The Philosophic Grammar of American Languages, as Set Forth by Wilhelm von Humboldt

On reading the memoir, I was so much impressed with the acuteness and justness of its analysis of American verbal forms that I prepared the translation which I now submit.

In the more recent studies of the American verb which have appeared from the pens of Friedrich Müller, J. Hammond Trumbull and Lucien Adam, we have the same central element of speech subjected to critical investigation at able hands. But it seems to me that none of them has approached the topic with the broad, philosophic conceptions which impress the reader in this essay of Humboldt’s. Although sixty years and more have elapsed since it was written, I am confident that it will provide ample food for thought to the earnest student of language.

On the Verb in American Languages. By Wilhelm von Humboldt

Translated from the unpublished original. By D. G. Brinton, M.D.

You recently had the goodness to give an appreciative hearing to my essay on The Origin of Grammatical Forms.

I desire to-day to apply the principles which I then stated in general to a particular grammatical point through a series of languages. I choose those of America as best suited to such a purpose, and select the Verb as the most important part of speech, and the central point of every language. Without entering into an analysis of the different parts of the verb, I shall confine myself to that which constitutes its peculiar verbal character – the union of the subject and predicate of the sentence by means of the notion of Being. This alone forms the essence of the verb; all other relations, as of persons, tenses, modes and classes, are merely secondary properties.

The question to be answered is therefore: —

Through what form of grammatical notation do the languages under consideration indicate that subject and predicate are to be united by means of the notion of Being?

I believe I have shown with sufficient clearness that a language may have a great diversity of apparent forms, and may express all grammatical relations with definiteness, and yet when taken as a whole it may lack true grammatical form. From this arises an essential and real graduated difference between languages. This difference, however, has nothing to do with the question whether particular languages employ exclusively agglutination or inflection, as all began with agglutination; but in the languages of the higher class, it became in its effects on the mind, identical with inflection.

As languages of the higher class, one has but to name the cultivated idioms of Asia and Europe, Sanscrit, Greek and Latin, in order to apply to them the above statement. It is still more necessary, however, to understand thoroughly the structure of those languages which are on a lower plane, partly because this will convince us of the correctness of the classification, partly because these tongues are less generally known.

It is enough to take up some single leading grammatical relation. I select for this purpose the verb as the most important part of speech, with which most of the others come into relation, and which completes the formation of the sentence, the grammatical purpose of all language – and often embraces it wholly in itself. But I shall confine myself solely to that which makes the verb a verb, the characteristic notation of its peculiar verbal nature. In every language this point is the most important and the most difficult, and cannot be made too clear to throw light upon the whole of the language. Linguistic character can be ascertained through this point in the shortest and most certain manner.

The verb is the union of the subject and predicate of the sentence by means of the notion of Being; yet not of every predicate. The attribute which is united to the substance by the verb must be an energic one, a participial. The substance is represented in the verb as in motion, as connecting the Being with the energic attribute. By means of this representation, and the peculiar nature of the attribute, the verb is distinguished from the mere logical copula, with which it is liable to be confounded if these ideas are not understood. If the verb is explained merely as a synthesis of Being with any other attribute, then the origin of the tenses cannot be wholly derived from one idea, for the idea of time alone would allow only a three-fold distinction. Moreover, in such case the true and efficient nature of the verb is misunderstood. In the sentence, “The man is good,” the verb is not a synthesis of the adjective “good” with the substantive, but it is a participial of the energic attribute “to be good,” which contains a condition, having beginning, middle and end, and consequently resembles an action. Fully analyzed, the sentence would be, “He is being good.” Where the substantive verb stands without a visible predicate, as in the sentence, “I am,” then the verb “to be” has itself as the object of a synthesis, “I am being.” But as rude nations would find this difficult to comprehend, the verb “to be” is either entirely lacking, as in many American languages, or else it has an original material sense, and is confounded with “to stand,” “to give,” “to eat,” etc., and thus indicates Being as identical with the most familiar occupations.

The subject, the substance represented as in action, may be one independent of the speakers, or it may be identical with one of them, and this identity is expressed by the pronouns. From this arises the persons. The energic attribute may exert its action in various manners in the substance or between two substances; this gives rise to the forms or classes of verbs. Their action must be confined to a given point or period of time. The Being may be understood as definite or indefinite, etc., and in this is the origin of modes. Being is inseparably connected with the notation of time. This, united with the fixation of the point or period of time of an action, forms the tenses. No verb, therefore, can be conceived as without persons and tenses, modes and classes; yet these qualities do not constitute its essence, but arise from the latter, which itself is the synthesis brought about by the notion of Being. The signs of these qualities must be made to appear in the grammatical notation of the verb, but in such a manner that they appear dependent on its nature, making one with it.

The energic attribute, which aids in forming the verb, may be a real movement or action, as going, coming, living, working, etc., or merely a qualitative Being, as a being beautiful, good, mortal, or immortal. In the former case, we have a real attributive verb, in the latter a substantive verb, in which an attribute is considered as at rest, hence as an adjective. Although in both cases the nature of the verb is the same, yet in many languages this difference leads to a corresponding variety in grammatical notation.

In accordance with these ideas culled from universal grammar, the forms of the conjugations in the various languages will now be considered.

I have taken as a basis for this investigation as many American languages as I thought sufficient for the purpose, and as would not make the survey oppressive by their number; but as I do not name all of them, and pay still less attention to pointing out in what other groups of languages the peculiarities named occur, it must be understood that what is here said is not intended as a characterization of American languages. This is reserved for another study.

In order to judge how closely these languages approach grammatical perfection in this point, we must take as our criterion that condition of speech where there is a class of words, which possess verbal power, and are at the same time separated by a definite form from all other parts of speech. With reference to this condition as the highest, we must arrange in various grades all other structural forms or paraphrases of the verb.

The notion of Being, which constitutes the basis and the essence of the verb, can be indicated either,

1. As expressed independently.

2. As incorporated in the verbal form as an auxiliary verb.

3. As included in the verbal form merely as an idea.

The differences of the languages under comparison can be appreciated most correctly by means of these three headings; but it must not be forgotten that any language may use the first and one of the last two methods, and that in languages which have a substantive verb conjugated with and without auxiliary verbs, all three may be employed.

I.

When the notion of Being is expressed independently

I must except from this class all instances where the substantive verb is formed from a radical, inasmuch as this root, like any other, must assume the verbal form, and thus come under one of the two other divisions. In such case it expresses the notion of Being, either by an auxiliary, as in the German Ich bin gewesen, or simply in the form, as, I am. When it is remembered that the substantive verbs of all languages are derived from concrete conceptions and impart to these merely the general notion of Being, the above becomes still more obvious.

Now if there is no root-form for the substantive verb, and yet it is expressed independently, and not by another verbal form, this can only be done either by the position of the governing and governed words, or by linguistic elements which are not properly verbs, but only become so by this use. In the former case the substantive verb is merely understood, in the latter it appears in a definite word, but without a fixed radical.

1. When the notion of Being is understood

One of the most common forms of sentences in American languages is to bring together an adjective and a substantive, the substantive verb being omitted.

Mexican: in Pedro qualli, the Peter (is) good.

Totonaca: aquit chixco, I (am) a man.

Huasteca: naxe uxum ibaua tzichniel, this woman (is) not thy servant.

In the Mixteca language such expressions have a peculiar arrangement. The adjective must precede the substantive, or rather the predicate must precede the subject, as in the reverse case the words are understood separately, and are not connected into a sentence: quadza ñaha, the woman is bad; ñaha quadza, the bad woman.

In the language of the Mbayas, a sentence can be made with any verb by dropping the verbal affixes, by transposing a letter characterizing the nouns as such, appending an adjective suffix, and uniting this with an independent pronoun. The grammars of this language call this form a passive, but it is just as much a neuter, and is not a verb but a phrase. From iigaichini, to teach, we have n-iigaichin-igi, taught, and as first person e n-iigaichin-igi, I am taught. The initial n which accompanies all nouns in this language, is merely the possessive pronoun of the third person, added according to the usage of many of these tongues to leave no noun without a possessive; the termination igi is a particle which indicates the place where anything remains. Literally, therefore, eniigaichinigi means, I (am) the stopping-place of his teaching, i. e., one who is taught. All affixes of mode and tense, however, may be united to this phrase, so that thus it approaches a verb.

Regarded apart from the changes through tenses and modes, the union of the subject and predicate with the substantive verb omitted, is admirably adapted to express the conjunction of two words in one idea, and as the languages which make use of it also possess the ordinary forms of conjugation, they thus possess a special expression for both the forms of verbs above referred to. We shall note this particularly in the Beto language.

When the subject is not an independent part of speech, but an affixed pronoun, the analogy of this method of notation to a verbal form increases. For this is present even when no characteristic of a tense is added, simply by the union of an attribute and a pronoun. It should be remarked once for all, however, that too much weight must not be attached to whether these elements form one word or not, as this is not an infallible criterion.

The verb cannot be considered to be present as a separate part of speech, when a verb can thus be made out of any word, not merely those stamped as verbs, but also out of those which bear the express characteristics of nouns; and therefore I include all these cases in the class under consideration. For in all these languages there is in fact no verb, but only separate elements of speech with the verb omitted. Such cases are, however, interesting, as showing the gradual approach to the verb, and the effort of the instinct of language to arrive at grammatical form.

The independent personal pronoun rarely makes an element of verbal form, as in speaking it is generally worn down to an affix. When it is used to form a verbal expression, the difference of the elements is apparent. Thus, in the Carib,

In that tongue, however, this placement is not applicable to every noun, but only after certain definite verbal forms, especially in negative expressions.

The Lule language confines this notation to participials, and expresses by it the condition of the action and also its time; mil quis amaiciton,

The affixed pronouns are either special, confined to these expressions, or if elsewhere in the tongue, are not employed with verbs, or not in this manner; or they are the pronominal affixes of the verb itself.

The Maya or Yucatecan language has a special pronoun which added to any noun forms a sentence with it, and possesses the power to add the idea of the verb; Pedro en, I am Pedro. But when it stands alone, without a predicate, it loses this power, as en alone does not mean, “I am.”

In the Beto language there is, indeed, no special pronoun of this kind, as the one used is also a possessive. Its position, however, makes the difference. When it is prefixed, it is the possessive, but when suffixed it carries with it the power of the verb: humani rru, man I (am); fofei rru, bad I (am). In a similar manner this tongue forms a substantive verb, ajoi rru. The meaning of the root is not given, but it seems to mean something present, at hand. It is suggestive that in these phrases the accent is always on the pronoun, as if to signify that that is the important element.

It is very common in American languages to find the noun and the verb using the same pronouns, with the former to indicate possession, with the latter the subject. This might be explained by supposing that the action is regarded as the possession of the agent. But it is simpler to suppose that in each case the connection of the person with the noun and the verb is in the thoughts, and this relation is recognized in expression.

In this way the Mbaya language has a sort of descriptive conjugation; connecting the participles with possessive pronouns; i-iligodi, I (am) explaining; but no doubt less definitely, “my explaining,” “I to explain.”

The language of the Abipones slightly alters the possessive pronouns in some persons and uses them in a similar manner: ri-aal, I am lazy; yo-amkata, he is good.

When the verbal pronoun is used in such expressions, it is entirely identical with the verb.

This is the case with the Mexican, where the verbal pronoun united to the participle forms a sentence: ni-tlaçotlani, I (am) a lover. This expression differs from the present indicative only in the form of the root-word, ni-tlaçotla; but it cannot form another tense or mode. The grammarians call such an expression a tense indicating habit. This, however, would not be a tense but a mode, and, in fact, the term rests on a misunderstanding. That such expressions indicate habit is shown by the fact that they do not apply, like the present of the verb, to the temporary action, but convey that it is a custom, or a business; not that I am loving just now, but that I am habitually a lover.

An entirely similar instance occurs in the North Guaranay language, which also permits, besides the regular conjugation, a union of the root of the verb with a pronoun, the verb being omitted. The grammarians of that tongue say that this adds extension and emphasis to the sense of the verb. The real difference, however, is that this procedure treats the verb as a noun, and the extension comes from considering the action expressed by the verb to have become a permanent quality; a poro iuca, I kill men (ordinary conjugation); xe poro iuca, I (am) a man-killer (form with the possessive pronoun); I kill men as my business.

In both these languages, therefore, what have been represented as peculiar and separated forms, tenses indicating habit, or forms of extension, are simply erroneous explanations of quite simple constructions. In Mexican the correctness of this explanation is confirmed by the forms of the vocative, which are identical with this supposed tense, in ti tlatlacoani, O thou sinner; literally, thou who (art) a sinner.

In the above examples the verbal power lies in the pronouns. But the Mbaya language constructs verbal sentences by adding the sign of the future to any adjective without a pronoun. This sign is de, or before a vowel d: de liidi, it will be pleasant to the taste; d otiya, he will be fat. I do not find other examples, and am uncertain whether other tenses and modes are thus formed. In that case the pronouns would have to be added, and the expression would lose its peculiarity, which is that the tense sign alone carries with it the notion of Being.