Sanskrit and Zend. As the leading features of this system will be brought before my readers in the next section, it is useless to enter here into details. It is an elevating spectacle to see the long life of a man of science consecrated to carrying out a deep and fertile idea, and producing a work of such magnitude and lasting value, exactly at the time when it was wanted.

5. EUGÈNE Burnouf.

What Grimm did for the Teutonic languages, Eugène Burnouf, hurried away by a premature death since the first publication of this Essay, achieved for the languages of Persia. As this subject will soon be treated fully in a report upon the latest results of the Persian researches, it will suffice here to exhibit the table of correspondence which Burnouf has established for Sanskrit and Zend in his profound works, particularly his "Commentaire sur le Yaçna” (Vol. I. Paris, 1835). The Sanskrit S (sharp dental) becomes H in Arian (as in Greek and Kymri).

To this period belong also the writings of Julius Klaproth, whose "Asia Polyglotta " (1823, 2nd ed. 1829) contains valuable materials for the lexicographal comparison of many of the Turanian languages, and important researches concerning the Tatar and Turkish tribes. His neglect, however, of the grammatical element (a reaction against Schlegel) is a great drawback to the usefulness of the philological researches of this acute and ingenious man.

THIRD PERIOD OF LINGUISTIC SCIENCE.

WILLIAM VON

HUMBOLDT

AND HIS POSTHUMOUS WORK ON THE DIVERSITY OF THE FORMATION OF HUMAN LANGUAGE.

THE desiderata of Bacon, and the general plan and fundamental views of Leibnitz had, to a considerable degree, been carried out in the course of this century. By a rare combination of philosophical thought, of philological accuracy, and of linguistic research, a method had been established for analyzing a given language and detecting its affinities with another of the same family. By this process, in the Semitic, and still more so in the Japhetic languages, the general observations of preceding philosophers on the characteristics and relative advantages or imperfections of the languages of mankind had become entirely obsolete, from being in part incomplete, in part erroneous, and scientifically speaking, inaccurate. The great desideratum then was, as a first step, to unite and examine philosophically and methodically all the different forms of human language with a full knowledge of all the modern discoveries. This want was admirably supplied by the immortal posthumous work of William von Humboldt (1835), the introduction to his analysis of the Kawi language (1836). Its title is, "On the Diversity of the Formation of Human Language, and its Influence on the Intellectual Development of Mankind." Beginning with the simplest elements of speech, the illustrious author gradually proceeds to the construction of a sentence, as the expression of intellect and thought. He then shows that the Chinese is a perfect form in its kind. In examining, explaining, and comparing the different means used by different nations to render single words susceptible of signs, destined to mark their posi

tion in a sentence, he shows that all accomplish this, more or less imperfectly, with the exception of the Sanskritic family, in which he gives the prize to the language of the Hellenes. Thus he is brought at last irresistibly to the result, that the Chinese language and the Sanskritic family represent the two extremes of all known formations of speech. With respect to the Semitic languages, he considers them as standing on the same line with the Sanskritic, in consequence of their decided tendency towards the system of inflexional forms; other formations necessarily occupy, according to him, a place between those

two extremes.

In following out this great plan of comparative philosophy respecting all the different phenomena of language, he does not enter into a particular consideration of the historical problem which is to occupy our attention. He considers it possible that the different classes of formations constitute, as it were, the stages of a continual development. It is also possible, he conceives, that such different formations may be accompanied by historical affinities, arising out of a common origin. But, he adds, this must entirely depend upon historical research*: a question into which he does not enter, nor the method of such an inquiry. He not only abstains from the historical investigation, but seems to declare, in another passage, that a complete and satisfactory classification of all languages is an impossibility, on account of the numberless varieties of formationst. In another, later passage of his work, he expresses his doubts whether there may not be a radical connexion between the Chinese and Burmese languages ‡, and gives some remarkable instances even of grammatical affinities. Now, should such a radical affinity be established, it is clear that an immense step would have been made towards proving that the languages of the great majority of mankind have a common

*

§ 5. p. xxxv. § 7. p. lxiii.

† § 24. p. cccxlvi.

Ibid.

p. ccclxxxviii.

origin. Humboldt, therefore, was far from denying such a possibility. Under these circumstances we think it safest to express the final result of his researches in the very words of the concluding sentence of his great work. These remarkable words are as follows:

"The result of what has been developed hitherto is this, as far as the expression of grammatical relations by particular signs, and the syllabic extent of words is concerned. If we consider the Chinese and the Sanskrit languages as the extreme points, there is in the other languages lying between those points, whether they keep the syllables separate, or attempt imperfectly to amalgamate them, a gradually increasing tendency to make the grammatical expression more visible, and to unite syllables to words more freely."

To have established this great result by a scientific method, with copious, sound, and thoroughly digested materials, constitutes, in my opinion, the lasting value of a work, which claims, besides, an eminent rank as being the concentration of the thoughts and researches of a man of sound judgment and profound learning, who had dedicated a great part of his active life partly to speculations on language in general, partly to a critical and detailed analysis of a variety of tongues. As to its bearing upon the great historical problem before us, although, as already observed, the author purposely refrains from entering upon the general question of the original unity or diversity of races and languages, his work will nevertheless be found to point out the most valuable landmarks for all who are bold enough to sail on this wide and dangerous ocean. Its researches belong to the Calculus sublimis of linguistic theory. It places William von Humboldt's name in universal comparative ethnological philology by the side of that of Leibnitz.

FIRST SECTION.

THE LINGUISTIC AND ETHNOLOGICAL RESULTS OF THE MOST RECENT IRANIAN, SEMITIC, AND CHAMITIC RESEARCHES OF HISTORICAL PHILOLOGY.

WITH William von Humboldt's posthumous work commences a new period for the great science of the philosophical history of language, and consequently of the primitive stage of our race's development. In my lecture of 1847, I accordingly stopped there, contenting myself with illustrating single points of my theory and method of historical investigation and the general results I had ventured to draw from those materials, by reference to some isolated researches of the last few years, including my own.

Since that time six years have elapsed, and much of what was then mere conjecture has been matured into a demonstrable fact. Whatever has been successfully achieved during that time on behalf of linguistic science, has been obtained by the critical method of philological analysis. While valuing all materials for what they are worth. I do not think that crude glossaries of languages not understood are proper materials for comparative ethnology, and I consider all conjectures and systems built upon such materials as below scientific consideration.

It is also, I think, now more generally acknowledged, and indeed proved by incontestable facts, that all critical philological results must remain incomplete and imperfect until they become historical. All philology must end in history; but the historical