in which he undertook to define the nature and aim of his study, and to vindicate its dignity and independence. This dissertation was followed by another in 1821,"On the Province of the Historian," which, in spite of its title, was scarcely anything more than a preliminary to his linguistical researches; a generalized and independent development of that element of the science which approximates it closely to that of history. The third dissertation, read in 1822, exhibits a still more decided influence of the Sanscrit, and the importance of this language for the study of the philosophy of human speech. It was "On the Origin of the Grammatical Forms, and their Influence on the Development of our Ideas,"+ an essay abounding in significant suggestions in reference to the historical growth, the internal structure and the general nature of human speech. These three dissertations may be looked upon as the first expansion and intensification of the views advanced but fragmentarily some years before in his prospectus relative to the Basque, and in the preface to his translation of the Agamemnon of Eschylus. They imparted to his studies a more philosophical direction, and led the way to more important and decisive results. For if to the English and the French be due the credit of having first brought the languages of the East to the knowledge of Europe, it was reserved for Humboldt to win Germany the honor of linking that knowledge to the ultimate and highest interest of man, and to transform it from ideal points of view into the order and consistency of a universal science.

But notwithstanding these advances, no one was more conscious than Humboldt himself of the insufficiency of the results, and of the necessity of a more extended examination of details in order to secure a safe basis for his theory. It is on this account that, when he found that the Chinese stood in contradiction to some of the general principles of his dissertations, he did not shun the labor of undertaking the study of this language likewise, and of comparing its peculiar characteristics with the results of his previous researches. It was then that the celebrated letter originated, addressed to Abel-Rémisat, Sur la nature des formes grammaticales en général, et sur le génie de la langue Chinoise en particulier: an essay in which his previously developed views concerning the nature of the grammatical forms, the origin, development, and general structure of language, are partly corrected, partly exWerke, vol. 3, p. 241, seq.

Werke, vol. 1, p. 1, seq.

panded, or more clearly defined, with perpetual reference to the apparently abnormal and irregular character of the Asiatic idiom in question.

But there was still another point which yet awaited his attention. This was "The Nature of Writing, and its relation to Language in general." This subject suggested itself naturally to him from Champollion's new discoveries respecting the hieroglyphics, to the study of which he applied himself with great assiduity, and, in connection with it, also to the Coptic. That his researches in this direction, too, were as successful as they were earnest and profound, is manifest from several important dissertations, which he read before the Academy, chiefly during the years 1824 and 1825. They are entitled: "On the Phonetic Hieroglyphics of Champollion the Younger;" "On four Egyptian Lionheaded Statues ;""On the Relation between Writing and Speech;" and lastly, his incomplete paper, "On Alphabetic Writing, and its Connection with the Structure of Language."*

We have now seen enough of Humboldt's proceeding to perceive that he aimed at nothing short of the universality and necessity of fundamental law in his linguistical studies. He could not, therefore, consider his position safe as long as there were other languages, or groups of them, to be examined, and compared with the results already obtained, in reference to a philosophical theory of speech. It was thus that, in 1827, he commenced a new series of researches ; and this time they were directed to the group or groups extending from Sumatra to Easter Island, and from New Zealand to the Sandwich Islands; in other words, to the groups of the entire Asiatic and Australian island-world, in which he suspected the existence of a middle ground between the domains of the Hindu and the American languages. We thus find him already, in 1828, reading before the Academy an essay "On the Language of the South Sea Islanders," and between 1829 and 1831 applying himself with new zeal to the study of the Mexican.t

But he soon relinquished the American languages to younger investigators, to concentrate his attention within the limits of a narrower sphere, to the point where he suspected a contact between the civilization of the Hindus and

*Werke, vol. 7, p. 294; and earlier separate edition, Paris, 1827. Its date is March, 1826.

Werke, vol. 6, p. 488, seq.; vol. 4, p. 302, seq.; vol. 6, p. 426, seq.; vol. 6, p. 526, seq.

the members of the Malay group. Such a point of contact he found in the island of Java, in which there were manifest traces of Hindu influence, and where this influence culminated in the Kawi language, a peculiar, learned idiom, in its character and object similar to the Sanscrit.

The new study was soon attended with results, and he read an essay " On the Kawi Language" before the Academy as early as January, 1831. And so great an importance did he attach to this ancient idiom, as a point of departure for a more extended survey of the Malay group, that he resolved to subject both its grammatical and lexical elements to a minute analysis. His plan in doing so was, in the first place, to prove it to be the result of the epoch at which Hindu culture flourished on the island, and then, by an elimination of the Malay element, to make the latter the basis for an intended examination of the remaining idioms akin to it. It was to this study that Humboldt devoted the whole of the remainder of his life, and it gave rise to his great work "On the Kawi Language," the completion of which, however, he never lived to see, and which, as we have it now, was edited for the Academy by Bushmann, after the author's decease. The long and profoundly philosophical introduction to this work, however, as well as the portion of it which treats of the connection between India and Java, both of them have, fortunately for us, received the final touches of the author himself.*

Before speaking of this introduction, which embodies all the results of Humboldt's researches in that direction, it is necessary to notice a few more minor treatises composed between the years 1827 and 1829, all of them relating to important points connected with the theory of speech. The first of them is a fragmentary dissertation "On the Dualis,"+ read before the Academy in 1827, in which the method of his science is developed with a clearness and precision such as none of his previous writings ever exhibited, and in which the grammatical form in question is analysed in its most intimate connection with the nature of speech. The next is the no less remarkable and profound essay "On the Affinity between the Local Adverbs and the Pronoun in certain languages," in which the origin and function of the pronoun are

* Uber die Kawi-Sprache auf der Insel Java. 3 vols. Berlin, 1836–1339, Werke; vol vi., p. 562, seq.

In the Academy's Abhandlungen, Berlin, 1830.

discussed, with illustrative examples from the Japanese, the Armenian, and the Tongish. In 1828, he read before the French Institute a dissertation "On the Affinity of the Pluperfect, the Reduplicating Aorist, and the Attic Perfect of the Greek, with the Tense-formation of the Sanscrit ;" and the same year we find an English essay "On the best means of ascertaining the Affinities of Oriental Languages," addressed to Sir Alexander Johnston, and read before the Royal Asiatic Society of London on the 14th of June. He had, in a similar manner, some time before, read an essay "On the Nature of the Verb," based upon observations on the American languages; and there are a number of other dissertations of a similar character which have thus far never appeared in type.

From all the documents which we have here enumerated it would not be difficult to construct a system of the philosophy of languagé as designed by Humboldt. But the author has exempted us from this labor by drawing the balance of the results of his immense researches, in the admirable introduction already spoken of, in which, with a profundity and acuteness rarely equalled, he discusses "The Structural Differences of Human Speech, and their Influence on the Intellectual Development of the Human Race."‡ Here, once more, he, as in fact in all his previous studies, links language and linguistics not merely to ethnography, but to the historical development of the race, to civilization, culture, and, in fine, to the highest problems connected with the nature of man. It is this introduction, therefore, which, in an attempt to expound Humboldt's system, shall have to form the main basis for our proceeding, while, at the same time, it will be our duty not to neglect such hints as we may be able to glean from the remaining portions of his writings already mentioned.

In this exposition of his system, Humboldt discusses the question concerning the origin, the definition, and the essential nature of language, the process of speech as exhibited in its constitutive elements, in articulation, the relation between thought and sound, the formation of roots, words, and grammatical forms, &c.; then, further, the organic principle and character of language, the idea of language, and the classifica

1836.

Reprinted in Werke, vol. vi., pp. 1-425. Original quarto edition, Berlin,

tion of languages; and, finally, the development of language in history, and its relation to the latter. We shall now endeavor to give a brief outline of Humboldt's answers to all these questions.

The question concerning the origin of language is older than the time of Plato's Cratylus, from which we perceive that at that time it was current even among the sophists, and must have been so before them among the philosophers. It recurs again in Aristotle and his followers, and among the moderns in Bacon, Hobbes, Berkeley, Leibnitz, Spinoza, and the French materialists of the last century. Ail these philosophers agree in attributing a human origin to language, and they only differ in their method of accounting for it; some making it a matter of necessity and invention, others the natural result of our physical and intellectual organization. Among the Germans, the empirical psychologists of the eighteenth century attempted a pragmatical explanation, and like many of their predecessors, made human speech the result of arbitrary convention, while their opponents, the theologians, vindicated for it no less. august an origin than a divine one, by claiming the Divinity himself as the first teacher of the human race. The latter of these hypotheses obtained very extensively, until the time of Herder, who supplanted it by investing its human origin with a profounder significance.. According to him, language is neither so far above man that the Divinity should have been necessary to invent it for him, nor so far below man that the brute should have been able to invent it. It is the necessary and conjoint result of sensibility and reflection, both of them acting upon the basis of man's natural organization and of his connection with the external world. It is his reflection that has converted the sounds of nature into significant signs, and invested them with a human element; and it is, again, his reflection which, in connection with feeling, has converted the forms and colors of the external world into sounds of speech.

Language is thus neither the inevitable product of mere physical organization, nor the mechanical manifestation of mere emotion, and least of all is it the result of arbitrary social convention; but it originates in the depths of the conscious human soul, and constitutes the distinctive characteristic of our race ab extra, as reason does that from within. Now, although Herder's theory was much more poetical than philosophical, Humboldt could yet scarcely do otherwise than follow in its traces, and develop more consistently