This method of coining words is, however, by no means universal in American languages. It prevails in most of those in British America and the United States, in Aztec and various South American idioms; but in others, as the dialects found in Yucatan and Guatemala, and in the Tupi of Brazil, the Otomi of Mexico, and the Klamath of the Pacific coast, it is scarcely or not at all present.
Another trait, however, which was confounded with this by Mr. Du Ponceau, but really belongs in a different category of grammatical structure, is truly distinctive of the languages of the continent, and I am not sure that any one of them has been shown to be wholly devoid of it. This is what is called _incorporation_. It includes in the verb, or in the verbal expression, the object and manner of the action.
This is effected by making the subject of the verb an inseparable prefix, and by inserting between it and the verb itself, or sometimes directly in the latter, between its syllables, the object, direct or remote, and the particles indicating mode. The time or tense particles, on the other hand, will be placed at one end of this compound, either as prefixes or suffixes, thus placing the whole expression strictly within the limits of a verbal form of speech.
Both the above characteristics, I mean Polysynthesis and Incorporation, are unconscious efforts to carry out a certain theory of speech which has aptly enough been termed _holophrasis_, or the putting the whole of a phrase into a single word. This is the aim of each of them, though each endeavors to accomplish it by different means. Incorporation confines itself exclusively to verbal forms, while polysynthesis embraces both nouns and verbs.
Suppose we carry the analysis further, and see if we can obtain an answer to the query,—Why did this effort at blending forms of speech obtain so widely? Such an inquiry will indicate how valuable to linguistic search would prove the study of this group of languages.
I think there is no doubt but that it points unmistakably to that very ancient, to that primordial period of human utterance when men had not yet learned to connect words into sentences, when their utmost efforts at articulate speech did not go beyond single words, which, aided by gestures and signs, served to convey their limited intellectual converse. Such single vocables did not belong to any particular part of speech. There was no grammar to that antique tongue. Its disconnected exclamations mean whole sentences in themselves.