Volume 23 1914 > Volume 23, No. 91 > Polynesian philology, a reply to Mr. Edward Tregear, by Sidney H. Ray, p 154-158
                                                                                             Previous | Next   

- 154

IN “The Journal of the Polynesian Society,” No. 89 (March, 1914), Mr. Tregear severely criticises the remarks I made in No. 88 (December, 1913), with regard to some supposed connections of Maori and Indian words published by Mr. F. W. Christian in the same Journal (No. 86, June, 1913).

Mr. Tregear suggests that the difference in the view of these words taken by Mr. Christian and myself has arisen over “the word ‘compare’ used in a distinct sense by one and in a general way by the other.” But Mr. Christian did not use the expression ‘compare’ at all. He calls the words he cites “Hindustani cognates of the Maori,” and the dictionary meaning of ‘cognate’ is “born of the same stock, related to in origin.” He speaks of the Sanscrit origin of Maori words (Kumara, totara, mâmari, wahine, and whenua) and ‘origin’ means “source or beginning.” He speaks of “Indian root words” which seem to him to be very “faithfully reproduced in Maori” and allied dialects. He does not ask us to compare the Sanskrit bhek, etc., with the Maori wheke, but gives the words as examples of “Indian cognates of the Maori,” and calls them examples of “new derivations which go to form the truth of Fornander and Tregear's theory of at least a partially Aryan Maori.” All this surely implies that in Mr. Christian's opinion there is a relationship between Maori and the Indian languages, and my note was intended to show that the method used by Mr. Christian was unscientific and that the likenesses of the words he connected was not evidence of an identity of origin.

My quotation from Sayce that “to compare words of different languages together because they agree in sound is to contravene all the principles of scientific philology” is called a platitude by Mr. Tregear, and he unsparingly condemns those who are guided by sound only in making comparisons. He agrees that “more than correspondence in sound and sense is needed, grammatical affinities count far more, and there are certain letter-changes to be considered under Grimm's law.” But Mr. Tregear would restrict these philological principles to the great historical languages (whatever they may be) and quotes from the Etymological Dictionary of Dr. Skeat examples - 155 of uncertain derivations. He calls these guesses because links are missing in the chain. He forgets that if some links in a chain be lost, and the pattern (structure and phonology) be known, we may form a very good idea of the character and appearance of the missing links. But, if we have two chains of different pattern (i.e., different structure and phonology) we can have no possible idea of any links which may have connected them, nor can we have any idea as to whether they were ever connected at all.

In the second part of my quotation from Sayce that “agreement of sound is the best possible proof of the want of connection,” Mr. Tregear has overlooked the fact that this was said of different languages, and tries to prove the statement false by citing the German hund and English hound, the Latin vir and Irish fear, Persian dokhter and German tochter, and asking whether their likenesses in sound and sense are proofs that there is no relationship between German and English, Latin and Irish, Persian and German. These likenesses are not proofs in themselves. They may be accepted because German and English, Latin and Irish, Persian and German have proved, as Mr. Tregear says, “by grammar and the coincidence of thousands of words” and by the establishment of the phonetic laws which govern the likenesses, to belong to the same family of languages. In the Nuba language of the Nile valley, uri is ‘black,’ ur ‘head,’ ngei ‘here,’ tona ‘his,’ and these words are almost identical with the Samoan uli, ulu, and nei, and the Maori tona, with the same meanings, but they have no value to show a likeness between Nuba and Polynesian, because Nuba and Polynesian have not been proved to belong to the same stock. It is the proof of grammatical likeness which alone makes the comparison of words admissible.

Mr. Tregear states that it was the likeness in sound and sense between a word in one language and a word in another which led the first discoverers to dream of the subject. But discoverers do not dream, and dreams are not facts. It was dreamed that Greek was Hebrew read backward, that Bask or Dutch was spoken in Paradise, that language was invented by Egyptian Gods, and Sanskrit was forged by the Brahmans. It was not until Hervas had proved the Semitic dialects to be alike in grammar, and the grammatical and phonological relationship of the Aryan tongues had been established by Grimm and Bopp, that the foundations of philological science were laid.

It is this failure to recognise the difference in the grammar of the Indian languages and Maori or Peruvian, which nullifies all Mr. Christian's comparisons. If he can prove that these have related grammars and can show phonetic laws existing between them, we may regard his connections as possible, but otherwise, I maintain that the whole series of supposed relationships is mere guesswork.

- 156

But Mr. Tregear pleads that oceanic linguistic comparisons should be made by methods different from those by which comparisons are made in the ‘historical’ and literary languages of other parts of the world. “Iron regulations” need not be observed, nor “strict observance of rules.” In other words, oceanic philology must have laws of its own, unlike those which govern the science elsewhere. But if philology is to be subject to special methods in Oceania, why not Biology and Mathematics?

Mr. Tregear would apparently substitute the geographical for the historical connection of languages. But geographical so far as it relates to words merely shows that the speakers have been in contact and borrowed words from one another when they had no suitable terms of their own for the object or action to be named. His example of the word for “dog” in the Aryan languages seems to imply a derivation of the English word “hound” through the German hund and Greek kuon from the Sanskrit s'van (not ovan), as though the word had been passed on from one of these languages to the other instead of each having its own history and descent from an ancient language which was not Sanskrit, but a dialect of the ancient Aryan of which Sanskrit is also a descendant.

With regard to the examples of Mr. Christian's supposed cognates which I noticed in my criticism, Mr. Tregear deals only with the word bhek “frog,” which Mr. Christian connected with the Maori word wheke “a cuttle fish.”

I condemned this and similar words in general terms as being of no value for comparison because they are onomatopœic and might well occur in totally unrelated languages. Mr. Tregear suggests that I have been misled, because the Sanskrit bheka, “frog,” is derived from the root bhi. Now there is nothing inherently improbable in the origin of the root bhi from the croak bhek, of a frog. In some Sanskrit dictionaries the word for “frog” bheka is given as onomatopœic, but Hindu grammarians require every word to be derived from a root. The roots are merely the abstractions which underlie certain sets of words, that is, they represent the general idea present in the derivations and are purely a device of the grammarian by which he can determine and classify the words which were formed before the roots.

In order to connect the frog with the cuttle fish, Mr. Tregear has given the root bhī a meaning which is the reverse of that found in the Sanskrit dictionaries. It does not mean “causing fear,” but “to fear” or “fearing.” From this is derived by means of the nominal Suffix aka, with guna (vowel change) of the ī the noun base bheka bhī + aka) meaning “the fearer,” “the one that fears,” like nayaka (nī + aka) “a leader” from the root nī. Thus bheka is the creature that fears, and means not only a frog, but also a timid man. All the primary derivatives as e.g. bhiru, timid, contain the idea of “timidity, - 157 fearing,” quite agreeing with the behaviour of the croaking amphibian. To express the causing of fear, aya is added to the root, hence bhayam, fear, what causes fear, bhayayati, he causes fear. There is also the impersonal noun bhetavya, one to be feared. It should be noted that none of the words which imply “causing fear” contain k, which appears to be essential in all Mr. Christian's words. The Sanskrit dictionaries appear to ignore the equivalent for “cuttle fish.” The “bone” was regarded as indurated sea-foam phenah, or samundra-phenah (samudra, ocean) from the root phan. The Malay for “cuttle fish” is ikan-gurita, and a similar name is current all over Indonesia and Melanesia.

Mr. Tregear quotes Mr. Christian's connection 1 of the Maori whai “sting ray” with the Persian word for “fairy” as a “brilliant example.” He states that the Malays “call the skate or ray pari, the fairy.” Now this must be merely a conjecture of Mr. Christian's. In Malay the words are pāri (2) ‘skate,’ and pĕri or pări (*) a nymph, two perfectly distinct words. The Malay pāri is unquestionably related to the Philippine pagi, Borneo pahi, Moluccan hali, ali, Melanesian vari, fai, vai, and Polynesian fai, whai and vai. In Malay it is usually found with the prefix ikan, fish, as ikan-pāri, where pāri has no reference to “wings,” but is solely due to the Malay idiom of prefixing classifying words as e.g. ikan-hiyu shark, ikan-lidah sole, ikan-merah red-fish. The Malay words for “wing” sayap or kepak, or for “flying” terbang do not appear in the name of the only fish which has apparent wings or seems to fly, the ikan-bilalang, the (English) flying-fish, literally the grasshopper-fish.

Mr. Tregear asks: “If the word (pāri or pări) came from Asia why should research leave it on the Malay shore? The Malays and Javans were well acquainted with Persian and Arabic poems and folk-lore.” But the fact that Malays and Javans borrowed pări for “fairy” in stories from Persian and Arabic does not prove that Indonesians or Polynesians who did not borrow the stories borrowed the word as a name for the skate when they already had the different word pāri for the name of the fish. Malays and Javans also borrowed in their Hindu tales the Sanskrit word vidyādhari, (meaning a female demi-god) and used it in the forms bidadari (Malay and Makassar), widadari (Javanese) for nymph, fairy. But there is no indication that bidadari or pări ever got further east than the stories.

From the Maori whai as the skate to the the winged or fairy-fish of Mr. Tregear's supposition is a far cry and “beyond philological grounds.”

- 158

In conclusion, the true affinities of the Polynesian languages will never be ascertained by a mere comparison of words. The essential unity of the Austronesian (i.e., Indonesian, Melanesian and Polynesian) depends for its proof on a substantial identity of grammatical form and expression in the three divisions. When it can be shown that this common foundation of grammar is comparable with Aryan, then it will be permissible to compare words in the two families of speech. When both are shown to agree in grammar with Peruvian or any other tongue we may begin to consider whether their words are related by the operation of phonetic laws.

1   This, though referred to by Mr. Tregear, is not found among Mr. Christian's Hindustani cognates. I have not been able to refer to it.
2   We regret we cannot reproduce the Malay letters, through absence of type.—EDITOR.