The subject is usually a pronoun inseparably connected, or at least included within the tense-sign; to this the nominal subject stands in apposition. Both subjective and objective pronouns are apt to have a different form from either the independent personals or possessives, and this difference of form may be accepted as a priori evidence of the incorporative plan of structure—though there are other possible origins for it. The tense and mode signs are generally separable, and, especially in the compound tense, are seen to apply not only to the verb itself, but to the whole scope of its action, the tense sign for instance preceding the subject.
Some further observations will set these peculiarities in a yet clearer light.
Although in polysynthesis we speak of prefixes, suffixes, and juxtaposition, we are not to understand these terms as the same as in connection with the Aryan or with the agglutinative languages. In polysynthetic tongues they are not intended to form words, but sentences; not to express an idea, but a proposition. This is a fundamental logical distinction between the two classes of languages.
With certain prefixes, as those indicating possession, the form of the word itself alters, as in Mexican, amatl, book, no, mine, but namauh, my book. In a similar manner suffixes or postpositions affect the form of the words to which they are added.
As the holophrastic method makes no provision for the syntax of the sentence outside of the expression of action (i. e., the verbal and what it embraces), nouns and adjectives are not declined. The “cases” which appear in many grammars of American languages are usually indications of space or direction, or of possession, and not case-endings in the sense of Aryan grammar.
A further consequence of the same method is the absence of true relative pronouns, of copulative conjunctions, and generally of the machinery of dependent clauses. The devices to introduce subordinate propositions I have referred to in a previous essay (above, p. [346]).
As the effort to speak in sentences rather than in words entails constant variation in these word-sentences, there arise both an enormous increase in verbal forms and a multiplication of expressions for ideas closely allied. This is the cause of the apparently endless conjugations of many such tongues, and also of the exuberance of their vocabularies in words of closely similar signification. It is an ancient error—which, however, I find repeated in the official “Introduction to the Study of Indian Languages,” issued by our Bureau of Ethnology—that the primitive condition of languages is one “where few ideas are expressed by few words.” On the contrary, languages structurally at the bottom of the scale have an enormous and useless excess of words. The savage tribes of the plains will call a color by three or four different words as it appears on different objects. The Eskimo has about twenty words for fishing, depending on the nature of the fish pursued. All this arises from the “holophrastic” plan of thought.
It will be seen from these explanations that the definition of Incorporation as given by M. Lucien Adam (quoted above) is erroneous, and that of Professor Müller is inadequate. The former reduces it to a mere matter of position or placement; the latter either does not distinguish it from polysynthesis, or limits it to only one of its several expressions.
In fact, Incorporation may take place with any one of the six possible modifications of the grammatical formula, “subject + verb + object.” It is quite indifferent to its theory which of these comes first, which last; although the most usual formula is either,
subject + object + verb, or,