We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 24 December 2009
I. Rarotongan is a Polynesian language spoken on the island of Rarotonga in the Cook Islands. A phonemic analysis yields a nine-term consonant system comprising four voiceless plosives (bilabial, dental, velar, glottal), three nasals (bilabial, alveolar, velar), a bilabial voiced fricative, and an alveolar flapped ‘r’. There is the usual Polynesian five-term vowel system (close front, close back, mid front, mid back, open central). All vowels may be either short (one mora) or long (two morae). Long vowels and short-vowel diphthongs behave phono-logically as dissyllables and are so treated in this paper. Syllable structures are limited to V and CV.
1 Glottalization and vowel lengthening occur in emphatic speech, and glottalization is frequent as an attack feature of phrase-initial vowel. These are not marked in the orthography (except, of course, in so far as they are implied by the punctuation and the general sense of the text).
2 See IV and V below.
1 -ua/-rua and -tou appear to be related to rua ‘two’ and toru ‘three’.
2 Some of these positions have wider syntactic validity, but they are set up here merely in order to state the distribution of the different pronominal forms, including their colligation with the personal article. (‘Colligate’ and ‘colligation’ are here applied to forms as well as categories; cf. Burton-Page, J., ‘Compound and conjunct verbs in Hindi’, BSOAS, XIX, 3, 1957, p. 476, n. 1).Google Scholar
3 A personal noun is defined as a member of a class having a distributional scatter over Positions 1, 2, 3, 4, 5, 6, and immediately preceded by the personal article in Positions 1 and 4. This class includes personal names, names of months, mea ‘so-and-so, whats-his-name’, ngāti ‘tribe’, and the interrogative ʔai ‘who ?’.
4 See XIV below, where the neutral complexes are discussed.
1 The syntax of the English translation carries no implications as to the syntax of the Rarotongan. In Texts 2 and 4, Rarotongan subject and agent are translated by English object and subject respectively.
1 The Rarotongan verb suffixes are: -a, -ngia, -ia, -ʔia, -kia, -mia, -na, -ria, -tia.
1 There may also be double apposition, e.g.
Kua ʔaere maiʔa Pā rāna ko Tere (v V d) [p P] [II] [ko P] Pā came with Tere.
The particle ma may be substituted for ko in this construction:
Kua ʔaere mai ʔa Pā rāua ma Tere.
1 The form of the first singular pronoun in this position is given here as ku. There are historical and comparative reasons for assuming that ku in this position may represent an earlier *aku: iāku being < iā *aku with reduction of the three-morae open vowel sequence to two morae. These are comparative considerations, however, and there would appear to be little justification for complicating a synchronie analysis by setting up yet a third first-singular form aku.
1 These observations are made solely with reference to Rarotongan, where the 0/A dichotomy often has semantic force. In a language where the meaning (everyday sense) of the distinction had been lost and there was left a purely grammatical system in which some nouns colligated with O-particles in all contexts and the remainder always with A-particles, the phenomenon would obviously be best dealt with in terms of nominal gender. It may be that in present-day Rarotongan (and widely in Polynesia) we have the linguistically interesting situation of a gender-system in the making—a half-way stage, where the O/A distinction, while apparently no longer semantically relevant in all contexts, has nevertheless not yet ossified into the purely mechanical colligation of given noun-class with given particle-class. Evidence that the O/A distinction is not felt to be compulsory in all contexts is provided by the existence of a pre-nominal possessive complex which is neutral to this distinction (see XIV below).
2 So, for instance, Churchward, S., Samoan grammar, 1951, pp. 25–6.Google Scholar For a somewhat different approach to the O/A-particles in Hawaiian as ‘markers of alienability’, see Pukui, M. K. and Elbert, S. H., Hawaiian-English dictionary, p. xix.Google Scholar Compare also the treatment of Tongan ‘subjective and objective possessives’ in Churchward, C. M.'s Tongan grammar, pp. 78–87 and 93–5.Google Scholar
1 Ngā nearly always means ‘a pair, a couple’, but very occasionally it is used in the sense of ‘just a few’.
1 # is the exponent of nominal plurality.
1 In this construction the nominal particle i, which marks the object (ʔare), is frequently omitted, e.g. Nāna i ʔoko mai te ʔare [nā II] (v V d) [c C]. It is also dropped when the object piece is placed before the verbal piece, an inversion which is frequent in interrogative sentences, e.g. Nā ʔai te ʔare i ʔoko mai? [nā P] [c C] (v V d)? ‘Who bought the house?’
1 e.g. Samoan
1 ʔou colligates with the ō morph and ʔau with the ā morph.
1 The first-cited morph (e.g. ua) marks the dual, the alternative (e.g. tou) marks the plural.
2 rua colligates with ko and tou with kou.
3 lua colligates with ʔo, and kou with ʔou.
4 The Maori forms are taken from W. L. and Williams, H. W., First lessons in Maori, 1940, p. 17Google Scholar, and the Hawaiian forms from Pukui, M. K. and Elbert, S. H., Hawaiian-English dictionary, p. xix.Google Scholar
