From J. Stanley Gardiner (1898), "The Natives of Rotuma," Journal of the Royal Anthropological Institute 27:500-503.

The Rotuman language is not an isolated tongue, but a member of a widespread family of languages, extending throughout Polynesia. To the ear it sounds perhaps, considerably different, owing, to a peculiarity, in the fact that the Rotumans have a great tendency to transpose the last letter, a vowel, with which all their words should end, to the middle of the word. Mr. Hale in the Wilkes Expedition Report remarks [fn.Loc. cit., pp.469 et seq.]: "A general law seems to be that when a word stands by itself, not followed by another on which it depends, it must terminate in a vowel; and this appears to be the original, and proper form of most of the words; but when combined, in any way whatsoever, with other words all alteration takes place by which the concluding syllable is transposed or contracted, as that the consonant shall be the final letter." Thus in ordinary conversation the name Rotuma is often turned into Rotuam; hoga is always hoag; the word oipeluga, a club, I have heard pronounced as oipeluag and oiapelug. In the text, I have as far as possible kept the words as pronounced, but in the short vocabulary (App. II) I have tried to spell the words in accordance with their original pronunciation. Another remarkable thing is the great facility with which the Rotumans will coin a word for anything new; peculiarities of the animal or thing will be taken, and from these a name made, somewhat in the German fashion. Thus the scorpion is known at one end of the island as the mamasse, the animal which eats at the tail, and at the other end as the monpuoga, the animal which eats the puoga, a small worm in the bananas.

For the purposes of comparison, I compared a rough vocabulary, which I first made, of about two hundred and fifty words with the words of the same meaning in Fijian and Samoan. Of these I found that twenty-nine words were related to both Fijian and Samoan, and evidently were derived from the same roots, ten to Fijian alone and thirty-three to Samoan only. The Samoan I obtained from the Rev. George Pratt's dictionary, but the Fijian by natives, who, knowing several dialects, especially searched for words related to the Rotuman. With the Gilbert islands there were supposed in Fiji to be considerable resemblances; I could make no general comparison, but the few words, which were the same, were all of general distribution through the whole Pacific, or else comparatively recently introduced, names of weapons, instruments, etc. Compared with Malay, by means of Dr. Bikker's vocabulary, there is scarcely a trace of resemblance to be found.

It must be noted, though, that formerly in Rotuma there was a language spoken, considerably different from the present one; in it are most of their songs, and a few phrases from it are still used, but their meaning has been lost. In addition, there was a peculiar language, or rather set of phrases, used to and in speaking of the sou; and other chiefs. These have been lost owing to the coming of the missions and the abolition of the sou. It was suggested by several of the old men that the change of language was due to the coming of the Niuafoou people to the island. In the vocabulary the words, given in the Wilkes Report, are inserted for comparison in brackets where different; they may possibly throw some light on this old language, as many are quite different from the terms I found in use.

While Samoan has fourteen letters in its alphabet--a, e, f, g, i, l, m, n, o, p, s, t, u, vÑ-it is necessary to give the Rotuman four more, k and r being found as well as t and l, and the h being often sounded very distinctly, while in such a word as sosoghi, sister or brother, it is scarcely aspirated at all, and such a word as haharagi, young, in the method adopted by the Rev. George Pratt for Samoan, would certainly be spelt 'a'aragi. In set speeches all words commencing with an h have it very distinctly aspirated. The letter j must also be added to indicate a sound resembling ch, ts, and the English j about equally. It occurs in the names of many places on the island, but is otherwise very uncommon; examples are Juju, Atja, etc., also nuju, the mouth. Vowels are pronounced as in the continental method. I know of no meaning dependent on the quantity, but it is a mark of respect, when speaking to a chief, to lengthen all or the chief vowels of each of the substantives, thus laying great stress on them. G is always nasal, and pronounced ng. All the other letters are pronounced as in English.

The same vowel is not generally repeated in a word without a break between, unless the word is a compounded one, as solgaasta, the north wind; saaraara, a centipede; huneele, the beach. In these cases each vowel is distinctly pronounced. The diphthongs are ai, as in tekaikai, a shell; au, as in rau, tobacco; oi, as in hoina, a wife; ou, as in filou., the head. Other vowels occurring together are pronounced each separately; thus haephaep, the hand, is ha-ep-ha-ep, apioiitu, a priest, a-pi-oi-i-tu. I am not really certain that any of the diphthongs are properly so, as in speaking slowly many are broken up into their component vowels. For emphasis almost any word may be repeated, But the repetition often changes the meaning; thus the terms manu and huf are applied broadly to many small flying animals, but manumanu is a bird, and hufhuf a bat.

The accent properly, in the Polynesian group of languages, is placed on the penultimate syllable, and this rule holds for Rotuma, except that when a chief is being spoken to it is often thrown on the first syllable. The transposal of the last vowel, too, often throws it on the last syllable. It is in no case thrown on the vowel, thus transposed.

In Rotuman there is no article definite or indefinite. The Wilkes Report gives ta, one, definite and indefinite, used for that, as opposed to ti, this, both being used as postfixes.

The names of natural objects, such as trees and animals, are mostly simple and indigenous to the island, or to Polynesia; to these must be added such simple manufactured articles, as the people may be supposed to have known, before they migrated to Rotuma. Compounded nouns usually indicate that the article, animal, or tree has been but recently introduced; exception however must be taken to articles of food or manufactured articles which have been brought by natives of other islands with their own names. The verbs and the nouns, or perhaps adjectives, for similar meanings are the same.

Number does not properly exist. For the plural numerals, or words such as imply a number, are used. Taucoko (pronounced tauthoko), a Fijian word, is now applied to people, while atakoa is applied especially to animals. Tene, many, is used generally for inanimate objects, such as stones, trees, etc. The Wilkes Expedition Report gives also maoi, many.

Gender is formed by the affix of fa, man or male, and honi, woman or female, usually shortened to hon or hen. In most cases among the larger animals, the male and female have separate names.

Case is indicated by prepositions. The genitive may be indicated by one of the possessive pronouns. K, or ka, is used as a prefix, and applies especially to movements, such as entering and leaving a house; it is particularly employed where an adjective is used. Se implies the act of moving forward to a place, and e the act of movement from a place.

The adjectives as a rule follow the noun. The numerals do not go above kiu, 10,000; they are almost for the smaller numbers identical with those of Samoa and Fiji. The pronouns are given fully in the appendix; compared to the rest of the language, their formation is very complete.

The tenses of the verbs are formed much in the same way as in many other Polynesian dialects. Past tenses are very generally formed by the addition of an adjective, used as an adverb, thus:Ñ

Lao .... .... .... .... .To go.
Gou lao .... .... .... I go
Gou la lao .... .... I will go.
Gou lao vahia .... .. I went.

The passive voice is usually formed by changing to the active. The Wilkes Expedition Report says, "The directive particles mai and atu are found in Rotuman under the forms m' and ato (or at'), suffixed to the verb. Thus lao or la, which signifies to go or move, becomes laato, to go away, leum, to come."

The affirmative adverbs are o, ou, and e, and negative igikei. Igikei is also used for not, but kat is a more polite term. "The negatives are kat (or kal) and ra, the first of which usually precedes the verb, and the second follows." [fn. Wilkes Expedition Report.]