Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 19, 1886
This text is also available in PDF
(466 KB) Opens in new window
– 482 –

Art. LXIV.—The Track of a Word.

[Read before the Wellington Philosophical Society, 4th August, 1886.]

In seeking to attract attention to the immense geographical district over which a word may be in use, and to the very great periods of time during which a word must necessarily have existed, I would confine myself mainly to a record of the facts concerning it collected by modern science, and leave for discussion the points arising from such record. The word I propose to examine is the Maori noun mata, which means “the eye,” or “face.” This word has been often commented upon as one which maintained itself most purely and with little phonetic variation among the dialects spoken in Polynesia; but I believe that the full significance of its very extraordinary diffusion over a large area of the world's surface has not been sufficiently observed or commented on. We will now, with the aid of a map, pass along a track where this word, sometimes in a form exceedingly pure, sometimes corrupted almost beyond recognition, may be found in the spoken languages of mankind at the present moment.

Leaving New Zealand and moving to the northward, we arrive at the Fijian Group, the natives of which, although not Polynesians, retain in their language many Polynesian words, and these in great purity. Here we find it mata, as in Maori; thence journeying eastward to Samoa, it is mata; at Rarotonga and Mangaia (Cook's and Hervey Islands), it is mata; at Tahiti (Society Islands), mata; at Nukuhiva (Marquesas Islands), mata; at Easter Island, mata; at Hawaii (Sandwich Islands), mata. This course has passed through the principal Polynesian islands, and before proceeding further I must digress for a brief space to notice the dialectical change producing the variant k of the Hawaiian. The change from t to k seems at first sight to be peculiar, and to those who have not made the transference of sounds a specialty of study appears almost impossible. But it is by no means confined to the Polynesian; in many languages far more advanced this letter-change occurs: in the Latin, Basculi and Bastuli, Vectones and Vettones; in Danish, mukke, for English “to mutter,” and laktuk (Latin lactuca), for English “lettuce;” in Greek we find the Doric makes ὅκâ for ὃτε, τζνo03C2 for κζιν03C2; the modern French of low-class Canadians gives mékier for métier, moikié for moitié, according to Professor Max Müller,* on whose choice of this word mata as a text I shall have much to say at a future time; but here it is only necessary

[Footnote] * Müller, “Science of Language,” 2nd series, p. 168.

– 483 –

to remark that in Polynesian the t to k transfer is exceedingly well marked, and that it is, even now, changing and spoiling the Samoan vernacular speech. The real k of the western dialects is, in Hawaiian, Tahitian, and Samoan, either lost altogether, or replaced by a kind of soft catch of the breath; the k which appears in the Hawaiian being the Maori and Tongan t. Thus, the Hawaiian kai is the Maori tai, the sea; the Maori kai, food, being represented by the Hawaiian ai. The Hawaiian kii, a carved image, is the Maori tiki, the medial k being lost, the t replaced by k.

Having thus shown that the change of mata to maka is the regular transfer of sound which should be looked for, I will now resume—the digression having been necessary, as we shall find that the t to k sound is not confined to the language of the Sandwich Islands. Returning to Fiji, we pass westward, first to Rotumah, where “the eye” is matho; then to the New Hebrides, where, among a Papuan population, many colonies of the fairer-skinned race have been planted. Here we find at Malicolo, maitang; at Tikopia, mata. At Santa Cruz, maku is the face; at San Christoval, ma is face; at Vaturana, mata; at New Georgia, mata. In New Ireland, the eye is matak; at Port Praslin, mata; in New Guinea (Triton Bay), matatongo; (Onim) matapatin. We find at Gilolo (Galela), mata. Four dialects of the Celebes give the eye as mata; Borneo (medial, near Labuan), mata. Ceram's seven dialects yield matamo, mata, matacolo, mata, matanina, matara, and matan. Timor gives mata; Savu (S.W. of Timor) is mata; Java = moto; Sumatra, mata, although in South Sumatra matty. The Malay proper is mata; the Dyak is mata. Let us now take a long flight to the westward, to the island of Madagascar. Concerning the Malagasy I shall say little, as it is a well-known fact (whatever may be the origin) that the language possesses very many words akin to the Malay, and which have no representatives on the African coast near at hand. Of these words, one is maso, the eye; the root, mat, having apparently passed through the change (so common in all languages) from hard k to soft c or s, thus: mat, mak, mac, maç, mas (= maso). We will now return to the vicinity of New Caledonia, and pass to the Marianne Islands, where at Guam we get mata; at Chamori, mata; at Ulea, matai; at Satawal, metal. In the Pelew Islands we get the corrupt form muddath; but in the Tagal of the Philippines matá. In Formosa, macha; in the Loo-Choo Islands, mi; and again in Japan, mi. We shall probably trace, as we go on, how this curious variant mi has arisen. The Ainios, or aborigines of Japan, have no representative word; neither have the Coreans, or the Kamschatkans, nor any tribes north of this point.

The Chinese (Canton) have the word as mok, so also the Tonquin gives mok, and Cochin-China mok; but Cambogia has

– 484 –

mat, Pegu = mot, the Ka dialect mot. The word in the Burmese proper is myitsi; and in Aracan myitsi; but the Palaong (S.E. of Bhamo) use metsi. Between the Burmese proper and the Siamese are the Karens, two of whose dialects (Sgau and Pwo) give me; a third closely allied is the Thoung-lhú = may. Taking Muneepoor as a centre, we have the Koreng, mik; Songpu, mhik; Luhuppa, mik; North Tankhul, amicha; Khoibu, mit; Maring, mit; Kapwi, mik; Maram, mik. The Siamese has no cognate word, but it is a very remarkable thing that one of the Siamese (or Thay) tribes which fought its way into Assam, and settled there, has the form pure as the Polynesian, viz., matta. We now pass into Assam through the varieties of Jili; Singpho, mi; Kakhyen, mi; Deoria Chutia, mukuti. In Assam, to the East, are the Mishmi tribes, in one dialect of whose speech (the Mijhu) we find mik; in Central Assam, the Mikir = mek. Of the Naga forms, (among five dialects showing no affinity), the Mithan yields mik; Tablung, mik. Entering Nepaul, one division, the Kirata tribes, (Kirata proper), gives mak; the Limbu, mik; the Lepcha spoken in Sikim yields amik. Among those peoples called the Broken Tribes are Vaya, mek; Chepang, mik; Dhimal, mi; Bodo, mogon; Garo, mikran. The Magars, who inhabit the lower levels of the Himalayan slopes, use mi (in Murmi); so also the Gurung on the higher slopes have mi. The Bramhú, a dialect of a degraded people, gives mik; the Nepaul proper in its purest form being mikha. The Pahri, one of the Broken Tribes, has mighi; but others of these tribes, the Darahi and Kuswar, while using ankhi for “eye,” (of Sanscrit derivation), call the head mud. In Bengal, at Aracan, the dialect used by Moslems, (called Ruinga), uses mata for head; the Hindu dialect, (called Rossawn), uses mustok. In Central India, the Sontal call the eye met, while the word for head with the Pakhya is mauto, and the Tharu is mudi. I am aware of the affinity between the last few words for “head” and the Sanscrit word for “face,” etc., but shall not in this paper touch the subject of the Aryan languages.

We now pass across the Indian frontier into Thibet—the land of the Bhot, or Bhotiya. To the south, near Nepaul, we find that the Serpa word is mik, and in other dialects mi. The rude tribes (called barbarians by Chinese) in the south-east of Thibet use, in the Changlo, ming; in the Gyarung, tai-mek. In Rampur (Milchan) we find mik; while a provincial dialect (Theburskud) has ; and the Sumchu is mi. In the eastern Bhot of the Takyal it is mido; and in the proper Thibetan dialect, as at Ladak, it is written mig, and pronounced mik. If we pass from the Indian frontier, across Afghanistan, to Persia, we find that in Persian mata is the face, and that in Arabic mata is also the face. This, in the widespread speech of the Arab, carries us to the Mediterranean Sea and the Indian

– 485 –

Ocean. I will not pursue the word further to the westward, or open up the great question as to its appearance in the European languages.

We have thus followed mata, the “eye,” or “face,” in its various changes, through almost every possible corruption to which a word is liable, (always excepting the peculiar Semitic formation about radix), and have seen that this word can have entered into, or departed from, the Asiatic mainland by three gateways—viz., 1st, by China and Japan; 2nd, by the Malay Peninsula; 3rd, by the Arabian route, past Madagascar. Some of the languages I have referred to are mere barbarous dialects, of which I have been able to gather about fifty words of each for purposes of comparison; but these share with others, (Burmese dialects, for example, of which I have been able to compare 500 principal words with Polynesian), in that they have no other apparent resemblance except in this most persistent word.

I must not omit to notice one other point before concluding—viz., that the zigzag course we have followed by no means defines the vast area covered even by the modern use of this word. Leaving out the Australians, the Papuans, and most Melanesians to the south; to the north excluding the Tungus languages, the Mongols, Samoyeds, the Turkish forms of Northwest Asia, Finns, Laps, etc.; and also the Dravidian tribes of Southern India: then, (with these exceptions), from Central Asia to the south of New Zealand; from near the shore of Africa to islands near the coast of America, this word has vitality. We trace it in spite of every disguise it assumes, aided by one slight change after another, but with the track still remaining visible. To use an oft-quoted word-example, no one in his senses would compare the French jour (a day), with the Latin dies (a day), unless he could track (either historically or geographically) its changes through dies, diurno, giorno, jour. So no one would compare the Thibetan mi with the Polynesian mata, if it was not that we could trace it step by step at the present hour through mat, mak, mik, mi. But, (an important “but,”) every now and then we have been refreshed on our search by the pure word starting up anew, (as in Northern India, matta), and at the very extremity of our journey by the reversion to the pure mata of Persia and Arabia.

The questions to be considered as resulting from this inquiry are these:—


Did the Polynesians bring this word from the mainland, either by China, Malacca, or the Arabian Gulf? Or,


Did they give the word to the mainland through either of these paths? Or,


Is this word a living sole-survivor (an “apteryx of language”), lingering in districts all over the south of the great Asiatic continent?

– 486 –

If the last be the case, why, against all rules known at present to philologists, should this vital word be shared by the inflected languages of Persia and Arabia, the agglutinative speech of Thibet and Malaysia, and the monosyllabic tongues of China and its islands?