CHAPTER XII. SURVEY OF LANGUAGES. The Northern and Southern Divisions of the Turanian Class. WE E have now examined the five classes of the UralAltaic family, the Samoyedic, Tungusic, Mongolic, Turkic, and Finnic. The Tungusic branch stands lowest; its grammar is not much richer than Chinese, and in its structure there is an absence of that architectonic order which in Chinese makes the Cyclopean stones of language hold together without cement. This applies, however, principally to the Mandshu; other Tungusic dialects spoken, not in China, but in the original seats of the Mandshus, are even now beginning to develop grammatical forms. The Mongolic dialects excel the Tungusic, but in their grammar can hardly distinguish between the different parts of speech. The spoken idioms of the Mongolians, as of the Tungusians, are evidently struggling towards a more organic life, and Castrén has brought home evidence of incipient verbal growth in the language of the Buriäts and of a Tungusic dialect spoken near Nyertchinsk. This is, however, only a small beginning, if compared with the profusion of grammatical resources displayed by the Turkic languages. In their system of conjugation, the Turkic dialects can hardly be surpassed. Their verbs are like branches which break down under the heavy burden of fruit and blossom. The excellence of the Finnic languages consists rather in a diminution than increase of verbal forms. The Tcheremissian and Mordvinian languages, however, are extremely artificial in their grammar, and allow an accumulation of pronominal suffixes at the end of verbs, surpassed only by the Bask, the Caucasian, and those American dialects that have been called polysynthetic. In declension also Finnish is richer even than Turkish. These five classes constitute the northern or UralAltaic division of the Turanian class. South-Turanian Languages. The languages which I formerly comprehended under the general name of South-Turanian, should, for the present at least, be treated as independent branches of speech. My work, thirty-five years ago, was that of a bold, perhaps a too bold pioneer. The materials then accessible were extremely scanty, rough-hewn, and often untrustworthy. We have learnt more caution since, and know that we have to account, not only for points of similarity, but for dissimilarities also, before we can speak with authority on the genealogical relationship of languages. I do not mean to say that my rough classification of these South-Turanian languages has been proved to be altogether wrong, but I am quite ready to admit that what is not proven' in linguistic science should be treated, for the present at least, as non-existent. Otherwise there is considerable danger of hasty conclusions impeding the free and untrammelled progress of scien tific inquiry. I still hold, for instance, that Tibetan and Burmese, or what I called the Gangetic and Lohitic languages, show traces of relationship which have to be accounted for, and which induced me to comprehend them under the common name of Bhotîya languages. I likewise hold that Siamese and what I called the Taic languages are closely connected with Chinese, and that both the Bhotiya and Taic groups point to a common origin with Chinese, though at a more distant period. The future will show whether I have guessed rightly or wrongly, for I cannot claim for my classification of these languages more than a hypothetical character. In the presence of scholars who have since made a special study of Chinese, Siamese, Tibetan, and Burmese, it would be unbecoming on my part to offer any opinion on the ultimate issues of these great linguistic problems which still await their final solution, and I gladly leave these matters to younger and stronger hands.1 For our own immediate purposes there is no necessity why we should extend our survey of languages beyond Europe and Asia. The principles of the Science of Language, with which alone we are concerned, have hitherto been elucidated almost exclusively by students of the Aryan, Semitic, the Chinese, and the Ural-Altaic, and the Malayo-Polynesian languages. This is, no doubt, an imperfection, but such imperfections exist in all sciences. Science can only advance step by step, and nowhere is this more true than 1 I give at the end a tabular survey of these North and South Turanian Languages, referring for further particulars to my 'Letter on the Turanian Languages,' published in 1854. in the Science of Language. Even after new clusters of languages have been explored and arranged into families, it will always remain extremely difficult, if not impossible, for one scholar to control the whole of the ever widening field of linguistic knowledge. There are, however, some excellent books in which the researches of scholars in different fields of human speech have been catalogued; and I can strongly recommend two works by Frederick Müller to those who wish to make themselves acquainted with the latest advances in linguistic and ethnological science, Grundriss der Sprachwissenschaft, Wien, 1876–1888, 4 vols. ; and Allgemeine Ethnographie, Wien, 1879, 1 vol. It may be useful, however, for our own purposes to add a short list of such languages and families of languages as have by this time been reduced to some kind of order, because some of them have to be used by ourselves from time to time in order to illustrate important features in the growth and decay of human speech. Tamulic Languages, Tamil, Telugu, Canarese, and Malayalan, constitute a well-defined family, with its smaller dialects, such as Tulu, and the vernaculars spoken by the Todas, Gonds, Uraon-Kols, Rajmahals, and, we may safely add, by the Brahuis. They occupy nearly the whole of the Indian peninsula, while dialects such as those of the Gonds, Uraon-Kols, Râjmahals, and Brahuis, scattered in less accessible places in the North, indicate the former more extended seats of the Tamulic or Dravidian race, before it had to make room before the |