being well. This does not, as you might suppose, make the language ambiguous; for the Indians speak and understand each other with great precision and clearness.

I have yet to answer your question about the f and w. There are in the Delaware language no such consonants as the German w, or English v, f, or r. Where w in this language is placed before a vowel, it sounds the same as in English; before a consonant, it represents a whistled sound of which I cannot well give you an idea on paper, but which I shall easily make you understand by uttering it before you when we meet. I am, &c.

LETTER XVI.

TO MR. HECKE WELDER.

Philadelphia, 31st July, 1816.

DEAR SIR--I have received with the greatest pleasure your two favours of the 24th and 26th inst. ; the last, particularly, has opened to me a very wide field for reflection. I am pursuing with ardour the study of the Indian languages (I mean of their grammatical forms) in all the authors that I can find that have treated of the subject, and am astonished at the great similarity which I find between those different idioms from Greenland even to Chili. They all appear to me to be compounded on a model peculiar to themselves, and of which I had not before an idea. Those personal forms of the verbs, for instance, which you mention in your letter of the 20th of June, I find The generally existing in the American languages. Spanish-Mexican Grammarians call them transitions, but they are not all equally happy in their modes of explaining their nature and use. The word "transition," however, I think extremely well chosen, as it gives at

once an idea of the passage of the verb from the pronoun that governs to that which is governed, from "I love" to "I love you." The forms of the Indian verbs are so numerous, that a proper technical term is very much wanted to distinguish this particular class, and I adopt with pleasure this appropriate Spanish name, at least, until a better one can be found.

I am sufficiently satisfied from the examples in your last letter that the Indians have in their languages "roots," or radical words from which many others are derived; indeed, I never doubted it before, and only meant to shew you by the instances of Father Sagard, and Lord Monboddo, what false ideas the Europeans have conceived on this subject. The various meanings of the word "wulit" and its derivatives, obtained, as you have shewn, by easy or natural transitions from one kindred idea to another, are nothing new in language. The Greek has the word "kalos,” which in its various meanings is very analogous to "wulit." Instances of similar "transitions" from different European idioms might be cited without end. There is one in the French, which strikes me at this moment with peculiar force. In that language, an honest man is "just" in his dealings and a judge in his judgments; but a pair of shoes is so likewise, when made exactly to fit the foot, and by a natural transition, when the shoes are too tight, they are said fo be too just (trop justes.) A foreigner in France is reported to have said to his shoemaker, complaining of the tightness of a pair of new made shoes: "Monsieur, ces "souliers sont trop équitables," I remember also an English song, beginning with the words "Just like love," where you see the word "just" is employed without at all implying the idea of equity or justice. But justice is strict, exact, correct, precise, and therefore the word just is employed for the purpose of expressing these and other ideas connected with that to which it was first applied.

I have made these trite observations, because I am

well aware that many a priori reasoners would not fail to find in so many words of different meanings derived from the same root, a proof of the poverty of the Indian languages. They would say that they are poor, because they have but few radical words, a conclusion which they would infallibly make without taking the pains of ascertaining the fact. If they were told that the Greek (the copiousness of which is universally acknowledged) has itself but a comparatively small number of roots, they would not be at a loss to find some other reason in support of their pre-conceived opinion. I have read somewhere (I cannot recollect in what book.) that there was not a greater proof of the barbarism of the Indian languages, than the comprehensiveness of their locutions. The author reasoned thus: Analysis, he said, is the most difficult operation of the human mind; it is the last which man learns to perform. Savage nations, therefore, express many ideas in a single word, because they have not yet acquired the necessary skill to separate them from each other by the process of analysis, and to express them simply.

If this position were true, it would follow that all the languages of savage nations have been in the origin formed on the same model with those of the American Indians, and that simple forms have been gradually introduced into them by the progress of civilisation. But if we take the trouble of enquiring into facts, they will by no means lead us to this conclusion. It is not many centuries since the Scandinavian languages of the North of Europe were spoken by barbarous and savage nations, but we do not find that in ancient times they were more comprehensive in their grammatical forms than they are at present, when certainly they are the least so, perhaps, of any of the European idioms; on the other hand, the Latin and Greek were sufficiently so by means of the various moods and tenses of their verbs, all expressed in one single word, without the use of auxiliaries; and yet

these two nations had attained a very high degree, at least, of civilisation. I do not, therefore, see as yet, that there is a necessary connexion between the greater or lesser degree of civilisation of a people, and the organisation of their language. These general conclusions from insulated facts ought constantly to be guarded against; they are the most fruitful sources of error in the moral as well as in the natural sciences. Facts ought to be collected and observations multiplied long before we venture to indulge in theoretical inferences; for unobserved facts seem to lie in ambush, to start up at once in the face of fine spun theories, and put philosophers in the wrong.

I wish very much that some able linguist would undertake to make a good classification of the different languages of the world (as far as they are known) in respect to their grammatical forms. It was once attempted in the French Encyclopedia, but without success, because the author had only in view the Latin and Greek, and those of the modern languages which he was acquainted with. His division, if I remember right, was formed between those idioms in which inversions are allowed, and those in which they are not. Of course, it was the Latin and Greek on the one side, and the French, Italian, &c. on the other. This meagre classification has not been generally adopted, nor does it, in my opinion, deserve to be. A greater range of observation ought

to be taken.

I do not pretend to possess talents adequate to carrying into execution the plan which I here suggest; but I beg you will permit me to draw a brief sketch of what I have in view.

I observe, in the first place, in the eastern parts of Asia, a class of languages formed on the same model, of which I take that which is spoken in the empire of China, as it stood before its conquest by the Tartars, to be the type. In this language, there is but a very small num

ber of words, all monosyllables. As far as I am able to judge from the excellent grammars of this idiom of which we are in possession, the words convey to the mind only the principal or leading ideas of the discourse, unconnected with many of those accessary ideas that are so necessary to give precision to language, and the hearer is left to apply and arrange the whole together as well as he can. It has but few or no grammatical forms, and is very deficient in what we call the connecting parts of speech. Hence it is said that the words spoken are not immediately understood by those to whom they are addressed, and that auxiliary modes of explanation, others than oral communication, are sometimes resorted to, when ambiguities occur. As I am no Sinologist, I will not undertake to say that the description which I have attempted to give of this language, from the mere reading of grammars and dictionaries, is very accurate, but I venture to assert that it differs so much from all others that we know, that with its kindred idioms, it deserves to form a genus in a general classification of the various modes of speech. From its great deficiency of grammatical forms, I would give to this genus the name asyntactic.

My second class of languages would consist of those which possess, indeed, grammatical forms, sufficient to express and connect together every idea to be communicated by means of speech, but in which those forms are so organised, that almost every distinct idea has a single word to convey or express it. Such are the Icelandic, Danish, Swedish, and even the German and English. Those forms of the nouns and verbs which are generally called declensions and conjugations, are in these languages the result of an analytical process of the mind, which has given to every single idea and sometimes to a shade of an idea, a single word to express it. Thus when we say "of the man," here are three ideas, which in the Latin are expressed by one single word hominis." In the locution "I will not," or "I am not willing,” and