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

The Othomi language makes use in such expressions not only of the pronouns but of all the affixes of the verb, and conjugates a noun together with its article, treating it as a verbal radical: qui-no-munti-maha, Thou wert the enriched. Here no-munti is “the enriched,” and all the remaining syllables are verbal inflections. Sandoval, who wrote a grammar of the language, explains no as an auxiliary verb; but with the noun he calls it an article, as it is, and he evidently misunderstood the expression. It is wholly a verbal, but as this procedure can be applied to any noun whatever, such an expression is far removed from a real, well-defined verbal form.

The same language has another peculiar form with the possessive, which can only be explained by supplying an omitted verb. Na nuhti means “my property;” but if to this is added the abbreviated pronoun used as a verbal affix, na-nuhti-gā, the words mean, “this property belongs to me,” or, “my property is it, mine.”

In the grammatically obscure consciousness of these people, the ideas of verbal and merely pronominal expression are confounded, as also in the Brazilian language, where “my father” and “I have a father” are expressed by the same word.

The advantages which these languages derive from the formation of sentences with the verb omitted are two.

They can change any noun into a verb, or at least they can treat it as such. It is true that this can also be done by a substantive verb when one is found, but as the languages in question unite the noun to the verbal flexions, their freedom is much greater.

The second advantage is, that when it is desirable to discriminate clearly between the two kinds of verbs, the one which has at base an energic attribute, the other which merely expresses the relation of predicate to subject, a thing to its qualities, this end can be much better reached by the process described than even by the substantive verb, which, by its full verbal form, always recalls the action of an energic attribute.

Many of the languages named include in these expressions particles of time, thereby obscuring the distinction referred to. But in others this is not the case. Thus in the Maya and Beto there are two conjugations, one with the pronoun without time particles, and one with them; and as in both these tongues the present of the true conjugation has a characteristic tense sign, a separate aorist of the present is formed by the other conjugation, which our cultivated tongues cannot express so conveniently.

2. When the notion of Being is expressed by a special word, but without a phonetic radical

Although the assumption here expressed sounds at first rather enigmatical, yet one can soon see that if the notion of Being is to be conveyed without a phonetic radical, it can only be done through the sign of the person, that is, in the pronoun, with or without a tense sign. This is actually the case in two languages, the Maya and the Yaruri.

We have already seen that in the Maya there is a special pronoun which unites a predicate to the idea of person into one sentence. There is also another which by itself conveys the idea of the verb, and of which each person has the signification both of the pronoun and the substantive verb, “I” and “I am,” “thou” and “thou art,” etc. Not only is it so used in the present, but it can take the signs of the tenses. It is distinguished from the pronouns previously referred to in the first and second persons of both numbers only by a prefixed t, as follows:

This similarity leads to the thought that a true phonetic radical may exist in this t, and may induce us to consider this word not as a pronoun but as a substantive verb. But this makes no difference. The fact remains that the word is used both as a simple pronoun and also as a substantive verb. In the translation of the Lord’s Prayer, the word toon is a simple pronoun. If t is a radical, it may just as well come from the pronoun. Some languages offer clear examples of this. In the Maipure the expression for the third person singular recurs with all the other persons, as if this sound meant the person, the man generally, and the first and second persons were denoted as the “I-person,” “thou-person,” etc. In the Achagua language the same radical occurs in all the pronouns, but does not, as in the Maipure, stand alone for the third person singular, but in it, as in the other persons, appears as an affix.

At any rate, this pronoun answers, in the Maya, all the purposes of the substantive verb, and there is no other in the language.

It is quite intelligible that in the conceptions of rude nations the idea of an object, and especially of a person, cannot be separated from the idea of his existence. This may be applied to the forms of expression above mentioned. What seems a violent and ungrammatical omission of the verb, is probably in those people an obscure association of thoughts, a non-separation of the object from its being. Probably it is from the same source that in some American languages every adjective is so considered that it includes not the idea alone, but the expression, “it is thus, and thus constituted.”

In the Yaruri language the absence of a phonetic radical meaning “to be” is yet more apparent. Each person of the pronoun is a different word, and they have no single letter in common. The pronoun which has verbal power is almost identical with the independent personal pronoun. The tense signs are prefixed to it. Thus, que, I am; ri que, I was, &c. This ri, however, is merely a particle which expresses that something is remote, and corresponds with our “from.” Ui-ri-di, there was water there, literally “water far is” (from us is). The subjunctive of this substantive verb is given as ri, “if I were.” This means, however, “in,” and is a particle. The notion of Being is added, as in the pronoun; and the ideas, “in the being,” and “if I were,” pass into each other.

Strictly speaking, both the verbal notations here expressed are identical with those already mentioned. Here also the verb is supplied by the mind. The difference is that in the latter case the pronouns alone signify being, and contain this notion in themselves, whereas in the other cases this notion arises from the conjunction of subject and predicate. Then also in the Maya language there is a special pronoun for this sole purpose. As far as the forms go, they entirely resemble those of a true verb, and if que and ten are regarded as mere verbs substantive, one who did not examine their elements would take them to be true verbs like the Sanscrit bhū, the Greek ειμι, and the Latin sum. The example of these languages thus teaches that in the analysis of the substantive verbs of other tongues it is not necessary that a common phonetic radical need be employed.

In the Huasteca language the substantive verb is replaced by affixing a tense sign to the independent pronouns; naua itz, I was, tata itz, thou wert, etc. But the case is not the same. The pronoun receives the verbal power by the suffix itz, and this appears only in later times to have become a sign of the preterit, and in an earlier period to have had a general sense. The mountaineers who seem to have retained the older forms of the tongue use the itz, not only in the preterit, but in the present and future. It was doubtless the expression of some general verbal idea, as, to be, to do, etc.

II.

The notion of Being is incorporated with the verb as an Auxiliary

Auxiliary verbs are used only for certain tenses, or form the entire conjugation. The former arises from accidental causes having relation only to these tenses, not to the verb in general. The latter readily arises when a substantive verb offers an easy means of conjugation by uniting with another verb. Sometimes the conjugation by means of an auxiliary shows that the linguistic sense of a notion sought something beyond the person and tense signs to express the verbal power itself, and therefore had recourse to a general verb. This can, indeed, only be constituted of those elements and a radical; but the want in the language is thus supplied, once for all, and does not return with every verb.

An excellent example of this is furnished by the Maya conjugation. In an analysis of it we find an element that neither belongs to the root, nor is a person, tense or mode sign, and when their varieties and changes are compared, there is evident throughout a marked anxiety to express the peculiar verbal power in the form of the verb.

The conjugation in the Maya language is formed by affixing the pronouns and mode and tense signs to the stem. The pronoun is, according to a distinction to be noted hereafter, either the possessive pronoun or that one which, without verbal power in itself, yet receives it when a predicate is attached to it to form a sentence.

Besides this, the suffix cah accompanies all verbs in the present and imperfect; and the suffix ah accompanies all transitive verbs through the remaining tenses, except the future. Present, 1st person, sing., canan-in-cah, I guard; imperf. 1st pers. sing., canan-in cah cuchi; perf., 1st pers. sing., in canan-t-ah. In is the possessive pronoun, cuchi the sign of the imperfect, t in the perfect is a euphonic letter.

The idea of transitive verbs is here taken somewhat narrower than usual. Only those are included which govern a word outside of themselves. All others are considered intransitive, even those which of themselves are active, but either have no expressed object (as, I love, I hate, etc.), or the word which they govern is in the verb itself, as in the Greek οικοδομεο, οικουρεω. As these can govern a second accusative, the object incorporated in the verb is included in the idea they express.

The tenses of the intransitive verbs, except the present and imperfect, while they drop ah and the possessive pronoun, are formed with that pronoun which forms sentences with a predicate.

There are cases where not only the present omits cah, but where the stem, if it ends in ah as is often the case, drops it, and substitutes ic. The signification then alters, and indicates an habitual action or quality. As ic is the sign of the gerund, this change appears to be the transformation of the verb into a verbal, and to effect this, it must be united to that pronoun which serves as the substantive verb; ten yacunic, I love, properly, I am loving (habitually).

What cah and ah mean by themselves, we are not informed. Where cah is attached to the stem of some verbs it signifies intensity. Ah is as a prefix the sign of the male sex, of the inhabitant of a place, and of names derived from active verbs. Hence it seems to have meant at first person, man, and later to have become a pronoun, and finally an affix. It is noteworthy that the same difference exists between ah and cah, as between en and ten. The c may therefore be a radical sound. In the conjugation, cah is treated wholly as a verb. For in this the possessive pronoun is always prefixed; and as in the present and imperfect it is placed after the stem of the verb and before cah, it is evident from the difference between the two forms canan-in-cah and in-canan-t-ah, that in the former cah, and in the latter canan, are regarded as the verbs. Canan-in-cah is precisely as the English “I do guard.”

Cah is consequently a true auxiliary verb; ten, when it appears in conjunction with en must have the notion of Being understood: ah appears to be of similar nature, but as it appears only in the conjugation of transitive verbs, it is a verbal sign, and thus receives its verbal power. That cah and ah do really possess this powever is evident from the fact that they are never used whenever either of the pronouns which are always associated with the notion of Being is present.

Except in the future of transitive verbs, there is no instance in the conjugation where the stem of the verb is not accompanied by one of these four syllables, all of which indicate Being, and all of which have the force of auxiliary verbs.

The future of transitive verbs not only does not take any of these syllables, but even rejects ah when it is the terminal syllable of the stem. In this case no other termination replaces it. On the contrary, all other verbs receive a new suffix in their future, varying as they are of one or many syllables. The nature of these suffixes has not been explained.

The definite results of this analysis are as follows:

1. The Maya language possesses in its conjugation, besides the inflection syllables of the persons and tenses, another element, which, except in the simple future of transitive verbs, distinctly carries with it the notion of Being; in the future of most verbs there is such an element, but of unknown origin, and it only fails in the future of one class of verbs.

2. This language displays an effort to express, besides the other purposes of the verb, particularly its synthetic power, which is all the more apparent as it uses different means in different cases, but all designed to accomplish the same purpose.

The Yaruri language constructs the whole of its conjugation in a yet simpler manner by means of an auxiliary verb.

The union of the pronoun and the tense sign which, as we have already seen, forms the substantive verb, affixed to the stem, completes the inflections of the one and only conjugation of attributive verbs, except that the independent pronouns are prefixed. Neither the stem nor the auxiliary words suffer any changes, except the insertion of an n in one person. The union remains, however, a loose one, and when person and tense are manifest by the connection, the auxiliary verb is omitted. This happens in certain verbs ending in pa. These, contrary to the usual rule, change in the perfect this termination to pea, by which the tense is made apparent, and as the person is evident from the prefixed personal pronoun, the auxiliary can be dropped without danger of obscurity.

The formation of certain tenses by means of auxiliaries is also frequent in American languages.

An optative of this nature in the Lule language has already been mentioned.

In the Mixteca tongue the imperfect is thus formed from the present, which carries with it the personal sign, and the perfect without its personal sign, a proceeding which, however rude and awkward it may be, shows a just appreciation of the peculiarity of this past tense, which expresses an action as going on, and therefore present in past time. The expression of continuous action is placed first, “I sin,” then this is more precisely defined by the mark of past time, “this was so;” Yo-dzatevain-di-ni-cuvui. Yo is the sign of the present, ni of the preterit, di is the pronoun; the other two words, to sin and to be: “I was sinning.”

The sign of the present, yo, is probably an abbreviation of the verb yodzo, I stand upon or over something, and so there is a second auxiliary in the sentence. This may often be a means of discovering the origin of tense signs, as, especially in American tongues, tenses are often formed by the union of verbs, as also occurs in Sanscrit and Greek.

The Othomi distinguishes certain past tenses, which, however, are separated by other characteristics, by a prefixed xa, which is called the third person singular of a substantive verb. As these tenses are precisely those in which the action must be completed, the perfect, pluperfect and future perfect, not, however, the imperfect and past aorist, such a connection is very suitable. Of this verb we have only xa, and there is another substantive verb gui, which itself takes oca in its conjugation.

The Totonaca language unites the perfect, in the person spoken of, with the third person singular of the future of the substantive verb, to form a future perfect. This is no completed form, but only an awkward sequence of two verbs; yc-paxquilh-na-huan, literally, “I have loved, it will be,”=“I shall have loved.”

In similar manner the substantive verb is used to form a tense of the subjunctive.

The sign of both the perfects in this tongue is the syllable nit, and niy means “to die.” It is not improbable that this affix is derived from this verb. Death and destruction are suitable ideas to express the past, and some languages employ negative particles as signs of the preterit. In the Tamanaca this is not exactly the case, but the negative particle puni added to a word which signifies an animate thing, intimates that it has died; papa puni, the deceased father, literally, “father not.” In the Omagua tongue the same word signifies old, dead, and not present.

In the Maipure and Carib tongues the negative particles ma and spa are also the signs of the preterit. Bopp’s suggestion that the Sanscrit augment was originally a privative finds support in this analogy. Yet I would not speak conclusively on this point, as probably that, the Greek augment ε, and the Mexican o, are only lengthened sounds, intended to represent concretely the length of the past time. At any rate one must regard the negation as an actual destruction, a “been, and no longer being,” not as simply a negation of the present.

III.

The notion of Being is present in the Verbal form only in idea

In this case the verb consists only of the stem, and the person, tense, and mode signs. The former are originally pronouns, the latter particles. Before they are worn down by use to mere affixes, the three following cases may arise:

1. That all three of these elements are equally separable and loosely connected.

2. That one of the two, the person or the tense and mode signs, obtains a closer connection with the stem, and becomes formal, while the other remains loosely attached.

3. That both these are incorporated with the stem, and the whole approaches a true grammatical form, although it does not fully represent it.

Case 1st