Fruits of the Discovery of Sanskrit.

455 stantine and Justinian; an object of interest to every people that in succession gained the empire of the East; kept back with almost frantic jealousy by the Brahmins of later times from the eyes of Europeans;-this ancient, sacred, and wonderful language at length found its way into the hands of western scholars, and, by its obvious affinity to Greek and Latin, threw open a new world to the researches of philology, and created an epoch in the history of science, the greatness of which is even yet unmeasured. We must refer our readers to Mr. Müller's lectures for particulars as to the way in which Sanskrit became grammatically known in England, Germany, and elsewhere, by the labours of Sir William Jones, Carey, Wilkins, Colebrooke, Bopp, and others. What we have now to do with is the revolution which the knowledge of it wrought throughout the whole sphere of linguistic study. It put an end to the reign of guess and hap-hazard. It showed that there were certain fixed laws under which languages agreed or differed. It drew the line between the apparent and the real, the organic and the accidental, connexion of tongues. It cleared away a whole troop of misconceptions as to the origin and historical development and relations of the various forms of human speech. Finally, it laid the basis of a new and scientific analysis and classification of languages; and for the first time approached, on the side of reason and matter of fact, the great theoretical questions in which the study of language culminates.

One of the first results of the inquiries, to which the causes above recited gave birth, was the discovery of what is known as Grimm's law of the interchange and correspondence of the consonantal elements of language. It had often before been remarked by philologists, that certain words, roots, and inflections, in this or that language, answered to similar forms in some other language. It was all but inevitable, that the classical scholar should observe the likeness in build and meaning, which sisto in Latin bears to histemi in Greek, and that nouns in um in the one language were very commonly represented by nouns in on in the other. In like manner, the student of Hebrew, with his Arabic lexicon in hand, could hardly fail to see that the yalad of the former was the walada of the latter; and that there was a strange literal conformity between the personal prefixes and affixes of the verbs in the two tongues. But that resemblances like these were the effect of causes having their seat in the constitution of human nature, and that they were as much reducible to law as any other phenomena coming under the cognizance of science, was a fact which Grimm was the first distinctly to enunciate and formulize. At the waving of his magic wand the primary elements of speech arranged themselves

1

after the order and plan of their natural genesis by the organs of speech; and it was shown from an ample induction of particulars, that where agreements such as those just named were found to subsist between languages, they were due either to interchanges among letters of the same organ, or to permutatious of letters of one organ for those of another, under definite and even rigorous linguistic law. The letters t and d, for example, are both of the order called dental; and, on comparing the English and German languages, it will be seen that, where we use the latter of the two, our neighbours are fond of employing the former. Thus tag, tief, tod, traum, taube, tochter appear on this side the Channel as day, deep, death, dream, dove, daughter; and on the same principle the drei, dick, dorn, durch, and dank of the one language are pronounced three, thick, thorn, through, and thanks in the other. This is a case in which letters of the same organ regularly answer to one another. A kindred phenomenon may be illustrated by the letters k, t, and p. These, though of different organs, all stand in the same relation to their respective classes, and are found to change among themselves, within certain limits, as well as with the consonants of the several groups to which they belong. Thus the Aramæan and Ethiopic languages exhibit a k in certain verbal affixes, which in Hebrew appears as t; and there are Greek and Latin forms, both roots and derivatives, in which a t or a p stands as the representative of a Sanskrit k. In like manner aspirated and unaspirated letters may be observed to follow determinable rules of mutation; and what one language writes dh or bh, another will write d or b. These are simple examples of a large and often subtle system of literal correspondence, which springs out of the very nature of languages, and of itself removes the study of speech from the domain of accident and mechanical contrivance to that of science and law. There are savage tribes in the Fiji Islands, which are said to modify their language by design, in order to conceal the meaning of what they say from their enemies. They wish to ' hide their mind,' as they themselves express it, from the people of Bau, the dominant state, and they alter the Bau dialect for this purpose. The missionaries believe that such is the origin of several of the numerous Fijian dialects; and it is possible that something of the kind may obtain elsewhere. But this is a fact of quite a different class from the phenomena just spoken of; and no amount of supposable artifice could account for appearances, which are clearly the result of family relationship, and of the plastic power of mind acting freely under the impulse of its own native forces.

Closely connected with Grimm's law of consonant equivalents

Linguistic Relationship defined.

457

-a law, which Bopp and others have carried much beyond its original circle,-is another important doctrine, also the fruit of the discovery of Sanskrit, and of the scientific researches of the last few years, the doctrine of the true kinship of languages. Till within this time, nothing could be more arbitrary or absurd than the manner in which scholars were accustomed to judge of the agreement or disagreement of tongues. The men and bears which children see in the clouds are not more unlike their prototypes in nature, than the features of resemblance, which were often imagined to subsist between various languages, were unlike the correspondences of a real and demonstrable affinity. On this subject comparative philology is already in a position to speak with great confidence. It has been erroneously supposed, that while difference of vocabulary is a sure sign that languages are foreign to each other, the possession of words in common is an infallible proof of kindred. Neither of these canons will hold. The Tatar languages, as Mr. Müller shows at large in his elaborate and masterly essay upon them and their fellows, are endlessly diversified in their vocables. They change both in time and space, like the sands of an estuary. Yet these languages are indisputably of the same great stock. Similar to this is the case of the multitudinous aboriginal languages of the American Indians; as to whose organic unity there is evidence without end, but of which, it is said, scarcely any two have a word in common. On the other hand, nothing is more certain, than that tongues may use the same words, fewer or more, as joint property, and yet be as unlike in their fundamental character as two geological deposits which happen to have boulders of the same granite or limestone imbedded in them. The language of the ruling dynasty in China, the so-called Mandarin dialect, contains numerous words which are equally shared by it and by the Chinese proper; but it is demonstrably not a tongue of the same description with that of the Chinaman. It is Tatar, sprinkled over and flavoured with forms derived from the speech of the subject race. Mr. Müller refers to the Turkish, as another case in point. Turkish is a Tataric or Turanian language. This can be proved as clearly as any fact in chemistry or optics. But, in the course of its life, it has picked up a multitude of words belonging to the Arabic and Persian,-languages of quite a different character; and it has so bedizened itself with their costumes, that a superficial observer might easily take it for a Semite or an Aryan. 'At the present moment,' Mr. Müller writes, 'the Turkish language, as spoken by the higher ranks at Constantinople, is so entirely overgrown with Persian and Arabic words, that a common clod from the country understands but little of the Osmanli,

though its grammar is exactly the same as the grammar which he uses in his Tataric utterance.' Were identity of words sufficient ground for arguing the family connexion of tongues, it could easily be proved that Sanskrit and Hebrew are brother and sister; that the language of Homer and Thucydides was the language of the Pharaohs; and that the speech of the factory girls of Manchester, in our own times, is one and the same with that of the boatmen of the Canton river, the people of the Friendly Islands, and the wandering tribes of the Arabian desert. It is not the dress, the look, the external form of a language, which determines its lineage and relationships. A beard is not an infallible sign of a Frenchman. Ethiopian serenaders are not always Negroes, though they carry black faces. And it would be precarious to affirm either that a bramble and a cactus are of the same order of plants because they are both prickly, or that beetles and butterflies should be classed together, because both have wings. Build, structure, organization,—this is the true basis of a scientific arrangement and grouping of phenomena in all departments of nature; and it is the only rational test and guide which the comparative philologist can use in his peculiar province. A multitude of accidental causes,-contiguity, commerce, conquest, religion, literature,-may import into a language words and forms which are none of its proper heritage; but these are only proselytes, and they no more affect the constitution and essential character of the tongue into which they are admitted, than the velvet is changed by the beads with which it is embroidered, or the chalk-rock by the fossil flints which are scattered through its substance. Mr. Müller's position, that there is no such thing as mixture in language, may seem at first sight to be a bold one; but it is perfectly defensible when the terms are explained. That which constitutes the individuality of language does not alter. There may be functional changes; there may be dismemberment, or even death; but, as long as life lasts, the body and soul remain the same, and nothing can transmute them into what they were not in the beginning. We repeat it; organization is the rule of judgment as to the affinity of languages. If their primary elements are of the same description, if the structure of their syllables follows the same laws, if their root-words are identical in meaning and composition, if their processes of grammatical derivation run in a common track, if their inflexional forms play a like part, and are traceable to the same parentage, if their methods of syntactical expression and arrangement are substantially one, then it is safe to conclude, that the tongues we bring into comparison are cognates; and the degree in which these

Applications of the Doctrine.

459

resemblances are found to obtain will determine their kinship as nearer or more remote. On the other hand, we may safely conclude, that languages which fail to exhibit these and the like organic correspondences, are not bound together by near family ties.

Let the Fijian and Arabic languages, for example, be brought into comparison, with the view of ascertaining whether or not they are related. Here the unlikeness of the vocabularies is a fact which presents itself at once. Any such similarity of words as obtains in Dutch and English, or in Manx and Erse, is plainly wanting. But Hottentots and Circassians are both men, though their visage and frame are so dissimilar. And the languages in question may be one in structure, though their outward appearance be diverse. Let us see. The Fijian alphabet has few sounds, and those chiefly vowels and nasals; that of the Arabic is large, and abounding in gutturals. Any such correspondence of letters, as comes of the operation of Grimm's law, is nowhere to be traced. The syllable in Arabic can never consist of a single vowel; it is often composed of a consonant followed by a vowel and another consonant. Under no circumstances does Fijian admit the latter arrangement; while nothing is commoner than the occurrence of a vowel as a separate word. The roots of the Arabic are mostly tri-consonantal. The Fijian is a stranger to any such principle. The verb system of the Arabic is remarkable for the wealth of its forms, and the number and regularity of its inflexious. The Fijian uses the verb, as it does the noun, without subjecting it to the smallest change, either by addition, subtraction, or any other process. An elaborate syntax belongs to the Arabic. The expression of ideas in Fijian is as simple and inartificial as can well be imagined. What is our finding, then? Fijian and Arabic are not related languages.

Suppose another case. Let the question of the kinship of the Sanskrit and Greek be raised. Here we are struck at once by the circumstance, that, whatever the difference of their vocabularies may be, they have many words in common. Nor can this be explained, as is the presence of Persian and Arabic words in the Turkish. They are too various in their kind, too equal in their distribution, and often too near the heart of the respective languages, to be accounted for by the theory of external contact. Thus, then, we have a presumption of organic oneness. But what is the testimony afforded by their structure? The Sanskrit alphabet is wider than the Greek; but the sounds of each are mostly of the same classes. Throughout the two languages there is the systematic correspondence of letters, of which Grimm's law is the formal exposition. In both, the constitution