates some future time, 211 indicates a certainty that the action will be accomplished. Ia e gagea na de nui ‘he will definitely climb the tree’.
213 nga- Examples: seenganga ‘kind of craziness’ (senga ‘crazy’); daaligidanga ‘way of chipping with an adze’ (daaligi ‘to chip at’); daulanga ‘place for beaching’ (dau ‘to beach’); sulumanga ‘place where the sun sets’ (sulu ‘dive’); alohanga ‘way of paddling’ (alo ‘to paddle’); milosanga ‘way of flowing’ (milo ‘to flow’). Nearly every A and V stem can take-nga, at least in some idiolects. Other allomorphs are sometimes selected by some individuals who would argue that the allomorph-nga suffixed to that stem is wrong, e.g. huudanga ‘way of singing’ (while some say huaanga). Some stems, all Nukuoro would agree, select more than one allomorph of 213 with no difference in meaning between the forms, e.g. anumanga (or aanunga) ‘way of dancing’. Some stems select more than one allomorph of 213, entailing some difference in meaning, e.g. holo ‘swallow’, hoolonga ‘way of swallowing food’, holomanga ‘throat’; unu ‘drink’, uununga ‘way of drinking’, unumanga ‘a group of people drinking coconut toddy’.
-nga occurs in slot 10 with bases and with certain minor morphemes with which it combines to form a base surrogate. daa ‘beat’, ngadaa ‘difficult’; iho ‘down’, ngaiho ‘downtown’; dala ‘tame’, ngadala ‘a species of fish, reputed to be very tame’.
2.6.11. Decade 220.
221. ge (‘different’) is a postposed minor morpheme occurring in sequence with bases of all classes. It is usually best translated ‘different’ or ‘of a different sort’.
221 + N: tangada ge ‘stranger’ (tangada ‘person’).
221 + A: uliuli ge ‘a strange sort of black’.
Only occasionally does ge combine idiomatically with bases.
hano ge ‘go away from’ (hano ‘go’), i.e. ‘go, not (as usual) returning’.
221 + V: dui ge ‘sew in a different fashion’ (dui ‘sew’).- 223
2.6.12. Decade 230.
231. ina (‘to it’) is a postposed minor morpheme. It is considered divisive for V bases, whose distribution is otherwise identical to A bases. Go ai ne boogia ina? ‘who caught (grabbing) it?’; Go au ne daala ina ‘it was I who left it free’.
While 211 most usually follows a V in sequence with ina, this is not necessarily the case. Sigo ina doo uga ‘pull your fishing line in (any which way)’; Sigoa ina doo uga ‘pull your fishing line in (the speaker having a clear idea of how this should be done)’.
2.6.13. Decade 240. This decade contains postposed minor morphemes which occur in sequence with bases of all classes.
241. iho ‘down’.
242. age ‘up’.
240 + V: gage age ‘climb up’; maanadu age ‘think up’ (or ‘decide by oneself’); maanadu iho ‘plan’ (or ‘decide in a group’); tilo age ‘look up at’; tilo iho ‘look down upon’.
240 + N: muna age ‘word from below’; muna iho ‘word from above’.
240 + A: doo iho ‘to fall’ (doo ‘change position’); mao age ‘further up’ (mao ‘far’); mao iho ‘further down’; ssui iho ‘getting wet from the top down’ (ssui ‘wet’); ssui age 'getting wet from the bottom up; bobo iho ‘rotten from the top down’ (bobo ‘rotten’); bobo age ‘rotten from the bottom up’.
2.6.14. Decade 250. This decade contains peripheral minor morphemes of very high frequency. Members of this decade indicate relative direction.
251. mai ‘to me’, ‘to us’.
252. adu ‘to you’, ‘away from me’.
253. ange ‘to him’, ‘to them’, ‘away from you and me’. maanadu mai ‘think of me’; maanadu adu ‘think of you’; maanadu ange ‘think of him’.
Some base stems are obligatorily paired with a member of 250. These appear to have been originally minor morphemes, although they may be the remnants of longer forms. gamai ‘give me’, ‘give us’; gavadu ‘give you’; gavange ‘give him’, ‘give her’, ‘give them’, ‘give’.
It should be understood that the pronominal referent is conceived of rather differently from English. e.g.: Au ga humai ‘I shall come to us’ (the speaker thinking of himself as already in the place to which he is coming) is sometimes used in place of au ga hanadu ‘I shall go to you’.
In relating an event which does not involve the listener, once a third person is designated, that person tends to become ‘you’ (adu). For example, in telling me about a picnic which I did not attend, someone might say, “au gu seesee adu . . .” (‘I walked to it’). Thus adu is the contrary of mai, as the following expression indicates: hano ge adu ‘go differently from where one is’ (i.e. ‘go away’). Ange, from this point of view, is the contrary of mai + adu. While mai and adu indicate relatively finite directions, ange often is simply neutral with respect to direction. tala ‘tell’; tala ange ‘relate’, ‘recount’; basa ‘speak’, basa ange ‘answer’.- 224
After V bases denoting change of state, ange is augmentative and mai is diminutive. uea ‘start something’, uea ange ‘increase’, uea mai ‘decrease’.
After adjectival expressions, 250 may indicate perspective. When followed by 82 (i), (A + 250) indicates comparison. laanui mai ‘appearing, from where we are, to be big’; Au e madua ange i de ia ‘I am older than he’.
In an expression where location is indicated, 250 may indicate relative location. de hale mai la ‘the house next door closer to us than that one’; hanonga mai ange nei or hanonga ange mai nei ‘the very next time’.
Ange may also mean ‘in addition’ or ‘again’. dahi hanonga ange ‘one more time’ or ‘another time’; goe ange hogi ‘you again also’ (i.e. ‘you too’).
Ange followed by gi, m- or i can mean simply ‘to be given to’. ange gi Soses ‘(letter addressed) to Soses’.
Ange following 21 + V can mean ‘never’. Au gu de gai ange de baibu ‘I will never smoke’.
Following certain verbal expressions, ange means ‘about’, ‘concerning’. Au ne midia ange hanu daala lagolago anaboo ‘I dreamed about lots of dollars last night’.
Members of this decade can occur in sequence where several of the above meanings are compatible. Ga-v-ange ange ange hanu gai gi tau soa ‘give (out) again (to) some food to our friends’; gauligi mai ange la i de ia ‘the next youngest to him’.
2.6.15. Decade 260.
261. ai is a postposed minor morpheme occurring in sequence with bases of all classes.
Following verbal expressions it simply focuses the action. One might say au e lodo ‘I want to’ or au e kino ‘I hate it’ in answer to questions, but they sound vaguely incomplete, as though the object of one's desire or hatred were about to follow. The addition of ai to the phrases above make them sound more complete. In many cases, ai may not be omitted from a well-formed sentence. e.g. De hee ai tau hai e hai ai? ‘what (else) shall we (two) do then?’ (having tried something that didn't work). In this sentence, the final ai may not be omitted. If the first ai is omitted, the sentence would mean ‘how are we going to do this thing?’.
Se me aha? ‘what is its use?’ (it obviously having some use); Se me aha ai? ‘what is its use?’ (its usefulness not being clear); De hau la ona e kaianeane ai ‘that lei is something that he is proud of’; Gu aha gu modu ai doo vae? ‘why is your foot cut?’.
Ai does not occur in sequence with 211.
Ai enters into combination with a number of preposed minor morphemes, forming base surrogates. Niai ‘whose’, goai ‘who?’, iai ‘there is’.
In the construction (N + ai), ai has the sense ‘finally’. Bonibei gu de maua i de basa mai; go Luugu ai iainei e basa mai ‘Ponape couldn't talk to me (on the radio); in the end it was Truk which talked to me’.- 225
2.6.16. Decade 270. This decade contains postposed minor morphemes of very high frequency. Members of this decade indicate absolute location.
In sequence with N and A, the location is spatial; in sequence with verbal expressions, the location is temporal. de hale nei ‘this house’; de hale na ‘that house near you’; de hale la ‘that house over there’. laanui nei ‘is this big?’; laanui na ‘is it big?’; laanui la ‘is that big?’. laanui nei de me nei? ‘is this big enough for that purpose?’; laanui la de me nei ‘this thing is big in comparison with that one’. au ga hano nei ‘I am going now’ (at this time); au ga hano na ‘at the time when I go’; au ga hano la ‘at the time when I went’; go au e gagea nei ‘am I supposed to climb (right now)?’; au e gagea na ‘I will climb for sure’. go au e gagea la ‘was it I who was supposed to climb (in the past)?’. It is not possible to say, *go au e gagea na or *au e gagea nei or *au e gagea la, as these forms would be redundant.
It will be noted from the examples involving verbal expressions that, whereas nei indicates definite present time and la definite past time, na indicates indefinite time, hypothetical, future, or unknown. From this point of view, nei and la are contraries and na is the contrary of them both taken together. In many cases it appears that na indicates progressive action (cf. Maaori ana), but it does so in just about the same sense as the English ‘going to . . .’, a vaguely future idea. Goai na e vaahano? ‘who (among us) is able to go?’; (Ga) hano na goe gi hee? ‘where are you going?’.
2.6.17. Decade 280.
281. donu is a postposed minor morpheme occurring in sequence with all classes of bases. It is usually best translated idiomatically by ‘real’ or ‘really’.
281 + A: mahamaha donu ‘really beautiful’; lagolago donu ‘really a lot’.
281 + V: e hai donu ‘really going to do it’; dangia donu ‘really pull!’.
281 + N: daina donu ‘full siblings’; mamu donu ‘authentic fish’.
Donu occurs in the nucleus as an A with the meaning ‘understand’. Goe gu donu mai? ‘do you understand me?’.
2.6.18. Decade 290. The members of this decade are postposed minor morphemes occurring in sequence with all base classes.
291. lo like ai is a kind of predicate complement meaning ‘in it’, ‘about it’, ‘for it’, or some such. In some sentences it appears to substitute for a verbal object (i + B), where the whole expression is not used. au e kino ai lo ‘I hate it’; haia ina lo oo lodo ‘do it as you wish’; au de hai haia ngau ‘excuse me’; au de hai haia lo au ‘excuse me for it’ i.e. ‘forgive me’.- 226
Lo is a polite substitute for gi. hano gi haia dagu hegau ‘go and do my work’; hanno lo haia dagu hegau ‘please go and do my work’.
Both lo and a phrase for which it might substitute occur together, additional clarity or graciousness of expression being gained thereby. gamai ai lo dagu beebaa ‘give me my book’; gu gohu lo de moana ‘the sea is becoming dark’; au gu maingaa ange lo dagu mamu ‘I just missed getting a fish’.
Lo followed by te is surely a contraction of lodo (Nl) + de . . .
293. hee ‘where?’, ‘what kind?’, ‘which?’.
293 + N: tangada hee? ‘person from where?’
293 + A: mmea hee? ‘what kind of red?’
293 + V: maga hee? ‘what manner of throwing?’
2.6.19. Decade 300. This decade contains postposed minor morphemes which invariably terminate a contour word. Like the members of decade 90 they occur together. Members of this decade occur in sequence with bases of all classes.
301. hu ‘occasion when’.
Goi kii hu goe i de mmahi i golu. ‘You are now the strongest of you two’.
Hanno, gai au e tali ai na hu goe. ‘Go, but keep in mind that I will be waiting for you (to return)’.
Hu is also used in constructions to indicate supposition or especially when following an imperative form, to indicate continuing action.
Goai ai hu ne humai la? ‘Who might it have been who came in that case?’
Daalia hu au. ‘Keep waiting for me’.
Donu hu is an idiom meaning ‘only’: Kana iho donu hu gi de goe! ‘Mind only your own business’. Preceded by the name of a person or society hu means ‘just like . . .’: Gelenisi hu ‘just like the Kapingamarangi people (i.e. inquisitive)’.
303. hogi ‘also’.
305. saunoa ‘completely’.
hu hogi saunoa, hu saunoa hogi, saunoa hu hogi, or any two member fraction are possible combinations. Other permutations are not possible. The above three phrases are considered semantically identical by informants. Ia e savini hu hogi saunoa ‘he also ran completely at that time’.
(to be continued)
1 Christian, Frederick William. “Nuku-oro vocabulary.” Jour. Poly. Soc., 7 (1898), pp. 224-'2.
2 Kubary, J. Beitrag zur Kenntniss der Nukuoro oder Monteverde-Inseln. Mittheilungen der Geographischen Gesellschaft in Hamburg, 16 (1900), pp. 71-138. Eilers, Anneliese. “Nukuor”, pp. 157-318 of Die Inseln um Ponape. Ergebnisse der Sudsee-Expedition 1908-10, serie 11B, band 8, G. Thilenius, ed. 1934.
3 Elbert, Samuel H. Kapingamarangi and Nukuoro word-list. (Mimeographed.) U.S. Navy Department. 1946.
4 Biggs, Bruce. “The Structure of New Zealand Maaori.” Anthropological Linguistics, 3 (1961), no. 3, pp. 1-54.
5 Elbert, op. cit.
6 Biggs op. cit., p. 17.