Rapanui is sometimes transcribed with long vowels; however, it is not clear that these are distinct from double vowels. Similarly with diphthongs: All vowel sequences are found. There are no final consonants in Rapanui. Syllables are restricted to the form vowel and consonant-vowel.
Within Eastern Polynesian, it is closest to Marquesan morphologically, although its phonology has more in common with New Zealand Māori, as both languages are relatively conservative in retaining consonants lost in other Eastern Polynesian languages.
Like all Polynesian languages, Rapa Nui has relatively few consonants. Uniquely for an Eastern Polynesian language, Rapa Nui has preserved the original glottal stop of Proto-Polynesian. It is a VSO language. Specific Rapa Nui features also include the change of use of anaphoric 'ai' to being a post verbal marker as well as the non-usage of any transitive suffixes, thus making it an ergative language and unlike any other Eastern Polynesian language which are accusative.
The most important recent book written about the language of Rapa Nui is Verónica du Feu's Rapanui (Descriptive Grammar) (ISBN 0-415-00011-4).
Very little is known about the Rapa Nui language in pre-contact European times. This is because there is no written documentation that demonstrates the phonology (sound system), morphology (word building system), syntax (how sentences are generated) or semantics (what words mean to native speakers). Through comparative linguistics however linguists are able to make confident assertions as to how the language may have sounded by comparing grammatical features found in Rapa Nui today to that of other Polynesian languages. They are also able to compare Rapa Nui vocabulary with that of other Polynesian languages to reconstruct Proto–Rapa Nui vocabulary. For example, the Rapa Nui word for 'to speak' kōrero is found in New Zealand Māori (kōrero), Cook Islands Māori and Hawaiian ('ōlelo) thus we are able to assert that in Proto-Rapa Nui and most likely in Proto–Eastern Polynesian, that the word for 'to talk' was *kōrero. However, since there is no written documentation verifying this an asterisk is placed before any reconstructed item to show that it is unattested.
The majority of Rapa Nui vocabulary is inherited directly from Proto–Eastern Polynesian. There some changes in semantics and some uniquely Rapa Nui words but the majority of the vocabulary has cognates elsewhere in Polynesia. This is not the result of borrowing but of genuine inheritance. Of course due to extensive borrowing from Tahitian there now exist two forms for what is the same word in the Proto language. For example, Rapa Nui has 'ite (Tahitian—'to know') and tike'a for 'to see' which in Proto-Eastern Polynesian *kiteqa. The *t and *k in the Rapa Nui form has undergone metathesis while in Tahitian the *k has changed to glottal stop. The final syllable in Tahitian has been lost. There are also hybridized forms of words e.g. haka'ite 'to teach' from haka (causative suffix—inherited from Proto–Polynesian *faka) and 'ite (to know—borrowed from Tahitian).
According to some research, the list is a misunderstanding, and the words are not related to numbers at all. It is speculated that the Spanish showed the western numerals to the islanders who did not understand their true meaning, but likened them to some other abstractions. For example, the "moroqui" for number eight (8) would have actually been "moroki", a small fish that is used as a bait, since "8" can look like a simple drawing of a fish. This initial contact with writing could have resulted in invention of the Rongorongo script.
Englert recorded vowel length, stress, and glottal stop, but was not always consistent, or perhaps the misprints make it seem so. He indicated vowel length with a circumflex, and stress with an acute accent, but only when it does not occur where expected. The glottal stop /ʔ/ is written as an apostrophe, as it is today, but is often omitted. The velar nasal /ŋ/ (now "ng") is sometimes transcribed with a "g", but sometimes with a Greek eta, "η", as a graphic approximation of "ŋ".
Comments on S. R. Fischer's 'Mangarevan Doublets: Preliminary Evidence for Proto-Southeastern Polynesian'. (Comments)
Jun 01, 2002; Ssteven Roger Fischer (2001) has recently published arguments for an early Eastern Polynesian protolanguage distinct from...