Sanskrit, and a word of the same language is the only one in use for the numeral ten. "It is on the same principle that I account for the existence of a similar class of Malayan words in the Tagala of the Philippines, although the whole number of Malayan words does not exceed one fiftieth part of the language. "In the Maori or New Zealand, the words forehead, sky, great, stone, point, to drink, to die, are Malay or Javanese; yet of these two tongues there are not a hundred words in the whole language. "As to the personal pronouns, which have often been referred to as evidence of a common tongue, in as far as concerns the language under examination, they are certainly the most interchangeable of words, and cannot possibly be received as evidence. Some of them, for example, are found in the Polynesian dialects, where, in a vocabulary of five thousand words, a hundred Malayan terms do not exist. "The numerals must surely be considered as out of the category of early invented words, for they imply a very considerable social advancement, and seem to be just the class of words most likely to be adopted by any savages of tolerable natural capacity. The Australians are not savages of such capacity, and although with the opportunity of borrowing the Malayan numerals, they have not done so, and in their own languages count only as far as 'two.'' All these principles thus laid down by Mr. Crawfurd, are, of course, liable to considerable limitation, according to the language and people which form the subject of our researches; yet, as a general thesis, it must no doubt be admitted that mere similarity of words does not prove the common origin of languages. It follows, on the other hand, that mere dissimilarity of words does not prove the absence of * In a Tagala Dictionary of 16,482 words, published by Father Juan de Nouda, Mr. Crawfurd discovers not more than an original connection of languages. As these points has been discussed before, we shall proceed at once to give what Mr. Crawfurd considers as the safest test of a common origin of languages. He says, "The words which appear to me most fit to test the unity of languages are those indispensable to their structure, which constitute, as it were, their framework, and without which they cannot be spoken or written. These are the prepositions, which represent the cases of language of complex structure, and the auxiliaries, which represent times and moods. If a sentence can be constructed by words of the same origin in two or more languages, such languages may be safely considered as sister-tongues, to be, in fact, dialects, or to have sprung from the same root. In applying this test, it is not necessary that the sentence so constructed should be grammatical, or that the parties speaking sister-tongues should be intelligible to each other. The languages of the south of Europe can be written with words common to them all, derived from the Latin without the assistance of any of the foreign words which all of them contain. The common stock, therefore, from which they are derived is Latin, and they are sister-tongues. English can be written with great ease with words entirely Anglo-Saxon, and without any French words, although French forms a sixth part of the whole body of its words, but no sentence can be constructed consisting of French words only." So far as this is meant as an acknowledgment that grammatical elements are the only safe basis for a classification of languages, nothing could be said against it. But first of all, languages do borrow even prepositions and conjunctions. In Turkish, every preposition in the true sense of the word, I mean every preposition standing before the noun which it governs, is Persian, Turkish* prepositions being always placed after the noun. Many conjunctions in Turkish are of Persian and Arabic origin.+ Secondly, sentences can be constructed in English, consisting of French, .i e. Latin, words only. If I say "avarice produces misery," every word is Romanic, but it does not follow that, therefore, English is a Romanic language. In fact, the single letter s, used as the exponent of the third person sin *Cf. Redhouse, Grammaire de la langue Ottomane, § 994. gular, is sufficient to stamp the language in which such a sentence can be framed as non-Romanic. Nothing, therefore, but grammatical forms can settle the relationship of languages definitely, and even grammatical forms have occasionally been transferred from one language into another. But in no instance has an entire grammatical system, a complete set of terminations of declension or conjugation, been appropriated by a foreign tongue, and where these terminations coincide as a whole, we may be sure that we have to deal with cognate idioms. Next to the evidence of grammatical terminations, come pronouns, then numerals; then conjunctions and prepositions; and, lastly, words expressive of the simplest ideas and the most common objects of every day's life. There are instances where even such words as father and mother, brother and sister, have been replaced by foreign appellations, or by words newly formed in members of the same family of languages. But, on the whole, owing to the familiar and frequent use of these words, people are unwilling to part with them and afraid to replace them by foreign terms not intelligible at first to the whole community. The Saxons learned to use many foreign words, yet their household words remained on the whole Saxon. So did their numerals without exception, so did their pronouns, and so did in the highest degree their grammatical terminations. But although we cannot agree with the somewhat too general principles by which Mr. Crawfurd tests the relationship of languages, we shall give the results to which his method has led him with regard to the Polynesian languages. "Applying this test to the Malayan languages, it will be found," Mr. Crawfurd maintains, "that a sentence of Malay can be constructed without the assistance of Javanese words, or of Javanese without the assistance of Malay words. Of course, either of these two languages can be written or spoken without the least difficulty, without a word of Sanskrit or Arabic. The Malay and Javanese, then, although a large proportion of their words be in common, are distinct languages, and as to their Sanskrit and Arabic element, they are extrinsic and unessential. When the test is applied to the Polynesian languages, we find an opposite result. A sentence in the Maori and Tahitian can be written in words common to both, and without the help of one word of the Malayan which they contain, just as a sen tence of Welsh or Irish can be constructed without the help of Latin, although of this language they contain at least as large a proportion of words as Maori or Tahitian do of Malayan. The Maori and Tahitian are therefore essentially the same language, and their Malayan ingredient is extrinsic." The Malayan races, according to Mr. Crawfurd, have diffused themselves, and the civilization which they attained by self-derived culture, from two distinct and independent centres. "The Malayspeaking Malays from the rich table-lands of the interior of Sumatra, -Sumatra, which, from its physical gifts, and large proportion of coast line abutting on placid seas, would be at once seized on by the geographer as a focus of civilization. And the Javanese-speaking Malays from Java, an island not less richly endowed in physical advantages." Into the question of the common origin of the language of the Malay, the inhabitants of the more eastern islands of the South Sea, the Negritoes, and Haraforas, we cannot enter at present, though we believe that Humboldt's work has laid open so many traces of relationship, that even after his theory with regard to a distant connection of the Malay with Sanskrit and the other Arian languages is dropped, much remains to encourage the comparative philologist to work that mine of philological research which the genius of Humboldt has opened, but not yet exhausted. The formal coincidences between the Malay and Taï grammar here pointed out for the first time, furnish a link between Asia and Polynesia, which, even by itself, is strong enough to hold two of the mightiest chains of languages together; the Nomads of the sea, extending from the cast coast of Africa to the west coast of America; the Nomads of the Continent swarming from the south-east to the north-west of Asia. But further researches will strengthen this link, and add new traces of their common origin, though we have hardly a right to expect many, considering that we have to deal with languages, in which grammatical elements, are, as it were, at the mercy of every speaker, in which roots are of the vaguest character, and can, by means of accents and determinate syllables, be made to express every conceivable shade of meaning-languages which had received no individual impress before their first separation, and have grown up since under the guidance of but few logical or grammatical principles, so as to make us sometimes doubt whether we should call them works of art or products of nature, or mere conglomerates of an irrational chance. While in political languages, comparative philology has to establish a principle by which to account for coincidences such as Asmi, I am, of the Veda, and Esmi, I am, used by the Lithuanian peasant of the present day, a principle must be found in nomadic dialects to account for differences such as we find between Mandshu and Finnish, Chinese and Tibetan, the Taï and Malay languages. These differences must be explained by analogies to be derived from American, Indo-Chinese, or Siberian idioms, where we still meet with tribes who, after a short separation, have become unintelligible to one another, and where but few traces remain in their idioms to enable the philologist to discover the common basis whence all proceeded. Unless such principles can be established, all attempts to prove the common origin of nomadic languages will fail. To transfer the rules of Arian or Semitic plilology to this vast field of linguistic research, would betray an utter ignorance of the nature of language; it would be, as it has been well expressed, like cutting stones with razors. To consider the few remaining coincidences between such idioms as the result of accident, would be a view incompatible with the philosophy of language, which allows indeed casual parallelisms between dialects no longer connected by any ties of relationship, but distinguishes carefully between these, the result of mere accident, and other congruences, which, though few in number and small in extent, could not, like the segments of a circle, coincide without the admission of a common centre whence all proceeded, and from which their various distances must be measured. EIGHTH SECTION. Tamulic Class. § 1. Early Traces of the Tamulic Nishâdas. We now return through the valley of Asam to where the Brahmaputra joins his sister, the Gangâ. It is here, on the coast of Bengal that we meet with the first historical traces of the Tamulic languages. |